Full text of "Chemistry and Technology of Explosives - Volume III"

See other formats

Chemistry and Technology
of Explosives

Vol. Ill

TADEUSZ URBANSKI

Department of Technology, Politechnika
Warszawa

Authorized translation by
MARIAN JURECKI

edited by
SYLVIA LAVERTON

PERGAMON PRESS

OXFORD • LONDON • EDINBURGH • NEW YORK
TORONTO • SYDNEY • PARIS • BRAUNSCHWEIG

PWN - POLISH SCIENTIFIC PUBLISHERS

WARSZAWA

Pergamon Press Ltd , Headington Hill Hall, Oxford

4 & 5 Fitzroy Square, London W. 1

Pergamon Press (Scotland) Ltd., 2 & 3 Teviot Place, Edinburgh 1

Pergamon Press Inc., 44-01 21 st Street, Long Island City, New York 11101

Pergamon of Canada, Ltd., 6 Adelaide Street East, Toronto, Ontario

Pergamon Press (Aust.) Pty. Ltd., 20-22 Margaret Street, Sydney, N.S.W.

Pergamon Press S. A. R. L., 24 rue des Ecoles, Paris 5 e

Vieweg & Sohn GmbH, Burgplatz 1, Braunschweig

PANSTWOWE WYDAWNICTWO NAUKOWE

PWN - Polish Scientific Publishers

Warszawa

Scan by Amenf
Interim image release.

First English edition 1967

Title of the original volume
Chemia i technologia materialow wybuchowych

Library of Congress Catalog Card No. 63-10077

\±

Printed in Poland (D. U. A. M.)
1494/67

CONTENTS

Page

Preface to Volume III xiii

Part 1

NITRAMINES

CHAPTER I. GENERAL INFORMATION

Structure and chemical properties of nitramines 1

Preparation of nitramines 8

Direct nitration 8

Indirect nitration 10

Nitramines as explosives 13

' Literature 13

CHAPTER II. ALIPHATIC NITRAMINES AND NITRAMIDES

Nitramine (nitramide) 15

Methylnitramine 16

Methylenedinitramine 17

Ethylenedinitramine 18

Other nitramines deriving from ethylenediamine 20

Nitrocyanamide 21

Nitroguanidine 22

Physical properties 23

Chemical properties 25

Explosive properties 29

The preparation of nitroguanidine 31

Nitrourea 33

Other aliphatic nitramines 34

Dinitrodimethyloxamide (MNO) 34

Dinitrodiethyloxamide 35

Dinitrodimethylsulphamide 36

Nitrodiethanolamine dinitrate (DINA) 36

Dinitrodi-(^-hydroxyethyl)-oxamide dinitrate (NENO) 37

Dinitrodimethyldiamide of tartaric dinitrate 37

Dinitrodi-(/?-hydroxyethyl)-sulphamide dinitrate 38

Literature " 38

V J CONTENTS

CHAPTER III. AROMATIC NITRAMINES

Tetryl 40

Nitration of dimethylaniline 41

Nitration of dinitromethylaniline 44

General rules for the preparation of tetryl 47

Physical properties • 48

Chemical properties 51

Explosive properties 53

Toxicity 56

Tetryl manufacture 56

Homologues and analogues of tetryl 62

The polycyclic analogues of tetryl 68

Nitramino-esters of nitric acid 70

Nitraminonitrophenols 72

Nitramino-azoxy compounds 73

Nitro methylene blue 73

Literature 74

CHAPTER IV. HETEROCYCLIC NITRAMINES

Cyclonite 77

Physical properties 78

Chemical properties 80

Explosive properties 84

Toxicity 86

Cyclonite manufacture 87

1. The action of nitric acid on hexamine 87

2. Preparation of cyclonite from hexamine, nitric acid and ammonium nitrate 105

3. Preparation of cyclonite from sulphamic acid, formaldehyde and nitric acid .... 107

4. Preparation of cyclonite from paraformaldehyde, ammonium nitrate and acetic an-
hydride 109

5. Preparation of cyclonite from hexamine dinitrate, ammonium dinitrate and acetic
anhydride Ill

The theory of cyclonite formation by methods 4 and 5 113

Octogen 117

Homocyclonite 119

Nitro derivatives of melamine 120

Nitrosamines 121

Literature 125

Part 2
PRIMARY EXPLOSIVES: INITIATORS
CHAPTER I. GENERAL INFORMATION
Literature 131

CHAPTER II. FULMINIC ACID AND ITS SALTS

Fulminic acid 132

■ MlKBiwy fulminate . . . . . . . . . .,«.«.. , . . . . . . . . . . 135

CONTENTS vii

Physical properties 136

Chemical properties 139

Explosive properties 146

Toxicity 149

Mercury fulminate manufacture 149

Other salts of fulminic acid 1 57

Literature 158

CHAPTER III. HYDRAZOIC ACID AND ITS SALTS

Hydrazoic acid 161

Lead azide 169

Lead azide manufacture 178

The continuous method of lead azide manufacture (according to Meissner) 179

Silver azide 182

Cupric azide 185

Other metal azides 185

Organic azides 191

Literature 196

CHAPTER IV. OTHER INITIATING EXPLOSIVES

Diazo compounds 201

Dinitrobenzenediazo-oxide (dinitrodiazophenol) 201

The derivatives of aminoguanidine 206

Tetrazene 206

Nitrosoguanidirie 210

Cyanamide salts 211

Nitrocyanamide salts 211

Nitrophenol salts 212

Lead picrate 212

Lead styphnate 213

Other styphnates 220

Lead dinitroresorcinate 220

Nitrosophenol salts 221

Nitramine salts 221

"Isonitramine" (nitrosohydroxylamine) salts 221

Salts of metazonic acid 224

Salts of oxalic acid 224

Peroxides 225

Acetylene and its salts (acetylides) 227

Cuprous acetylide 227

Silver acetylide 229

Various initiators 229

Nitrogen sulphide 229

Nitrogen selenide 229

Salts of thiocyanic acid 230

Complex salts 230

Silver perchlorate 232

Initiating compositions 232

The preparation of primer compositions 235

Compositions for explosive rivets 240

Literature 240

v jij CONTENTS

Part 3
COMPOSITE EXPLOSIVES

General information 245

CHAPTER I. HIGH EXPLOSIVES

Fusible explosives 247

Mixtures of nitro compounds 247

Mixtures with ammonium nitrate 253

Manufacture and selection of fusible mixtures 255

The phlegmatization of fusible mixtures 257

Semi-fusible and infusible explosives 258

Mixtures with nitrates— mainly with ammonium nitrate 259

Mixtures with aluminium and other metals 266

Mixtures with chlorates and perchlorates 274

Mixtures with potassium and ammonium perchlorates 278

Plastic explosives 281

Incompatibility in explosive mixtures 283

Literature 285

CHAPTER II. LIQUID EXPLOSIVES

Historical 288

Mixtures with nitrogen dioxide, nitric acid and tetranitromethane 288

Mixtures with hydrogen peroxide 290

Mixtures with liquid oxygen (Oxyliquits) 290

Liquid rocket propellants— propergols 291

Mixtures with nitrogen dioxide 291

Mixtures with nitric acid 292

Hydrogeji peroxide H 2 2 299

Hydrazine 305

1,1-Dimethylhydrazine (UDMH) 308

Mixtures with liquid oxygen and ozone 309

Nitric esters 309

Ethylene oxide 310

Attempts to increase the energy of liquid mixtures for rocket propulsion 310

Mixtures with powdered metals 311

Boron, silicon and beryllium compounds 311

Organometallic compounds 312

Fluorine and its derivatives 312

Mixtures with perchloric acid 313

Reactions of free atoms or radicals 316

General considerations 316

Final remarks 318

Literature 319

CHAPTER III. BLACKPO WDER

Historical - 322

Composition of blackpowder 324

CONTENTS IX

Modified blackpowder 330

Theory of the burning of blackpowder 335

Explosive properties of blackpowder 340

The manufacture of blackpowder 342

Raw materials 342

Milling the ingredients 345

Mixing the ingredients 349

Pressing 352

Corning . .- 354

Finishing 356

Blending 359

Cannon powder 359

Safety in blackpowder factories 361

Literature 363

CHAPTER IV. COMPOSITE PROPELLANTS FOR ROCKETS

General information 365

Mixtures with the salts of perchloric acid 367

Mixtures of perchlorates with elastomers. Thiokol propellants 368

The technology of the manufacture of rocket charges containing composite propellants with

thiokol 373

Mixtures of perchlorates with other elastomers 380

Mixtures of perchlorates with plastics 380

Mixtures with ammonium nitrate 383

New method of mixing ingredients of composite propellants 389

Various composite propellants and their characteristics 392

Mixtures with ammonium picrate 393

Explosive properties of composite propellants 393

Literature 393

CHAPTER V. MINING EXPLOSIVES

Research on the safety of mining explosives 396

Safety explosives before World War I 403

Conditions of shotfiring in mines 406

Mining explosives used during World War I 408

Research after World War I 409

General consideration on safety of explosives 413

Fundamental components of mining explosives 420

Oxygen carriers 421

Active ingredients and combustibles 423

Oxygen balance 423

Inert ingredients increasing safety 427

Inert neutralizing agents

Tests for mining explosives 433

Transmission of detonation 433

Sensitiveness to detonation 434

Power of explosives 438

Safety tests with methane and coal-dust 439

Application of statistics to gallery testing of explosives 445

X CONTENTS

Stability of mining explosives 446

Mining explosives used in various countries 446

Belgium 447

Czechoslovakia 448

France 451

Germany 455

Great Britain 461

Hungary 468

Japan 468

Poland 475

U.S.A 480

U.S.S.R 484

Combined blasting and water infusion for coal breaking 489

Liquid oxygen explosives (Oxyliquits) 491

Some other peaceful applications of explosives 495

Literature 495

CHAPTER VI. THE MANUFACTURE OF MINING EXPLOSIVES

The manufacture of ammonium nitrate explosives 498

Ammonium nitrate-fuel oil mixtures 508

The manufacture of dynamites 511

The manufacture of chlorate and perchlorate explosives 520

Cardox, Hydrox and Airdox cartridges 521

Literature 526

CHAPTER VII. SMOKELESS POWDER

Historical 528

Properties of smokeless powder 532

Physical properties 532

Explosive properties 532

Mechanical properties 543

Flash and methods for suppressing it 544

Smoke formation 548

Erosiveness of smokeless powder 548

Stability of smokeless powder 550

Stability tests 557

Stabilization of smokeless powder 559

Stabilization with diphenylamine 559

Inorganic stabilizers 563

Organic stabilizers 564

Apparent stabilizers 567

Literature 567

CHAPTER VIII. THE MANUFACTURE OF SMOKELESS POWDER

Introduction 570

Nitrocellulose powder 571

Nomenclature 571

Manufacture of nitrocellulose powder 573

CONTENTS Xi

The preparation of nitrocellulose mixtures 582

Partial dissolution of nitrocellulose 583

Shaping the dough 590

The stabilization of an unstable powder 632

Ball-grain powder 632

Nitrocellulose bulk powder (Schultze powder) 640

Double base powders 641

Nitroglycerine powders with a volatile solvent 642

Solventless nitroglycerine powders 644

Solventless powders with a low content of nitroglycerine 652

The manufacture of solventless powder in German factories 660

Solventless powder in Japan 663

Flashless charges and fiashless powders 663

Smokeless powder with penthrite 670

Smokeless powders containing nitroaliphatic compounds 671

Smokeless powders for rockets - 671

Cast double base propellants 675

General safety considerations in the manufacture of smokeless powder 682

Literature 686

Author index 689

Subject index 702

Errata to Volumes I and II 715

PREFACE TO VOLUME III

The manuscript of Volume III was submitted to the publishers in 1960. Owing to
the rapid progress in the many branches of chemistry and chemical technology
that form the scope of this volume, some paragraphs became obsolete and it has
therefore been necessary to add some new items and to rewrite certain paragraphs.

The necessary information obtained from the current literature has been sup-
plemented with data from a number of colleagues who kindly made available some
less accessible material. It is my pleasant duty to express my thanks to Prof. S.
Claesson (Uppsala), Prof. M. A. Cook (University of Utah, Salt Lake City), Prof.
W. Cybulski (Mikolow, Poland), Dr. L. Deffet (Sterrebeck, Belgium), Dr. R. W.
Van Dolah (Pittsburgh, Pa.), Dr. A. G. Grenier (Dow Chemical International,
Midland, Michigan), Prof. J. Hackel (Warsaw), Prof. M. Kryszewski (Lodz), Dr.
J. Meissnef (Frankfurt a./M.), Prof. H. Sudo (Tokyo), Dr. A. Wetterholm (Gyttorp,
Sweden) and to the firms: Draiswerke G.m.b.H., Maschinenfabrik (Mannheim-
Waldhof), Olin Industries, Inc. (East Alton, 111.), H. Orth G.m.b.H. Fachbiiro fiir
Verfahrenstechnik (Ludwigshafen-Oggersheim, Pfalz), Werner & Pfleiderer, Maschi-
nenfabriken und Ofenbau (Stuttgart).

Some of the problems, such as safety in coal mines, merit a special monographic
treatment which should go far beyond the scope of this volume. Therefore safety in
coal mines was tackled only from the point of view of the composition of coal mine
explosives.

Another rapidly developing branch of applied science is theory and practice of
rocket propulsion. Only general information on composition of rocket propellants
is given in the book. This is justified, as several special books on rocket fuel are now
available.

My thanks are due to the translator, Mr. M. Jurecki and to Mrs. Sylvia Laverton,
F.R.I.C. for tidying up the English text and to Mrs. A. Malawska, M.Sc. for her
skilled editorial work.

Author

[xiii]

^P»-»*.l.

Part 1
NITRAMINES

CHAPTER I

GENERAL INFORMATION

Nitramines are substances which contain a nitro group bonded to the nitrogen
atom: )>N— N0 2 .

In formal terms all nitramines may be regarded as derivatives of the simplest
inorganic nitramine, NH 2 N0 2 . If only one hydrogen is replaced by an alkyl or aryl
group, the resulting substance is a primary nitramine, if two hydrogens are replaced,
secondary nitramines are formed:

R \ R \

>N— N0 2 >N— NO z

W Ri/

Primary nitramine Secondary nitramine

Nitramines may also be said to include nitramides (primary and secondary)
which differ from nitramines proper in that one of the groups (R) is acyl or
sulphonyl.

STRUCTURE AND CHEMICAL PROPERTIES OF NITRAMINES

The existence of a bond in nitramines linking the nitrogen atom of the nitro
group with that of the amino group is proved by the formation of hydrazine deriva-
tives by the reduction of nitramines. Secondary nitramines give a particularly high
yield of hydrazine derivatives.

There exists a group of compounds isomeric with primary nitramines, which also
yield hydrazine derivatives when reduced. These compounds are nitrosohydroxylamine
derivatives :

R— N— OH

I
NO

[1]

2 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

This structure was deduced from their mode of preparation: nitrosohydroxyl-
amines are obtainable by nitrosation, nitramines by nitration. The physical prop-
erties of primary nitramines are entirely different from those of nitrosohydroxyl-
amines, so that phenylnitramine C 6 H 5 NHN0 2 , for example, differs radically from
phenylnitrosohydroxylamine

C 6 H 5 NOH

I
NO

Traube's "isonitramines", which are described later in this volume, are nitroso-
hydroxylamines.

The existence of an N— N bond in|fjrimary nitramines is also proved by the
fact that these substances are obtained by the oxidation of diazo compounds. Ad-
ditional evidence that nitramines contain a nitro and not a nitrite group and hence
have a structure different from R— N— ONO is provided by the fact that nitramines
are relatively resistant to alkali, whereas nitrous esters are highly unstable.

These inferences concerning the structure of nitramines, based on their chemical
properties, are confirmed by the data obtained by X-ray analysis of the simple
nitramines: dimethylnitramine and ethylenedinitramine. In particular, Costain and
Cox [1], and Llewellyn and Whitmore [2], established that the grouping

C \ /°

>N— N<

W X>

is planar. Bond angles and interatomic distances are shown in Fig. 1.

CH 3

Fig. 1. Structure of the nitramino group, with dimethylnitramine as an example

(not to scale).

Thus, in principle, the dimensions of the nitro group in both aromatic and
aliphatic nitro compounds are identical.

Nitramines dissolved in water, alcohol or dioxane, give a broad ultra-violet
absorption band, the maximum of which lies between 225 and 240 fi.

R. N. Jones and Thorn [3] adduce the two following typical absorption curves
for nitramines: a primary — ethylenedinitramine — Fig. 2 and a secondary —
2,5-dinitro-2,5-diazahexane (according to R. N. Jones and Thorn) — Fig. 3. They
have also been investigated by Baly and Desch [4], Franchimont and Backer [5],
Carmack and Leavitt [6] and Corey, Dekker, Malmberg, Le Rosen, and Schroeder [7].

Jll_ ,.

WMUWI ...

NITRAMINES — GENERAL INFORMATION

2200 2600 „ 3000

Wave Length (A)

Curve A. Solvent, ethanol

Curve B. Solvent, 0.2 N sodium

hydroxide

Fig. 2. Spectrum of typical primary nitr-

amine (ethylenedinitramine), according to

R. N. Jones and Thorn [3].

4.0

- B,y

3.6

3.2

^ 2.8

2.4

2.0

1.6

\ -

1.2

2200 2600 o 3000

Wave length (A)

— Curve A. Solvent, dloxane

— Curve B. Solvent, 0.2 N sodium

hydroxide

Fig. 3. Spectrum of typical secondary
nitramine (2,5-dinitro-2,5-diazahexane), ac-
cording to R. N. Jones and Thorn [3].

The ultra-violet absorption spectra of nearly 60 nitramines were examined by
R. N. Jones and Thorn, who drew up an interesting rule about the coefficient of
extinction for nitramines.

According to this rule, compounds which contain one primary nitramine group
show a coefficient of extinction in the maximum of the absorption curve, e max , equal
to approximately 7000. If the molecule contains n primary nitramine groups, e max
increases in proportion to the number of nitramine groups, i.e. £ max =7000 n. This
empirical rule is valid for values of n ranging from 1 to 3. For secondary nitr-
amines, £ max =5500. If the number of secondary nitramine groups is n, e max = 5500 n.
This rule is valid for values of n ranging from 1 to 4. If both primary and secondary
nitramines occur together in a molecule, e max is the sum of the appropriate mul-
tiples of 7000 and 5500.

The infra-red spectra of nitramines show approximately the same frequencies
as ordinary nitro compounds (Lieber, Levering and Patterson [8], Salyamon and
Yaroslavskii [9], Salyamon and Bobovich [9] and Bellamy [10]).

On the basis of these authors' results, Bellamy states that the nitro group in
nitramines has the following vibration frequencies:

asymmetric 1587-1530 cm~i

and symmetric 1292-1260 cm-'.

A CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

According to Bellamy, these values may be considered as the average for all
nitramines except nitroguanidine which, as the investigations of Lieber and his
co-workers and of Kumler have shown, has a very high frequency of asymmetric
vibration, i.e. ranging 1655-1620 cm- 1 .

Similar deviations are shown by nitrourea. Presumably these are caused by the
existence of tautomeric forms.

Nitramines show no basic properties whatever — indeed, primary nitramines
have distinct acidic properties and can form salts with alkalis. Conversely, nitr-
amides may be more strongly acidic than carboxylic acids, as, for example, nitro-
urethane, which is a stronger acid than formic acid.

Primary amines react slowly with ammonia in a benzene medium, to form am-
monium salts. Hence Hantzsch [11] assumed that primary nitramines (I), like pri-
mary and secondary nitroparaffins, are pseudoacids and react in a tautomeric aci-
form (II)

*° x°

R— NH— N<f & R— N=N<^

I II

This view was generally accepted. It was based, however, not so much on ex-
perimental evidence as on Hantzsch's personal authority. In point of fact Euler [12]
found that the rate of formation of the ammonium salt is by no means as slow as
Hantzsch believed and expressed doubt as to the existence of the supposed tauto-
merism. No further evidence confirming the existence of the aci-form was forth-
coming until the O-alkyl derivatives, e.g. O-methyl-methylnitramine (III),

*°
CH 3 — N=N<C

\OCH 3

III

isomeric with dimethybitramine (Gillibrand and Lamberton [13]), were obtained.
That the compound (III) actually has such a structure is proved by the fact
that on hydrolysis with 40% sulphuric acid two molecules of methyl alcohol are
formed:

*°

CH 3 -rN=N<> -> N 2 + 2CH 3 OH

j 'X)CH 3

OH J H

One of the characteristics of nitramines is the ease with which they decompose

in sulphuric acid. Primary nitramines undergo decomposition with particular ease;

alcohol is formed and N 2 is evolved on boiling in dilute (2%) sulphuric acid (van

Erp [14], Backer [15]):

RNHNO2 -> ROH+N 2

Secondary nitramines are more stable towards sulphuric acid and decomposition
rarely occurs until 40% or more sulphuric acid is used at a temperature of 100°C.

NITRAMINES— GENERAL INFORMATION

Concentrated nitric acid decomposes some primary nitramines; methylnitramine,
for example, is decomposed by anhydrous nitric acid, even at a temperature below
0°C, to form methyl nitrate and N 2 (Franchimont [16]).

Rearrangement of primary and secondary aromatic nitramines occurs in an acid
medium with the formation of C-substituted aniline derivatives (Bamberger and
Landsteiner [17]).

Dinitrophenylmethylnitramine, for example, on treatment with nitric and sul-
phuric acid, is rearranged to form trinitro-N-methylaniline :

/N0 2
N< NHCH 3

\ X CH 3

/ \— N0 2 -» 2 N— |j / \— NO2

N0 2 N0 2

Hughes and Ingold [18] suggest the following explanation of the rearrangement:
R

^N-O h® r \f n-o

II - ii — II

o kJ o

R\©/H R ^ M / H

This rearrangement, which is of great importance for the manufacture of tetryl,
will be illustrated when the production of this substance is discussed.

The majority of aromatic-aliphatic nitramines undergo denitration on heating
with phenol, especially in the presence of sulphuric acid. Tetryl, for example, under-
goes the following reaction:

no 2

nhch 3

2 N-/ V N0 2 pheno1 , 2 N-^ V-NO2
H 2 S0 4

no 2 no 2

Nitramines dissolved in concentrated sulphuric acid give a blue colour with
diphenylamine reagent.

With primary nitramine sulphuric acid may promote the separation of nitric
acid, which results in extensive decomposition of the substance. This will be dis-
cussed in more detail when dealing with nitroguanidine (p. 26).

6 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

There is as yet no evidence that in the presence of sulphuric acid secondary
amines lose nitric acid which would be capable of nitrating phenol. Attack by con-
centrated sulphuric acid, presumbaly loosens the linkage between the nitrogen
atoms, leading to the expulsion of the nitronium ion NOf , which is a nitrating
agent.

Most nitramines are fairly resistant to alkalis. Some of the primary amines, e.g.
the simple aliphatic ones, are not susceptible to decomposition even under the
influence of a hot, 20% solution of potassium hydroxide (van Erp [14]). On the
other hand, Hantzsch and Metcalf [19] found that N-nitraminoacetic acid is de-
composed by sodium hydroxide. According to Barrott, Gillibrand and Lamberton
[20] most primary amines undergo decomposition on treatment with a solution
containing 0.8-8% of NaOH at a temperature of 95°C. The reaction proceeds ac-
cording to the formula:

R v Rv

>CH— NHN0 2 -+

\ R \

>CH— NHN0 2 -> >0=0 + N 2 + H 2

1/ Rl/

The concentration of alkali required depends on the properties of the radicals
R and R 1 . The more electrophilic the radicals and the more acidic the nitramine,
the easier the course of reaction. Secondary nitramines are decomposed by an aqueous
solution of sodium hydroxide. The reaction conditions, including the concentrations
of NaOH solutions differ according to the substance. Van Erp and Franchimont [21]
found that the reaction proceeded by the following mechanism:

/N0 2
^CH 2 Ri

RN/ -» R— N=CH— Ri + HNO3

H 2
RNH 2 + RiCHO

Various products are formed by the reduction of nitramines, depending on the
reaction conditions.

Vigorous reduction may involve the rupture of the N—N linkage with the formation
of amine and ammonia. Milder reducing agents yield different products, including
hydrazine derivatives, e.g.:

K R \

>N— N0 2 -+ >N— NH 2

1/ Ri/

Such a reduction may occur quantitatively and hence may be utilized for ana-
lytical purposes. According to Cope and Barab [22], the Schulze-Tiemann (FeCl 2
+ HC1) or Lunge (Hg+H 2 S0 4 ) methods are suitable for this purpose. In both
cases the reaction proceeds as follows:

\N— N0 2 -> ^>NH + NO

NITRAMINES — GENERAL INFORMATION 7

The reaction of primary aromatic nitramines with nitrous acid is specific and
leads to the formation of diazonium nitrate (Bamberger [23]). According to Stevens
[24] this reaction is also a kind of reduction, and presumably may be represented
as follows:

/NO h 2 o
ArNH— N0 2 + HN0 2 -> Ar— N< ► ArNH— NO + HNO3

X N0 2 I

[ArN 2 ]© NO© + H 2

Primary nitramines react with diazomethane to yield N-methyl or O-methyl
derivatives.

Thus, methylnitramine is converted into dimethylnitramine and phenylnitramine
nto phenyl-O-methylnitramine:

C 6 H 5 N=N<*

i^0^ti3
Ilia

As early as 1910, Franchimont[16] observed that primary nitramines (and nitro-
paramns) react with formaldehyde and secondary amines (e.g. piperidine). The forma-
tion of aminomethylnitramines (IV) then occurs:

N0 2
/Ri I /Ri

RNH— N0 2 + CH 2 + HN< -> RN— CH 2 — N<

\R X R

It was shown (Woodcock [25]) that intermediate hydroxymethyl compounds
may be formed as for example those of type (V)

N0 2

N— CH 2 OH

I
(CH 2 )»

I
N— CH 2 OH

I
NO2

V
(n= 1, 3 or 4)

According to Lamberton and his co-workers [26], equilibrium is established in an
aqueous solution:

N0 2

RNHNO2 + CH 2 & RN— CH 2 OH
VI

The hydroxymethyl derivative (VI) is, however, more stable in an acid medium.
In a neutral medium the equilibrium shifts markedly to the left.

8 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

PREPARATION OF NITRAMINES

Nitramines and nitramides arise in various ways depending on the amines and
amides subjected to nitration. There are direct and indirect nitration methods.

DIRECT NITRATION

Direct nitration with anhydrous nitric acid (98%) can be accomplished most
conveniently in the presence of monosub%tituted N-alkylamides. A secondary nitr-
amide is then formed:

hno 3 /N0 2 v

RCO— NHRi > RCO— N< (1)

\Ri

One of the direct methods of nitrating primary amines is based on this reaction.
An amine is acylated to form a primary amide which is in turn nitrated according
to reaction (1) and then hydrolysed to release the acyl group. This procedure will be
discussed more fully in later chapters.

Generally, reaction (1) is not successful with non N-substituted, i.e. primary
amides, most of which undergo decomposition when nitrated. Primary aliphatic
amines also decompose under the action of concentrated nitric acid.

Similarly, primary aromatic amines undergo complex reactions when heated with
nitric acid. The amine derivatives of anthraquinone, pyridine and thiazole are excep-
tions : the amino group in these compounds is decomposed by the nitrating mixture.
This was noticed by Scholl [27], who proposed the following method for nitrating
)?-aminoanthraquinone :

Chichibabin [28] and Razorenov observed that the amino derivatives of pyridine
are nitrated in a similar way. For example, a mixture of nitric and sulphuric acids
converts a-amino-pyridine into a-nitramino-pyridine.

Ganapathi and Venkataraman [29] found that aminothiazole and its homologues
can be nitratedin the same way, using a double excess of nitrating mixture.

-j N R-jr-

R — n N R— N

^NHNOs

NITRAMINES— GENERAL INFORMATION 9

In addition to the compound (I), a certain amount of a compound (II) nitrated
in the ring is formed when a stoichiometric quantity of HN0 3 is used. As the investi-
gations of Wright et al. [30] subsequently showed, the compound (II) is also obtain-
able when aminothiazole is treated with a nitrating mixture containing 10-30% of

water.

R— ii N

I \\

o 2 n/ / \ S ' / \nh2
II

At an earlier stage Orton [31] found that some aniline derivatives which are
difficult to nitrate in the ring, are liable to form N-nitramines when treated with
nitric acid and acetic anhydride. For example, he prepared the corresponding nitr-
amine from 2,4,6-tribromaniline :

NHN0 2

I
Br— f\~ Br

. V

The nitration of secondary amines by the above method is generally success-
ful:

R\ HNO3 R\

>NH > >N-N0 2 (2)

Ri/

R \
Ri/

It is noteworthy that the conversion of the group ^>NH into ^>N — N0 2 is not
always practicable. As early as 1916 Franchimont and Dubsky [32] called attention
to the fact that if the group ^>NH has the properties of an imido group (for example
— CO — NH — CO — ), it is not susceptible to nitration, while the same group with
amido characteristics (for example — CO — NH — CH 2 — ) is readily nitrated. De-
veloping this observation, Wright et al. [33, 34] were led to the discovery that very
weakly basic amines with the )>NH group (e.g. in the form of — CH 2 — NH— CH 2 ) are
nitrated more readily than strong bases with the same group. Strong bases require
the addition of a catalyst for nitration (for example, ZnCl 2 , HC1; this will be discus-
sed in later chapters) whereas weak bases can be nitrated without a catalyst. Amines
of the type R— CH 2 — NH— CH 2 — R, for example, where R = CN, COOH or CONH 2 ,
i.e. amines of a considerably diminished basicity, are nitrated relatively easily.

In some more complex cases, secondary amines can be nitrated with nitrating
mixtures. Diethanolnitramine dinitrate (DINA), for example, can be prepared
by the action on diethanolamine of a mixture of nitric acid, nitric anhydride and
zinc chloride (this is dealt with in more detail on p. 10).

Some amines undergo direct nitration during nitrolysis ; nitration is then followed
by partial degradation involving the cleavage of the bond between the carbon and
nitrogen atoms (p. 12).

10 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

INDIRECT NITRATION

Nitration by the "dehydration" of the amine nitrate

This method is commonly applied to the nitration of primary amines (3) and
in particular the preparation of nitroguanidine, nitrourea etc., as well as in the
nitration of secondary amines (4):

e e -h 2 o

RNH3NO3 > RNH— N0 2 (3)

R \ e e -h 2 o R \

(4)

j*\ e e -h 2 o «-\

>NH 2 N0 2 *■ >N— N0 2

Ri/ Ri/

It was applied by Bamberger and Kirpal [35] in the preparation of dimethyl-
nitramine and nitropiperidine with a rather poor yield. They used acetic anhydride
as a dehydrating medium.

Wright et al. [33, 36] found that the yield of nitramines was considerably improved
by adding zinc chloride or hydrogen chloride to the acetic anhydride. Under such
conditions dimethylamine yields 65% of dimethylnitramine. It is not impossible
that the essence of the action of these additions is to reduce the basicity of the amine
and thus facilitate the introduction of a nitro group in accordance with the rule
outlined above.

Another method of "dehydration" of the amine nitrate is based on treatment
with concentrated sulphuric acid. This is used commercially in manufacturing such
primary amines as nitroguanidine (p. 31).

Nitration of primary amines by acylation

This method involves acylation of the primary amine by introducing an acetyl
or oxalyl group followed by nitration of the secondary amine so formed in accord-
ance with the reaction (1). The product thus nitrated undergoes alkaline hydro-
lysis to yield a primary nitramine:

/N0 2 x no 2

RCON<( -+ RCOOH + HN< (5)

x Ri \Ri

Very frequently the transition through urethanes is employed by treating the
primary amine with chloroformate. The N-substituted urethane so obtained is
nitrated by substituting the free N-hydrogen and then subjecting the product to
alkaline hydrolysis which results in the formation of the salt of a primary nitr-
amine and a base. The free nitramine is obtained by acidification.

This type of reaction can be illustrated by the nitration of methylamine according
to Franchimont [37] (6):

„ cicoocjHj HNO3

CH3NH2 + CH3NH— COOC 2 H 5 > CH 3 — N— COOC 2 H s *

NITRAMINES— GENERAL INFORMATION 11

2NH 3 Q ffi

* CH 3 — N=N02 NH4 + NH 2 — COOC2H5 (6)

HCl
CH3NH— N0 2 + NH4CI

The method was later worked out by Brian and Lamberton [38] to produce
previously unobtainable nitramines.

The formation of nitramines through chloramines

This method was originally suggested by Berg [39]. It consists in acting with
silver nitrate on chloramines:

RNHC1 + AgN0 3 -> RNHNO2 + AgCl

Wright et al. [40] elaborated a new method of preparing nitramines by acting on
chloramines with nitric acid in the presence of acetic anhydride. A typical example
is the preparation of sec-butylnitramine (III) (Smart and Wright; Suggitt, Myers
and Wright [40]) from dichloramine (I):

C 2 H S C 2 H 5 C 2 H 5 CI C 2 H 5

I I HN0 3 I I H 2 I

H— C— NH 2 -+ H— C— NCI 2 > H— C NN0 2 > H— C— NHNO2

I I +(CHjCO) 2 I

CH 3 CH 3 CH 3

I II

H 2

CH 3
III

C 2 H 5 C 2 H S

I I

C=0 -<- C=NN0 2

I I

CH 3 CH 3

» V IV

The intermediate (II) is unstable and is hydrolysed in water to form (III) and
(IV). The latter (2-butanenitramine) is also unstable and decomposes to butanone-2
(V).

The intermediate formation of chloramine explains the catalytic action of hydro-
chloric acid in the nitration of amines, as mentioned above. The following reaction
mechanism was drawn up by Wright [41]

2HC1 + 2HN0 3 + 3(CH 3 CO) 2 ->- 2CH 3 COOCl + N 2 3 + 4CH 3 COOH (a>

R 2 NH + CH 3 COOCl -> R 2 NC1 + CH 3 COOH (b)

R 2 NC1 + HN0 3 -> R2NNO2 + HOC1 (c>

*HOCl + (CH 3 CO) 2 -> CH 3 COOCl + CH 3 COOH (d>

Thus hydrocliloric acid reacts in the presence of nitric acid to yield chlorine
acetate (a)— a compound with cationic chlorine. The latter in turn forms a chlor-
amine (b) which is nitrated to a nitramine (c).

The mechanism of the nitration of amines has not yet been explained fully.
Wright et al. [34, 42] suppose that the nitration of secondary amines and probably of

12 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

secondary amides takes place by the formation of intermediate complexes between
the amine and nitric acid. It is probable that the bond N— N is formed, followed
by the loss of HOX (7):

R\ ©e/O e R x e e/O e _ HOX R x ® y o

>N + N< & >N— N< > >N— N< (7)

X OH X OH

In catalysed reactions X represents CI, in uncatalysed reactions it represents H.
This scheme, however, has the disadvantage of ignoring the influence of the
nitronium ion (NOf ) on the reaction, whereas, as expounded in the chapter on
nitration theories (Vol. I), the nitronium ion is of enormous importance for such
a reaction. In this connection Lamberton [43] suggests alternative schemes which
appear more probable. Scheme (8) leads to the formation of a nitramine and
scheme (9) to a salt of nitric acid:

\n + NO© ?* N— N0 2 ?± \n— N0 2 + H© (8)

H H

\n + HN0 3 & N HON0 2 ?± \nH 2 + NO© (9)

H H

Where there are strong bases which are nitrated with difficulty, reaction (9)
predominates over reaction (8). At the same time Lambeiton called attention to the
reversibility of reaction (9). Indeed, it is known that nitramines such as nitroguani-
dine or nitrourethanes exhibit nitrating properties in the presence of sulphuric acid,
thus behaving as if they can split off the nitronium ion or the nitric acid molecule.

Nitration by nitrolysis

The term "nitrolysis" suggested by Linstead [44] is usually applied to a nitrating
mechanism in which both the rupture of C— N bond and the formation of a nitr-
amine occur simultaneously with the formation of alcohol which subsequently under-
goes esterification (10):

R2NCH2R1 + HON0 2 -> R 2 N— N0 2 + HO— CH 2 Ri

HNO3 (10)

N0 2 — O— CH2R 1

Nitrolysis may also proceed without giving rise to alcohol in accordance with
eqn. (11). Nevertheless, a nitric ester is formed by the possible action of the NOf ion
on a free alkyl cation:

R 2 N— CH2RI + NO© -> R 2 — N— CH 2 Ri -> R 2 N— NO z + CH 2 Ri

,e (ID

no 2

NO?

OjN— 0-CHjR»

NITRAMINES— GENERAL INFORMATION 13

As these equations suggest, the nitration of an amide may lead to a nitramine
or a nitramide, according to reaction (12a) or (12b):

X— CO— NR 2 -> X— COOH + R 2 N— N0 2 (12a)

\ /N0 2

X— CO— N< + ROH (12b)

The nitration of hexamethylenetetramine, which contains the grouping

— CH 2 — N— CH 2 —

I
CH 2

to produce cyclonite, is also a nitrolysis reaction. In addition to cyclonite, a nitrate
of methylene glycol is also formed as a result of the cleavage of one of the three
linkages between each nitrogen and carbon atom:

ON0 2

CH 2

^0N0 2

Other methods for the preparation of nitramines

Some nitramines may be prepared without treating amines with nitric acid.
The classical example is the so-called "E-method" of cyclonite preparation in which
a nitramine is formed by dehydration of a mixture of paraformaldehyde and am-
monium nitrate, i.e. without using either amine or nitric acid (this will be discussed
more fully on p. 109). When a nitramine is required with a non-nitrated aromatic
ring which readily undergoes nitration with nitric acid, Bamberger's method [45],
involving the oxidation of diazo compounds (13), may be applied.

C 6 h 5 N=NOK K3Fe(CN> > C 6 H 5 NK-N0 2 H ° > C 6 H 5 NH-N0 2 (13)

NITRAMINES AS EXPLOSIVES

Nitramines differ from nitro compounds in possessing a somewhat better oxygen
balance, due to the fact that the group N— N0 2 gives twice the volume of nitrogen as
the group C— N0 2 . On the other hand nitramines have a worse oxygen balance
than nitric esters.

As regards explosive strength, nitramines occupy a position midway between
nitro compounds and nitric esters. They also hold a central position regarding other
properties, such as chemical stability and sensitiveness to impact and friction.

LITERATURE

1. W. Costain and E. G. Cox, Nature 160, 826 (1947).

2. F. J. Llewellyn and F. E. Whitmore, /. Chem. Soc. 1948, 1316.

3. R. N. Jones and G. D. Thorn, Can. J. Research 27 B, 828 (1949).

4. E. C. C. Baly and C. M. DESCH, /. Chem. Soc. 93, 1747 (1908).

14 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

5. A. P. N. Franchimont and H. J. Backer, Rec. trav. chim. 32, 327 (1913).

6. M. Carmack and J. J. Leavitt, /. Am. Chem. Soc. 71, 1221 (1949).

7. B. B. Corey, A. O. Dekker, E. W. Malmberg, A. L. Le Rosen and W. A. Schroeder,
Report No. P.B. 18856, Dept. of Commerce, Washington D.C.

8. E. Lieber, D. R. Levering and L. J. Patterson, Anal. Chem. 23, 1594 (1951).

9. G. S. Salyamon and N. G. Yaroslavskii, Sbornik Statei Obshch. Khim. 2, 1325 (1953);
G. S. Salyamon and Ya. Bobovich, ibid. 2, 1332 (1953).

10. L. J. BELLAMY, The Infra-red Spectra of Complex Molecules, Methuen, London, 1958.

11. A. Hantzsch and F. E. Dollfus, Ber. 52, 258 (1902>

12. H. Euler, Ber. 39, 1607 (1906).

13. M. I. Gillibrand and A. H. Lamberton, J. Chem. Soc. 1949, 1883.

14. H. van Erp, Rec. trav. chim. 14, 40 (1895).

15. H. J. Backer, Die Nitramine-Ahrens Sammlung 18, Stuttgart, 1912.

16. A. P. N. Franchimont, Rec. trav. chim. 29, 296 (1910).

17. E. Bamberger and K. Landsteiner, Ber. 26, 490 (1893).

18. E. D. Hughes and C. K. Ingold, Quart. Rev. 6, 34 (1952).

19. A. Hantzsch and W. V. Metcalf, Ber. 29, 1680 (1896).

20. J. Barrott, M. I. Gillibrand and A. H. Lamberton, J. Chem. Soc. 1951, 1282.

21. H. van Erp and A. P. N. Franchimont, Rec. trav. chim. 14, 224 (1895).

22. W. C. Cope and J. Barab, J. Am. Chem. Soc. 38, 2552 (1916); see also K. Lehmstedt and
O. Zumstein, Ber. 58, 2024 (1925).

23. E. Bamberger, Ber. 30, 1248 (1897).

24. T. S. Stevens, according to A. H. Lamberton [43].

25. D. WOODCOCK, /. Chem. Soc. 1949, 1635.

26. A. H. Lamberton, C. Lindley, P. G. Owston and J. C. Speakman, J. Chem. Soc. 1949,
1641.

27. R. Scholl et al, Ber. 37, 4427 (1904); R. Scholl and A. Krieger, ibid. 37, 4681 (1904).

28. A. E. Chichibabin and B. Razorenov, Zh. Russ. Khim. Obshch. 47, 1280 (1915).

29. K. Ganapathi and A. Venkataraman, Proc. Indian Acad. Sci. 22 A, 343 (1945).

30. J. B. Dickey, E. B. Towne and G. F. Wright, /. Org. Chem. 20, 499 (1955).

31. K. J. P. Orton, /. Chem. Soc. 81, 806 (1902).

32. A. P. N. Franchimont and J. V. Dubsky, Rec. trav. chim. 36, 80 (1916).

33. W. J. Chute, G. E. Dunn, J. C. Mackenzie, G. S. Myers, G. N. R. Smart, J. W. Suggitt
and G. F. Wright, Can. J. Research 26 B, 114 (1948).

34. G. E. Dunn, J. C. Mackenzie and G. F. Wright, ibid. 26 B, 104 (1948).

35. E. Bamberger and A. Kirpal, Ber. 28, 462, 535 (1895).

36. W. J. Chute, K. G. Herring, L. E. Toombs and G. F. Wright, Can. J. Research 26 B, 89
(1948).

37. A. P. N. Franchimont, Rec. trav. chim. 13, 308 (1894).

38. R. C. Brian and A. H. Lamberton, /. Chem. Soc. 1949, 1633.

39. A. Berg, Ann. Chim. [7] 3, 358 (1894).

40. J. C. Mackenzie, G. S. Myers, G. N. R. Smart and G. F. Wright, Can. J. Research 26 B,
138 (1948); G. S. Myers and G. F. Wright, ibid. 26 B, 257 (1948); G. N. R. Smart and G. F.
Wright ibid. 26B, 284 (1948); J. W. Suggitt, G. S. Myers and G. F. Wright, /. Org. Chem
12, 373 (1947).

41. G. F. Wright in H. Gilman's Organic Chemistry, Vol. IV, p. 951, J. Wiley, New York, 1953.

42. G. S. Myers and G. F. Wright, Can. J. Research 26 B, 257 (1948).

43. A. H. Lamberton, Quart. Rev. 5, 75 (1951).

44. R. P. Linstead, unpublished; quoted by A. H. Lamberton [43].

45. E. Bamberger and L. Storch, Ber. 26, 471 (1893); E. Bamberger, ibid. 27, 359 (1894V
55,3383(1922). .'

CHAPTER II

ALIPHATIC NITRAMINES AND NITRAMIDES

NITRAMINE (NITRAMIDE)

NH 2 -N0 2
m.p. 72-73°C (decomp.)

This is one of the simplest nitramines obtained by Thiele and Lachman [1] by
the decomposition of nitrourethane :

NH 2 NHNO2 NHNO2 NHNO2

I I KOH I acid I

CO — v CO ► CO > CO — -> NH 2 N0 2 + C0 2 (14)

I I (alcohol) I I

OC2H5 OC2H5 OK OH

The structure of the compound was the subject of controversy for some time [2].
However, the latest experiments of Clusius [3] with 15 N-labelled nitramine have
confirmed Thiele's formula [4] :

NH2NO2 ?* NH=NO(OH)

A carefully-devised method for preparing it was announced by Marlies, La Mer
and Greenspan [5]. According to Bell and Wilson [6] the product contains some
acidic impurities (about 1 %), possibly unchanged nitrourethane.

This substance is an intermediate product of the decomposition of the important
explosive nitroguanidine. It is also present in an aqueous solution of nitrourea or a
sulphuric acid solution of nitrourea (Davis and Blanchard [7]).

Nitramine is believed to occur in an ammonium nitrate solution in an excess
of concentrated sulphuric acid as a result of the dehydration of this salt:

NH4NO3 -> nh 2 no 2 + H 2 (15)

Davis and Abrams [8] report the following experimental observations in support
of this supposition. On heating a solution of ammonium nitrate in sulphuric acid
to 150°C, nitric acid cannot be distilled, but nitrous oxide is evolved, probably from
the decomposition of nitramine. If, however, the solution is kept for a long time
between 90 and 120°C, nitric acid can be obtained by distillation. The authors'

[15]

16 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

conjecture is that the addition of water to the nitramine takes place according to the

reaction (16):

NH2NO2 + H 2 -> NH 3 + HNO3 (16)

Nitramine has explosive properties but it is not of any practical value for many
reasons, primarily because of its high reactivity which impairs its chemical stability.
It decomposes at a temperature as low as its melting point. At room temperature
it decomposes slowly, to form nitrous oxide and water. On heating to 60-65°C
decomposition occurs in an aqueous solution. It decomposes explosively on contact
with concentrated sulphuric acid. *

Bell and Caldin [9] and Caldin and Peacock [10] investigated the decomposition
kinetics of nitramine under the influence of alkali in various solvents. Dimethylaniline
was used as a base. According to Bell [1 1] the decomposition of nitramine proceeds
through the formation of the aci-form:

*°

NH2NO2 -> hn/^

The reaction with a base B would then take the following course:

B + HN=NOOH -+ BH© + N 2 + OH e -> B + H 2 + N 2

The rate and activation energy of decomposition depend to a great extent on the
type of solvent used.

METHYLNITRAMINE

CH3NHNO2

m.p. 38°C

This is a powerful explosive, stronger than tetryl but weaker than cyclonite.
It is, however, of no practical value chiefly because its preparation is too expensive,
requiring first the conversion of methylamine into urethane and then into its nitro
derivative. On hydrolysis the latter yields methylnitramine. Similarly, the hydrolysis
of dinitrodimethyloxamide (p. 35) leads to the formation of methylnitramine.

Methylnitramine is very readily soluble in water, alcohol, chloroform and benzene
but is less soluble in ether. It is a strong acid which easily forms salts, including
explosive ones. It is not decomposed by boiling water, even in the presence of alkalis,
but it is liable to destructive distillation yielding dimethylnitramine (CH 3 ) 2 NN0 2 ,
m.p. 57°C, methyl alcohol, nitrous oxide and many other products.

Methylnitramine decomposes explosively in contact with concentrated sulphuric
acid. It is evolved when aniline reacts with tetryl, a diphenylamine derivative (p. 51)
is produced simultaneously. Methylnitramine reacts with picryl chloride to form
tetryl. The structure of tetryl (p. 40) was first proved by this synthesis.

ALIPHATIC NITRAMINES AND NITRAMIDES 17

METHYLENEDINITRAMINE

NHN0 2

I
CH 2

I
NHN0 2

m.p. 101°C

This substance was isolated in the form of its barium salt by Hirst et al. [12]
when investigating the nitration of hexamethylenetetramine to cyclonite. They found
that hexamethylenetetramine, when dissolved in nitric acid at 40°C, yields the product
(II) which is hydrolysed by barium hydroxide to form the barium salt of methylene-
dinitramine. From this the free nitramine may be obtained:

N0 2
I
N— CH 2 .NHNO2

/ \ Ba(OH) 2 /

H 2 C NH2NO3 y CH 2

\ / and acidification \

N— CH 2 ntinv 2

I
N0 2

II I

Brian and Lamberton [13] accomplished the synthesis of methylenedinitramine
with methylene-bis-N-acetamide (III) :

no 2

.NHCOCH3 .N— COCH3 hydrolysis NHN0 2

/ / Ba(OH) 2 /

2CH3CONH2 + CH 2 -> CH 2 -> CH 2 > CH 2

\ \ and acidification \

\NHCOCH3 \N— COCH3 \nhno 2

N0 2

This synthesis served as a basis for the confirmation of the structure of methylene-
dinitramine.

Methylenedinitramine undergoes decomposition under the influence of strong
acids and strong bases. At a pH of about 1.0 and 10.0, however, it is fairly stable
although it decomposes readily when the pH ranges between 3 and 8, the maximum
of the decomposition rate occurring at pH 5.4 (Lamberton et al. [14]). Decomposition
is accompanied by the evolution of nitrous oxide and formaldehyde.

The methylenedinitramine homologues with the general formula:

.NHN0 2

(CH 2 )«
\nhno 2
are far more stable.

Traube [15] prepared methylenedi-isonitramine, the isomer of methylenedinitr-
amine, in the form of a sodium salt, by the action of nitric oxide on acetone, in the pres-

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ence of sodium alcoholate As T. Urbanski et al. [16] showed, this substance can also
be obtained by using paraldehyde instead of acetone. The structure of the 'soni^o
group has now been established as nitrosohydroxylamine. This substarT nd its
salts, which according to T. Urbanski et al. possess initiating propel ^ £
more fully described later (p. 221). p^pcrues, will be

ETHYLENEDINITRAMINE

CH 2 NH-N0 2

CH 2 NH-N0 2
m.p. 175-176°C

Franchimont and Klobbie [17] prepared ethylenedinitramine (EDNA Haleit^
by the nitration of ethylene urea a-imid^UH™^ -1 l^^A, Haleite)

sulphuric acids. (According to S hweZ fwtZ * ""^ ° f "** Wd

the action of uri on ethyUS ^^Z^^^^ *

with mixed nitric acid and acetic anhydride:

CH 2 NH 2 OC 2 H 5 CH,— NH ru 1

( 5 <~h 2 NH CH 2 _N CH 2 NH-NO z

I + CO -> x co

CH 2 NH 2 OC 2 H 5 CH 2 -NH

+ C0 2 (17)
1
CH 2 NHN0 2

N0 2

m.p. 132-132.5°C m.p. 211-212°C

Z 2 f™lr f * ™ S ' able ""' "-*'«**-* by heating i„ water
™.h nii a iLta S ( * 8 VTb!; f T r t rraa ' e °" e,h *"« U ^™ »<« Nation

CH 2 NH 2 e ! cooc ! H 5 CH 2 NH.COOC 2 H 5 ^CH 2 N.COOC 2 H 5 14% NHj C h 2NHN 2
CH 2 NH 2 CH 2 NH.COOC 2 H 5 CH.N-COOC.H, > CH 2 NHN0 2 ™

N0 2
m.p.H0°C m.p.82-83°C

^s-^r^ssfir 4 ^"*

ALIPHATIC NITRAMINES AND NITRAMIDES 19

They found that a good yield of ethylenedinitramine may also be obtained from
ethylenediamine through diacetylethylenediamine (ethylene-bis-acetamide). The nitra-
tion of the latter involves the use of nitric acid (98 %) mixed with acetic anhydride :

N0 2

I
CH 2 NH 2( ch 2 CO) 2 oCH 2 NHCOCH3 hno 3 CH 2 -N— COCH3 , 4% nH , CH 2 NHN0 2

CH 2 NH 2 " CH 2 NHCOCH 3 (CH 3 CO) 2 o > CH 2 — N— COCH3 " CH 2 NHN0 2

N0 2
m.p. 136°C

It is important to bear in mind that ethylene-bis-acetamide (earlier prepared by
A. W. Hofmann [20]) cannot be nitrated with nitric acid alone or nitric and sulphuric
acids. The product of nitration is readily hydrolysed by the action of 30% NaOH
solution or ca. 15% NH 3 solution at room temperature.

The same authors found that ethylenedinitramine may be obtained in a similar

way by promoting the transient formation of ethylene oxamide from ethylenediamine

and ethyl oxalate :

NO 2

CH 2 NH 2 COOC 2 H 5 CH 2 — NH— CO H N0 3 CH 2 — N— CO 14% NH , CH 2 NHN0 2

I +1 -H I ► I I > I

CH 2 NH 2 COOC 2 H 5 CH 2 — NH— CO (CH 3 C0) 2 CH2 _ N _co CH 2 NHN0 2

N0 2
m.p. 275-285°C m.p. 197-198°C

Ethylene oxamide was prepared earlier by van Alphen [21]. It cannot be nitrated
either with nitric acid alone (98%) or with mixed nitric and sulphuric acids.
Ethylenedinitramine is produced on a technical scale in the following way:
One part of ethylene urea is introduced at a temperature not higher than 10°C
into ten parts of a mixture consisting of:

HNO3 15.4%
H 2 S0 4 74.0%
H 2 10.6%

After the last portion of ethylene urea has been added, the solution is poured into
ice water. The nitroethylene urea thus precipitated is filtered, carefully washed and
thrown into boiling water. On hydrolysis carbon dioxide is evolved. Boiling is con-
tinued until all the gases have been removed, and then the solution is cooled down.
Ethylenedinitramine crystallizes in the rhombic system as white, lustrous crystals,
s.g. 1.75 which after filtration are washed with cold water and dried at 50°C.

Hale [22] recommended ethylenedinitramine for use as a high explosive. It is
insoluble in ether, but soluble in nitrobenzene and dioxane. The solubility of ethylene-
dinitramine is given in Table 1.

Ethylenedinitramine is non-hygroscopic and picks up only 0.01 % of moisture in
damp conditions at room temperature. It is a strong acid and easily forms salts. The

20 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

potassium salt can be recrystallized from alcohol. The silver and lead salts are
highly sensitive to impact (they have a sensitiveness similar to that of mercury fulmi-
nate), but have no initiating properties. Ethylenedinitramine is not explicitly toxic.

Table 1
Solubility of ethylenedinitramine

Temperature

Solubility in 100 g of

°C

water | 95% alcohol

25
50
75
95

0.3
1.25
4.95
16.4

1.25
3.45
10.1

Its chemical stability is fairly high and only an insignificant amount decomposes
on prolonged boiling in water. Boiling in dilute sulphuric acid causes decomposition
with the evolution of nitrous oxide, acetaldehyde and ethylene glycol.

Hale [22] reports that the ignition temperature of ethylenedinitramine is 180°C,
i.e. similar to that of nitroglycerine. On heating at 120°C its stability is of the order of
that of tetryl.

Ethylenedinitramine is a powerful explosive:

Heat of explosion 1 267 kcal/kg

Volume of gases ( V ) 908 l./kg

Rate of detonation at a density of 1 .55 7750 m/sec

A. LB. Robertson [23] found that under a pressure of 100 mm of nitrogen at
174-178°C the thermal decomposition of EDNA is a first order reaction. The activa-
tion energy is 30.5 kcal and log 5= 12.8.

The substance possesses quite uncommon and valuable explosive properties.
It is more powerful than tetryl, and considerably less sensitive to impact (as sensitive as
picric acid). However, its acidic properties limit its use to a great extent. In this
respect it resembles picric acid. Even so ethylenedinitramine, under the name of
Haleite, has been accepted in the United States as a military explosive. During
World War II, production in that country was carried out by the method outlined
above according to eqn. (17)

OTHER NITRAMINES DERIVING FROM ETHYLENEDIAMINE

A series of new nitramines, the ethylenediamine derivatives, was prepared by
Picard and Meen [24] by the action of acetone cyanohydrin on ethylenediamine and
its analogues (i.e. diethylenetriamine etc.), followed by the nitration of the products

ALIPHATIC NITRAMINES AND NITRAMIDES 21

obtained with mixed nitric acid and acetic anhydride. In the presence of CH 3 COCl
or HC1 or ZnCl 2 , nitration gives yields as high as 76%. The simplest example of
these compounds is the product (II), prepared by the following steps.

H 3 C CN 3 H 3 C CH 3
\ / \ /

C \ C \

NH N— N0 2

I I

CH2NH2 H 3 C OH CH 2 CH 2

\/ I I

+ C -> CH 2 -> CH 2

/\ I I

CH 2 NH 2 H 3 C CN NH N— NO2

c / ex

/ \ / \

H 3 C CH 3 H 3 C CH 3

m.p. 54.5-56°C m.p. 212°C (decomp.)
I II

Higher homologues of ethylenedinitramine

NH-N0 2

(CH 2 )„

I
NHN0 2

are crystalline substances with the following melting points:

m.p.

at «=3 69°C

n=4 163°C

n=5 60°C

The explosive properties of these compounds have not been examined.

NITROCYANAMIDE

/NH-N0 2

m.p. 137-138°C

Harris [25] isolated this substance by the action of anhydrous hydrogen chloride
on the solution of the silver salt of nitrocyanamide in acetonitrile. It has explosive
properties but it is of no practical value due to the difficulty of preparing it. On the
other hand its salts may be of practical value. These salts and their preparation will
be discussed in the chapters dealing with initiating explosives (p. 211).

22 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

NITROGUANIDINE

NHNOJ

NH 2

C=NH

C=NN0 2

NH 2

(tautomeric forms)
m.p. 232°C and 257°C

Jousselin [26] prepared nitroguanidine by the action of anhydrous nitric acid or
sulphuric acid on guanidine nitrate. The preparation of this substance by the action
of sulphuric acid has been developed as an industrial method for the production of
nitroguanidine. The method described by Marqueyrol and Loriette [27] follows
somewhat different principles. It consists in acting with anhydrous nitric acid on
guanidine sulphate which, in turn, is obtained on treating dicyandiamide with sulphuric
acid.

Attention has been focussed on the explosive properties of nitroguanidine since
the beginning of the present century. Proposals were made for its use as a com-
ponent of various high explosive mixtures, e.g. fusible ternary mixtures containing
ammonium nitrate and guanidine nitrate (Albit) apart from using nitroguanidine
itself. Before World War I, detonating fuses filled with nitroguanidine were used in
French mines.

As a high explosive nitroguanidine had limited application until World War II,
when it acquired a considerable significance owing to the fact that flashless and rela-
tively non-erosive powders containing nitroguanidine, nitrocellulose, nitroglycerine and
nitrodiethyleneglycol, were employed very widely. As early as 1901 Vieille [28] point-
ed out the negligible erosive properties of nitroguanidine as a component of propel-
lant powders. Since that time interest in this substance as a component of propellant
explosives has continued to increase. At first, however, nitroguanidine found no
practical application since it cannot form a solution with the colloidal propellant
and because it remains "foreign" , to this colloid it makes the propellant brittle. This
is particularly evident in nitrocellulose propellants. Recchi [29], however, called atten-
tion to the fact that the incorporation of nitroguanidine into a totally colloidal
nitroglycerine propellant is possible without much detriment to its elasticity and
mechanical strength. His idea was put into practice when: (1) the production of
nitroguanidine from atmospheric nitrogen, starting from cyanamide, was developed
and (2) nitrodiethyleneglycol came into use as a component of totally colloidal
"double base" propellants, these being notable for their greater elasticity and me-
chanical strength as compared with propellants containing nitroglycerine.

ALIPHATIC NITRAMINES AND NITRAMIDES 23

PHYSICAL PROPERTIES

Nitroguanidine exists in two crystalline forms. The a-form results from the action
of sulphuric acid on guanidine nitrate followed by the precipitation of the product
with water. This form crystallizes from water in long, fairly flexible needles.

The /?-form is produced either alone or together with some of the a-compound,
by the nitration of the mixture of guanidine sulphate and ammonium sulphate which
results from the action of sulphuric acid on dicyandiamide. The /?-form crystallizes
from water in thin, elongated plates. It is converted into the a-compound by solution
in sulphuric acid and precipitation with water. Both forms of nitroguanidine melt at
the same temperature. Several authors quote different melting points: 232, 246,
257°C.

The two forms appear to differ slightly in their solubility in water, neither form
being converted into the other. At 25 and 100°C the solubility of the a-form is 4.4 g/1.
and 82.5 g/1. respectively. Between these temperatures the /?-form appears to be more
soluble.

The problem of preparing nitroguanidine in finely powdered form is of great
importance, since this is the only form suitable for incorporation into colloidal
propellants (nitroglycerine or nitrodiethyleneglycol powders). Rapid cooling of the
aqueous solution of nitroguanidine produces very small crystals, but they are still too
coarse for use as a component in propellants. The desired fine powder may be obtained
by spraying hot nitroguanidine solution onto a cooled, metallic surface, by allowing
the spray to drop through a tower in counter current to a stream of cold air and
finally, by allowing the product to crystallize from solutions containing substances
which regulate the size of the crystals as they are formed.

Pritchard and Wright [30] have described their method for preparing fine-crystal-
line, free-flowing nitroguanidine. They prepared a hot aqueous saturated nitroguani-
dine solution which was poured into cold methanol. Ninety per cent of the nitroguani-
dine precipitated in fine crystals. From the solution containing 10% of nitrogu-
anidine, methanol was distilled off. The remaining aqueous solution was used again
for dissolving nitroguanidine. The best ratio (by volume) of water to methanol lies
between 1 : 2 and 1:1.

The apparent densHy of the crystals is 0.96, whereas that of ordinary commercial
nitroguanidine is about 0.25 and that of the product rapidly crystallized from methanol
is about 0.40.

The solubility of nitroguanidine in organic solvents is limited. Desvergnes [31]
determined its solubility in various solvents : water, acetone, methyl and ethyl alcohols,
ethyl acetate, ether, benzene, toluene, pyridine, chloroform, carbon tetrachloride
and carbon sulphide. In all these liquids the solubility of nitroguanidine is negligible,
the highest value-for pyridine-being 1.75 g/100 ml at 19°C.

According to Pritchard and Wright 1 part of nitroguanidine is dissolved in 1 1
parts of water at boiling point and in 375 parts of water at 20°C. The greatest increase
of solubility lies between 90 and 100°C.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Nitroguanidine dissolves in concentrated acids yielding labile salts. Its solubility
in sulphuric acid has been reported by Davis [32] (see Table 2).

Table 2
Solubility of nitroguanidine in sulphuric acid

Concentration of

Solubility of nitroguanidine in 100 ml of acid

sulphuric acid

°/

at0°C

at 25°C

5.8

10.9

3.4

8.0

2.0

5.2

1.3

2.9

0.75

1.8

0.45

1.05

0.30

0.55

0.12

0.42

T. Urbanski and Skrzynecki [33] examined a number of systems containing
nitroguanidine and found the following binary and ternary eutectic mixtures— Table 3.

Table 3
Eutectic mixtures with nitroguanidine

Content of nitro-

Freezing point

Components

guanidine in eutectic

of eutectic

°C

Nitroguanidine +

ammonium nitrate

131.5

Nitroguanidine +

guanidine nitrate

166.5

Nitroguanidine +

guanidine nitrate +

ammonium nitrate

17.5 and
22.5 of
guanidine nitrate

113.2

The ultra-violet absorption spectrum of nitroguanidine was first examined by
Baly and Desch [34]. Figure 4 shows a curve plotted according to the investigations
of R. N. Jones and Thorn [35]. In a neutral solution with an aqueous solvent the
curve shows two maxima at about 210 and 265 /*. The absorption curve is unaffect-
ed by the addition of hydrochloric acid but under the influence of IN NaOH
solution the two maxima are converted into one, at about 250 p. These changes
may be caused by the tautomeric modifications of nitroguanidine. However, McKay,

ALIPHATIC NITRAMINES AND NITRAMIDES

Picard and Brunet [36] suggest that nitroguanidine may be a resonance hybrid.
The structure of nitroguanidine will be discussed later. In the infra-red, nitroguan-
idine gives an absorption band of asymmetric vibrations of the N0 2 group which

^
&

4.0

II!"

,■■■■■% "A

3.6

- /'

3.2

2.8

vs. 1

2.4

! ! 1 1

2200 2600

Wave length/A)

Carve A. Solvent, water

— »- B. Solvent, N hydrochloric acid

— "- C. Solvent, N sodium hydroxide

Fig. 4. Spectrum of nitroguanidine, according to R. N. Jones and Thorn [35].

deviates considerably from the average values of the N0 2 group in nitramines
(approximately 1635 cm' 1 instead of normal value 1560 cm- 1 ). The existence of
tautomeric forms may explain this devation (Bellamy [37]).

CHEMICAL PROPERTIES

Jousselin [26] wrongly ascribed to nitroguanidine the formula of an N-nitroso
compound. Later Pelizzari [38], Franchimont [39] and Thiele [40] considered it to
be nitramine. Franchimont proposed a nitroimine structure (II) (p. 22) and later
Thiek deduced that it has the structural formula of a primary nitramine (I) (p. 22)
by virtue of its ability to form salts. Much later, T. Urbanski, Kapuscinski and
Wojciechowski [41] prepared the complex mercuric salt by the action of mercuric
nitrate on an aqueous solution of nitroguanidine and examined its explosive prop-
erties. Nitroguanidine does not form salts with other metals. Later on Wright et al.
[42, 43] once more expressed the opinion that nitroguanidine has the nitroimine
structure (II) taking the form of nitramine only when influenced by alkalis. This
new structure was based on the fact that unaltered nitroguanidine is precipitated
from solution in concentrated alkalis, not a nitroguanidine salt. Similarly according
to these authors potentiometric titration of freshly-prepared nitroguanidine solution
in alkalis is indicative of a lack of acidic function. This agrees with the well-known
observation that metallic vessels containing nitroguanidine do not corrode. Moreover,

26 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Wright, Barton and Hall [42] found, that on keeping a nitroguanidine solution for
a long time (10-20 hr) in dilute alkalis (0.1 N solution of NaOH) under reduced
pressure in order to remove volatile by-products, such as ammonia, potentiometric
titration demonstrates that the dissolved substance behaves as an arid. According
to the authors, this is caused by the conversion of the nitroimine form (II) into the
nitramine form (I) under the influence of an alkaline medium. Later Kirkwood and
Wright [44] and Kumler and Sah [45] came to the conclusion, on the basis of dipole
moment measurements, that nitroguanidine has the structure of a nitroimine or
forms a resonance hybrid. This view was also confirmed by de Vries and Gantz [46].
On examining the absorption spectrum in the infra-red, Kumler [47] found that a
hydrogen bond occurs in nitroguanidine and in a series of its derivatives:

NH 2

HN N=0

\h-o/

Nitroguanidine has weakly basic properties and this accounts for its ability to
form salts with concentrated acids, e.g. it forms a sulphate with concentrated sulphuric
acid. Nitroguanidine is hydrolysed on heating with concentrated sulphuric acid
evolving nitrous oxide and carbon dioxide, the former probably derived from hydro-
lysis of nitramine and the latter from hydrolysis of cyanamide. The latter also yields
ammonia on decomposition.

NH 2 — C— NHN0 2 -> NH 2 N0 2 + NH 2 CN (19)

+ 2H 2

N 2 + H 2 2NH 3 + CO2 (20)

A solution of nitroguanidine in concentrated sulphuric acid also undergoes
hydrolysis after standing for some time at room temperature. Then when diluted
with water it no longer gives a precipitate of nitroguanidine.

A freshly prepared solution of nitroguanidine in sulphuric acid contains no
free nitric acid, but in the presence of substances which are readily nitrated it behaves
as if this were so, e.g. the solution nitrates phenol, acetanilide and cinnamic acid
and in the presence of mercury reacts in a nitrometer with the evolution of nitric
oxide in the same way as nitric acid. Hence in certain cases a solution of nitroguanidine
in sulphuric acid may be utilized as a nitrating mixture.

One explanation is that in the presence of these substances nitramine is hydrolysed
to form nitric acid [48].

NH 2 N0 2 + H 2 -> NH 3 + HNO3 (21)

Another, more recent explanation, ascribes to nitroguanidine the ability to liberate
the NOf ion under the influence of sulphuric acid [48].

The decomposition of nitroguanidine by the action of ammonia in aqueous
solution also proceeds according to equations (19) and (20).

ALIPHATIC NITRAMINES AND NITRAMIDES

Barton, Hall and Wright [42] found that the action of alkalis on nitroguanidine
involves hydrolysis with the formation of ammonia, nitrourea and the products of
the decomposition of nitrourea.

Nitroguanidine demonstrates high stability in aqueous solution on boiling, but
on long-continued boiling evolves small amounts of ammonia, possibly due to decom-
position according to equation (22) :

NH 2 — C— NH— N0 2 -> NH 3 + HNCN— N0 2
NH Nitrocyanamide

(22)

The ammonia so formed causes the decomposition of the nitramine that results

from reaction (19).

According to equation (19) decomposition also occurs on boiling nitroguanidine
in an aqueous solution of ammonium carbonate, with liberation of nitrous oxide
and ammonia. The latter combines" with the cyanamide also resulting from reaction
(19) and guanidine carbonate is formed in almost quantitative yield.

In the presence of primary aliphatic amines in aqueous solution nitroguanidine
undergoes decomposition according to equation (22). Ammonia is then evolved, and
nitrocyanamide combines with the amine to form alkylnitroguanidine, e.g. N-methyl-
N'-nitroguanidine :

HNCN— N0 2 + CH3NH2 -> CH 3 NH— C— NH— N0 2

(23)

Wright and McKay [49] suggest a different reaction mechanism for the N-methyl-
N'-nitroguanidine: in the first stage it involves the addition of methylamine to the
nitroimine form of nitroguanidine, and this is followed by the liberation of ammonia
and the formation of methylnitroguanidine (I).

Methylnitroguanidine so produced (I) is capable of further reaction with methyl-
amine to yield dimethylguanidine (II).

These reactions may be characterized by the following mechanism (Wright [50]):

NH 2

I
CH 3 — NH 2 + N0 2 — N=C— NH 2 -

CH3-

1
-N-

1
— C—

-N-

-NO2

H 2 N

1 !

H NH

I II
CH 3 — N— C— N— CH 3 «-

I
H

NH2NO2

28 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Dimethylguanidine is formed by the liberation of nitramine. The process is ir-
reversible.

The structure of this substance was proved by the facility with which it can be
hydrolysed to amine and nitrous oxide (24) :

CH3NH— C— NH— N0 2 + H 2 -> CH3NH2 + NH 3 + N 2 + C0 2 (24)

The reaction (24) indicates that the alkyl and nitro groups in N-alkyl-N'-nitro-
guanidine are linked with different nitrogen atoms.

The same N-alkyl-N'-nitroguanidines are obtainable by the nitration of alkyl-
guanidines (Davis and Elderfield [51]).

With diamines such as ethylenediamine, addition to nitroguanidine leads only
to the evolution of ammonia, so that the course of reaction is somewhat different.
This will be considered later.

When heated with an aqueous solution of hydrazine (reaction (25)) nitroguan-
idine yields N-amino-N'-nitroguanidine (II) (Philips and Williams [52]), a white,
crystalline substance of marked explosive properties, m.p. 182°C:

NHN0 2 NH-N0 2

C=NH + NH 2 NH 2 -> C=NH + NH 3 (25)

NH 2 NH-NH 2

I II

The substance (II) is converted by the action of nitrous acid into nitroguanyl
azide (Ha) at 0°C or into nitraminotetrazole (lib) (Lieber et al. [53]) at 70°C.
Both substances are explosive:

NH— NO 2 N N

1 y

C=NH 2 NNH— C

I \

N3 NH— N

Ha lib

Under the influence of diamines, such as ethylenediamine, nitroguanidine yields
cyclic compounds of type III with evolution of ammonia; they are liable to the
further nitration through type (IV) up to cyclic nitramides of type (V) (McKay and
Wright [54]).

/ N H 2 NH

H 2 C/ H 2 N H 2 C/ \

I + >C— NH— N0 2 -> I C— NH-NOj ->

H 2 C X NH^ H 2 C1 /•

\nh 2 \n

ALIPHATIC NITRAMINES AND NITRAMIDES 29

N0 2 N0 2

I I

H 2 C/ \ H 2 C/ \

C— NH- N0 2 -> I C=0

/" H 2 a /

N X N

N0 2

rv v

On reduction, nitroguanidine is converted first into nitrosoguanidine and then
into aminoguanidine i.e. guanylhydrazine. The latter is used for the manufacture
of tetrazene (p. 206), and in organic chemistry to form crystalline derivatives from
aldehydes and ketones, just as semicarbazide forms semicarbazones.

Nitroguanidine and nitrosoguanidine both give a blue colour with diphenylamine
in concentrated sulphuric acid anvi both give the characteristic reactions described
below:

(1) To an approximately 25% solution of nitroguanidine in cold water some
drops of saturated ferric ammonium sulphate solution are added, then a 24% solu-
tion of NaOH. The filtered solution has a red colour resembling that of fuchsine.

(2) Nitroguanidine is dissolved in acetic acid, diluted to approximately 10%,
treated with zinc dust in the cold, set aside for thirteen to nineteen minutes and
filtered; 6% cupric sulphate solution is thesi added to the filtrate. The solution turns
intensely blue and on boiling, becomes turbid, gives off gas and deposits a precipi-
tate of metallic copper. If silver acetate is added instead of cupric sulphate a precipi-
tate of metallic silver is deposited on boiling.

EXPLOSIVE PROPERTIES

Nitroguanidine decomposes immediately on melting, evolving ammonia and water
vapour and forming solid products. According to Davis and Abrams [8], among the
products resulting from the decomposition of nitroguanidine the following substan-
ces occur : nitrous oxide, cyanamide, melamine (from the polymerization of cyanamide),
cyanic acid (from the decomposition of nitrocyanamide), cyanuric acid (from the
polymerization of cyanic acid), and ammeline (VI) and ammelide (VII) (from the
co-polymerization of cyanic acid and cyanamide). The decomposition of the above
substances involves the formation of carbon dioxide, urea, nitrogen, hydrogen cyanide,
cyanogen and compounds not yet fully defined, such as melam, melem, and mellon
probably containing condensed triazine rings.

NH NH

II II

/ c \ / c \

HN NH HN NH

II II

OC C=NH OC CO

X N X X N /

H H

VI VII

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

At a temperature below the melting point nitroguanidine is said to be relatively
stable— more so than nitric esters and similar in stability to aromatic nitro com-
pounds.

Attention has been paid to nitroguanidine as an explosive since Vieille [28]
found that the gases from the decomposition of nitroguanidine are less erosive
than those from the decomposition of other explosives of comparable power (Table 4).

Table 4
Explosive properties of substances containing nitroguanidine

Substance

Charge
g

Pressure,,

kg/cm 2

Erosion
perf

Specific

pressure /

Nitroguanidine

3.90

2020

2.3

9000

Explosive gelatine

3.35

2460

31.4

10,000

Ballistite with

addition of 57%

nitroglycerine

3.55

2450

24.3

10,000

Nitrocellulose rifle

propellant B 7

3.55

2240

6.4

9600

Vieille expressed the opinion that nitroguanidine is less erosive than other ex-
plosives of the same power due to its low temperature of explosion. On the basis
of experiments with a manometric bomb Patart [55] calculated the following data
for nitroguanidine as an explosive:

temperature of explosion 907°C
covolume 1.60

specific pressure / 7140m

Such a low temperature of explosion was very surprising and the author appears
to have been in doubt as to its validity. Indeed, Muraour and Aunis [56] have shown
that the temperature of explosion of nitroguanidine may, in fact, be much higher.
They pointed out that nitroguanidine ignites with difficulty and undergoes incomplete
explosive decomposition. That is why the explosion temperature, as reported by
Patart, is so low.

Taking into account the results of experiments conducted with a manometric
bomb as well as the chemical composition and specific heat of the products of de-
composition Muraour and Aunis calculated the following values for the explosion
of nitroguanidine:

temperature of explosion 2098°C
covolume 1.077

specific pressure / 9660 m

Urbaiiski and Kapuscinski [57] found the following values for the explosive
properties of nitroguanidine (Tables 5 and 6).

aliphatic nitramines and nitramides
Table 5 Table 6

Effect of the compression on the
density of nitroguanidine

Pressure

Density

kg/cm 2

of loading

220

1.18

345

1.28

695

1.40

1040

1.48

1385

1.52

1730

1.57

2080

1.61

2775

1.65

3465

1.69

4160

1.71

4855

1.75

Rate of detonation of
nitroguanidine

Density
of loading

Rate of detonation

in an iron pipe

27/34 mm dia.

0.80

4695

0.95

5520

1.05

6150

1.10

6440

1.20

6775

1.30

6195

1.40

3300

1.45

2640

—

Cook [58] reports the rate of detonation to be 5460 m/sec at a density of 1.0.

Nitroguanidine may be regarded as an explosive which is powerful, but dif-
ficult to detonate. This accounts for the considerable fall in the rate of detonation
under increased density of loading. The diameter of loading also exerts a great
influence on the rate of detonation, behaviour which is also characteristic of explo-
sives which detonate with difficulty. An explosive of density 0.95 in a pipe of 20 mm
inner diameter gave a rate of detonation of 4340 m/sec.

According to T. Urbariski, Kapuscinski and Wojciechowski [41] the silver and
mercuric complex salts of nitroguanidine are more sensitive to impact than nitro-
guanidine itself, e.g., a weight of more than 10 kg must fall 100 cm to explode nitro-
guanidine, but the mercuric salt detonates when struck by a 10 kg weight falling only
12.5 cm.

THE PREPARATION OF NITROGUANIDINE

Bourjol [59] reviewed various methods of preparing nitroguanidine from guani-
dine nitrate and sulphuric acid, and also carried out extensive experiments to find
the most convenient laboratory method of carrying out the reaction.
The main features of the reaction, according to Bourjol are:
(1) The rate of reaction depends on the ratio H 2 S0 4 : (H 2 S0 4 + H 2 0). The
higher the ratio the faster the reaction. The rate is considerably reduced when the
concentration of sulphuric acid falls to 82% H 2 S0 4 , and the reaction practically
stops at 79-80% H 2 S0 4 . The concentration at the beginning and the end should
be 94-95% and 85-88% respectively.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(2) The rate of reaction depends on the size of the guanidine nitrate crystals.
It is strongly advisable to grind the guanidine nitrate before introducing it into the
sulphuric acid.

(3) It is advisable to use enough sulphuric acid to dissolve the guanidine nitrate
completely.

(4) The temperature during the reaction should be kept below 30°C. On the
other hand, too low a temperature may not be advisable as the solubility of guanidine
nitrate may be reduced and the reaction may be too slow. It is advisable to maintain
a temperature of 20-25°C at the beginning of the reaction and to raise it at the end
to 35-40°C, but not higher.

(5) The reaction solution should be diluted to 15% H 2 S0 4 keeping the tempera-
ture below 30°C $o ensure full precipitation of the nitroguanidine. The product
should be washed with an aqueous solution of ammonium carbonate and then with
water at 15-25°C.

According to Bourjol the yield is 92.8%.

Cave, Krotinger and McCaleb [60] worked out a general method for preparing
explosives in the form of fine crystals. It consists of introducing a hot solution into
a cold diluting liquid.

In the case of nitroguanidine, a hot aqueous solution was introduced into cold
methanol' Still finer and more uniform crystals were prepared by introducing a cold
solution in n-butanol into carbon tetrachloride. The results are given in Table 7.

Table 7

Solvent

Diluting
liquid

Average
crystal size

Limits of
the crystal size

Ratio
(length)
(width)

Water
n-Butanol

Methanol
Carbon tetra-
chloride

55
1.5

3-155
0.2-4.0

30
11

The method of manufacturing nitroguanidine adopted in Germany during
World War II consists of the addition of guanidine nitrate to a nitrator filled with
98% sulphuric acid, while the temperature is maintained below 45°C by cooling.
The nitroguanidine sulphate is formed very rapidly and nitroguanidine is then
precipitated by introducing the contents of the nitrator into a diluter containing
water, the mother liquor and the wash water at 0°C. The suspension of nitroguanidine
in 20% sulphuric acid is separated from the latter by centrifuging, washed with water
and is centrifuged again to a 25% content of water.

Crude, acidic nitroguanidine is dissolved in a boiling mixture of water and the
mother liquor from previous crystallization. To 1 part of nitroguanidine 14-16 parts
of solvent are added. The solution is neutralized with ammonia, filtered and chilled
by injecting the hot (approximately 100°C) solution into a vessel under reduced
pressure: this results in lowering the temperature of the solution to 45°C. A crystalline

■**a*Maa^MM

ALIPHATIC NITRAMINES AND NITRAMIDES 33

suspension of the product is formed and separated in a centrifuge. Thus nitroguanidine
containing 6% of water is obtained. The mother liquor is returned for re-use.

Nitroguanidine occurs in a fine-crystalline form in which it is suitable for the
manufacture of flashless propellant. A different form of nitroguanidine is used as
a high explosive. When it is to be compressed, its solution is rapidly evaporated
under reduced pressure to form a specially fine-crystalline product.

Nitroguanidine intended for use in a fusible mixture with trinitrotoluene takes
the form of fairly large, very regular crystals. For this purpose a colloidal substance
is added to the nitroguanidine solution, which is allowed to crystallize slowly.

In the method outlined above 136 kg of guanidine nitrate and 300 kg of 98%
sulphuric acid are used to produce 100 kg of nitroguanidine.

Another method of nitroguanidine preparation is that of Marqueyrol and Loriette
mentioned already [27]. According to Aubertein [61], here nitroguanidine is formed
by the following reactions: dicyandiamide is heated with 60% sulphuric acid at
150°C to form guanidine sulphate which however is not isolated. Instead, the reaction
mixture is treated directly with a 30% excess of anhydrous nitric acid at 25°C. Nitro-
guanidine, m.p. 257°C, is prepared in this manner in a 91% yield.

NITROUREA

NHN0 2

c=o

I
NH 2

m.p. 159°C (decomp.)

Nitrourea, like nitroguanidine, is prepared by the action of sulphuric acid on urea
nitrate. It was recommended as an explosive by Badische Anilin und Soda-Fabrik
in 1915 [62] but without success, as it was not sufficiently stable. In the presence
of water it decomposes at a little above 60°C with the evolution of nitrous oxide.

Davis and Blanchard [7] found that an aqueous solution of nitrourea or its
solution in concentrated sulphuric acid is hydrolysed to nitramine and cyanic acid
according to the equation:

nhno 2

CO ^z± NH 2 N0 2 + HNCO
NH 2

The reaction is reversible since nitramine in aqueous solution combines with
cyanic acid to reform nitrourea. Nitrourea is decomposed by gaseous ammonia,
a reaction which, according to Watt and Makosky [63] proceeds as follows:

H 2 NCONHN0 2 — ► HOCN + N 2 + H z O + NH 3
HOCN + NH 3 — ► NH4OCN — > NH 2 CONH 3

34 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The cyanic acid can react with ammonia to produce ammelide (I) and ammeline

(II):

HO N NH 2

\ y \ /

3HOCN + NH 3

3HOCN + 2NH 3

C C

I II

N N

\ /
C

I
OH

I
HO N NH 2

\ y \ /

c c

I II

N N

% y

NH 2

In addition, a polymer with the empirical formula (CsH^OsN,),, is produced.

Reaction with liquid ammonia at — 33°C probably proceeds in a manner similar
to that described by Davis and Blanchard: it is believed that nitramine (decomposing
into N 2 and water) and urea (apparently from cyanic acid and ammonia) are
formed.

The action of ammonia on nitrourea does not lead to the formation of nitro-
guanidine.

Urbahski, Kapuscinski and Wojciechowski [41] showed that nitrourea is a more
powerful explosive than nitroguanidine. Its lead block expansion is 310 cm 3 .

As a primary nitramine, nitrourea can form salts. The potassium, silver, mercuric
(Thiele and Lachman [64]) and ammonium (Hantzsch and Wiegner [65]) salts are
described in the literature.

T. Urbanski et al. [41] found that the silver and mercuric salts are much more
sensitive to impact than nitrourea itself, but have no initiating properties.

OTHER ALIPHATIC NITRAMINES

DINITRODIMETHYLOXAMIDE (MNO)

/N0 2
CO-N<

^CH

I /CH 3

CON<

\no 2

m.p. 124°C
This substance was prepared by Franchimont [66] by the action of anhydrous

ALIPHATIC NITRAMINES AND NITRAMIDES 35

nitric acid on N-dimethyloxamide. The latter is obtained readily by the interaction

of methylamine with methyl or ethyl oxalate:

N0 2
I
COOR CONHCH3 CON— CH 3

I + 2NH 2 CH 3 — > I — > I

COOR CONHCH3 CON— CH 3

N0 2

The product of nitration of dimethyloxamide is soluble in nitric acid and is
separated by pouring the solution into water. It decomposes on treatment with
concentrated sulphuric acid or on boiling with aqueous ammonia or barium hydroxide
solution, forming the corresponding methylnitramine salt. Similarly, long-continued
boiling in water results in complete decomposition, with the formation of oxalic
acid and methylnitramine.

In spite of being hydrolysed so readily its chemical stability is exceptionally high.
Haid, Becker and Dittmar [67] report that dinitrodimethyloxamide, like trinitro-
toluene, tetryl and penthrite, does not evolve oxides of nitrogen on being heated at
100°C for 30 days.

T. Urbanskj [68] found the substance to be only slightly sensitive to impact: it
does not explode under the impact of a 5 kg weight falling 90 cm. The rate of
detonation of dinitrodimethyloxamide was determined by T. Urbanski in a tin plate

tube 21 mm dia.:

density 0.87 6500 m/sec

density 1.22 6440 m/sec

density 1.33 7130 m/sec

The lead block expansion was 370 cm 3 .

T. Urbanski examined the possibility of mixing this substance with penthrite
(PETN) and picric acid to lower the melting points of these explosives.

Dinitrodimethyloxamide forms eutectic mixtures with penthrite and picric acid,

as follows :

with 37% penthrite m.p. 100.5°C

with 45% picric acid m.p. 78.6°C

DINITRODIETHYLOXAMIDE

/N0 2
CON<

\C 2 H 5

/C2H5

CON<

\no 2

m.p. 34-35°C

In general this substance has the same chemical properties as its dimethyl homo-
logue, described above. It is a weaker explosive than the dimethyl derivative and
shows only slight sensitiveness to impact (less than trinitrotoluene). It gives a lead
block expansion of 220 cm 3 .

36 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

DINITRODIMETHYLSULPHAMIDE
/N0 2

S CH 3

so 2

/CH 3

1 /^

m.p. 90°C

This substance was prepared by Franchimont [69] from dimethylsulphamide
obtained by the interaction of methylamine and sulphuryl chloride:

no 2

I
.NHCH 3 /NCH 3

S0 2 C1 2 + 2NH2CH3 — ► so 2 — > so 2 "

\nhch 3 \nch 3

no 2

The nitration is carried out with a large excess of anhydrous nitric acid (10 parts
of acid to 1 part of sulphamide). The product is precipitated by pouring the nitric
acid solution into water.

It dissolves with difficulty in water, moderately well in chloroform and benzene,
and readily in hot alcohol. Its ignition temperature is 160°C.

It is a very powerful explosive, as reported by Naoum and Meyer [70]. It gives
a lead block expansion of 395 cm 3 , i.e. similar to that of tetryl, but has the disadvantage
of being highly sensitive to impact.

NITRODIETHANOLAMINE DINITRATE (DINA)

,CH 2 CH 2 ON0 2

N0 2 — N

\CH 2 CH 2 0N0 2
m.p.49.5-51.5°C

In the inter-war period a certain interest was taken in the nitric esters of amino
and amido alcohols. The simplest of these was the product of nitration of diethanol-
amine, i.e. nitrodiethanolamine dinitrate.

The method of preparation was described by Wright, Chute, Herring and Toombs
[71]. They treated diethanolamine with a mixture of nitric acid and acetic anhydride
in the presence of hydrogen chloride as a catalyst. Instead of hydrogen chloride,
its salts such as zinc chloride may be used. The yield amounts to 90%, but is much
lower without a catalyst.

The substance is a very powerful explosive, similar to nitroglycerine in this respect.
It is capable of gelatinizing nitrocellulose and hence can be used instead of nitrogly-
cerine in propellants.

ALIPHATIC NITRAMINES AND NITRAMIDES 37

DINITRODI-(>HYDROXYETHYL)-OXAMIDE DINITRATE (NENO)

NO z
I
CO— N— CH 2 CH 2 ON0 2
I
CO— N— CH 2 CH 2 ON0 2

I
N0 2

m.p. 88°C
I

This substance was prepared by Herz [72] who recommended it as an explosive.
It is obtainable from oxalic ester by the following reactions :

COOC2H5 2NH 2 CH 2 CH 2 OH CONHCH 2 CH 2 OH nanuton

I > I > I

COOC2H5 CONHCH 2 CH 2 OH

In explosive power, it occupies an intermediate position between penthrite and
tetryl. Its lead block expansion is 450 cm 3 .

According to Domafiski and Mieszkis [73] the rate of detonation, at a density
of loading of 0.93 in a paper tube 10 mm dia., is 5200 m/sec (under the same con-
ditions the rate of detonation of penthrite was approximately 6000 m/sec). Cook [58]
found the rate of detonation to be 5530 m/sec at a density of loading of 1.0. The
substance is similar to tetryl in sensitiveness to impact. Its chemical stability is slightly
less than that of tetryl. Its ignition temperature ranges from 165 to 170°C.

In spite of many advantages, this substance has not achieved practical applica-
tion due to its high cost.

DINITRODIMETHYLDIAMIDE OF TARTARIC DINITRATE

N0 2

I
CO— N— CH 3

I
(CHON0 2 ) 2

CO— N— CH 3
I
N0 2

m.p. 114°C(decomp.)

T. Urbanski [74] prepared this substance by the nitration of the dinitrodimethy'-
diamide of tartaric acid with a mixture of nitric acid and acetic anhydride at a temp-
erature below — 2°C. It is capable of gelatinizing nitrocellulose.

It is a very powerful explosive (the rate of detonation at a density of 0.80 in a
cartridge 10 mm dia. is 4060 m/sec, and the lead block expansion is 390 cm 3 ) but it
is not sufficiently stable since its ignition temperature is only slightly above its melting
point. It is also very sensitive to impact— like nitroglycerine.

38 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

DINITRODI-(jS-HYDROXYETHYL)-SULPHAMIDE DINITRATE

N0 2

,N— CH 2 CH 2 ON0 2

S0 2

^N— CH 2 CH 2 ON0 2 1

I
N0 2

Herz [72] described the preparation of this substance by the action of ethanolamine
on sulphuryl chloride, followed by the nitration of the dihydroxyethylsulphamide so

obtained.

In explosive properties and explosive power it is similar to dinitrodi-(/?-hydr-

oxyethyl)-oxamide, described above.

LITERATURE

1. J. Thiele and A. Lachman, Ann. 288, 267 (1895).

2. A. Hantzsch, Ann. 292, 340 (1896).

3. K. Clusius, Helv. Chim. Acta 44, 1149 (1961).

4. J. Thiele, Ann. 296, 100 (1897).

5. C. A. Marlies, V. K. La Mer and J. Greenspan, Inorganic Synthesis, Vol. I, p. 68, McGraw-
Hill, New York, 1939.

6. R. P. Bell and G. L. Wilson, Trans. Faraday Soc. 46, 407 (1950).

7. T. L. Davis and K. C. Blanchard, J. Am. Chem. Soc. 51, 1790 (1929).

8. T. L. Davis and A. J. J. Abrams, /. Am. Chem. Soc. 47, 1043 (1925); Proc. Amer. Ac. Sci. 61,
437 (1926).

9. R. P. Bell and E. F. Caldin, Trans. Faraday Soc. 47, 50 (1952).

10. E. F. Caldin and J. Peacock, Trans. Faraday Soc. 51, 1217 (1955).

11. R. P. Bell, Advances in Catalysis 4, p. 170, Interscience, New York, 1952.

12. E. L. Hirst et al., according to A. H. Lamberton, Quart. Rev. 5, 75 (1951).

13. R. C. Brian and A. H. Lamberton, /. Chem. Soc. 1949, 1633.

14. A. H. Lamberton, C. Lindley and J. C. Speakman, J. Chem. Soc. 1949, 1650.

15. W. TRAUBE, Ann. 300, 81 (1898).

16. T. Urbanski and J. Zacharewicz, Wiad. Techn. Uzbr. 18, 16 (1932); T. Urbanski and T.
Wesolowski, ibid. 18, 28 (1932).

17. A. P. N. Franchimont and E. A. Klobbie, Rec. trav. chim. 5, 280 (1886); 7, 17, 239 (1888).

18. C. E. Schweitzer, /. Org. Chem. 15, 471 (1950).

19. W. E. Bachmann, W. J. Horton, E. L. Jenner, N. W. MacNaughton and C. E. Maxwell,
J. Am. Chem. Soc. 72, 3132 (1950).

20. A. W. HOFMANN, Ber. 21, 2333 (1888).

21. J. VAN ALPHEN, Rec. trav. chim. 54, 937 (1935).

22. G. C. Hale, /. Am. Chem. Soc. 47, 2754 (1925); U.S.Pat. 2011578 (1935).

23. A. J. B. Robertson, Trans. Faraday Soc. 44, 627 (1949).

24. J. P. Picard and R. H. Meen, Can. J. Chem. 30, 102, (1952).

25. S. R. HARRIS, /. Am. Chem. Soc. 80, 2302 (1958).

26. L. Jousselin, Compt. rend. 85, 548 (1877); 88, 814, 1086 (1879); Bull. soc. chim. France [2], 30,
186 (1878).

27. M. Marqueyrol and P. Lombtte, Swiss Pat. 87384 (1917).

ALIPHATIC NITRAMINES AND NITRAMIDES 39

28. Pi VlElLLE, Mem. poudres 11, 173 (1901).

29. V. Recchi, Z. ges. Schiess- u. Sprengstoffw. 1, 286 (1906); VI Intern. Chem. Congress, Rome
2, 580 (1906).

30. E. J. PRITCHARD and G. F. WRIGHT, Can. J. Research 25 B, 257 (1947).

31. L. DESVERGNES, Mem. poudres 19, 217 (1922); Chimie et Industrie 22, 37 (1930).

32. T.L.DAVIS, /. Am. Chem. Soc. 44, 868 (1922); The Chemistry of Powder and Explosives,
J. Wiley, New York, 1943.

33. T. Urbanski and J. Skrzynecki, Roczniki Chem. 16, 353 (1936).

34. E. C. C. Baly and C. M. Desch, /. Chem. Soc. 93, 1747 (1908).

35. R. N. JONES and G. D. THORN, Can. J. Research 27 B, 828 (1949).

36. A. F. McKay, J. P. Picard and P. E. Brunet, Can. J. Chem 29, 746 (1951).

37. L. J. Bellamy, The Infra-red Spectra of Complex Molecules, Methuen, Londoxi, 1958.

38. G. Pelizzari, Gazz. chim. ital. 21, II, 405 (1891).

39. A. P. N. Franchimont, Rec. trav. chim. 10, 231 (1891); 13, 308 (1894).

40. J. Thiele, Ann. 270, 1 (1892).

41. T. Urbanski, Z. Kapuscinski and W. Wojciechowski, Wiad. Techn. Uzbr. IV, 35 (1935).

42. S. S. Barton, R. H. Hall and G. F. Wright, J. Am. Chem. Soc. 73, 2201 (1951).

43. A. A. Amos, P. D. Cooper, E. Nishizawa and G. F. Wright, Can. J. Chem. 39, 1787 (1961).

44. M. W. Kirkwood and G. F. Wright , J. Org. Chem. 18, 629 (1953).

45. W. D. Kumler and P. P. T. Sah, /. Org. Chem. 18, 669, 676 (1953).

46. J. H. de Vries and E. S. C. Gantz, /. Am. Chem. Soc. 76, 1008 (1954).

47. W. D. Kumler, /. Am. Chem. Soc. 75, 3092 (1953); 76, 814 (1954).

48. T. UrbaNskI, Teoria nitrowania, PWN, Warszawa, 1954.

49. A. F. McKay and G. F. Wright, /. Am. Chem. Soc. 69, 3028 (1947).

50. G. F. Wright in H. Gilman'S, Organic Chemistry, Vol. IV, p. 951, J. Wiley, New York, 1953.

51. T. L. Davis and R. C. Elderfield, /. Am. Chem. Soc. 55, 731 (1933).

52. R. Philips and J. F. Williams, /. Am. Chem. Soc. 50, 2465 (1928).

53. F. Lieber, E. Sherman, R. A. Henry and J. Cohen, /. Am. Chem. Soc. 73, 2327 (1951);
F. Lieber, E. Sherman and S. H. Patinkin, ibid. 73, 2329 (1951).

54. A. F. McKay and G. F. Wright, /. Am. Chem. Soc. 70, 430 (1948).

55. G. PATART, Mem. poudres 13, 153 (1905/6).

56. H. MURAOUR and G. Aunis, Mem. poudres 25, 91 (1932/33).

57. T. UrbaNski and Z. Kapuscinski, Wiad. Techn. Uzbr. 38, 525 (1939).

58. M. A. COOK, The Science of High Explosives, Reinhold, New York, 1958.

59. G. Bourjol, Mem. poudres 32, 11 (1950).

60. G. A. Cave, N. J. Krotinger and J. D. McCaleb, Ind. Eng. Chem. 41, 1286 (1949).

61. P. Aubertein, Mem. poudres 30, 143 (1948).

62. Badische Anilin u. Soda-Fabrik, Ger. Pat. 303929 (1915).

63. G. W. Watt and R. C Makosky, Ind. Eng. Chem. 46, 2599 (1954).

64. J. Thiele and A. Lachman, Ann. 288, 267 (1895).

65. A. Hantzsch and G. Wiegner, Z. physik. Chem. 61, 485 (1908).

66. A. P. N. Franchimont, Rec. trav. chim. 2, 96 (1883); 4, 196 (1885); 13, 308 (1894).

67. A. Haid, F. Becker and P. Dittmar, Z. ges. Schiess- u. Sprengstoffw. 30, 66, 105 (1935).

68. T. UrbaNski, Wiad. Techn. Uzbr. IV, 3 (1935).

69. A. P. N. Franchimont, Rec. trav. chim. 2, 96 (1883).

70. Ph. Naoum and K. F. Meyer, Ger. Pat. 505852 (1929).

71 • W. J. Chute, K. G. Herring, L. E. Toombs and G. F. Wright, Can. J. Research 26 B, 89
(1948).

72. E. Herz, Brit. Pat. 367713 (1932).

73. T. Do'manski and K. MlESZKIS, Wiad. Techn. Uzbr. 44, 309 (1939).

74. T. UrbaNski, Roczniki Chem. 16, 334 (1936).

CHAPTER III

AROMATIC NITRAMINES

TETRYL

The key representative of aromatic nitramines is the trinitro derivative of phenyl-
methylnitramine— tetryl. This is 2,4,6-trinitrophenylmethyhiitramine or picrylmethyl-
nitraraine or N-2,4,6-tetranitro-N-methylaniline.

CH 3x yt
\ N /

N0 2

Tetryl is a widely-used explosive. It is also known under the names of Pyronite,
Tetrylit, Tetralite, Tetralita. In England it has been known for some time as Composi-
tion Exploding— CE. The incorrect name— tetranitromethylaniline— may sometimes
be encountered in the literature.

Tetryl, known since 1877, has been used as an explosive since 1906. During
World War I because of its explosive power and sensitiveness to initiation it was
employed in the filling of blasting caps and as a "priming" to fill detonating gains
(boosters). It is still used for the same purpose, although to a much lesser extent
since the introduction of PETN and cyclonite. During World War II tetryl was also
utilized as a component of high explosive mixtures.

Mertens [1] was the first to obtain tetryl in 1877 by the action of fuming nitric
acid on a dimethylaniline solution in sulphuric acid or by boiling a dinitrodimethyl-
aniline solution in fuming nitric acid. Soon afterwards Michler et al. [2] prepared
it by the action of fuming nitric acid on quaternary dimethylaniline salts. Neither of
these authors gave the correct structure of the product. Romburgh [3] later explained
more precisely the conditions of tetryl formation from methyl- and dimethylaniline.
He also postulated and then proved its structure by synthesizing it from potassium
methylnitramine and picryl chloride:

AROMATIC NITRAMINES

H 3 C

N0 2

\ N /
I
CH 3 2 N— /\- N0 2

N0 2

NITRATION OF DIMETHYLANILINE |

Nitration of dimethylaniline results in the oxidation of one of the methyl groups
to the carboxyl group which is not strongly linked to nitrogen and is readily split off
as carbon dioxide. Thus, as the nitration of dimethylaniline proceeds, gases consisting
of NO and N0 2 (from the reduction of nitric acid) and of C0 2 (from the oxidized
N-methyl group) are evolved abundantly.

The reaction mechanism was originally represented as follows:

H 3 C^ /CH 3 H 3 C^ ^CHa

NO,

-H 3 i

COOH

-N0 2

H 3 C V ,H

0- 1

N0 2
IV

N0 2

H 3 C X /H

\ N /

H3C \ z** 02

Indeed, all the above indicated intermediates with the exception of the hypothetical
product (III) have been isolated from the reaction mass.

The more recent work of Clarkson, Holden and Malkin [4] shows, however, that
the reaction proceeds somewhat differently. In fact, the dinitroderivative (II) which
undergoes demethylation to the substance (IV) is formed first, and then the nitramine
(VI) is isomerized (p. 5) to the trinitro derivative (V), before the latter is nitrated
to tetryl:

H 3 C^ ,no 2

H 3 <\ .H

\ N /

H 3 C^ /N0 2

42 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The last step (V) to (I) is reversible: if tetryl is dissolved in concentrated sulphuric
acid and allowed to stand the N-nitro group is expelled and trinitromethylaniline (V)

is formed (cf. p. 5).

The empirical equation for the preparation of tetryl from dimethylaiuline is as

follows :

C 6 H 5 (CH 3 )2 + IOHNO3 -> C 6 H 2 (N0 2 ) 3 N(N0 2 )CH 3 + 6NO2 + C0 2 + 8H 2

Hodgson and Turner [5] examined the action of various concentrations of nitric
acid alone on dimethylaniline, by nitrating 5 g test specimens of the latter on a
laboratory-scale.

They prepared tetryl by treating dimethylaniline with a 20-fold volume of nitric
acid, s.g. 1.52 at -5 to 0°C.

They also obtained a lower-nitrated product by employing nitric acid, s.g. 1.42
at 0°C. Later authors (Clarkson, Holden and Malkin [4]) found that this product
was N-2,4-trinitroethylaniline. At room temperature the reaction proceeds violently
and decomposition with a tendency to explosion readily occurs in the nitrator.
Nitric acid of lower concentrations, s.g. 1.34 and 1.254, gives 2,4-dimethylaniline
'at 0°C in quantitative yield. At a higher temperature they obtained a mixture of
this compound with 2,4-dinitromethylaniline.

The oxidizing action of nitric acid predominates when a still lower concentration
is used-specific gravity 1.12. In this case, originally described by Mertens [6] (who
suggested a wrong structure) and fully described by Romburgh [7] 3,3',5,5'-tetra-
nitrotetramethylbenzidine is formed as well as 2,4-dinitrodimethylanihne. Nitric acid,
s.g. 1.024 does not react with dimethylaniline, but in the presence of nitric oxides,
or NaN0 2 , leads to the formation of />-nitrosodimethylaniline.

Recently, T. Urbanski and Semenczuk [8] made it clear that tetryl may be safely
prepared by nitrating dimethylaniline with nitric acid, s.g. 1.40. Essential safety
precautions for the procedure are :

(1) The use of a large excess of nitric acid

/ nitric acid = cfl- 40 by weight \
I dimethylaniline J

(2) The conduct of the reaction in two stages.

In the first stage dimethylaniline is dissolved in nitric acid, the temperature not
being allowed to exceed 7°C. It is then gradually raised to 80°C. When the vigorous
reaction has subsided, the mixture is heated at 90 C C.

On cooling, the solution deposits crystals of tetryl. To obtain complete precipita-
tion of the product, water may be added to the solution. The yield of tetryl is 78%.

The reaction can also be carried out by dissolving dimethylaniline first in an
excess of nitric acid, s.g. 1.40 (e.g. in the weight ratio of nitric acid to dimethylani-
line of approximately 15 : 1) then adding excess nitric acid, s.g. 1.50 to this solution,
until the final weight ratio is approximately 25 : 1, finally proceeding as above. In
, mjfc 129.5°C is obtained in approxinoately 83% yield.

AROMATIC NITRAMINES

This method is suitable for continuous operation in a series of small nitrators.
A higher temperature is maintained in each successive nitrator, e.g. from 5 to 90°C.
By thus dividing production among a large number of reactors, an acceptable degree
of safety is achieved.

T. Urbaiiski and Semenczuk found that tetryl prepared with nitric acid alone
is of high purity, possibly because it is not contaminated with 2,4,6-^ initro-N-methyl-
aniline formed from tetryl as a result of the loss of the N-nitro group on heating
with sulphuric acid, present in an ordinary nitrating mixture.

Developing their method further, Semenczuk and T. Urbanski [9] devised a
way of nitrating dimethylaniline with nitric acid alone in the presence of an organic
solvent inert to nitric acid. This has two advantages: by diluting nitric acid with
a solvent, the course of the reaction is moderated; and by using a relatively low-
boiling solvent, e.g. chloroform, a "thermostatic" medium is created, the upper
temperature of which is limited by the boiling point of the solvent.

As early as 1938, Shorygin and Topchiyev [10] nitrated dimethylaniline with a
solution of nitrogen dioxide in chloroform, but they were only able to obtain 4-nitro-
dimethylaniline with a small amount of 3-nitiodimethylaniline. No demethylation
of the N-dimethylamino group occurred.

Semenczuk and T. Urbanski found that the following solvents may be used:
dichloromefhane, chloroform, carbon tetrachloride and tetrachloroethane. Their
nitration method consisted of introducing a solution of dimethylaniline into fuming
nitric acid diluted with the same solvent, maintaining a temperature of approxi-
mately 5°C. The mixture was then warmed cautiously to 40°C. On approaching
this temperature, strong evolution of nitrogen dioxide occurred and the temperature
rose. When chloroform or dichloromethane was used, the solvent distilled off at
61°C. The remaining solution, free from solvent, was warmed to 80°C until a light
orange colour was established. When the reaction was complete, water was added
to precipitate tetryl. The yield was high, 98% of theoretical and the purity of the
product was very satisfactory, m.p. 129°C. When carbon tetrachloride .was used,
it was removed either by decantation or by distillation at about 77°C. Higher boiling
solvents such as tetrachlorethane should be removed only by decantation, after
which the acid layer should be diluted with water as described above.

Semenczuk and T. Urbanski also showed that dimethylaniline may be nitrated
to tetryl by a mixture of nitric acid with acetic acid or acetic anhydride. As yet the
only other results mentioned in the literature with reference to this method are
those of Orton [11]. He asserted that dimethylaniline gives no N-nitro derivative
when reacted with a mixture of nitric acid with acetic anhydride or acetic acid; he
obtained only 2,4-dinitromethylaniline.

Semenczuk and T. Urbanski employed various proportions of nitric acid, s.g.
1.50, to acetic anhydride or acetic acid. The best results were obtained when the
volume ratio was used

nitric acid 50

acetic anhydride or acetic acid 50

44 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The spontaneous reaction which occurred after introducing dimethylaniline into
the nitrating mixture caused the tempeiature to rise to about 40°C. After the reaction
subsided, it was necessary to warm the reaction mixture and to keep it at 80°C
until the reaction was complete.

Water was added to the cooled reaction solution and pure N-2,4,6-tetranitro-
methylaniline precipitated. The yield was 90% of theoretical.

Thus, Orton's. statement should be limited to the specific experimental condi-
tions which he himself used.

NITRATION OF DINITROMETHYLANILINE
The conditions of nitration of dinitromethyl aniline with nitric acid were recently
studied by Lang [12]. He explained that in order to introduce the N-nitro group
it is necessary to use nitric acid with a concentration of more than 70%. The nitration
then proceeds as follows:

H 3 C X X H

\n/

H 3 C X ,NO,

A-NO,

70% HNQ 3

A-™,

N0 2
IV

no 2

In the presence of nitrogen dioxide, the reaction proceeds differently, and the
N-nitroso derivative is formed. When 99% nitric acid is used together with N0 2 , a
mixture of di- and trinitro nitroso derivative is produced, while 40% nitric acid
gives only a dinitro derivative

H3< \ M / H H3C. ,NO H 3 C X .NO

N0 2

-N0 2 ^XHNO^+no^ |^-N0 2 + 2 N-VVn0 2

N0 2

The nitration of 2,4-dinitromethylaniline to tetryl with nitric acid alone was
also studied by Issoire and Burlet [13]. These authors found that nitration may be
carried out by means of nitric acid, s.g. 1.44 (75% concentration) and over. When
using nitric acid, s.g. 1.44-1.46 (75-80% concentrations) the temperature of nitra-

* AROMATIC NITRAMINES 45

tion should be maintained above 70°C. The final temperature may be lower if the
final concentration of acid in the nitrator is higher than 85%. For instance when
using. 95% nitric acid (s.g. 1.50) in such quantity that by the end of nitration the
concentration is less then 85%, it is sufficent to maintain a temperature of 50°C
during that latter period.

The authors obtained a quantitative precipitation of tetryl by diluting the con-
tents of the nitrator to a 50-55% concentration of HN0 3 . Tetryl is practically
insoluble in such dilute acid hence the precipitation is quantitative.

Tetryl prepared in this way was contaminated by the following substances:

(1) chlorbdinitrobenzene occurring in dinitromethylaniline obtained by the
action of methylamine on chlorodinitrobenzene;

(2) 2,6-dinitromethylaniline which also contaminates technical dinitromethyl-
aniline.

Issoire and Burlet found that 2,6-dinitromethylaniline is nitrated with much
greater difficulty than the 2,4-isomer. To obtain tetryl from the 2,6-isomer, nitration
should be carried out with nitric acid, s.g. 1.50, i.e. a concentration higher than
95%.

The evolution of heat resulting from the introduction of the nitro group into
methylaniline was calculated by Garner and Abernethy [14] as follows. On transi-
tion of:

methylaniline to p-nitromethylaniline 36.4 kcal/mole,

p-nitromethylaniline to 2,4-dinitromethylaniline 25.2 kcal/mole,

2,4-dinitromethylaniline to 2,4,6-trinitromethylaniline 11.9 kcal/mole,

2,4,6-trinitromethylaniline to tetryl 1.0 kcal/mole.

By-products formed during the preparation of tetryl

Apart from tetryl a number of other substances may be found in the products
of the reaction. One of them is 2,3,4,6-tetranitrophenylmethylnitramine, wi-nitro-
tetryl (VII). This compound was originally reported by Romburgh [15] as accom-
panying tetryl. Romburgh originally thought that m-nitrotoluene is particularly
readily formed when tetryl is prepared from methylaniline, due to the fact that the
nitro group is directed to the meta position by the methylamino group:

H3C M H 3 Cs X H H 3 C V /N0 2 H 3 C\ /N0 2

O z N— f^V- N0 2 H 2 Q 2 N— /\~ N °2
■NO, I I— N0 2 > L J— OH

A nitro group at the meta position is readily hydrolysed in water to the phenolic
group with the formation of the nitro derivative of N-methyl-m-aminophenol (VIII).
Since technical dimethylaniline usually contains a certain amount of methylaniline,

46 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

its nitration product always contains a small quantity of the product (VII). This
substance may be removed by boiling in water in order to hydrolyse it to the com-
pound (VIII).

Later Romburgh and Schepers [16] explained that the substance (VII) is also
formed from dimethylaniline if nitration is carried out in the presence of a large
excess of sulphuric acid (20-fold with respect to dimethylaniline). It is evident that
the presence of this substance is undesirable owing to the poor stability of the nitro
group in the meta position and to the formation of metal salts of the substance
(VIII) which are sensitive to impact.

According to Desvergnes [17] methylaniline does not form m-nitrotetryl when
nitrated first with dilute acids and then with more concentrated ones.

Recently, Bogdal and D. Smolenski [18] made an extensive investigation of the
conditions of formation of m-nitrotetryl. Contrary to the previous work of Rom-
burgh [15] they found that m-nitrotetryl was formed in larger quantity when dimethyl-
aniline was nitrated. Methylaniline yielded a smaller amount of m-nitrotetryl under
identical nitration conditions.

According to these authors the following conditions lead to the formation of
m-nitrotetryl:

(1) If nitrating mixtures (nitric and sulphuric acids) are used for nitrating di-
methyl- and methylaniline dissolved in sulphuric acid, the percentage of m-nitro-
tetryl is higher than when nitrating with nitric acid alone: the figures observed
were 39 and 16% respectively.

(2) As the concentration of nitrating mixture falls, the percentage of m-nitro-
tetryl is considerably reduced, e.g. when the water content is 20-25%, the percentage
of m-nitrotetryl falls to below 1%. This agrees with Desvergnes [17].

(3) A lower temperature of nitration favours the formation of m-nitrotetryl,
e.g. when dimethylaniline is nitrated at 65 and 0°C the corresponding percentages
of m-nitrotetryl are 11 and 30%, respectively.

Bogdal and Smolenski [18] also found that the nitration of methylaniline with
less concentrated nitration mixtures readily led to formation of oxidation products.

The other substances which are formed apart from tetryl are the benzidine
derivatives (IX), (X), and (XI).

NO 2 N0 2

j 1

N0 2 N0 2

H3C \ J—\ J~\ / CUi

>N^f y-f >— n<
H 3 cy f^ ^^ CH 3

H3 °\ J~^ s\ / CH3

N0 2 N0 2

IX, decomposition at 272°C

X, decomposition above 200°C

N0 2

h 3 c Xn 1— .
o 2 n/ 1=

.I /CH 3

N0 2

XI, m.p. 222°C (with decomp.)

AROMATIC NITRAMINES 47

These substances are insoluble in benzene and hence easily removable from
tetryl by crystallization in this solvent. They are fine-crystalline products of a yellow-
ish colour. They increase with the amount of water contained in the nitrating acid.
Michler and Pattison [19] proved that N-tetramethylbenzidine is formed by heating
dimethylaniline with sulphuric acid.

The same reaction undoubtedly occurs when dimethylaniline is nitrated. Mertens
[6] isolated all three substances from the reaction of nitric acid with dimethylaniline
and the substances (X) and (XI) by the treatment of methylaniline with nitric acid.
Van Romburgh [7, 20] elucidated the structure of these compounds, in particular
the position of nitro groups.

The presence of undesirable impurities such as m-nitrotetryl and substances
which are insoluble in benzene necessitates fairly laborious purification of tetryl.
The by-products are formed at the very beginning of the nitration process, before
the formation of dinitrodimethylaniline (II). Nitration of highly pure dinitrodi-
methylaniline yields tetryl containing no such impurities.

GENERAL RULES FOR THE PREPARATION OF TETRYL

In technical operations dimethylaniline is employed as a starting material chiefly
because it is obtainable more easily and is much cheaper than methylaniline. Moreover,
tetryl prepared from dimethylaniline is purer than that from methylaniline. Hence
the preparation of tetryl from dimethylaniline is more economic, in spite of the
greater consumption of nitric acid by dimethylaniline which uses up nitric acid to
oxidize one of the N-methyl groups. The yield is not higher than 80%, due to the
side-reactions.

Since methylamine prepared from methyl alohol and ammonia has become
available commercially, the preparation of tetryl from dinitromethylaniline obtained
from chlorodinitrobenzene and methylamine has been widely used.

CI NHCH 3

^\— NO? ^ \— NO?

" 2 ™> CH > 1 II 2 HNO > _, tetryl

N0 2 N0 2

The yield of nitration of dinitromethylaniline amounts to 95%, hence this method
is the more economic.

Nitration is usually carried out in such a way that the solution of dimethyl-
aniline in sulphuric acid is introduced into a nitrating mixture rich in nitric acid.
The nitration reaction proceeds vigorously, and it is therefore most important to
control the reaction temperature. Formerly it was believed that for safety's sake
the lowest possible temperature of nitration should be maintained. Later it became
clear that at such a temperature nitration is not brought to completion and a large
quantity of incompletely nitrated products accumulates, which may lead to an

48 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

accident since at a certain moment these products begin to react, emitting a great
amount of heat and so creating the possibility that the contents of the nitrator may
explode. Thus, too low a temperature of nitration of dimethylaniline is considered
to be unsafe.

The nitration of dimethylaniline is now conducted at 68-72°C, with very vigorous
stirring. It is usually carried out by pouring gradually a sulphuric acid solution of
dimethylaniline sulphate into the nitrating mixture. Due to the high temperature
and vigorous stirring the nitration reaction proceeds at once so that there is no
danger of the accumulation of under-nitrated products.

Some of the older factory regulations drew attention to the possibility of resini-
fication of the product when a high nitration temperature was used and therefore
recommended maintaining a low temperature, i.e. 30-45°C. It was found, however,
that resinification of the product can be avoided by very vigorous stirring at high
temperatures and by employing a nitrating mixture fairly rich in nitric acid, e.g.
66.5% HN0 3 , 16% H 2 S0 4 , 17.5% H 2 0.

The nitration method described above is particularly suitable for use in a con-
tinuous system.

The purification of tetryl aims at removing by-products such as tetranitro deriva-
tive (VII), substances insoluble in benzene and the spent acid occluded by the crystals.
The product is washed with cold water and then treated with hot water. This brings
about the conversion of compound (VII) to (VIII)— the latter is soluble in hot
water. The tetryl is then dissolved in benzene and insoluble constituents removed
by filtration. The resulting solution is washed with water until it is completely free
from acid. Alternatively, tetryl may be dissolved in acetone, precipitated with water,
and finally deacidified.

PHYSICAL PROPERTIES

Tetryl crystallizes in the form of crystals which are colourless immediately after
preparation and crystallization but which rapidly turn yellow under the influence
of diffused light. The technical product is usually pale yellow.

Chemically pure tetryl melts at 129.45°C although for technical purposes a
melting point of 128.8 or 128. 5°C is acceptable.

Its crystals belong to the monoclinic system. The product used in industrial
practice should be crystalline in form and be easily pourable into a mould for com-
pression. According to Davis [21] a product with mixed, i.e. relatively large and
small crystals (Fig. 5), is best suited for this purpose.

Such a product may be prepared by various methods, e.g. by mixing a coarse
crystalline substance derived from crystallization in benzene with a fine crystalline
one obtained by the precipitation of tetryl with water from an acetone solution.
Another method (according to Crater [22]) consists of pouring the benzene solution
into water heated to above the boiling point of benzene. Alternatively, crystalliza-
tion from dichlorethane (according to Rinkenbach and Regad [23]) may produce
an acceptable form of tetryl.

AROMATIC NITRAMINES

The specific gravity of the crystals is 1.73, whereas the product, when fused and
poured into a mould, solidifies to a mass with a density of 1.62. Under a pressure
of 2000 kg/cm 2 a density of 1.71 can be attained.

Fig. 5. Tetryl in a free flowing form, according to Davis [21].

The specific heat of tetryl is as follows (C. A. Taylor and Rinkenbach [24]):

at 0°C 0.213 cal/g°C

at 20°C 0.217 cal/g°C

at 50°C 0.223 cal/g°C

at 80°C 0.228 cal/g°C

at 100°C 0.231 cal/g°C

atl20°C 0.234 ca1/g°C

Belayev and Matyushko [25] give 0.225 as the specific heat of tetryl.

Tetryl has a heat of fusion of 20.6 kcal/kg, a heat of combustion of 854.3 kcal/
/mole, hence the calculated heat of formation AH t is +7.5 kcal/mole (Garner and
Abernethy [26]) or +23.7 kcal/mole (Kast [27]).

According to Prentiss [28] the thermal conductivity of tetryl at 25°C is 0.00088.
Belayev and Matyushko [25] give 0.00023.'

Tetryl is virtually insoluble in water. It dissolves moderately well in concen-
trated mineral acids, but in spent acid its solubility is barely 0.3%. Conversely, con-
centrated nitric acid is a good solvent for tetryl. When a solution in concentrated
nitric acid is diluted slowly with water, for instance by placing it in a moist atmo-
sphere, gradual precipitation of tetryl occurs. Tetryl dissolves very readily in acetone.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Its solubility in benzene varies [29] depending on whether the substance is heated
with benzene to a given temperature (A) or the solution, saturated at a high tem-
perature, is cooled to a given temperature (B). The corresponding values are given
in Table 7.

Table 8
Solubility of tetryl in benzene

Solubility

(A) Heating in 100 g of
benzene

(B) Cooling a saturated
solution

Temperature °C

3.9
10.2

5.5
12.2

7.4
14.9

9.7
18.25

13.25
22.5

The big difference between the two states of equilibrium is presumably accounted
for by the formation of tetryl solvates with benzene at higher temperatures and
their greater solubility in benzene.

The solubility of tetryl in other solvents based on the data of various authors,
mainly those of C.A.Taylor and Rinkenbach [24] is tabulated below [21]:

Table 9
Solubility of tetryl (g in 100 g of solvent)

Temper-
ature

Water

Alcohol

95%

Carbon
tetra-

Chloro-
form

Dichloro-
ethane

Carbon
disul-

Ether

°C

chloride

phide

0.0050

0.320

0.007

0.28

1.5

0.0090

0.188

0.0058

0.366

0.011

0.33

0.0120

0.273

0.0065

0.425

0.015

0.39

0.0146

0.330

0.0072

0.496

0.020

0.47

0.0177

0.377

0.0075

0.563

0.025

0.57

3.8

0.0208

0.418

0.0080

0.65

0.031

0.68

0.0244

0.457

0.0085

0.76

0.039

0.79

0.0296

0.493

0.0094

0.91

0.048

0.97

0.0392

—

0.0110

1.12

0.058

1.20

7.7

0.0557

—

0.0140

1.38

0.073

1.47

0.0940

0.0195

1.72

0.095

1.78

—

0.0270

2.13

0.124

2.23

—

0.0350

2.64

0.154

2.65

18.8

—

0.0440

3.33

0.193

—

0.0535

4.23

0.241

—

0.0663

5.33

0.237

—

0.0810

—

0.0980

—

0.1220

—

0.1518

—

100

0.1842

—

AROMATIC NITRAMINES 51

CHEMICAL PROPERTIES

Tetryl is highly resistant to the action of dilute mineral acids. For instance long-
continued boiling with dilute sulphuric acid fails to decompose it, whereas in con-
centrated sulphuric acid the N-nitro group is split off with the formation of trini-
tromethylaniline (N-methylpicramide) (V) and nitric acid, as shown by Davis and
Allen [30] and later confirmed by Lang [12]. This reaction proceeds fairly quickly
at elevated temperatures (e.g. 60°C and above) but slowly at room temperature.
A solution in sulphuric acid, when brought into contact with mercury, reacts as
nitric acid: in the presence of mercury the N-nitro group is reduced to nitric oxide.
This enables tetryl to be determined quantitatively in the nitrometer.

When tetryl is boiled with a solution of sodium carbonate or a dilute (e.g. 2%)
aqueous solution of potassium or sodium hydroxide, the nitramino group is hydro-
lysed as follows [3, 6, 31]:

h 3 c Xn/ no 2

hoh 2 N— f \— N0 2
> + NH 2 CH 3 + HN0 2

The products obtained are picric acid (in the form of the corresponding picrate),
methylamine, and nitrous acid (as nitrite).

On heating with alcoholic ammonia a picramide is formed [3]:

H3C. ,N0 2

\ N /

NH 2

NH 3 2 N-/S— N0 2

no 2

Tetryl combines with aniline in benzene solution even at room temperature to
form 2,4,6-trinitrodiphenylamine (XII) and methylnitramine :

N0 2 N0 2

° 2N -C^- N < N02 +H2N -0 - ° 2N -C^- NH <I> + HN < N02

no 2 no 2

XII
m.p. 179.5-180°C

The substance (XII) is precipitated as red crystals, and methylnitramine can be
extracted from the solution with water.

The conversion of tetryl into picric acid or methylpicramide (V) may also proceed
under milder conditions. Thus Farmer [32] and Desvergnes [33] found that picric

5 2 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

acid can be formed simply by heating tetryl for a prolonged period of time, e.g.
at 120°C. According to some authors heating in high-boiling solvents (e.g. in xylene)
leads to the conversion of tetryl into N-methylpicramide (V) and picric acid, together
with resinous products and unidentified crystalline products.

When heated with phenol, tetryl is converted into trinitrophenylmethylamine
(V) (Romburgh [7]; cf. p. 5).

With tin and hydrochloric acid, tetryl is reduced and hydrolysed to form 2,4,6-

triaminophenol.

Tetryl combines with an excess of sodium sulphide to form a 13% solution.
Even at room temperature the nitro groups are reduced with the formation of a
non-explosive substance. This reaction is exploited for the destruction of waste tetryl.

Tetryl is reduced at 80-90°C by the action of a 10% sodium sulphite solution
to form a non-explosive product. In a similar manner tetryl reacts with sodium thio-
sulphate to yield yellow-coloured, unidentified products.

Yefremov et al. [34] studied systems containing tetryl by thermal analysis and
found that it forms additive compounds with phenanthrene, fluorene or retene in
the mole ratio 1 : 1 which do not melt uniformly. It also forms an additive compound
in the same ratio with naphthalene, m.p. 86.8°C.

According to Yefremov and Khaibashev [35], Giua [36], C.A.Taylor and
Rinkenbach [37], tetryl and trinitrotoluene form an additive compound in the ratio
of one mole of tetryl to two moles of trinitrotoluene which readily dissociates and
melts at 61.1 °C (Taylor and Rinkenbach report m.p. of the addition compound

as 68°C).

Tetryl also forms an ordinary eutectic mixture with 76.5% trinitro-m-xylene,
m.p. 118.8°C and with 29.5% trinitroanisole, m.p. 22.8°C [35].

At room temperature tetryl appears to be perfectly stable. At 100-120°C the
methylnitramino group of tetryl decomposes slightly giving off a certain amount
of formaldehyde and nitrogen oxides. On studying the decomposition of tetryl
at 120°C under reduced pressure, Farmer [32] found that in 40 nr 1.5-3.0 cm 3 of
gases are evolved. He also examined its decomposition at lower temperatures and
found that the temperature coefficient of decomposition is 1.9 cm 3 /5°C. By extra-
polating the curve in the system: decomposition rate-temperature, Farmer calcu-
lated that a decomposition lasting 40 hr at 120°C would take 1700 years at 20°C.

Tests lasting for many years have shown that 20 years of storage, at room temper-
ature involve no discernable changes in tetryl, nor was any distinct decomposition
of tetryl observed at 65°C after 12 months; at 75°C after 6 months and at 100°C
after 100 hr.

The decomposition rate of tetryl increases sharply (approximately 50-fold) at
its melting point. If tetryl contains admixtures which lower its melting point, it
begins to decompose at a lower temperature corresponding to the melting point
of the mixture. For example, addition of TNT to tetryl markedly increases its rate
of decomposition at the melting point of the mixture. This is probably due to the
more energy rich and less stable liquid state (see Vol. II, p. 182). When studying

AROMATIC NITRAMINES 53

the decomposition kinetics of pure tetryl or of tetryl containing an admixture of
picric acid at 140-1 50 °C, Hinshelwood [38] came to the conclusion that the speed-
ing up of the reaction rate results in the formation of picric acid due to the hydro-
lysis of tetryl at that temperature. The larger the amount of picric acid so formed,
the higher is the decomposition rate of the molten tetryl.

Desvergnes [33] found that when tetryl having a m.p. 128.5°C was heated for
24 hr at 100°C its m.p. was lowered by 0.4°C after 600 hr of heating the m.p. had
fallen to 71°C.

On heating tetryl at 120°C at reduced pressure (a few mm Hg) the gas evolved
in the early stages of the experiment consisted almost entirely of nitrogen and
carbon monoxide.

Towards the end of the test appreciable quantities of carbon dioxide were pro-
duced. No nitric oxide was detected at any stage of the experiment.

According to van Duin [39] the ignition temperature of tetryl is 196°C; when
it is heated from 100°C at the rate of 20°/min, ignition occurs at 187°C.

T. Urbanski and Schuck [40] found that tetryl explodes when dropped onto
a heated copper surface:

at 236°C after 6.2 sec
at 260°C after 2.0 sec
at 280°C after 1.1 sec
at 302°C after 0.4 sec
at 310°C instantly

According to Fanner [32] the activation energy of the thermal decomposition
of tetryl £=60.0 kcal and log 5=27.5. Hinshelwood [38] reports similar values.
A. J. B. Robertson [41] found, however, that at higher temperatures (211-260°C)
£=38.0 kcal and log 5=15.4. According to Szyc-Lewanska [42] at still higher
temperatures (260-300°C) the activation energy is even lower, viz. E= approximately
20.0 kcal (based on the data given above [40]).

Roginskii and Lukin [43] found that tetryl is not liable to explode when heated
at 150°C in a sealed ampoule, though at temperatures above 150°C, e.g. between
150 and 170°C, explosion may ensue as result of chain reactions occurring during
decomposition on long-continued heating.

EXPLOSIVE PROPERTIES

Lenze [44] was the first to examine the explosive properties of tetryl. It is a more
powerful explosive than TNT (its strength, depending on the method of investiga-
tion applied, ranges between 110 and 130% of that of TNT). Its sensititiveness to
impact and friction, particularly to rifle fire, is higher than that of TNT.

The explosive decomposition of tetryl, like that of other explosives, depends
on the method of initiation, density etc. Haid and Schmidt [45] give the following
equation for the decomposition of tetryl, at a density of 1.56:

C24.4H17.4O27.9N17.4 = 5.59CO z + 10.85CO + 5.91H 2 + 1.89H 2
+ 0.03C m H„ + 0.27CH 4 + 0.6HCN + 0.58C 2 N 2 + 7.82N 2 + 5.8C

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The heat of explosion is 1095 kcal/kg, the volume of gases 750 l./kg, the
temperature of explosion 3530°C.

Kast [27] gives similar values: 1090 kcal/kg, 710 l./kg, 3370°C respectively.

According to Carlton Sutton [46], depending on its density tetryl may explode
or detonate with widely varying intensity and therefore with different heats of ex-
plosion or detonation:

at a density of 0.9 and less with 935 kcal/kg (explosion)

at a density of 1.1-1.3 with 1070 kcal/kg

at a density of 1.45-1.71 with 1160 kcal/kg (detonation)

The sensitiveness of tetryl to impact, as reported by several authors, may be
expressed as 70-80% of the impact energy necessary to cause the explosion of pic-
ric acid. * ■ a •

According to a number of authors, the explosive power of tetryl, determined in
the lead block, varies between 340 and 390 cm3, i.e. between 114 and 120% of the

value for picric acid.

The rate of detonation of tetryl is reported by various authors as follows:

at a density of 1.0 and 1.53

at a density of 1.0
at a density of 1.43
at a density of 1.63-1.65
at a density of 1.63-1.65
pressed pellets (no density
given)

5450 m/sec and 7215 m/sec
respectively (J. Marshall [47])
5600 m/sec (Cook [48])
6425 m/sec (Selle [49])
7200 m/sec (Kast [27])
7520 m/sec (R. Robertson [50])
7230 m/sec (Koehler [51])

E Jones and Mitchell [52] found that under the influence of a standard No. 6
detonator, a charge of tetryl at a loading density of 0.94 and 25 mm dia. initially
detonates at a low rate which increases after a certain distance (the authors give

no figures).

The explosive power of tetryl at a loading density of 0.3, was compared with
that of other explosives in a manometric bomb by Koehler [51] who determined the
following pressure values from which the temperatures of explosion have been
computed:

Table 10

Explosive

Temperature of
explosion
°C

Tetryl
Picric acid
Trinitrotoluene
Trinitrobenzene

AROMATIC NITRAMINES

When confined in a 24 mm diameter tube, tetryl is capable of burning. According
to Andreyev [53], at a density of 0.9, the rate of burning of tetryl can be expressed
as a function of temperature by the equation:

-=31.50-0.055 T n

The diagram (Fig. 6) shows the dependence of u m and — on temperature.

40 80 120 t°C
Fig. 6. Rate of burning of tetryl against initial temperature, according to Andreyev [53].

A very important property of tetryl is its sensitiveness to initiation by a primer—
hence its rapid rise in importance as an explosive for use in detonating caps, gains
(boosters) etc. Martin [54] gives the following figures comparing the sensitiveness
to initiation of tetryl and trinitrotoluene, under the influence of various primary
explosives (Table 11).

Table 11

Primary explosive

Minimum charge (grammes) for
detonation of

Trinitrotoluene

Tetryl

Mercury fulminate
Silver fulminate
Lead azide
Silver azide

0.36
0.095
0.09
0.07

0.29
0.02
0.025
0.02

W. Taylor and Cope [55] determined the minimum charge of a mixture of mer-
cury fulminate (90%) and potassium chlorate (10%) necessary to detonate mix-
tures of trinitrotoluene and tetryl (Table 12).

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Table 12

Composition of mixture

Weight of

per cent

initiator

Trinitrotoluene

Tetryl

100

0.25

0.22

0.21

0.20

100

0.19

TOXICITY

Tetryl is a toxic substance. Breathing its dust induces symptoms of poisoning.
A concentration of 1.5 mg/m? of tetryl dust in the air is reported noxious (Troup
[56]) but a lower concentration may be toxic. Tetryl has a particularly potent effect
on the skin, producing symptoms of an allergic character: the skin turns yellow and
dermatitis develops. Thus workers employed in production, especially those engaged
in handling tetryl, should be provided with protective clothing. Parts of the body
exposed to tetryl dust should be protected by barrier cream containing 10% of
sodium perborate. Daily baths are essential. The presence of tetryl dust in the air
often causes irritation of the upper respiratory tract. Tetryl poisoning is also ac-
companied by general symptoms such as lack of appetite, insomnia, giddiness etc.

The symptoms usually occur 2-3 weeks after beginning work with tetryl. In
many cases (60-68%) some adaptation occurs and the effects of the poison appear
less pronounced.

Witkowski et al. [57] report that in one factory in the U.S.A. during World War II
out of 1258 workers engaged in handling tetryl 944 fell ill, while in another, 404 out
of 800-900 were affected.

TETRYL MANUFACTURE

Nitration of dimethylaniline

According to Sokolov [58] dimethylaniline used for the manufacture of tetryl
should be a pale-yellow liquid, s.g. 0.955-0.960 at 15°C, b.p. 192-194°C, not less
than 95% of the substance distilling between 192.7 and 193.7°C. It must not contain
water or aniline. Only traces of methylaniline are permissible.

The nitric and sulphuric acids should be purified to the extent usually required
for nitrating acids. Nitric acid of 92-98% concentration, containing not more than
3% of nitric oxides is commonly utilized. The concentration of sulphuric acid may
vary between 96 and 99%. The nitrating mixture should be rich in nitric acid, i.e.
containing not less than 65% of HN0 3 .

AROMATIC NITRAMINES

The production of dimethylaniline sulphate

In the manufacture of tetryl, it is usual not to nitrate dimethylaniline directly,
but to dissolve it first in concentrated sulphuric acid and then to nitrate the dimethyl-
aniline sulphate so obtained. Straightforward nitration of dimethylaniline proceeds
so violently, that it can be carried out only under the special conditions described
on pp. 42-43. Many years' experience of tetryl manufacture has shown that the ratio
of sulphuric acid to dimethylaniline should not be lower than 3 : 1, a smaller amount
of sulphuric acid may be detrimental to the nitration process. (Nitration by a peri-
odic (discontinuous) method may cause ignition in the nitrator due to the fact that

Fig. 7. Flow diagram of plant for dissolving dimethylaniline in sulphuric acid, accord-
ing to Sokolov [58].

a portion of dimethylaniline was not previously converted into sulphate.) The
proportion of sulphuric acid to dimethylaniline should not however be too high,
e.g. a ratio of 100 : 1 has a detrimental effect on the yield of tetryl. While the di-
methylaniline is dissolving the temperature should be maintained between 20 and
45°C, but not higher, to avoid sulphonation of the benzene ring.

The lay-out of a plant for dissolving dimethylaniline in sulphuric acid (according
to Sokolov [58]) is shown, with some modifications, in Fig. 7. It comprises a dosing
tank (i) for sulphuric acid and dosing tank (2) for dimethylaniline. First, 14,400 kg
of 96% sulphuric acid are poured into reactor (3) followed by 1000 kg of dimethyl-
aniline maintaining the temperature between 25 and 30°C. The dimethylaniline is
poured in over a period of about 3 hr, and the contents of the reactor are then main-
tained for 30 min at 40°Cand finally cooled to 20°C. The solution of dimethylaniline
sulphate so obtained is pumped over to container (4), whence it flows down to
nitrator (J).

Before nitration the solution of dimethylaniline sulphate is tested for the pre-
sence of free dimethylaniline by treating a test sample with a large amount of water.
If free dimethylaniline is present the solution is cloudy. This test is of great im-
portance (see above).

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Nitration (see Bain [59] and Rinkenbach [29]). A continuous method of nitra-
tion at 70°C is the safest and most advantageous. The lay-out shown diagrammati-
cally in Fig. 8 is typical.

The nitrating acid is metered in the dosing tank (7) and the dimethylaniline
solution in the dosing tank (2). Both liquids are introduced into the nitrator I, their

Fig. 8. Flow diagram of plant for continuous nitration of dimethylaniline.

rate of inflow being so regulated that 15.4 parts of the dimethylaniline solution in
sulphuric acid mix with 9.2 parts of nitrating mixture composed of:

hno 3 67%

H 2 S0 4
H 2

16%

17%

The contents of the nitrator are heated to a temperature of 68°C and heating
is then discontinued and the temperature in the nitrator maintained at 60 to 72°C.
The nitrator is cooled externally, if necessary. For safe and efficient nitration very
vigorous stirring is essential to ensure that the reacting liquids are mixed almost
instantaneously. Since the mixture is kept in nitrator / for a very short period, reaction
may be incomplete. Its contents are discharged via an overfall to a larger reactor //,
also provided with a stirrer. In reactor II the same temperature (70°C) is maintained
by heating. Here the reaction is completed and the liquid, together with the partly
crystallized product, is allowed to run into crystallizer (3) in which the whole is
cooled to 20°C and afterwards discharged to the vacuum filter {4). Tetryl is collected
on the filter and the spent acid is passed on to be .denitrated.

AROMATIC NITRAMINES

Here, as in other continuous processes involving rather risky exothermic reactions,
the following precautions for ensuring the safety of personnel are of utmost im-
portance:

(1) A stainless steel cooling coil should be fitted inside the nitrator to make
possible a rapid heat removal to enable the contents to be cooled very rapidly (the
coil is not shown in Fig. 8).

(2) Vigorous and reliable stirring is essential. The stirrer should be provided
with a spare d.c. motor driven by a battery, to replace the main motor in case of
damage. A compressed air pipe should also be fitted to supply air to the nitrator
as a last resort.

Fig. 9. Flow diagram of planJjNbr washing tetryl, according to Sokolov [58].

On nitration, carbon dioxide and nitrogen dioxide are evolved. These gases are
passed to water absorption towers where considerable amounts of nitric acid are
recovered.

Washing the tetryl. The filtered product is despatched in aluminium barrels
from the nitration department to a special room where it is poured into a wash-
ing tank (Fig. 9). The tank of 1350 1. capacity, fitted with a stirrer, may be
of wood lined with stainless steel. It is fed with water through pipe (/), and is heated
by direct steam injection through pipe (2).

The contents of the tank are stirred with water for 30 min and allowed to stand
for 30-40 min. The aqueous layer is decanted through the top valve (5) into a number
of settling tanks (6). The tetryl is washed first with cold water, and then by a suc-
cession of washings at increasing temperatures: 40, 60, 80, and 90 or 100°C.

After washing the product is discharged through a wide valve and channel (4)
to a vacuum filter (5). After filtration and the removal of water, wet tetryl contain-
ing 16-18% of water is passed on for crystallization. Tetryl is also collected from
the settling tanks every 10 days. This tetryl may contain sand and other mechanical
impurities, and must therefore be crystallized.

Washed tetryl should not contain more than 0.015% of acid (calculated as
H 2 S0 4 ). If the acidity is higher, the product is given an additional wash. Washed
tetryl crystallizes into yellow or reddish-yellow crystals.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Crystallization from benzene. For crystallization wet, freshly washed, deacidified
tetryl is used. It is introduced into container (7) (Fig. 10) equipped with a mixer,
a heating jacket and a reflux condenser (2). Container (7) is filled with benzene. The
whole is heated, while stirring, until the tetryl is dissolved. The solution is then
passed through the warmed filter (3). The products insoluble in benzene are retained

To the pump
Benzene to distillation-* — i
Fig. 10. Flow diagram of plant for the crystallization of tetryl.

on the filter. The clarified benzene solution, with an aqueous layer, is conveyed to
the warmed separator (4) in which the lower aqueous layer is separated. The aqueous
layer contains the residual acid washed out of the benzene solution of tetryl. The
benzene solution is transferred to the crystallizer (5) in which the crystallization of
tetryl takes place on cooling. The crystals of tetryl are then separated from the
mother liquor on the vacuum filter (6). Tetryl should not be separated in centrifuge,
as this is considered too dangerous.

The tetryl is then dried for 24 hr in a shelf drier at 55-60°C. This period, however,
may be reduced to 12 hr if a vacuum drier is employed.

All possible safety measures usually employed with sensitive and dangerous
explosives must be taken in the drier, e.g. the frequent dusting of heaters, sweeping
of floors etc.

The dried, crystalline tetryl is sifted on vibrating, well-earthed copper screens.

AROMATIC NITRAMINES 61

Crystallization from acetone. Tetryl may be crystallized from acetone by cool-
ing, but crystallization by precipitation with water is preferable, as this permits
removal of traces of acid and gives a very fine crystalline product which, when mixed
with the product prepared by crystallization from benzene, forms a free-flowing
mixture. The plant is, in principle, similar to that represented in Fig. 10, except that
separation is unnecessary and a much larger crystallization vessel is needed to hold
both the solution and the added water.

According to the Soviet standard specification WST 5, top-grade tetryl should
meet the following requirements:

Freezing point, min. 127.7°C

Moisture and volatile matter, max. 0.02%

Substances insoluble in acetone, max. 1 %

Acidity (calculated as H2SO4), max. 0.01 %

According to U.S. requirements [59], high grade tetryl should possess the fol-
lowing characteristics:

M.p. 128.5-129.1°C

Acidity not more than 0.08% (H 2 S0 4 )

120°C vacuum test not more than 4.0 cm 3 of gas evolved in 40 hr.

Nitration of dinitromethylaniline

According to the German (Griesheim) method [60] tetryl is manufactured in
two stages: first dinitromethylaniline is prepared and this is then nitrated.

Dinitrd|iethylaniline is produced by the reaction of chlorodinitrobenzene with
methylamine in the presence of sodium hydroxide. A solution is prepared consisting
of 300 1. of water, 1140 1. of 35% aqueous sodium hydroxide solution and 1225 kg
of 25% aqueous methylamine solution. This solution is added over a period of 12 hr
to a vigorously stirred suspension of 2000 kg of chlorodinitrobenzene in 1350 1. of
water heated to 95-100°C.

All is stirred for one hour longer and then the product is cooled (still stirring).
After being allowed to crystallize, it is centrifuged. The dinitromethylaniline so
obtained is nitrated in a similar manner to dimethylaniline. Nitration proceeds
without such an abundant evolution of gaseous products, it is less violent and may
be conducted at a lower temperature.

According to Desseigne [61] the reaction is carried out with a 20% excess of
HNO3 . The concentration of sulphuric acid employed for dissolving the dinitro-
methylaniline and in the nitrating mixture is so matched that the nitrating mixture
always contains 16% of water as the sulphuric solution of dinitromethylaniline is
poured into the nitrator during nitration. The composition of the nitrating mixture is:

HNO3 78%
H 2 S0 4 6%
H 2 16%

62 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

To dissolve dinitromethylaniline in sulphuric acid a mixture of 860 parts of
spent acid from a previous nitration is used, together with 168.5 parts of 95%
sulphuric acid and 117.5 parts of 20% oleum. The spent acid contains:

HN0 3 1.0%

N0 2 0.5%

H2SO4 82.5%

H 2 16.0%

The dinitromethylaniline solution in sulphuric acid is introduced slowly into
the nitrator containing the nitrating mixture. The addition takes about 1 hr and
the temperature of the mixture is maintained at 30°C.

The precipitation of tetryl begins about 10 min after the reaction has started.

After mixing the two solutions, the temperature is raised to 50-55°C and main-
tained thus for 40 min. The presence of crystals in the semi-liquid mixture results
in a violet colour which gradually changes to grey and then to yellow.
Nitration is judged to be complete when the colour turns pale-yellow. The
contents of the nitrator are then cooled to 20°C and the product is filtered off on
a vacuum filter. Spent acid (1290 parts) is drawn off, f of it returning to nitration
and ~ going to denitration.

The filtered tetryl is added to water (500 parts) so heated that on the addition
of tetryl its temperature is 50°C. The mixture is then stirred for 10-15 min, and
the tetryl filtered and washed with cold water until free of acid (helianthine test).
After washing, the tetryl is dried at 70°C.

From 100 parts of dinitromethylaniline 138.6 parts of tetryl can be obtained,
i.e. 95.3% of the theoretical yield. The m.p. of the tetryl so produced is 128.4°C.
If necessary, it can be re-crystallized.

The raw material consumption per 1000 kg of tetryl produced is:

720 kg of dinitromethylaniline

617 kg of 95% sulphuric acid

642 kg of 20% oleum

507 kg of nitrating mixture containing 87% of

HN0 3 and5%ofH 2
165 kg of 50% nitric acid.

HOMOLOGUES AND ANALOGUES OF TETRYL

2,4,6-Trinitro-3-methylphenylmethylnitramine (methyltetryl, tolyltetryl)

H 3 C V /N0 2
\ N /

2 N— f\~ NO2

-CH 3

no 2

m.p. 101-101. 5°C

AROMATIC NITRAMINES

Romburgh [62] prepared this product by the nitration of N-dimethyl-m-toluidine,
British authors [63] obtained it in the following manner during World War I using
the product of sulphitation of y-trinitroluene, the main isomer of a-trinitrotoluene
(Vol. I):

CH 3 CH 3

2 N

-SQ 3 Na

NH(CH 3 ) 2

2 N— j^\

1 V I-N(CH 3 ) 2

N0 2

.nh 2 CH 3

2 N— / N

CH 3

-NHCH3

N0 2

This course of reaction was confirmed by T. Urbanski and Schuck [40] who
also found that the final product is somewhat more stable to heat than tetryl. In-
stantaneous explosion, for example, occurred on contact with a metal surface heated
to 320°C (tetryl explodes at 310°C).

It is an explosive of approximately the same power as picric acid.

2,3,4,6-TetranitrophenyImethyInitramine and its derivatives

H 3 Q /N0 2
\ N /

1
>2N— f\- N0 2

Ul-N0 2

N0 2

m.p. 146-147°C

VII

According to Romburgh [64, 16] this substance is thought to be an impurity of
tetryl arising during the nitration of dimethylaniline and, with peculiar ease, when
methylaniline is nitrated. This is due to the fact that from the beginning of nitration
the nitro group may take the mete-position with respect to the methylamine group.

The formation of this substance by the nitration of m-nitromethylaniline was
examined by Blanksma [65] and the conditions of its conversion into a number
of other substances by the substitution of the nitro group at the mete-position with

64 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

the OH, OCH 3 , NH 2 , NHCH 3 groups were elucidated by van Duin and van
Lennep [66]:

<> r I II N /"~ 3

2 N-//)-NO, CH3
N0 2

These are the following compounds:

(VIII) 2,4,6-trinitro-3-methylnitraminophenol (m.p. 183°C)

(XII) 2,4,6-trinitro-3-methylnitraminoanisole (m.p. 96-97°C)

(XIII) 2,4,6-trinitro-3-methylnitraminophenetole (m.p. 98-99°C)

(XIV) 2 > 4,6-trinitro-3-methylnitraminoaniline (m.p. 188°C; under the prolonged influence of more
concentrated ammonia 2,4,6-trinitro-m-phenylenediamine is formed; Vol. I)

(XV) 2,4,6-trinitro-3-methymitramino-N-methylaniline (m.p. 190-192°C).

All these substances have been examined [66] with reference to their sensitive-
ness and stability (Table 13).

Owing to the presence of the phenolic group, the substance (VIII) forms explosive
salts. Its lead salt, however, is of no practical value owing to its hygroscopicity.

The substance (XV) is susceptible to further nitration resulting in the formation
of a dinitramine described below (so-called ditetryl).

AROMATIC NITRAMINES

Table 13

Sensitiveness to impact

Maximum height of drop (cm)

Ignition

Substance

not causing explosion

temperature
°C

Stability

2 kg | 10 kg

VIII

30-33

197

Stable at 95°C
for 3 days

XII

15-16

198

Decomposition
at 90°C after 2 hr

XIII

16-19

202

Decomposition
at 90°C after 2 hr

XIV

43^15

201

Stable at 95°C
for 30 days

Tetryl

50-51

196

Stable at 95°C

(standard) |

•

for 185 days

2,3,4,5,6-Pentanitrophenyl-N-methyInitramine(N-2,3,4,5,6-hexanitro-N-methylaniline)

2 N. y CH 3

\ N /

I
2 N— /\— N0 2
2 N— If— N0 2

NO z
m.p. 132°C (decomp.)

This substance was obtained by Blanksma [67] by the nitration of a mixture
of 3,5-dinitro-N-methylaniline and 3,5-dinitro-N,N-dimethylaniline.

It is unstable and decomposes readily above its m.p. or when boiled in water.

2,4,6-Trinitro-l ,3-di(methylnitramino)-benzene (ditetryl)

H3C. y N0 2

\ N /

2 N-/^-N0 2cH3

X N0 2
NO z
m.p. 206°C

Romburgh [68] was the first to prepare this explosive compound by the nitra-
tion of N,N'-dimethyl- or N.N'-tetramethyl-m-phenylenediamine. A more practical
method was developed by van Duin and van Lennep [66] who started from tetranitro-
phenyl-methylnitramine (VII) (p. 63):

66 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

H 3 C V /N0 2 H 3 Q /N0 2 H 3 C X /N0 2

\n/ \n/ x n/

1 I X

2 N— [TS— NO z NH * CH2 , 2 N-/^|— N0 2 HN °3 y 2 N-/yN0 2

-N0 2 I /J— NHCH 3 I J — N<

)-™> U-NHCH3 Y _N \n0 2

N0 2 N0 2 N0 2

Its chemical properties are very similar to those of tetryl.

Ditetryl is more sensitive to impact than tetryl (it is exploded by a 2 kg weight
falling 21-26 cm as compared with 49-51 cm for tetryl). Its ignition temperature
is 214°C (that of tetryl is 196°C). It is much less stable than tetryl: decomposition
occurred on heating at 95°C for 4 days (tetryl withstands 185 days heating [66]).

2,4,6-Trinitro-l ,3,5-tri(methylnitramino)-benzene (tritetryl)

HjG^ y N0 2

2 N°^ N

N-
/\

.:; 1

H 3 c/

N0 2

m.p. 280°C (decomp.)

This product was prepared by Blanksma [69] by the nitration of 2,4,6-trinitro-
1,3,5-trimethylaminobenzene. The latter was obtained from 3,5-dichloro-(or di-
bromo)-2,4,6,-trinitroanisole by the action of an alcoholic solution of methylamine.

T. Urbahski [70] reported a more convenient method of tritetryl preparation,
starting from symmetrical trichlorotrinitrobenzene :

NHCH 3

I
. NOz NHzCH,^ Q 2 N— j^\— N0 2

—CI H3CHN— L )— NHCH 3

N0 2 NO 2

m.p. 268-270°C

T. Urbafiski investigated its explosive power in a manometric bomb at a density
of 0.05 and found it to be more powerful than trinitrotoluene (by about 46%):

Trinitrotoluene — pressure 420 kg/cm 2
Tetryl — pressure 580 kg/cm 2

Tritetryl — pressure 613 kg/cm 2

"AROMATIC NITRAMINES 67

According to Medard [71] tritetryl gave a lead block expansion of 130 (taking
the value for picric acid as 100). Under a pressure of 1500 kg/cm 2 , a density of
1.43 was obtained.

2,4,6-Trinitrophenylethylnitramine (ethyltetryl)

H 5 Q

O z N-

_/\_ NO

]

NO z

i.p. 96°C

Romburgh [3] was the first to prepare this substance both by nitrating ethyl-
aniline and by nitrating diethylaniline. It is comparable to tetryl in its physical and
chemical properties. As an explosive it is weaker than tetryl. Its sensitiveness to
impact and its explosive power, measured in the lead block, are somewhat greater
than those of picric acid.

According to Medard [71] ethyltetryl gave a lead block expansion of 104 (taking
the value for picric acid as 100). Under a pressure of 1500 kg/cm 3 , a density of
1.63 was obtained.

2,4,6-Trinitrophenyl-n-butyInitramine (butyltetryl)

n-H 9 C 4v y N0 2

\ N /

2 N— f \— N0 2

N0 2
m.p. 97.5-98°C

This product has been prepared both by the nitration of N-n-butylaniline and
by the action of n-butylamine on chlorodinitrobenzene followed by nitration of
dinitro-N-n-butylaniline.

Tetryl and butyltetryl are alike in their physical and chemical properties. The latter
is notable for its low sensitiveness to impact, very similar to that of trinitrotoluene.
Since it is slightly more powerful than trinitrotoluene and at the same time highly
sensitive to detonation by mercury fulminate, it was suggested (Davis [72]) for use
in detonators, gains (boosters) and other initiating or priming charges.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

THE POLYCYCLIC ANALOGUES OF TETRYL
3,5,3',5'-Tetranitro-4,4'-di(methylnitramino)-benzophenone

N0 2 N0 2

H 3 C I / I CH 3

> N -< >- co -< >- N <
O2N/ ^j ^ ^ f X N0 2

N0 2 NO2

m.p. 200°C

Romburgh [73] prepared this substance by nitrating Michler's ketone. Gali-
nowski and T. Urbanski [74] obtained the same substance by nitrating auramine with
a mixture of nitric and sulphuric acids. On heating with 2% KOH the substance
was converted into the corresponding nitrophenol.

N0 2

H 3 <
H3O

N-

>-f

CH 3 H3C

" N Wo 2 N> N ^nr^ /y '
no 2 ° no 2

ch 3
no 2

no 2

ho_ x y~~ c_

No 2

no 2

>— OH
N0 2

3,5,3',5',3",5"-Hexanitro-4,4',4"-tri(methylnitramino)-triphenylcarbinol

H3C. /N0 2
\ N /

H 3 C.

2 N

N0 2 / N
2 N-

\/"

— N0 2

H3C/ \no 2

I, decomposition 228°C

Galinowski and T. Urbanski [74] prepared this substance by nitration of crystal
violet with a mixture of nitric and sulphuric acids. On heating with 2% KOH two
products were formed: a semi-quinone (II) and a phenol (III).

AROMATIC NITRAMINES

Compound (II), on treatment with concentrated nitric acid, adds a molecule of
nitric acid and reverts to compound (I):

N0 2

NO,

N0 2

NO 2
OH

III

H 3 a ,no 2

2 N— f\— N0 2

N0 2 ' "

H 3 C'

S N0 2

Hexanitrodiphenylethylenedinitramine (also called ditetryl*, bitetryl, or octyl)

2 N N0 2

I I
CH 2 — N— / S— N0 2

N0 2

I
CH 2 — N— / V-NO2

°& AoT

m.p. 218°C

* The same name is applied to the compound described on p. 65.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

It seems probable that this compound may find a practical application. It was
first prepared by Bennett [75] by the nitration of diphenylethylenediamine with a
mixture of concentrated sulphuric acid and anhydrous nitric acid at 30-35°C. Later
Cox [76] worked out a practical method for its use in detonators.

It is most conveniently prepared by the interaction of ethylenediamine and
chlorodinitrobenzene, followed by the nitration of the amine so obtained:

CH 2 NH 2 C

CH 2 NH 2 CI

— N0 2 CH 2 NH

N0 2
N0 2

-N0 2 CH 2 NH

N0 2

The substance is fairly stable. It resembles tetryl in its explosive power, but is
more like penthrite in its sensitiveness to impact.

According to Medard [71] compression under 1500 kg/cm 2 gives a density of
1.50. Its rate of detonation at a density of 1.60 is 7.350 m/sec and its lead block
expansion is 115 (taking the value for picric acid as 100).

NITRAMINO-ESTERS OF NITRIC ACID

Trinitrophenyl-y9-hydroxynitraminoethyl nitrate (pentryl, pentyl)

2 N V /CH 2 CH20N02

\n/

1
2 N— /\— N0 2

no 2

m.p. 126°C

This is a crystalline product readily soluble in most organic solvents, including
nitroglycerine. It was prepared and recommended as an explosive by Moran [77],
and later examined in detail by Clark [78] and Romburgh [79].

Pentryl has been prepared by two methods : (a) from aniline and ethylene oxide
(according to Herz [80]) and (b) from chlorodinitrobenzene and ethanolamine
(according to Moran) with the subsequent nitration of N-hydroxyethylaniline or
its dinitro derivative:

AROMATIC NITR AMINES 71

NH 2 NHCH 2 CH 2 OH

J H 2 C CH 2

(a) f \ \>"

N0 2

I
N— CH,CH 2 ON0 2

\
2 N— /\— N0 2

CI NHCH 2 CH 2 OH

i i

^ I )- N ° 2 NH 2 CH 2 CH 2 OH f V N °2 / N0 2

I I I

N0 2 N0 2 m.p. 222°C

When using method (b) a certain amount of the diphenylamine derivative (II)
is also formed as a by-product:

N0 2

y-No 2

N— CH 2 CH 2 OH

-N0 2

I
no 2

The specific gravity of pentryl is 1.82, and its apparent density 0.45. Under a
pressure of about 230 kg/cm 2 a density of 0.74 may be obtained.

In contact with a metal surface heated to 235°C, pentryl explodes in 3 sec. The
chemical stability of pentryl is considered satisfactory.

Pentryl is remarkable for its high explosive power which, according to various
authors, is equal to or slightly higher than that of tetryl. For instance it gives a
lead block expansion 20% larger than that of tetryl. At a density of 0.80 its rate
of detonation is 5400m/sec, that of trinitrotoluene being 4450 m/sec (for an
equal sized charge).

Pentryl is less sensitive to impact than tetryl (according to Clark, the maximum
height causing no explosion in the drop test is 50 cm for pentryl, whereas for tetryl
it is 27.5 cm and for picric acid 42.5 cm).

Medard [71] reports that pentryl, at a density of 1.56, detonates at the rate
of 7180 m/sec and its lead block expansion is 114 (taking the value for picric acid
as 100). Under a pressure of 1500 kg/cm 2 , a density of 1.68 is obtained.

The minimum initiating charges for pentryl as compared with other explosives
(according to Clark [78]) are listed in Table 14.

chemistry and technology of explosives
Table 14

Substance

Mercury fulminate
g

Lead azide
g

Pentryl
Tetryl
Picric acid
Trinitrotoluene

0.150
0.165
0.025
0.240

0.025
0.03
0.12
0.16

Hexaiu^rodiphenyl-/?-hydroxynitraminoethyl nitrate

O z N

2 N

— N0 2

2 2 N
CH 2 ON0 2
m.p. 184°C

According to Clark [81] this compound is obtained by the nitration of tetra-
nitrodiphenyl-/?-hydroxyaminoethane (II) which is formed as a by-product when
preparing dinitrotriphenyl-yS-hydroxyaminoethane from chlorodinitrobenzene and
ethanolamine (see above, p. 71).

The explosive properties of hexanitrodiphenyl-/?-hydroxynitraminoethyl nitrate
are similar to those of pentryl. It is slightly more stable on heating; its ignition tem-
perature lies between 390 and 400°C. It is somewhat less sensitive to impact than
pentryl and rather more powerful (by 3 %) in the lead block test. It requires a stronger
initiator than pentryl, tetryl or picric acid, but a weaker one than trinitrotoluene.

NITRAMINONITROPHENOLS

The only compound of this type described in the literature is 2,6-di-(methylnitr-

aminomethyl)-4-nitrophenol (I). The location of the C-nitro group has not been

determined so that a formula with N0 2 at the ortho position with respect to the

phenol group is also possible.

H 3 C J CH 3

>N— CH 2 — / \- CH 2 — N<

Q 2 W X N0 2

N0 2
I, m.p. 156°C

This compound has been described by Semenczuk [82]. It was prepared by the
nitration of 2,4,6-tri-(dimethylaminomethyl)-phenol (II) with nitric acid, s.g. 1.40,

AROMATIC NITRAMINES 73

at 40°C. The starting substance was obtained by the condensation of phenol with
formaldehyde and dimethylamine, according to Bruson and MacMullen [83].

(CH 3 )2NH 2 C-/\-CH 2 N(CH3)2

CH 2 N(CH 3 ) 2
II

Vigorous reaction with nitric acid, s.g. 1.40, e.g. at 80°C, for 2hr, causes the
substance (I) to undergo nitrolysis and nitration to form picric acid.

The substance (I) is a moderately powerful explosive: it gives an expansion of
250 cm 3 in the lead block. It is less sensitive to impact than trinitrotoluene. Its
ignition temperature (195-200°C) is about the same as that of tetryl. It forms in-
flammable salts. The lead, thallous and potassium salts burn violently with a sharp
report.

NITRAMINO-AZOXY COMPOUNDS

By the nitration of bis(4,4'-dimethylamino)-azoxybenzene T. Urbahski and
J. Urbanski [84] obtained a hexanitroazoxy derivative (azoxytetryl) of the structure
(I) i.e. 3,3',5,5'-tetranitro-bis(4,4'-nitromethylamino)-azoxybenzene

\— N=N— < V- N<

N0 2 NO z

i V^ --\ N02

° N0 2

m.p. 208-209°C (decomp.)

N0 2 ° N0 2

NITRO METHYLENE BLUE

A few nitro derivatives of methylene blue have been described. By introducing
one nitro group under mild conditions (nitric acid of ca. 20% and sodium nitrite)
methylene green was obtained [85].

According to Gnehm [86] methylene blue, when subjected to the action of nitric
acid (density 1.33, i.e. 52%) in presence of acetic acid diluted to 50%, yields
"dinitrodimethylthionine" nitrate of suggested structure (I).

!(N0 2 ) 2 I

NOf

74 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The position of two nitro groups was unknown.

According to the author mentioned, one of the two dimethylamino groups was
fully demethylated.

Experiments carried out by T. Urbafiski, Szyc-Lewahska and Kalinowski [87]
suggest the structure of 2,4-dinitro-3-(methyl)nitramino-7-(dimethyl)amino-5,5-
dioxyphenothiazine (II) for the Gnehm product. The phenol produced by alkaline
hydrolysis has the structure of 2,4-dinitro-3-hydroxy-7-(dimethyl)amino-5,5-dioxy-
phenothiazine (III)

no 2

(CHj)^ ^ \ S X Y X K " (CH 3 ) 2 /X/ ' X S / ^ X H

III

According to the same authors [87] energetic nitration of methylene blue may
yield 2,4,6,8-tetranitro-3,7-di(methylnitramino)-5,5-dioxyphenothiazine (IV). This
substance, when warmed with a 2% solution of NaOH, underwent the usual hydro-
lysis of the nitroamino group, placed in ortho position, to two nitro groups and
yielded phenol (V).

0;N N

,N X/V/ N0 2

02^

^ s /y\ N / H3
7 ^0 N ° 2

no 2

° 2N v\/ n \/y n ° 2

Ho/yx

/Y\)H

no 2

The product of Gnehm forms an intermediate stage between methylene blue and
product (IV).

The nitro compound (IV) possesses interesting properties: it burns readily
without melting.

LITERATURE

1. K. H. Mertens, Ber. 10, 995 (1877).

2. W. MlCHLER and F. Salathe, Ber. 22, 1790 (1889).

3. P. VAN Romburgh, Rec. tray. Mm. 2, 31, 103, 304 (1883); 3, 392 (1884); 8, 215 (1889).

4. C. E. Clarkson, I. G. Holden and T. Malkin, /. Chem. Soc. 1950, 1556.

5. H. H. Hodgson and J. Turner, /. Chem. Soc. 1942, 584.

6. K. H. Mertens, Ber. 19, 2123 (1886).

7. P. van Romburgh, Rec. trav. chim. 5, 240 (1886); 6, 251 (1887).

8. T. Urbanski and A. Semenczuk, Bull. Acad. Polon. Set., CI. Ill, 5, 649 (1957).

AROMATIC NITRAMINES 75

9. A. SemeNczuk and T. UrbaAski, Bull. Acad. Polon. Sci., CI. Ill, 6, 309 (1958).

10. P. P. Shorygin and A. V. Topchiyev, Zh. obshch. khim. 8, 981 (1938).

11. K. J. P. Orton, Ber. 40, 370 (1907).

12. F. M. LANG, Compt. rend. Ill, 1384 (1948).

13. J. ISSOIRE and G. Burlet, Mem. poudres 39, 65 (1957).

14. W. E. GARNER and C. L. Abernethy, Pwc. Roy. Soc. (London) 99, 213 (1925).

15. P. VAN Romburgh, Rec. trav. chim. 6, 251 (1887); 8, 273 (1889).

16. P. VAN ROMBURGH and Schepers, Verslag Gewone Vergader. Afdeel. Natuurk., Nederland.
Akad. Wetenschap. 22, 293 (1913).

17. L. DESVERGNES, Chimie et Industrie, 24, 785, 1304 (1930).

18. S. BOGDAL and D. SmoleNski, Bull. Inst. Appl. Chem. 10, 3 (1956); Vlth Jubilee Congress
of the Polish Chem. Soc, Warsaw, 1959, p. 180.

19. W. Michler and S. Pattison, Ber. 14, 2161 (1881).

20. P. van Romburgh, Rec. trav. chim. 41, 38 (1922).

21. T. L. Davis, The Chemistry of Powder and Explosives, J. Wiley, New York, 1943.

22. W. C. Crater, U.S. Pat. 1996146 (1936).

23. W. H. Rinkenbach and E. D. Regad, U.S. Pat. 1940811 (1934).

24. C. A. Taylor and W. H. Rinkenbach, /. Am. Chem. Soc. 45, 104 (1923).

25. A. F. Belayev and N. Matyushko, Dokl. Akad. Nauk. SSSR 30, 629 (1941).

26. W. E. Garner and C. L. Abernethy, Proc. Roy. Soc. (London) 99, 213 (1921).

27. H. Kast, Spreng- u. Ziindstoffe, Vieweg & Sohn, Braunschweig, 1921.

28. A. M. Prentiss, Army Ordnance 4, 242 (1923/24); Phys. Ber. 5, 1469 (1924).

29. W. H. RINKENBACH, Kirk & Othmer Encyclopedia of Chemical Technology, Interscience,
New York, 6, 51 (1951).

30. T. L. Davis and C. F. H. Allen, /. Am. Chem. Soc. 36, 1063 (1924).

31. A. P. N. Franchimont and H. J. Backer, Rec. trav. chim. 32, 327 (1913).

32. R. C. Farmer, /. Chem. Soc. 117, 1432, 1603 (1920).

33. L. Desvergnes, Mem. poudres 19, 217 (1922).

34. N. N. Yefremov and A. Tikhomirova, Ann. inst. anal. phys. chim. (U.S.S.R.) 3, 269 (1926);
4, 92 (1928); A. Bogoyavlenskii and N. N. Yefremov, ibid, 3, 299 (1926).

35. N. N. Yefremov and O. K. Khaibashev, Ann. inst. anal. phys. chim. (U.S.S.R.) 17, 130 (1949).

36. M. Giua, Gazz. chim. ital. 45, II, 32 (1915).

37. C. A. Taylor and W. H. Rinkenbach, Ind. Eng. Chem. 15, 73 (1923).

38. C. N. Hinshelwood, /. Chem. Soc. 120, 721 (1921).

39. C. F. van Duin, Thesis, Utrecht, 1918.

40. T. UrbaNski and M. Schuck, unpublished.

41. A. J. B. Robertson, Trans. Faraday Soc. 44, 667 (1948).

42. K. Szyc-Lewanska, unpublished.

43. S. Z. Roginskii and A. Ya. Lukin, Acta physicochim. U.S.S.R. 2, 385 (1935).

44. F. Lenze, Jahresber. Mil. Vers. Amt. 2, 6 (1895); 3, 54 (1895); 4, 10 (1897); F. Lenze and
H. Kast, ibid. 6, 21 (1899).

45. A. Haid and A. Schmidt, Z. ges. Schiess- u. Sprengstoffw. 26, 253, 293 (1931).

46. T. CARLTON Sutton, Trans. Faraday Soc. 34, 992 (1938).

47. J. Marshall, Ind. Eng. Chem. 12, 336 (1920).

48. M. A. Cook, The Science of High Explosives, Reinhold, New York, 1958.

49. H. Selle. Jahresber. Chem.-Techn. Reichsanstalt 8, 121 (1929).

50. R. Robertson, J. Chem. Soc. 119, 1 (1921).

51. A. Koehler, according to L. Desvergnes, Mem. poudres 19, 217 (1922).

52. E. Jones and D. Mitchell, Nature 161, 98 (1948).

53. K. K. Andreyev, Thermal Decomposition and Burning of Explosives (in Russian), Gosenergo-
izdat, Moskva-Leningrad, 1957.

76 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

54. F. Martin, Vber Azide u. Fulminate (Habilitationsschrift), Darmstadt, 1913.

55. W. Taylor and W. C. Cope, U.S. Bureau of Mines Techn. Paper 145 (1916); /. Soc. Chem.
Ind. (London) 35, 1181 (1916).

56. H. B. Troup, Brit. J. Ind. Med. 3, 20 (1946).

57. L. J. Witkowski, C. N. Fischer and H. D. Murdock,/. Amer. Med. Assoc. 119, 17 (1942).

58. A. M. SOKOLOV, Manual of Manufacture of Explosives, Ushakov and Lebedev (Ed.), (in Rus-
sian), Goskhimizdat, Moskva-Leningrad, 1934.

59. C. J. BAIN, Army Ordnance 6, 435 (1925/26).

60. I.G. Farbenindustrie Manufacture of Intermediates for Dyestuffs, BIOS Final Report No. 986,
Part II.

61. G. DESSEIGNE, Mem. poudres 28, 156 (1938).

62. P. van Romburgh, Rec. trav. chim. 3, 414 (1884).

63. Technical Records of Explosives Supply 1914-1918 No. 2; Manufacture of TNT, p. 25, HMSO,
London, 1920.

64. P. van Romburgh, Rec. trav. chim. 8, 274 (1889).

65. J. J. Blanksma, Rec. trav. chim. 21, 254 (1902).

66. C. F. van DuiN and B. C. R. van Lennep, Rec. trav. chim. 39, 145 (1920).

67. J. J. Blanksma, Rec. trav. chim. 21, 266 (1902).

68. P. van Romburgh, Rec. trav. chim. 7, 1 (1888).

69. J. J. BLANKSMA, Rec. trav. chim. 27, 23 (1908).

70. T. UrbaNski, Roczniki Chem. 17, 591 (1937). ^

71. L. Medard, Mem. poudres 33, 45 (1951).

72. T. L. DAVIS, U.S. Pat. 1607059 (1926).

73. P. van Romburgh, Rec. trav. chim. 6, 368 (1887).

74. S. Galinowski and T. Urbanski, /. Chem. Soc. 1948, 2169.

75. G. M. Bennett, /. Chem. Soc. 115, 576 (1919).

76. R. F. B. Cox, U.S. Pat. 2125221 (1938).

77. E. C. Moran, U.S. Pat. 1560427 (1925).

78. R. V. Le Clark, Ind. Eng. Chem. 25, 1386 (1933).

79. P. van Romburgh, Chem. Weekblad 31, 728 (1934).

80. E. Herz, Ger. Pat. 530704 (1930); Brit. Pat. 367713 (1930).

81. R. V. Le Clark, Ind. Eng. Chem. 26, 554 (1934).

82. A. SemeNczuk, Biul. WAT 5, XXII, 58 (1956).

83. H. A. Bruson and C. W. Mac Mullen, /. Am. Chem. Soc. 63, 270 (1941).

84. J. UrbaNski and T. UrbaNski, Bull. Acad. Polon. ScL, ser. chim. 6, 307 (1958); Roczniki
Chem. 33, 693 (1959).

85. Ulrich, according to G. Schultz, Farbstofftabellen, No. 1040, p. 451, Akademische Ver-
lagsges., Leipzig, 1931; Meister, Lucius & Bruning (Hochst), Ger. Pat. 38979 (1886).

86. R. Gnehm, J. prakt. Chem. 76, 407 (1907).

87. T. UrbaNski, K. Szyc-LewaNska and P. Kalinowski, Bull. Acad. Polon. Sci., ser. chim
7, 147 (1959).

CHAPTER IV

HETEROCYCLIC NITRAMINES

CYCLONITE

Cyclonite or cyclo-trimethylenetrinitramine (l,3,5-trinitrohexahydro-sym-triazine :
l,3 r 5-trinitro-l,3,5-triazacyclohexane) is a very

N0 2

/ N \
H 2 C CH 2

I I

2 N— N N— N0 2

x ch/

important explosive. It is also known as Hexogen, RDX and T4. It achieved great
importance during World War II as a constituent of many explosive mixtures from
which a high power was required. Cyclonite was used in detonators and primers,
and in detonating gains or boosters replacing tetryl. It was used extensively in mix-
tures with trinitrotoluene, as a so-called hexolite, semi-liquid, fusible explosive.
Mixtures of trinitrotoluene and cyclonite with aluminium, and plastic explosives
containing cyclonite were also used. Some of these contained ammonium nitrate.

Cyclonite was first prepared by Lenze [1]. The method of manufacture by the
nitration of hexamethylenetetramine nitrate with nitric acid is described in Henning's
patent [2] of 1899.

The author did not cite the product as an explosive — he recommended its use
in medicine — but in later patents [3] he proposed the use of cyclonite in the manufac-
ture of smokeless propellant. In 1921 Herz [4] modified Henning's method by ni-
trating hexamethylenetetramine itself, not its nitrate. Hale [5] described in detail the
preparation of cyclonite by nitrating hexamethylenetetramine, and reported its ex-
plosive properties.

Investigations carried out at that time revealed the outstanding value of cyclonite
as an explosive: its high chemical stability, which is not much lower than that of
aromatic nitro compounds, and its great explosive power, which considerably
surpasses that of aromatic nitro compounds, such as trinitrotoluene and picric acid.

[77]

78 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

PETN, which has about the same power, compares unfavourably with cyclonite,
the latter being less sensitive to mechanical stimulus and having a higher chemical
stability.

Thus during the inter-war period in countries with well developed chemical
industries, and in particular in Germany, Britain and U.S.S.R., new, more econo-
mical and safer methods of cyclonite preparation were worked out. During World
War II similar work was also carried on in the United States and Canada. German
methods became generally known after the war. Since it was found that the methods
elaborated in Britain, Canada and the United States were not essentially different
from those employed in Germany, the majority of them were also published. These
methods consist in principle of synthetizing cyclonite from various forms of formal-
dehyde (e.g. paraformaldehyde, hexamethylenetetramine or simple Schiff bases) and
from ammonium nitrate or hexamethylenetetramine nitrate as a source of nitramino
groups. They will be discussed in detail further on.

During World War II Germany produced 7000 tons of cyclonite a month; by
the end of the war the United States was manufacturing about 15,000 tons a
month.

PHYSICAL PROPERTIES

Cyclonite is a white substance, crystallizing in the orthorhombic system (Terpstra
[6]; Hultgren [7]). The refractive indexes n D for different axes are: a = 1.5775; /?
= 1.5966; y= 1.6015. The melting points reported by various authors are 202°C,
203. 5°C, 204. 1°C and 206-207°C.

The specific heat of cyclonite is 0.30 cal/g°C, its heat of combustion 2285 kcal/kg,
and its heat of formation — 96 kcal/kg, i.e. — AH ( =—2\3 kcal/mole. Cyclonite
is therefore an endothermic compound and this is one of the factors which
render it so highly explosive.

The vapour pressure of cyclonite at various temperatures (according to Edwards
[8]) is:

at 1 10.0°C 4.08 x 10"5 cm Hg
atl21.0°C 1.04x10-" cm Hg
at 131.4°C 2.57 x 10^4 cm Hg
at 138.5°C 4.00 x 10-4 cm Hg

These figures correspond with the empirical equation :

1
log p= 10.8? -5850-

where p denotes vapour pressure in cm Hg, and T is the absolute temperature.

Cyclonite crystals have a specific gravity of 1.820. When compressed the densities

given in Table 15 are obtained. A density of 1.73 may be obtained under a pressure

of 2000 kg/cm 2 provided that a desensitizer (e.g. wax) is added to reduce friction.

HETEROCYCLIC NITRAMINES

Cyclonite is practically insoluble in water. Majrich [9] reports its solubility as
0.01% at 0°C and 0.15% at 100°C. Cyclonite is soluble in concentrated nitric acid.
In sulphuric acid at concentrations above 70% solution is accompanied by decom-
position.

Table 15

Pressure kg/cm 2

Density

750

1.46

1000

1.50

1250

1.54

1500

1.56

1750

1.58

2000

1.60

In most organic liquids cyclonite dissolves with difficulty. The best solvent, from
which it can be crystallized, is acetone. Table 16 contains solubility data for cyclonite
(in g per 100 g of solution).

Table 16
Solubility of cyclonite in organic solvents (according to T. Urbanski and

KWIATKOWSKI [10]

Temper-
ature
°C

Methyl
alcohol

Ethyl
alcohol

Isoamyl
alcohol

Ethyl
ether

Acetone

Benzene

Toluene

Carbon
tetra-
chloride

0.140

0.040

0.020

4.18

0.016

0.180

0.070

0.023

0.050

5.38

0.020

0.018

—

0.235

0.105

0.026

0.055

6.81

0.045

0.020

0.325

0.155

0.040

0.075

8.38

0.055

0.025

—

0.090

—

0.480

0.235

0.060

10.34

0.085

0.050

—

0.735

0.370

0.110

12.80

0.115

0.085

0.005

15.27

—

1.060

0.575

0.210

—

0.195

0.125

0.007

64.5

1.250

—

0.880

0.320

—

0.300

0.210

0.015

78.1

1.180

—

79.5

0.400

—

0.500

—

0.295

—

0.850

—

0.465

—

100

1.325

—

0.640

—

110

1.900

—

0.980

—

120

—

2.990

—

131.6

3.870

The solubility of cyclonite (in g per 100 g of solution) in other solvents as report-
ed by various other authors is given in Table 17. It is insoluble in carbon disulphide,
soluble in hot aniline and phenol.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

T. Urbanski and Rabek-Gawroriska [11] found that cyclonite dissolves in molten,
highly-nitrated aromatic hydrocarbons, substituted urea derivatives, and camphor
to form eutectics of the composition given in Table 18. It is almost insoluble in
molten diphenylamine.

Table 17

Solubility of cyclonite in other organic solvents

Temper-
ature
°C

Chloro-
form

Methyl
acetate

Ethyl
acetate

Cyclo-
hexa-
none

Nitro-
benzene

Pyridine

Mesityl
oxide

20
25
50
97

0.008

2.95
6.0

0.517

12.7
27

1.5
12.4

1.60

3
12

Table 18
EuTEcnc mixtures with cyclonite

Content of cyclonite

Freezing point

Second component

in eutectic mixture

of eutectic

°/

°C

/7-Nitrotoluene

about 0.5

50.4

p-Nitroanisole

about 0.5

50.9

a-Nitronaphthalene

about 1 .5

55.4

m-Dinitrobenzene

85.5

a-Trinitrotoluene

2.5

78.6

1 ,3,5-Trinitrobenzene

about 3

113.8

Picric acid

112.9

Tetryl

118.1

sym-Dimethyldiphenylurea

112.4

.ry/M-Diethyldiphenylurea

70.4

Camphor

137.5

CHEMICAL PROPERTIES

Herz [4] first suggested a hypothetical structural formula for cyclonite. This
formula is now considered to be correct since it has been confirmed by a number
of methods of synthesis discussed later.

It has been established that cyclotrimethylenetrinitrosoamine (II), a product of
the nitrosation of hexamethylenetetramine, is oxidized with nitric acid (about 40%)
to cyclonite (I). This confirms the structure of cyclonite.

NO N0 2

i»

H 2 C

/ N \

ON— N

CH 2

H 2 C

/ N \

CH 2

N— NO

x ch/
II

2 N— N N— N0 2

x ch/

HETEROCYCLIC NITRAMINES 81

It also provides a method of obtaining a chemically pure preparation, free from
octogen. The reaction proceeds stepwise. It will be discussed more fully on p. 123
(Brockmann, Downing and Wright [12]).

However, attempts to prepare cyclonite from substances containing a sym-
triazine ring, e.g. from N,N',N"-trichloro-cyclotrimethylenetriamine (III) (obtained
by Delepine [13], by the action of hypochlorous acid on hexamethylenetetramine)
were a failure.

An attempt to prepare cyclonite by reacting the substance (III) with silver nitrite
also failed, as it resulted in total decomposition of the molecule.

H 2 C CH 2

CI— N N— CI

x ch/
III

According to Vernazza [14] cyclonite is decomposed at 25°C by concentrated
sulphuric acid, to form nitrogen and formaldehyde.
'The reaction is said to proceed as follows:

.OH
C 3 H 6 6 N 6 + m H 2 S0 4 -> 3 CH 2 + 2 S0 2 + 2 N 2 + (m-2) H 2 S0 4

\)NO

This reaction indicates why the results of analysis of cyclonite in the nitrometer
are too low.

SimeSek [15] doubted the correctness of the Vernazza equation and believed that
the decomposition of cyclonite by concentrated sulphuric acid at 2(M0°C may be
expressed by the following empirical equation :

C 3 H 3 6 N 6 + In H® -> 3 CH 2 + (3-«)N 2 + «NO® + nNH®

2 4

Simecek suggested the following mechanism for the decomposition reaction:

- CH 2 ffi -CH 2

I (b n h®(h 2 so 4 ) 1

— N— N0 2 < > — NH

q H®(H 2 so 4 ) ^ j ^ HSO® + NO© (26)

II a

CH 2

h«0 4

N 2 + CH 2 (28) J© HSO©

MH 2

\H,0

lib CH 2 + NHfHSO© (27)

According to reaction (26) concentrated sulphuric acid liberates the nitronium
ion from cyclonite (as from nitroguanidine, as reported on p. 26). Cyclotrimethylene-
triamine sulphate (II) is formed which is then converted to a SchifTs base sulphate.

82 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The latter is hydrolysed to formaldehyde and ammonium sulphate according to
reaction (27). Simultaneously reaction (28) takes place, involving those molecules of
cyclonite unaffected by sulphuric acid. Reaction (28) results in the formation of
nitrous oxide and formaldehyde.

Sime6ek also came to the conclusion that N 2 is not liberated in the nitrometer
during the decomposition of cyclonite since formaldoxime CH 2 =NOH is formed
in the presence of formaldehyde hence the analytical results are low.

It should be noted that the presence of water in the sulphuric acid promotes
the decomposition of cyclonite. This is particularly easy when the acid contains
1-15% of water. In a strictly anhydrous medium the course of the reaction is much
milder. Nitric acid which contains S0 3 does not cause the decomposition of cyclo-
nite.

According to Somlo [16] the action of a 4% solution of NaOH at 60°C produces
the total decomposition of cyclonite after 5 hr. Somlo also studied the decomposi-
tion of cyclonite by concentrated sodium hydroxide and found it to be complete
within 2-4 hr at 60°C. Among the decomposition products he detected nitrates,
nitrites, organic acids, ammonia, nitrogen, formaldehyde and hexamethylenetetra-
mine.

Epstein and Winkler [17] reported a preliminary examination of the kinetics* of
cyclonite decomposition proceeding by a homogeneous phase in acetone solution.

The kinetics of decomposition of cyclonite in an alkaline medium of sodium
methoxylate or lithium in a solution of methyl alcohol within the temperature range
19.0-44.93°C was investigated in detail by W. H. Jones [18]. He studied the course
of the reaction by determining the concentration of NO 9 ion with chloramine T,
by titrating with acid in the presence of various indicators and by determining the
unaltered cyclonite polarographically. On the basis of these analyses Jones drew
up four equations, according to which the decomposition reaction proceeds under
the influence of a methoxyl anion (OCH 3 ) e (A, B and C denote hypothetical inter-
mediate products)

Cyclonite + (OCH 3 ) e — > A + CH 3 OH + NO© (29)

A + (OCH 3 ) e > B e + CH3OH (30)

B 9 + (OCH 3 ) e > C 29 + CH3OH (31)

C 29 :5— > D 9 + NO e (32)

Reactions (30) and (31) proceed very rapidly; K t and K 2 are the constants of
reactions (29) and (32); C 29 is a strong base; B 9 and D e are weaker bases.
The equation for the rate of reaction as deduced by Jones is :

== = Ki{a— 3xi) (b—xi) + K 2 (xi—x 2 ) (33)

where a denotes the initial concentration of (OCH 3 ) e ions, b the initial concentra-
tion of cyclonite.

HETEROCYCLIC NITRAMINES 83

After time t:

x is the total molar concentration of the NOf ion; x x the molar concentration
of NOf from reaction (29); x 2 the molar concentration of NOf from reaction (32).
Jones also deduces the following equations :

- < ^=K l R(a-3b + 3R) (34)

dZ
-^=K,Z(ib-a + Z) (35)

-~=K^0Y-a){Y+2b- a y (36)

where after time t:

R is the molar concentration of cyclonite; Z the molar concentration of (OCH 3 ) e
(determined by nitration in the presence of phenolphthalein); Y=Z+C, i.e. total
basicity (determined by titration in the presence of bromothymol blue); C the molar
concentration of the C 2e ion.

The nature of the intermediate products of reaction was not determined by Jones.
Instead, he suggested that the intermediate product A is formed from cyclonite by
the elimination of HN0 2 and has the structure given below. He derived the hypo-
thetical structure of the ions B e . and C e in a similar manner.

Cyclonite is more stable to heat than penthrite and at a temperature above the
melting point of tetryl, e.g. at 140°C, also shows greater stability than the latter.

According to Avogadro [19] the ignition temperature of cyclonite is 215°C (this
author reports 185°C for penthrite). T. Urbanski and Pillich [20] found the following
values for ignition temperatures :

tetryl 203°C

penthrite 209°C
cyclonite 229°C

Heating began at 150°C, the rate of temperature rise was 107min. Ignition after
a 5 sec delay occurred at 260°C.

Many authors, viz. : Majrich [9], Tonegutti [21] and Haid, Becker and Dittmar
[22] point out that cyclonite is more stable than penthrite and tetryl.

A. J. B. Robertson [23] found that the decomposition of cyclonite at a temper-
ature above its melting point (between 213 and 299°C) proceeds as a first order reac-
tion. At 213°C, half the substance decomposes in 410 sec, and at 299°C in 0.25 sec.
The gaseous decomposition products contain chiefly N 2 , N 2 0, NO, CO and C0 2 .
According to A. J. B. Robertson the activation energy E of the thermal decomposi-
tion of cyclonite at these temperatures is 47,500 cal, log B= 18.5. The reaction rate
in the liquid phase is approximately ten times greater than in the solid phase (below
the melting point.)

T. Urbanski and Krawczyk [24] studied the stability of cyclonite alone and in
mixture with trinitrotoluene and found that:

84 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(1) Samples of cyclonite when heated for 62 hr at 120°C, or for 36.5 hr at 110°C
and subsequently for 60 hr at 120°C, contained no acidic decomposition products.

(2) In the Taliani test at 134.5°C cyclonite behaves like tetryl, i.e. it gives a
pressure of 11.5 mm Hg after 24 hr heating, as compared with trinitrotoluene which
gives 16 mm Hg.

(3) A mixture of 80% cyclonite with 20% trinitrotoluene decomposes somewhat
faster, after 24 hr producing a pressure of 39 mm Hg, possibly due to the fact that
the cyclonite is partly molten. In spite of this result the mixture should be consider-
ed more stable than tetryl.

According to Tabouis, Ortigues and Aubertein [25] cyclonite which has already
been subjected to the Abel heat test at 80°C, fails to withstand a repeated test, rapidly
darkening a starch-iodide paper. In the authors' opinion this is caused by the pres-
ence of traces of nitric acid in the crystals. On heating, this acid is liberated from the
crystals, causing the sample to fail when the test is repeated. Cyclonite which con-
tains more than 0.035% of HN0 3 (which is the maximum content permitted by
French specifications) may pass in the Abel test the first time. The authors suggest
that the traces of nitric acid may be removed by boiling cyclonite in an autoclave
at 140°C.

Majrich found that light has a negligible effect on cyclonite. T. Urbahski and
Malendowicz [26] reported that under the influence of ultra-violet irradiation cyclo-
nite changes colour from white to pale yellow, but undergoes no other alteration.
In particular, unlike nitric esters, neither NO, nor N0 2 is evolved from cyclonite.

EXPLOSIVE PROPERTIES

According to Avogadro [19] when exploded cyclonite decomposes with the evolu-
tion of CO, C0 2 , H 2 0, N 2 and an insignificant amount of H 2 . This author found
the composition of the explosion products (after cooling) to be as follows :

CO 25.22%

CO z 19.82%

H 2 16.32%

H 2 0.90%

N 2 37.83%

Avogadro calculated theoretically and found experimentally the following con-
stants:

heat of detonation 1 390 kcal/kg

gas volume 910 l./kg

"temperature of explosion 3380°C
specific pressure (J) 12,600 m

Other authors gave somewhat lower figures :

1370 kcal/kg (Tonegutti [27])
1359.5 kcal/kg (Medard [28])

HETEROCYCLIC NITRAMINES

The most extensive investigations were those of Apin and Lebedev [29]. They
examined the heat of detonation and the gas volume at different densities. The heat
of detonation increases from 1290 to 1510kcal/kg and the gas volume falls from
730 to 630 l./kg when the density is increased from 0.50 to 1.78 g/cm 3 . The variation

J 0.6

1.8

1.0 1.4

d(g/cm*)

Fig. 11. Heat of detonation of cyclonite against the density of the charge, according
to Apin and Lebedev [29]. Curve / — liquid water, curve 2— gaseous water.

of the heat of detonation with density is linear, as shown in Fig. 11. The composi-
tion of the gaseous products also changed with density: increase in density was
accompanied by an increase in carbon dioxide content.

Apin and Lebedev gave the following decomposition equation for cyclonite when
the density was 1.10:

C 3 H 6 6 N 6 -> 0.93CO 2 + 2.01CO + 2.13H 2 + O.O6CH4 + 3N 2 + 0.75H 2

Data concerning the rate of detonation of cyclonite reported by various authors
are tabulated below (Table 19).

The values are somewhat higher than those of penthrite.

Table 19
Rate of detonation of cyclonite Cm/sec)

Density of

T. Urbanski

Laffitte and

Kast [31]

and Galas [32]

Parisot [33]

Cook [34]

0.80

5000

0.92

—

5500

1.0

— .

6080

1.05

—

6000

1.35

7400

1.40

7550

1.45

7705

1.70

8380

The lead block expansion ranges from 450 to 520 cm 3 according to different
authors. The relative value taking picric acid as 100, is estimated at about 170.

Cyclonite is less sensitivite to impact than penthrite.

The following figures, based on the investigations of T. Urbanski [30], are charac-
teristic of the sensitiveness of cyclonite and tetryl to impact.

86 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The work required to cause :

10% of explosions 50% of explosions

is:

for penthrite
for cyclonite
for tetryl

0.11 kg/cm 2
0.14 kg/cm 2
0.56 kg/cm 2

0.20 kg/cm 2
0.22 kg/cm 2
0.92 kg/cm 2

According to other authors (e.g. Izzo [35]) the difference in sensitiveness between
cyclonite and tetryl is insignificant : as the following drop test figures, using a 2 kg
weight, indicate : 30-32 cm for cyclonite, and 40 cm for tetryl.

u
5.0

4.0

""3.0

2.0

***

< \

U -

*^"\

80 120 t°C

Fig. 12. Rate of burning of cyclonite against the initial temperature, according to

Andreyev [36].

Andreyev [36] reports that powdered cyclonite of density about 0.9 burns in
tubes of 5-6 mm dia., but when ignited at 20 C C cyclonite usually goes out.

Under a pressure of 800 mm Hg the relationship between the mass rate of the
burning of cyclonite and the temperature may be expressed by the equation:

u m = -0.0107 + 0.000182 T
1

= 43.92-0.0709 To

This relationship is illustrated in the diagram (Fig. 12).

The ease with which cyclonite can be detonated is markedly less than that of
penthrite, but it is more easily detonated than tetryl.

TOXICITY

No explicitly harmful effect has been noticed among workers employed in the
production or handling of cyclonite. Its toxicity appears to be considerably limited
by its poor solubility which prevents it entering the blood stream.

Nevertheless the danger of poisoning is always present wherever workers have
to deal with cyclonite dust, e.g. in drying and sifting operations and in measuring

HETEROCYCLIC NITRAMINES 87

the dry substance etc. It has been shown that breathing cyclonite dust gives rise to
tonic-clonic spasms. These symptoms occur after a few days of breathing the dust.
They last for 5-10 min, occurring intermittently and do not pass off at once when
the patient is removed from the atmosphere containing cyclonite dust.

The toxicity of cyclonite has led to suggestions for its use as a "selective poison".
For instance, it has been patented [37] as a rodenticide. According to the patent
specification the lethal dose for rats is 20 mg. It is much less toxic for domestic
animals and human being.

CYCLONITE MANUFACTURE

1. THE ACTION OF NITRIC ACID ON HEXAMINE
General information

The oldest and simplest method of preparing cyclonite is based on the introduc-
tion of hexamethylenetetramine (hexamine) into an excess of concentrated nitric
acid, s.g. 1.50-1.52, free of nitric oxides, at 25-30°C, thereafter pouring the whole
into cold water.

According to Hale [5] the reaction may be represented by the following equa-
tion: v

(CH 2 ) 6 N 4 + 4HN0 3 -+ (CH 2 • N • N0 2 ) 3 + 3CH 2 + NH4NO3 (37)

whereas Schnurr [38] formulates the reaction as follows:

(CH 2 ) 6 N 4 + 6HNO3 -> (CH 2 ■ N • N0 2 ) 3 + 6H 2 + 3C0 2 + N 2 (38)

It appears however that the reaction proceeds according to both equations
simultaneously since ammonium nitrate and formaldehyde are formed according
to equation (37) and C0 2 , N 2 and water according to equation (38). Some of the
methylene groups and nitrogen atoms of hexamine are therefore not utilized for
the production of cyclonite. The nitration of hexamine with nitric acid requires
from four to eight times the theoretical amount of nitric acid.

Apart from the main reactions (37) and (38), side reactions (39) and (40) also
take place.

Reaction (39) is a hydrolysis of hexamine resulting in the formation of formal-
dehyde and ammonia, and reaction (40) consists of the oxidation of formaldehyde
with nitric acid.

(CH 2 ) 6 N 4 + 6H 2 -> 4NH 3 + 6CH 2 (39)

2CH 2 + 2HN0 3 -> 2HCOOH + H 2 + NO + N0 2 (40)

Apart from the side reactions (39) and (40), others may occur which also result
in the formation of explosive substances other than cyclonite.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

This may be explained by considering the nitration of hexamine with nitric acid
as a stepwise degradation by nitrolysis, i.e. the nitration of amine involving the
stepwise cleavage of the bond between the nitrogen and carbon atoms. According
to Lamberton [39] this idea was first advanced by Linstead. The main work on this
subject has been done by British [39] and Canadian [40] authors.

According to the investigations of Hirst, Carruthers et al. [41] and Vroom and
Winkler [42], the action of nitric acid on hexamine dinitrate (IV) results in the for-
mation of the substance (V) by nitrolysis.

/ CH2 \

/ \e e

N NHNO3

\ /

CH 2 CH 2

N0 2

I
N

CH2ONO2

H 2 C

CH 2 CH 2

H 2 C

CH 2 CH 2

I
CH 2 CH 2

_ e
NHNO3

e
NHNO3

Alcohol groups are esterified with nitric acid. Further nitrolysis may cause the
cleavage of N— C bonds. The experimental data (Wright and Myers [43]) indicate
that the cleavage of bond A most probably occurs with the formation of the hypo-
thetical compound (VI). In turn the latter may undergo nitrolysis at position B to
form another hypothetical product (VII) [39, 40, 43, 44] and a known compound
(VIII) [39, 40, 44], i.e.

NOp

3 A

e \

CH 2 — NH CH 2

I I I

NN0 2 CH 2 NCH 2 ON0 2

I I I

CH 2 N CH 2

CH 2

N0 2

I
-N

N0 2

I
CH 2 — N K

I I * /CHzONOz

NNO,CH 2 N<

I £->|fl-H N CH 2 ON0 2
CH 2 N CH 2

t
C

NN0 2 CH 2

I \<-G
CH 2 N CH 2 ON0 2

t t
D C

VII

CH 2 ON0 2

I
NN0 2

I
CH 2 ON0 2

VIII

On nitrolysis of the bond C the compound (VII) finally gives cyclonite. If, how-
ever, the bond D is nitrolysed, a chain compound may be formed.

HETEROCYCLIC NITRAMINES 89

Another open-chain methylenenitramine may arise from the compound (VI),
if the bounds E and F are nitrolysed. The compound isolated and identified as
l,9-dinitroxy-2,4,6,8-tetranitro-2,4,6,8-tetrazanonane (IX) is then formed:

NO z N0 2 N0 2 N0 2

CH 2 — N— CH 2 — N— CH 2 — N— CH 2 — N— CH 2

ON0 2 t> N 2

The formation of this substance is favoured by a low nitration temperature.
It is unstable and highly sensitive to impact hence its presence in cyclonite is very
undesirable.

According to Singh [45] under the influence of the nitracidic ion (H 2 NO©),
hexamethylenetetramine first undergoes hydrolysis identical to that leading to the
formation of (VI). On hydrolysis the ion (Via) would be formed or on further hetero-
lysis the ion (VIb), then (Vic). The latter would undergo nitrolysis to form cyclo-
nite, methylnitramine (Villa) and formaldehyde:

N0 2 N 2

CH 2 -N CH 2 -N

I | /CH 2 OH© [ | ,CH 2 OH©

2 N-N CH 2 N< '■— ~+ 2 N-N CH 2 N /

I X CH 2 OH© - H2 ° || |\ CH ©

CH 2 -N— CH 2 2 tu 2 -^ in, 2

Via VIb

N0 2

I
CH 2 — N

I | y CH 2 OH©

2 N— N CH 2 N< 2

I I I X N0 2

CH 2 — N CH 2

VIc \ /CH 2 OH

* cyclonite + 2 N— N<( + CH 2

X CH 2 OH
Villa

Among the reaction products a cyclic ether is also present to which the structure
(X) is ascribed. It is possible that it arises either from the compound (VII) on nitro-
lysis of the two bonds D and G and the dehydration of the two alcohol groups so
produced, or directly from the compound (VII) on a sui generis nitrolysis

/\ H2 / CHz

2 N-N N-N0 2 + h 2 o 2 N-N \— NO a

H^ L >NH 2 CH 2 ON0 2 + | |

h ^c ch 2 h 2 c ch 2

\n/ \o/

CH 2 ONO z

VII v-

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The compound (X) (3,5-dinitro-l-oxa-3,5-diazacyclohexane) dissolves in hot
water and crystallizes from solution on cooling. Its melting point is 97°C.

Nitrolysis of the compound (V) may also lead to the formation of the substance
(XI) containing an eight-membered ring, called octcgen (HMX) (see p. 1 17) which
always accompanies cyclonite, but slightly reduces its power.

N<5 2

I
CH 2 — N— CH 2

I I

2 N— N N— N0 2

I I

CH 2 — N— CH 2

I
N0 2

Evidence of the presence of the compound (V) in the intermediate products
of nitration has been provided by Wright et al. [44] who found that the substance (XII)
(l,5-endomethylene-3,7-dinitro-l,3,5,7-tetrazacyclo-octane,or 3,5-dinitro-l,3,5,7-tetr-

az'abicyclo [3,3,1] nonane (DPT)) may be isolated in amount corresponding to
5-12% of the cyclonite by neutralization of the ammonia with spent acid.

CH 2

NN0 2

CH 2

DPT XII

m.p. 213°C

In turn this substance may undergo further nitrolysis which may lead, for ex-
ample, to the compound (XI) described above while nitrolysis of the bonds G and
H may yield the open-chain compound (IX), which is highly sensitive to impact,
and therefore very undesirable.

According to Wright et al. [44] DPT can also yield 3,7-dinitro-3,7-diaza-l,5-
dioxacyclo-octane (m.p. 263-264°C) under action of nitric acid (99%) and am-
monium nitrate:

CH 2 — O— CH 2

I I

2 N— N N— N0 2

I I

CH 2 — O— CH 2

Wright et al. [46] pointed out that apart from the compound (VII) nitrolysis
of hexamine may also lead to the transient formation of l-di(hydroxymethyl)-amino-
methyl-3,5-dinitro-l,3,5-triazacyclohexane (XIII) [(VI) is the dinitrate of (XIII)]
through the nitrolysis of the bonds K and M:

HETEROCYCLIC NITRAMINES

/CH 2
/ \

/
CH 2 CH 2

I
H 2 C CH 2 CH 2

sjsp

2 N

/N0 2

N— CH 2 — N
R

CH 2 — N— CH 2

\<- p
CH 2

2 N.

/N0 2

•N— CH 2 -

-N'

CH 2 — N

_n/

N0 2 |

CH 2 ON0 2

HOH 2 C CH 2 OH
XIII

CH 2

N— N0 2

I
CH 2 ON0 2

In Wright's opinion the existence of the transient compound .(XIII) is proved
by the formation of substance (IX) which can be isolated from the products when
cyclonite undergoes nitrolysis of the P and R bonds.

To protect the hydroxyl groups of the compound (VII) or (XIII) by acetylation
Wright cooled a solution after nitrolysis to - 55°C and then added acetic anhydride
at a temperature below -25°C. Apart from cyclonite he isolated l-acetoxy-7-
nitroxy-2,4,6-trinitro-2,4,6-triazaheptane (XIV). The addition of the solution from
nitrolysis to acetic anhydride gave a diacetyl derivative: l,7-diacetoxy-2,4,6-trinitro-
2,4,6-triazaheptane (XV) :

2 N.

N — CH 2 -

CH 2 — N'

/N0 2

-N

N0 2

2 N

\n-

-CH 2

CH 2 — OCOCH3

CH 2 — N

■/■

N0 2

-W

,N0 2

CH 2 — OCOCH3

CH 2 0NO 2
XIV

CH 2 OCOCH 3
XV

The nitration of cyclonite at very low temperatures also leads to the formation
of a series of other compounds. Some of them are converted into cyclonite by the
action of nitric acid at room temperature.

If the nitrolysis of hexamine dinitrate (IV) is conducted at a very low temper-
ature, e.g. -40°C, then, as reported by Hirst et al. [41] it is not cyclonite which is
formed, but a dinitrate of a dinitro derivative (la) (3,5-dinitro-3,5-diazapiperidinium
nitrate): N02

H 2 C

H 2

2 N— N NH 2 NQ 3
\ /
X CH 2

This was confirmed by Vroom and Winkler [42], who purified the substance (la)
by dissolving it in anhydrous 75% nitric acid at -20°C, and then adding ice water
at temperatures from -20 to -15°C. The purified substance has a melting point
of 98-99°C.

92 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The formation of this compound may be explained by the fact that at a low
temperature nitrolysis at point C (p. 88) is completed to produce amine nitrate and
alcohol (or its ester), hence the formation of compound (la).

The latter is unstable and undergoes decomposition in hot water with the evolu-
tion of formaldehyde, nitrous oxide and ammonium nitrate

la + H 2 -> 3CH 2 + 2N 2 + NH4NO3

Cyclonite (I) can be prepared by the action of concentrated nitric acid or acetic
anhydride on the compound (la) at room temperature. A N-acetyl derivative of
dinitro compound (XXIX) (p. 116) is obtained by the action of acetic anhydride
on the compound (la) in the presence of sodium acetate.

Vroom and Winkler found that under the influence of dilute alkali in an acetone
or ethanol medium substance (la) gives a compound which, according to the investiga-
tions of Chute, McKay, Meen, Myers and Wright [47] is bicyclic (Ic).

Vroom and Winkler believed substance (la) to be an intermediate compound
in the preparation of cyclonite by the nitrolysis of hexamine. This view was shown
to be incorrect by Wright, Berman and Meen [46] who proved that the substance
(la) cannot exist in the circumstances under which nitrolysis is usually carried out.
Therefore Wright et al. suggested that the substance (la) arises when the reaction
solution is diluted, viz. by the nitrolysis of the bonds C and P in the substances (VII)
and (XIII) respectively.

According to Dunning and Dunning [48] the nitration of hexamine dinitrate at
a temperatureof about - 30°C, followed by treatment of the product with ethyl alcohol
permits the isolation of the product of nitrolysis of (VI) at point B in the form of an ether
(lb), m.p. 115°C and the compound (IXa) with a chain structure

N0 2

/ N \
H 2 C CH 2

I I

2 N— N N— CH 2 OC 2 H 5
\ /
CH 2

It) IXa

In addition, a bicyclic compound (Ic), m.p. 136°C, is present in the products

of nitration at a low temperature, whereas nitration of the compound (lb) with

anhydrous nitric acid at a temperature of -30°C leads to the formation of an

ether (Id)

no 2 no 2

I I

/ N \ / N \

H 2 C ch 2 h 2 c ch 2

II II

2 n— n n— ch 2 — n n— n0 2

ch 2 , \;h 2

N0 2

1
NH-

-CH 2 -

1
-N— CH 2 -

HETEROCYCLIC NITRAMINES 93

N0 2 N0 2

! !

/ N \ / N \

H 2 C CH 2 H 2 C CH 2

II II

2 N— N N— CH 2 — O— CH 2 — N N— N0 2

\ / \ /

X CH 2 N CH 2

Cyclonite (I) arises from the nitration of compounds (lb), (Ic) and (Id) at room
temperature. The compound (IXa) also yields cyclonite on reaction with acetic
anhydride and paraformaldehyde.

According to Karpukhin and Chetyrkin [49] the nitrolysis of hexamine may
proceed with the formation of trihydroxymethylamine nitric ester (XVI):

(CH 2 ) 6 N 4 + 6HNO3 -> (CH 2 -N-N0 2 ) 3 + N(CH 2 0-N0 2 ) 3 + 3H 2 (41)

I XVI

This may result from the nitrolysis of the compound (VI) at the bond C. This
ester, like esters (VIII) and (IX), is unstable and readily decomposes. Finally, formal-
dehyde, split off from hexamine, may yield unstable methylene glycol nitrate (XVII)
in the presence of anhydrous nitric acid.

/ ON0 2
CH 2
\0N0 2

XVII

Owing to the presence of the unstable compounds (VIII), (IX), (XII) and (XIV),
various reactions, mainly exothermic, occur in spent acid after the nitration of
hexamine and the precipitation of cyclonite with water. Such reactions may lead to
explosion hence it is not surprising that the first attempts to manufacture cyclonite,
undertaken shortly after World War I, showed that the greatest difficulty in pro-
ducing cyclonite lies in the danger created by the spent acid. The presence of all
these products in spent acid makes its storage extremely dangerous. Some of these
by-products may also contaminate cyclonite, lowering its stability.

Searching for ways of getting rid of these products, a method was worked out by
which their decomposition was induced under strictly controlled conditions. De-
composition is caused, for instance, by pouring the mixture into hot water after
nitration. The amount of water and the temperature are coordinated so that a con-
centration of 50-55% HNO3 and a temperature of 70-90°C are maintained. Highly
pure cyclonite is precipitated and N0 2 evolved from the decomposition of all un-
stable products. This is the so-called "degassing process".

Owing to .the side-reactions, the yield of nitration of hexamine does not exceed
75-80% (when calculated according to equation (37) or (38)); 110-119 kg of cyclonite
can be obtained from 100 kg of hexamine.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Dunning, Millard and Nutt [50] studied the rate of nitration of hexamine
with various concentrations of nitrid acid, at 0°C and obtained the results, some
of which are given in Table 20 and the graph in Fig. 13.

96%H\Oj

°93%HN0 3

oS0%HNO3

85%HN0 3

30 40 50
Time ( mi n)

Fig. 13. Rate of nitration of hexamine at 0°C with various concentrations of nitric
acid, according to Dunning, Millard and Nutt [50].

Vroom and Winkler [42] also examined the kinetics of the nitration of hexamine to
cyclonite with various concentrations of nitric acid at 0°C and drew the following
general conclusions:

(1) Maximum yield (about 40% of theoretical calculated on the formaldehyde
used i.e. 80% of theoretical calculated on hexamine) can be obtained with all the
concentrations of nitric acid used: 88-97% (Fig. 14). The minimum molar ratio of
nitric acid to hexamine for maximum yield was found to increase from 26 : 1 with
97% acid to 110 : 1 with 88% acid.

10 20 30 40 50 60 70 80 90
Molar ratio of nitric acid to hexamine

Fig. 14. Effect of nitric acid-hexamine ratio on final yield of cyclonite, according to
Vroom and Winkler [42].

(2) The rate of nitrolysis increased rapidly as the molar ratio of nitric acid to
hexamine was increased and continued to do so after the molar ratio was raised
above that required for maximum yields (Fig. 15).

HETEROCYCLIC NITRAMINES

Table 20

Concentration of HNO3

Time,

99%

96% | 93% |

90%

85%

Yield of cyclonite [% of theoretical according to equation

(37) or (38)]

1.5

55.7

2.5

66.9

68.7

40.9

13.3

—

6.5

70.2

75.3

65.8

30.5

—

74.6

74.7

73.4

45.4

—

80.9

79.2

77.2

59.9

14.9

80.5

63.9

26.2

100

33.0

120

~ ; ~

32.6

For 85-96% HN0 3 at 0°C the authors deduce the following empirical equation:

According to the British data the heat of nitration of 1 kg of hexamine to cyclonite
is 277 kcal/kg. According to the German data (Schnurr [38]) it is about 500 kcal/kg.

20 40 60 80 100
Molar ratio of nitric acid to hexamine

Fig. 15. Effect of nitric acid-hexamine ratio on initial rate of cyclonite formation at
0°C, according to Vroom and Winkler [42].

Gilpin and Winkler [51] report the following heats of nitration when cyclonite
is prepared in different ways:

hexamine + nitric acid (97.5%) - JH=$8.0 kcal/mole
hexamine nitrate + nitric acid (97.5%) — AH= 69.2 kcal/mole
hexamine dinitrate+ nitric acid (97.5%) —AH= 41. 7 kcal/mole

The formation of hexamine dinitrate from hexamine and nitric acid proceeds
with a heat effect of — zf// = 33.5 kcal/mole: hexamine nitrate is also converted into
dinitrate with a heat evolution of —AH=\5.1 kcal/mole.

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

^.7

u ■— ^^^

^v.-

i >

100 90 80

Concentration of HNO3

Fig. 16. 7 — Nitration of hexamine at 20°C; 2— nitration of hexamine at — 35.5°C;
5— nitration of hexamine dinitrate at — 35.5°C, according to Dunning, Millard and
ft Nutt [50].

On the basis of these figures the authors infer that hexamine dinitrate is formed
at one stage of the nitration of hexamine to cyclonite.

Dunning, Millard and Nutt [50] published a graph (Fig. 16) showing the relation
between the heat of nitration of hexamine to cyclonite and the concentration of

I
Z2\

-3

f 2

If '\r

—5

~ 7 8

lOrnin

Time

Fig. 17. Temperature changes during the nitration of hexamine with nitric acid,
according to Dunning, Millard and Nutt [50]. 1— Anhydrous acid, nitration tempera-
ture 20°C; 2 — anhydrous acid, nitration temperature — 35.5°C; 3 — 96% acid, nitra-
tion temperature 20°C; 4 — 96% acid, nitration temperature -35.5°C; 5 — 90% acid,
nitration temperature 20°C; 6 — 90% acid, nitration temperature -35.5°C; 7—85%
acid, nitration temperature 20°C; 8 — 85% acid, nitration temperature -35.5°C.

nitric acid (within the range 80-99%). The graph expresses the integral heat of
reaction in kcal per mole of hexamine at 20 and — 35.5°C and of hexamine dini-
trate at -35.5°C.

HETEROCYCLIC NITRAMINES

Temperature changes during nitration with various concentrations of acid at
20 and -35.5°C are shown on another graph (Fig. 17). Alterations in the shape
of the curves are particularly marked when nitration proceeds at -35.5°C. During

-^ 10

\cQH : hexamine Curve
10-5 o

6-4 .

4-3 a

Temperature: upper curves l"c
lower — .— 30°c

30 40

Molar ratio of H.\o : , .- hexamine

Fig. 18. Effect of nitric acid-hexamine ratio on final yield of cyclonite in acetic acid,
according to Kirsch and Winkler [52].

nitration with 85% acid at -35.5°C the quantity of heat evolved corresponds to
the formation of hexamine dinitrate only.

It appears that the reaction stops at this stage under these conditions.

Kirsch and Winkler [52] studied the influence of acetic acid on the nitrolysis of
hexamine to cyclonite.

Experiments were made at 1°C and 30°C using various molar ratios of acetic
acid/hexamine and varying the nitric acid/hexamine ratio between 26 : 1 and 81 : 1.
Acetic acid was found to reduce the reaction rate and the yield of cyclonite (Fig. 18).
However even with the most dilute solution of hexamine in acetic acid (molar ratio
10.5 : 1) the final yield of cyclonite approached a maximum of 80% at a molar ratio
of nitric acid to hexamine of 48 : 1. In the absence of acetic acid this yield was
obtained when the molar ratio was only 26 : 1. It appears that some of the nitric
acid was used up in reacting with acetic acid and was therefore unavailable for
nitrolysis.

There are various methods of utilizing the spent acid remaining from the nitration
of hexamine. It is possible:

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(1) To distil off nitric acid and to utilize the ammonium nitrate remaining after
distillation for other purposes.

(2) To neutralize the acid with ammonia thus producing ammonium nitrate.

(3) To utilize the acid for preparing hexamine dinitrate and to nitrate the di-
nitrate, so obtained, to cyclonite. This method was studied by T.Urbahski [53] who
found, however, that the dinitrate is nitrated to cyclonite with some difficulty, part
of the dinitrate remaining unchanged in the product. Since the dinitrate is not suf-
ficiently stable, its presence in cyclonite is undesirable. Nevertheless, it may be used
up for a synthesis of cyclonite by adopting a combined method, using hexamine
dinitrate, ammonium nitrate and acetic anhydride. This will be discussed in later
sections.

A factory at Ayigliana (Italy) [54] used an entirely different method for processing
and utilizing spent acid containing 45% HN0 3 : after the precipitation of cyclonite
in the cold, the acid is subjected to distillation to recover the formaldehyde it con-
tains. It is important to bear in mind that the distillation of acid which has not
been passed through the "degassing process" is very dangerous even if carried out
at a low temperature under vacuum (40°C is the recommended temperature).

British method

Nitration. In the production method employed in a factory at Bridgwater [55],
the nitrator is fed continuously with nitric acid and hexamine and the product of
nitration together with the acid flows off, also continuously.

The nitrator of stainless steel, 90 cm long, 32 cm wide, 80 cm high is divided
by partition walls into three chambers. Each chamber contains a rotary high-speed
stirrer. The first chamber is equipped with three concentric cooling coils of stain-
less steel 16 mm in diameter with a cooling surface of 1.85 m 2 . The next two chambers
contain single coils. Gases are expelled to the absorption towers through pipes
leading from the lid of the nitrator. The pipe from the first chamber is equipped
with a sight glass for observing the colour of the gases. If the colour is brown, the
contents of the nitrator should be discharged forthwith into the drowning tank
under the nitrator. Hexamine stored on the floor above is introduced by means of
a screw conveyer into the first and second chambers of the nitrator through an
inlet 5 cm in diameter. The feed mechanism is so arranged that the second chamber
receives a quarter of the amount entering the first chamber. The rate at which the
total amount of hexamine introduced into the nitrator may be varied is from 56 to
170 kg per hour. The weight of nitric acid introduced into the first chamber through
a pipe 2.5 cm dia. is 12 times that of the hexamine.

The temperature in the first and second chambers of the nitrator should not
be allowed to rise above 25°C. Control is achieved by intensive cooling. In the
third chamber a temperature of 38°C is maintained by passing warm water through
the coil.

HETEROCYCLIC NITRAMINES

An overflow for the acid and the reaction product is located in the third chamber
about 58 cm from the bottom. The general view of the nitrator is shown in Fig. 19.
A drowning tank is situated under the nitrator and connected to it by a pipe 7.5 cm

Fig. 19. General view of a continuous nitrator for the preparation of cyclonite [55].

in diameter. The tank is filled with an aqueous solution of urea which, on reaction
with nitric oxides, considerably reduces the amount of brown fumes evolved when
the nitrator's load is discharged into it.

Dilution. The liquid leaving the nitrator through the overflow passes to the
diluter below. Here decomposition of unstable nitration products and precipitation
of cyclonite takes place. The diluter (Fig. 21), of dimensions 265x60x115 cm, is
divided into four chambers, each containing a rotary stirrer (195 r.p.m.) and a heat-
ing coil. The diluter is filled continuously up to a level of 65 cm with 55% nitric acid.
As the liquid containing more concentrated acid flows in from the nitrator water
is added to maintain this concentration. A temperature of 75°C is maintained in
the diluter by means of the heating coils.

100

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 20. Nitrator control panel. In the centre the drowning valve handle [55].

Fio. 21. General view of a diluter [55].

HETEROCYCLIC NITRAMINES

101

&.<*

■Wwi^-'W ■=* It™

ff| £|?^*ie^flf:^|t j«^if«if^^;*f»

ri I

Fig. 22. Stainless steel absorption towers for nitric oxides [55].

Fig. 23. Continuous vacuum filters (exterior view) [55],

102

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

During dilution nitric oxides are evolved abundantly (more than 7 kg/hr). These
oxides are extracted by means of a ventilator or steam ejector to a cooling tower
where they are brought to 20-30°C. The tower, which measures 1.8 m in diameter

Fig. 24. An interior view of the filter [55].

and 3.3 m in height, contains steel pipes 38 mm in diameter, cooled externally with
water. Here some of the vapour is condensed and nitric oxides pass on to the absorp-
tion towers (Fig. 22).

Ik & I

Fig. 25. Disk mill for cyclonite slurry [55].

Filtration. The suspension of cyclonite in nitric acid passes through a pipe 7.5
cm in diameter to a continuous vacuum filter (Fig. 23, 24) encased in stainless steel.
The casing is connected with the suction system that conveys nitric oxides to the
absorption towers.

The product, after filtration, is conveyed mechanically to another filter on which
the cyclonite is washed with cold water until deacidification is as complete as pos-

HETEROCYCLIC NITRAMINES

103

sible. Mechanical scrapers remove wet, washed cyclonite from the filter and it is trans-
ferred in trucks to be purified.

Spent acid and that from the absorption towers is distilled over sulphuric acid.

Purification. Crude cyclonite consists of crystals of various sizes. They still
contain 0.1-0.2% of nitric acid. For purification a suspension of cyclonite in water

Fig. 26. Cyclonite boiling vats [55].

is conveyed by suction via a stainless steel container to stainless steel disk mills
(Fig. 25) where the cyclonite crystals are finely ground. From the mill the ground
product suspended in water is pumped to wooden boiling vats (Fig. 26) 250 cm
high and 240 cm dia. The vats have sloping bottoms and are equipped with a stirrer
and a pipe supplying live steam. There are two valves in the sides: one, in the lowest
part of the vat, can be used for emptying it, the other, 105 cm from the bottom,
is used for decantating the liquid. Each load amounts to about 2500 kg of dry cyclo-
nite.

104 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The suspension of the ground cyclonite in water is allowed to stand in the vat
for | hr. The liquid is then decanted from above the cyclonite layer and allowed to
run through a filter which retains any finely-divided product carried away
by the water. After the liquid has been decanted the tub is filled with cold water,
its contents are stirred, allowed to stand again, and then decanted. This operation
is repeated three times after which the cyclonite is washed with hot water at 90-100
°C for 12 hr. After washing the liquid is decanted and the wet cyclonite is removed
from the tub to the filter. The filtered cyclonite containing 10% of water, goes on
to the department where explosive mixtures are prepared.

Material balance. To produce 1000 kg of cyclonite, 833 kg of hexamine and
8779 kg of HN0 3 are required; 3482 kg of dilute 55% HN0 3 are recovered plus
3429 kg of HNO3 from the absorption towers. Thus the net consumption of HN0 3
for nitration is 1868 kg. In addition, 490 kg of H 2 S0 4 are used for the concentra-
tion of HNO3.

German method

This has been called the "SH-method" and the cyclonite so obtained "SH-Salz"
after Schnurr [38] who developed the process in 1937-38.

Hexamine is added to 99% nitric acid in the proportion of 1 part of hexamine
to 8 parts of acid, in a nitrator with a capacity of 1.5 m 3 working on the batch
system. The temperature is maintained between 5 and 10°C in the nitrator by means
of a coil chilled with saline solution. Nitration takes 1 hr.

The contents of a number of nitrators then flow continuously through a series
of reactors in which the reaction mixture remains for 2 hr, during which the nitration
reaction is completed. In these reactors the temperature is kept at 15-20°C.

When reaction is complete the whole is introduced into a battery of six diluters
with water maintained at a temperature of 70-75°C. The first diluter has a capacity
of 3 m 3 , the following four 1.5 m 3 each, and the last 3 m 3 . Sometimes, to initiate
decomposition of the unstable nitration products, the presence of a small amount
of nitrogen dioxide is required. The amount of water added to the diluter should
be such that it maintains a 50% concentration of HN0 3 . At this concentration
cyclonite is fully precipitated. The suspension of cyclonite in acid then flows to a bat-
tery of coolers in which it is cooled gradually to temperatures of 50, 35, and 20°C.

Cyclonite is separated from acid on a vacuum filter and then washed with water.

Originally, the product was purified by boiling with water in an autoclave under
a pressure of 3.5 atm, at 140°C for 2 hr. Since this entailed the risk of explosion,
it was superseded by crystallization from acetone.

Nitric oxides evolved in the diluter are passed through a cyclone separator
which retains liquid droplets, and thence to absorption towers in which the 50%
spent acid is concentrated to yield 60% nitric acid.

To produce 1000 kg of cyclonite ("SH-Salz") by this process, 830-840 kg of
hexamine and 7100 kg of 99% nitric acid are required, and 5200 kg of HN0 3 are

HETEROCYCLIC NITRAMINES 105

recovered and condensed to its initial concentration. The net consumption of acid
is thus 1700 kg.

According to other data the raw material consumption per 1000 kg cyclonite is
as follows:

880 kg of hexamine
6800-7760 kg of 99% nitric acid

from which 1720-1850 kg is used in the reaction and

5080-5850 kg, which is recovered as 99% HNO3.

The German specification for cyclonite ("SH-Salz") is:

melting point

above 200°C

ignition temperature

215-230°C

apparent density

700 g/1.

loss of weight when dried

at 100°C for 5 hr should

not exceed

.0.1%

Abel heat test at 120°C 10 min without visible change
(in a paraffin bath) in the test paper (after 20 min
a slight coloration is admis-
sible)

An aqueous extract (obtained by boiling cyclonite with water) should have
a neutral reaction and contain no Cl e , SO^ 9 or NOf ions. Only traces of ammonia
and formaldehyde are admissible (the latter is determined by a fuchsin solution de-
colorized with S0 2 ).

The product should not contain more than 0. 1 % of material insoluble in acetone.
The acetone solution should not contain more than 0.2% of acid, calculated as
HN0 3 . An acetone solution precipitated with water should not contain SO| e ions.

Cyclonite intended for high explosive charges should pass through a 0.75 mm
mesh sieve. Cyclonite for caps and detonators should pass through a 0.60 mm. mesh
sieve.

If necessary, cyclonite may be mixed with montan wax.

Cyclonite lots comprised 2500-7500 kg: 12.5 kg in paper bags were packed in
cardboard boxes, with lids sealed by means of adhesive tape.

2. PREPARATION OF CYCLONITE FROM HEXAMINE, NITRIC ACID AND AMMONIUM

NITRATE

This method was worked out by Knoffler [38], and is known as the "K-method".

It is based on the fact that hexamethylenetetramine contains 6 methylene groups
per 4 amino groups, i.e. the number of amino groups is lower then that required
for obtaining two molecules of cyclonite. In order to supply the additional two
amino groups the appropriate amount of ammonium nitrate is added to the nitrating
nitric acid in accordance with the equation:

C 6 H 12 N 4 + 2NH4NO3 + 4HN0 3 ~> 2(CH 2 N-N0 2 ) 3 + 6H2O (42)

106 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

An increase in yield over that calculated on the basis of the hexamine used is
thus obtained. The reaction proceeds only at a high temperature (about 80°C).
This has gained favour because the reaction mixture of nitric acid, hexamine and
ammonium nitrate may be heated to the desired reaction temperature without
fear of an explosion, whereas a reaction mixture without ammonium nitrate may
be dangerous even at a temperature just above 25°C.

According to the investigations of T. Urbariski and Szyc-Lewanska [56], cyclonite
is also formed by the action of nitric acid and ammonium nitrate on hexamethylene-
triperoxidediamine which is discussed on p. 225.

It is assumed that cyclonite is formed from this substance since it contains the
grouping — CH 2 — N. Six methylene groups are attached to two nitrogen atoms.
The deficiency in nitrogen atoms is made good by ammonium nitrate, as described
above.

According to a description from a factory at Elsnig nitration is conducted in
reactors with a capacity of 500 1., in which 1 part of hexamine per 8.6 parts of nitric
acid plus the calculated amount of ammonium nitrate are added to 99% nitric
acid.

A temperature of 15°C is maintained in the nitrator by cooling and the reaction
mixture enters the reactor below, which is equipped with a number of vertical pipes
warmed externally with hot water. Here the reaction mixture is heated to 80°C and
this temperature is maintained for 30 min while the reaction (42) involving am-
monium nitrate takes place. The mixture is then introduced into the container
below where the whole is cooled to 20°C.

Cyclonite in approximately 90% yield is crystallized in the following way: the
product is separated from the spent acid on a rotary filter and after being washed
with water and neutralized with a 5% solution of sodium carbonate it is crystallized
from acetone.

In the method outlined above it is particularly difficult to utilize the spent acid
due to the considerable amount of ammonium nitrate it contains. The usual method
in which the nitric acid is recovered by distillation with sulphuric acid and condensa-
tion is not practicable since ammonium salts are deposited in the retorts, distillation
columns, pipes, valves etc.

The following method has therefore been developed: after the cyclonite has
been filtered off, the spent acid is cooled to - 12°C. Ammonium trinitrate is then
crystallized, separated in a centrifuge and recycled to the nitrator. After centri-
fuging the acid contains 10% of cyclonite and a considerable amount of ammonium
nitrate. It is further processed so as to yield cyclonite, nitric acid being converted
into ammonium nitrate.

The acid is allowed to run into a container of 3 m 3 capacity where it is neutral-
ized with gaseous ammonia while the temperature is raised to 70°C. The cyclonite
thus formed is separated on a vacuum filter, and the filtrate is cooled when about
two-thirds of the ammonium nitrate crystallizes. The latter is separated in a centrifuge,
and used in the preparation of explosive mixtures. The small amounts of hexamine

HETEROCYCLIC NITRAMINES 107

dinitrate and ammonium nitrate that remain in the liquor are recovered, as a mixture,
by concentrating the solution, after which they are returned to the nitrator.
To produce 1000 kg of cyclonite

480-500 kg of hexamine
4800 kg of ammonium nitrate
8600 kg of nitric acid
are required from which:

3600 kg of ammonium nitrate
7200 kg of nitric acid
are recovered.

The consumption of ammonium nitrate amounts to 1200 kg and of nitric acid
to 1400 kg.

3. PREPARATION OF CYCLONITE FROM SULPHAMIC ACID, FORMALDEHYDE AND

NITRIC ACID

A method for the preparation of cyclonite from sulphamic acid, formaldehyde
and nitric acid was developed in 1934 by Wolfram [38]. It is known as the "W-meth-
od" and the cyclonite so obtained is known as "W-Salz". The W-method is based
on the condensation of the potassium salt of sulphamic acid with formaldehyde
and the nitration of the condensation product ("white salt", a Schiff's base of the
type XVIII) with nitric acid.

Starting from sulphuric anhydride and ammonia the following series of reac-
tions was employed:

.ONH4
SO3 + 2NH 3 -> S0 2 (a)

^NH 2

.ONH4
S0 2
2S0 3 + 3NH 3 -> NNH (b>

SO2
\0NH4

NH 4 OS0 2 NH 2 + (NH 4 OS0 2 ) 2 NH + Ca(OH) 2 ->

(NH 2 S0 2 0) 2 Ca + CaS0 4 + 3NH 4 OH (c)

(NH 2 S0 2 0)2Ca + K2SO4 -> 2NH 2 S0 2 OK + CaS0 4 (d)

.OK

30°C / . x

S0 2 (e>

.OK
S0 2

\nh 2

+ CH2O

X)K
3S0 2

+ 3HN0 3

\n=ch 2

pH=5

\N=CH 2
XVIII

>- (CH 2 -N-N0 2 )3 + 3KHS0 4 (43)

108 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The ammonia evolved in reaction (c) during the preparation of calcium sulpha-
mate is recycled to take part in reactions (a) and (b). Since calcium sulphamate
dissolves fairly easily, it is transformed into the sparingly soluble potassium salt by
reaction (d).

The mechanism of these reactions has been examined by Binnie, Cohen and
Wright [57]. X-Ray investigations of the crystal lattice of this compound and cryo-
metric studies, have shown that the compound XVIII has the structure of a cyclic
trimer, i.e. that of 1,3,5-triazacyclohexanetrisulphonic acid (XVIIIa).

.CH 2

KO3S— N' \n— SO3K

I I

H 2 C CH 2

\ N /

I
SO3K

XVIIIa

On treatment with nitric acid under completely anhydrous conditions (80%
nitric acid and 20% S0 3 ) the compound (XVIIIa) gives cyclonite by substitution
of the sulpho groups with nitro groups (80% of the theoretical yield). The reaction
is interesting because the cyclonite is not decomposed in spite of the presence of
sulphuric acid in the nitrating mixture. The reason is that the medium is completely
anhydrous due to the presence of an excess of S0 3 .

For the nitration of compound (XVIII) the following nitrating mixture was
used at Kriimmel [38]:

HNO3 80-81%

H2SO4 4-5%

SO3 13-14%

N 2 4 1-2%

This mixture was prepared from 99% nitric acid and sulphuric anhydride. "White
salt" (XVIII) was introduced to the mixture at a temperature of 30°C. The heat
evolved during nitration (approximately 500 kcal per kg of cyclonite) was removed
by means of a cooling coil. The cyclonite so formed was partly suspended and
partly dissolved in the nitrating liquid. Addition of water completely precipitated
the product which was separated on a vacuum filter. The composition of the spent
acid was :

HNO3 23%

H 2 S0 4 13-14%

KHSO4 10-11%

H 2 52-54%

The cyclonite was then washed with water and after neutralizing the residual acid
with a 5 % solution of sodium carbonate it was recrystallized. Initially, cyclonite was
crystallized from nitrobenzene. However this proved to be dangerous due to the
high boiling point of the solvent; after a plant had been destroyed by an explosion,
crystallization from acetone was adopted. The spent acid is denitrated, and KHSO4.

HETEROCYCLIC NITRAMINES 109

recovered from the retort. The yield obtained ranked from 80 to 90%, depending on
the formaldehyde used.

This method proved less profitable than others, and was discontinued by the
end of World War II.

4. PREPARATION OF CYCLONITE FROM PARAFORMALDEHYDE, AMMONIUM
NITRATE AND ACETIC ANHYDRIDE

This method was worked out by Ebele [38], and known as the "E-method" in
Germany. The same method was also invented independently by Ross and Schiessler
[58] in Canada in 1940. In this process paraformaldehyde and ammonium nitrate
undergo dehydration under the influence of acetic anhydride with the formation of
cyclonite according to the equation :

3CH 2 0+3NH 4 N03+6(CH3CO)2^ (CH2NN0 2 )3 + 12CH 3 COOH (44)

The studies of Wright et al. [59], and Winkler et al. [60] led to the conclusion
that this method involves two essential steps.

The first is the synthesis of hexamethylenetetramine in acetic anhydride from
paraformaldehyde and ammonium nitrate :

6CH 2 + 4NH4N03 + 3(CH 3 CO)20-> C6H12N4+4HNO3 + 6CH3COOK (45)

The second step is the known nitrolysis :

C6H12N4 + 4HNO3 -> (CH 2 N-N02)3 + NH 4 N03 + 3CH20 (46)

The detailed mechanism of this reaction will be dealt with later on p. 113.

The advantage of this method lies in the fact that the preparation of cyclonite
. is accomplished without using hexamine and nitric acid so that the dangers of nitra-
tion are avoided. On the other hand acetic anhydride is required, which is rather
expensive.

This method is safe on condition that it is conducted as described below, by
adding the reactants to the previously warmed acetic anhydride. Otherwise, the
exothermic nature of the reaction may cause an explosion. It is therefore inad-
missible to mix the reactants and heat up the mixture, since this causes too violent
a reaction. Addition of boron fluoride to the mixture promotes the initiation of
reaction and increase its safety.

It appears that apart from cyclonite, octogen (XI), a N-acetyl derivative (XXX)
and a number of nitramines, mainly chain compounds, less stable than cyclonite,
are also formed by this method, in side-reactions. Hence the cyclonite obtained has
a relatively low melting point (190-195°C) and may be less stable than that prepared
by other methods. The conduct of the reaction in the presence of boron fluoride
reduces the number of by-products formed.

Since the majority of the chain compounds formed as by-products are more soluble
in acetic acid than cyclonite itself, contamination of the latter with these substan-
ces may be partly avoided by filtering off the cyclonite from the spent acid. The

HO CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

by-products can be recycled since some of them may be transformed into cyclonite
on treatment with acetic anhydride and ammonium nitrate under the conditions in
which the main reaction is conducted.

The by-products obtained during the reaction are discussed more fully on p. 113.

Manufacture at Bobingen [38,61] was on the following lines. A reactor of aluminium
or stainless steel (capacity 1.2 m^) is filled with acetic anhydride and then 0.4% of
BF 3 is added. Acetic anhydride is warmed to 60-65°C and at this temperature
ammonium nitrate and paraformaldehyde are added gradually. Due to the high
temperature and the presence of boron fluoride the reaction starts at once and heat
is evolved. The heating is then turned off and the temperature maintained by cooling,
within the range 60-65°C. The addition of the reactants requires approximately
6 hr, after which the contents of the reactor are cooled to 20°C. The precipitated
cyclonite is separated from the solution on a vacuum filter. The by-products remain
in the spent liquor.

160 kg of cyclonite are obtained from each reactor. The output from a number
of reactors, amounting to about 800 kg, is stabilized by boiling in an autoclave at
140°C.

The spent acid is distilled in order to separate acetic acid. The paste containing
cyclonite and by-products that accumulate on the bottom of the retort is removed
continuously, through a syphon overflow. The greater part of this mass (about
80%) is dissolved in acetic anhydride and returned to the reactor while the residue
(about 20%) is mixed with ammonium nitrate to make cheap explosives.

Acetic anhydride is prepared from distilled acetic acid by the ketene method
with ethyl phosphate as a catalyst. The water from the first cyclonite wash on the
vacuum filter contains on an average 20% of acetic acid which is recovered by
extraction with ethyl acetate.

The yield of cyclonite on a plant scale is 63-65% calculated as formaldehyde;
on a laboratory scale a yield of up to 80% may be obtained.

To produce 1000 kg of cyclonite requires

630-635 kg of paraformaldehyde

1 800 kg of ammonium nitrate

and 5000-5100 kg of acetic anhydride containing about 19 kg of boron fluoride.

From the spent acid

about 110 kg of the paste of cyclonite and by-products
and 41 50-4200 kg of acetic anhydride
are recovered.

Thus the consumption of acetic anhydride per 1000 kg of cyclonite amounts to
ca. 800 kg.

According to the German data the product obtained contains :

93.5% of cyclonite
6.0% of octogen
0.5 % of the acetyl derivative (XXIX)

HETEROCYCLIC NITRAMINES 111

5. PREPARATION OF CYCLONITE FROM HEXAMINE DINITRATE, AMMONIUM
DINITRATE AND ACETIC ANHYDRIDE

This method is a combination of the first and fourth methods. It was worked
out by W. E. Bachmann in the U.S.A. in 1941 and independently by Koffler [38, 62]
in Germany in 1943. In Germany it was known as the "KA-method".

Hexamine dinitrate is reacted with ammonium dinitrate in the presence of acetic
anhydride. Unlike the E-method no paraformaldehyde is used, all the necessary
methylene groups being provided by hexamine, and the additional amino groups
(as in the K-method) by ammonium nitrate. Nitric acid enters into reaction in
combination with hexamine and as ammonium dinitrate :

C 6 Hi 2 N 4 -2HN0 3 + 2(NH 4 NO r HN0 3 ) + 6(CH 3 CO) 2 0^

2(CH 2 N-N0 2 ) 3 + 12CH 3 COOH (47)

The yield by this method, calculated on the CH 2 groups in hexamine dinitrate,
amounts to 75-80%.

The advantage of the KA-method, as compared with the E-method, lies in the
fact that the methylene and part of the amino groups required are introduced
in a dehydrated form, hexamine being their source, whereas when paraformaldehyde
is used, dehydration is essential. Hence less acetic anhydride is used in the
KA-method than in the E-method.

At the Bobingen factory hexamine dinitrate is prepared by the action of 50%
nitric acid on hexamine at a temperature below 15°C, using spent acid from the
nitration of pentaerythritol. NH 4 N0 3 • HN0 3 is obtained by the reaction of equi-
molecular amounts of ammonium nitrate and concentrated nitric acid.

325 kg of acetic anhydride is mixed in the reactor with 1871. of filtrate from
the previous batch. While maintaining the temperature between 40 and 50 C C five por-
tions of 23.9 kg of ammonium dinitrate and of 22 kg of hexamine dinitrate are
added to the reactor, after which 271 kg of acetic anhydride are introduced followed
by a further 5 portions of ammonium dinitrate and hexamine dinitrate, as above.
Experience showed that a small excess of hexamine dinitrate is useful. After the
last portions of the reactants have been added, the contents of the reactor are heated
to 60°C and this temperature is maintained for 0.5 hr. This procedure requires 4 hr,
after which the whole is cooled to 20°C. Cyclonite, crystallized in 70-71% yield, is
filtered off. The melting point of the product ranges between 188 and 190°C.

Part of the filtrate is recycled, while most of it undergoes distillation. On distilla-
tion, a paste-like residue of highly impure cyclonite (m.p. 160°C — total melting)
is obtained. Under the reaction conditions this product may be transformed
into cyclonite, after it is returned to the reactor.

The filtered cyclonite is washed with water, to yield a 30% solution of acetic
acid from which concentrated acetic acid is recovered by extraction with ethyl
acetate.

112 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The consumption of ethyl acetate amounts to 1 kg per 100 kg of cyclonite. From
the recovered concentrated acetic acid 85% acetic anhydride is produced by the
ketene method.

To produce 1000 kg of cyclonite the KA-method requires

400 kg of hexamine
430 kg of ammonium nitrate
680 kg of 99% nitric acid
2400 kg of acetic acid (as anhydride)

from which about 1950 kg of acetic acid are recovered, i.e. the net consumption
amounts to about 450 kg.

During World War II W. E. Bachmann et al. [62, 63, 64] worked out a process of
cyclonite manufacture identical in principle to the above. Bachmann's idea was to
combine Ross's E-method with hexamine nitration. A semi-plant scale equipment
based upon the new "combined" Bachmann process came into operation at the end
of 1941. After the procedure for cyclonite production had been worked out, the
problem of the regeneration of acetic acid was solved; in the United States cyclonite
was produced mainly by this method.

The procedure worked out by Bachmann differs from that described above in
that the solution of 117 parts of 98% nitric acid in 508 parts of acetic anhydride is
first prepared while maintaining the temperature between 5 and 15°C.

The solution is then heated to 70-75°C and a mixture of 1 14 parts of ammonium
nitrate with 192 parts of hexamine dinitrate is introduced gradually in portions
(hexamine dinitrate is prepared beforehand by treating 65 parts of 70% nitric acid
with a solution of 40 parts of hexamine in 70 parts of water, at 15°C;the mixture is
then cooled to 5°C and the dinitrate is filtered off in 95% yield).

When the mixture of ammonium nitrate with hexamine nitrate is introduced to
a solution of nitric acid in acetic anhydride heat is evolved and the temperature
rises spontaneously. At that point heating should be stopped and the temperature
maintained within the limits of 73-78°C, by cooling. After the reactants have been
introduced the whole is stirred for 15 min longer, maintaining the temperature at
75°C by heating. The whole is then cooled to 60°C, and the product is filtered, washed
with acetic acid, and then with water. The melting point of the product thus obtained
is 203-204°C. 195-202 parts of cyclonite are obtained (61-63% of the theoretical
yield). The yield increases to 70-73.5% if the mixture is cooled after the reaction to
25°C, but the product then contains a lot of octogen, hence its melting point is
191-202°C.

The heat of reaction was determined by Gilpin and Winkler [51]. The reaction
of a solution of hexamine with Bachmann's reagent to yield cyclonite is exothermic:
-,d/f= 140 kcal/mole.

Hexamine mononitrate gives a heat effect of -AH =126 kcal/mole with Bach-
mann's reagent while hexamine dinitrate with the same reagent gives — AH
= 118 kcal/mole.

HETEROCYCLIC NITRAMINES 113

At the Elsnig factory the crystallization of cyclonite is accomplished as follows.
About 1 10 kg of cyclonite are introduced into a closed tank, with a capacity of
1000 1., equipped with stirrers and lined with a woollen filter cloth. Approximately
900 1. of acetone heated to 50°C are run into the tank to dissolve the cyclonite, after
which the solution filtering through the filter cloth is drained down into a 3000 1.
tank. (The filter cloth is changed every 10 hr). Here about 1350 1. of water is added
over a period of 5 min, while the temperature is maintained at 25°C, and cyclonite
is precipitated from the acetone solution in the form 'of fairly large crystals : approxi-
mately 90% of the total are longer than 0.1 mm. The precipitated cyclonite is sepa-
rated on a vacuum filter.

Acetone vapour is recovered in an absorber. The aqueous acetone filtered from
the cyclonite is purified and freed from water by distillation in a rectifying column.
The acetone losses amount to 7-8% per 100 kg of cyclonite.

According to data from German factories, desensitization ("phlegmatization")
of cyclonite is carried out as follows. A phlegmatizing tank, equipped with a heating
or cooling jacket, is filled with 450-500 1. of hot water (80-88°C). The stirrer (170
r.p.m.) is then started and 120 kg of recrystallized cyclonite is poured in. Molten
montan wax is then introduced in the proportion of 5-10% the weight of cyclo-
nite.

If a smaller amount of wax is used, a higher temperature should be maintained,
viz. addition of 5% wax requires about 88°C and 10% between 80 and 82°C.

After the wax has been added and incorporated, the cooling should start immedi-
ately by introducing cold water into the jacket. The overall procedure requires ap-
proximately 2 hr. The phlegmatized cyclonite is filtered on a vacuum filter, dried at
100°C and sieved.

To differentiate phlegmatized cyclonite from the non-phlegmatized product a
dye is added during mixing.

THE THEORY OF CYCLONITE FORMATION BY METHODS 4 AND 5

Examination of the mechanism of cyclonite formation has clarified the condi-
tions under which various products similar in structure to cyclonite are formed.
These products always accompany cyclonite prepared by method 5 and often that
prepared by method 4.

In method 5, as in method 1, nitrolysis of hexamethylenetetramine occurs, the
alcohol groups so produced being esterified with acetic anhydride and not with
nitric anhydride as in method 1. The most important by-product formed to the
extent of nearly 10% by method 5 is octogen (XI). It is produced as outlined above,
as a result of hexamine nitrolysis (p. 90).

In the presence of acetic anhydride hexamethylenetetramine dinitrate may be
transformed into the product (XIX), and then into (XX). These are the analogues
of substances (V) and (VI).

JJ4 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

™ 3 N0 2

e I

CH 2 — NH— CH 2 CH 2 — N £ x

III II ^/CH 2 OCOCH 3

2 NN CH 2 NCH 2 OCOCH 3 -» 2 NN CH 2 N<

III | | /-H \CH 2 OCOCH 3

CH 2 — N CH 2 CH 2 — N CH 2

t
M

XIX XX

The nitrolysis of the bond /in substance (XX) results in the formation of products
(XXI) l-acetoxymethyl-3,5-dinitro-l,3,5-triazacyclohexane and (XXII) bis(acetoxy-
rnethyl)-nitramine

CH 2 — NN0 2 CH 2 OCOCH 3

I I I

2 NN CH 2 NN0 2

CH 2 — N— CH 2 OCOCH 3 CH 2 OCOCH 3

t
K

XXI • XXII

Substance (XXI) may subsequently undergo nitrolysis at K to form the
nitramine ester l,7-diacetoxy-2,4,6-trinitro-2,4,6-triazaheptane (XXIII) (a chain
compound "BSX", m.p. 154-155°C):

N0 2 NO z N0 2

CH 2 — N— CH 2 — N— CH 2 — N— CH 2

I I

OCOCH 3 OCOCH 3

XXIII

This compound sometimes occurs in an appreciable quantity in cyclonite pre-
pared by methods 4 and 5. It is an undesirable by-product due to its high sensi-
tiveness to impact and to its relatively poor stability, which is lower than that of
cyclonite. Bachmann and Sheehan [62] found that stirring the reactants at a low
temperature (e.g. 0°C) and subsequently heating the whole to 75°C favours the
formation of the product (XXIII).

On the other hand, Wright et al. [47] found that the nitrolysis of hexamine with
nitric acid in the presence of acetic anhydride but in the absence of ammonium
nitrate involves a decrease of the yield of cyclic products. The amount of the chain
compound (XXIII) formed is then increased.

The fact that nitramino groups may arise under the influence of ammonium
nitrate and nitric acid in the presence of acetic anhydride is shown by the reaction
in which aminomethylnitramines (obtained from nitramine, formaldehyde and, say,
morpholine) are treated with acetic anhydride in the presence of ammonium nitrate
and nitric acid at 55°C (Lamberton et al. [65]):

HETEROCYCLIC NITRAMINES 115

N0 2 NO Z N0 2

I I /— \ I

NH N— CH 2 — N O N CH 2

I I \ / Ac.O, HNO. I I / \

(CH 2 )„ -> (CH 2 )„ / _^ 7 nh.no,. Si°c ^ N-N0 2 + 2 N-N O

NH N— CH 2 — N O N CH 2

I ! \_/ i

N0 2 N0 2 N0 2

\_/

Compound (XXIII) may undergo nitrolysis resulting in the formation of the
ester (XXIV) — l,5-diacetoxy-2,4-dinitro-2,4-diazapentane:

N0 2 N0 2

I I

CH 2 — N— CH 2 — N— CH 2

I I

OCOCH3 OCOCH3

XXIV

The chain compound XXV, l,9-diacetoxy-2,4,6,8-tetranitro-2,4,6,8-tetrazano-
nane m.p. 182.5-183. 5°C may also arise from compound (XX) by nitrolysis of the
bonds L and M.

N0 2 N0 2 N0 2 N0 2

I I I I

CH 2 — N— CH 2 — N— CH 2 — N— CH 2 — N— CH 2

I I

OCOCH3 OCOCH3

XXV

It is possible that there may not be time for the free alcohols (XXVI) and (XXVII)
to be transformed into esters (XIX) and (XX), but under the influence of ammonium
nitrate they will undergo "demethylation" with loss of formaldehyde molecules.

e
no 3

N0 2

CH 2 — NH— CH 2 < CH 2 — N

III II /CH 2 OH

2 NN CH 2 NCH 2 OH 2 NN CH 2 P N<

III I I \ I \CH 2 OH

CH 2 — N — CH 2 CH 2 — N CH 2

XXVI XXVII

If compound (XXVI) is "demethylated", the compound (XII) (DPT) will be
produced as a result of the additional nitration.

In method 5 ammonium nitrate not only supplies the necessary number of nitr-
amino groups, but also brings about the loss of molecules of formaldehyde from
the intermediate products of types (XXV) and (XXVII). If the time is too short
for the esterification of the alcohol (XXIII), nitrolysis of this compound occurs
first at the bond P (not at bonds L, Mand / as in the acetylated product (XX)) and
cyclonite is formed.

Experiments have shown that the chain compounds of types (XXIII), (XXIV)
and (XXV) or the compound (IX) described earlier may also undergo cyclization

1J6 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

under the influence of nitric acid, acetic anhydride and ammonium nitrate (Chapman

[66]).

On synthesis of cyclonite from paraformaldehyde, ammonium nitrate and acetic
acid (method 4), products with cyclic structure, chiefly cyclonite and octogen, may
arise due to the polymerization of the transiently-formed, hypothetical methylene-
nitramine (XXVIII) [67]:

CH 2 + NH4NO3 -» CH 2 =N— N0 2 + 2H z O
XXVIII

There is an alternative view (e.g. expressed by Wright et al. [59]) that hexamethylene-
tetramine is first formed and then nitrolysed. This assumption is said to be supported
by the separation from the reaction products of both methods 4 and 5 of the cyclic
products (XXIX) l-aceto-3,5-dinitro-l,3,5-triazacyclohexane and (XXX) 1-aceto-
3,5,7-trinitro-l,3,5,7-tetrazacyclo-octane which may also be obtained from hexamine
by treating it with nitric acid and acetic anhydride:

N0 2
y CH 2 I

/ \
2 N— N N— N0 2

I I

H 2 C CH 2

\ N /

I
COCH3

XXIX

The reaction products (in methods 4 and 5) also include other cyclic N-acetyl
derivatives: (XXXI) l,5-diaceto-3,7-endomethylene-l,3,5,7-tetrazacyclo-octane and
(XXXII) 1 ,5-diaceto-3,7-dinitro-l ,3,5,7-tetrazacyclo-octane :

N0 2

CH 2 N CH 2

III
CH3CON CH 2 NCOCH3

I I I
CH 2 N CH 2

XXXI

Under the influence of acetic anhydride and nitric acid compound (XXX) gives
(XXV) (Marcus and Winkler [68]).

The transient formation of hexamine in method 4 was confirmed by Winkler,
Gillies and Williams [60] (see equation (45)) who examined the reaction mechanism
of cyclonite preparation by method 4. They found that hexamine dinitrate is formed
at 35°C as an intermediate product. At the same time nitric acid is evolved hence
nitrolysis of hexamine dinitrate may occur, in other words the mechanism of cyclo-
nite formation would be similar to that of direct nitration of hexamine with nitric
acid.

CH 2 -

-N— CH 2

2 NN

NN0 2

CH 2 -

-N— CH 2

COCH3

XXX

CH 2 -

-N CH 2

CH3CON

NCOCH3

CH 2 -

-N — CH 2

N0 2

XXXII

HETEROCYCLIC NITRAMINES U7

This observation also appears to explain why an excess of paraformaldehyde
should have a harmful influence on the yield of cyclonite. In particular, Winkler
noticed that the addition of formaldehyde to a hexamine solution in aqueous acetic
acid causes the decomposition of hexamine which proceeds at a rate depending on
the ratio of formaldehyde to hexamine.

The formation of octogen in method 5 will be discussed below.

OCTOGEN

NOj

I
CH 2 — N— CH 2

I I

2 N— N N— N0 2

CH 2 — N— CH 2

I
N0 2

XI, m.p. 276-277°C

Octogen, known in Britain as HMX, i.e. cyclotetramethylenetetramine, 1,3,5,7-
tetranitro-l,3,5,7-tetrazacyclo-octane, is a white crystalline substance which occurs
under several polymorphous forms differing from one another in specific gravity
and sensitiveness to impact. Usually octogen is obtained in /? form, which is the
least sensitive to impact (Table 21).

McCrone [69] found that octogen is polymorphic and exists in four forms I {fi\
II (a), III (y) and IV (§). The first three are stable at room temperature, but HMX-IV
transforms very readily.

HMX-I can be prepared in different ways. Wright et al. [70] recommend prepara-
tion by cooling a hot saturated solution of octogen in acetone or acetonitrile. HMX-II
was prepared [70] by dissolving HMX in hot 70% nitric acid: crystals of HMX-II
precipitate on cooling. When HMX-I was dissolved by warming in 50% acetic acid,
HMX-III precipitated on cooling. HMX-III can also be obtained by steam distilling
off the solvent from a hot solution of HMX in water-saturated cyclohexanone.
When HMX-I was sublimed at 180°C, the product was HMX-IV.

All the crystalline forms were examined by Wright et al. [70]. They examined
X-ray diffraction pattern, infra-red spectra and dielectric constants. They concluded
that the HMX polymorphs are actually lattice-caged conformational modifications.

Octogen, like cyclonite is insoluble in water and non-hygroscopic. Its solubility
m organic liquids is similar to that of cyclonite. Octogen and cyclonite are almost
alike in chemical reactivity. They differ only in that octogen is more resistant to the
action of sodium hydroxide than cyclonite. This reaction forms the basis of one of
the methods of separating octogen from cyclonite. It consists in heating the mixture
of octogen and cyclonite with a hydroxide solution under such conditions that the
latter is decomposed while the former remains unaltered.

118

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Table 21
Properties of octogen (according to Rinkenbach [71])

Properties

Specific gravity
Relative sensitiveness to

impact

(cyclonite=180)
Stability of crystalline form

Form

1.87

1.82

1.77

325

stable

metastable

labile

Another method of separating octogen from cyclonite depends on the difference
in solubility of the two substances. Octogen is more easily soluble in 55% nitric
acid or 2-nitropropane than cyclonite. The mixture is warmed in nitric acid and
filtered and a mixture enriched in octogen is crystallized from the filtrate. Cyclonite,
being soluble in 2-nitropropane, is then extracted from the mixture with solvent,
while octogen remains undissolved. Octogen can be purified by crystallization from

70% nitric acid.

A mixture of cyclonite and octogen, rich in octogen, is best prepared by method
5, i.e. from hexamine nitrate, ammonium nitrate and acetic anhydride. A particularly
large amount of octogen is formed when the product is prepared by Bachmann's
method at temperatures ranging from 73 to 78°C and the mixture is then cooled to
25°C and filtered. This mixture m.p. 191-202°, contains approximately 10% of

octogen.

The ratio of cyclonite to octogen produced by method 5 was examined by Epstein
and Winkler [72]. They found that in general reduction in the amount of ammonium
nitrate used for the reaction reduces the yield of cyclonite and increases that of octogen.

The results of these experiments are given in a graph (Fig. 27) which shows that
the largest amount of octogen is formed when two moles of ammonium nitrate
react with one of hexamine.

Bachmann et at. [64] also described the conditions under which the substance (XII)
(DPT) can be prepared in a yield of about 20% from hexamine and nitric acid, in
the presence of acetic anhydride and acetic acid, at temperatures between 15 and
30°C. On nitrolysis with nitric acid in the presence of ammonium nitrate and acetic
anhydride at 60-65°C, this substance gives octogen in 80% yield.

As an explosive octogen is superior to cyclonite in that its ignition temperature
is higher (an explosion ensues in 5 sec at 335°C while with cyclonite this occurs at
260°C). The chemical stability of octogen is also superior. In a vacuum, at 120°C
in 40 hr octogen evolves 0.4 cm^ of gas (cyclonite 0.9 cm*) at 150°C it evolves 0.6 cm*
of gas (cyclonite 2.5 cm 3). Thus, at 150°C octogen possesses a stability of the same
order as trinitrotoluene or picric acid.

According to A. J. B. Robertson [23] decomposition of octogen at temperatures
above 280°C occurs as a monomolecular reaction. Activation energy £=52.7 kcal,
log fi= 19.7.

HETEROCYCLIC NITRAMINES

119

£0

1 1 — I

MUar ratios

• Final yield

Acetic Nitric Acetic /-/examine
acid acid anhydride

104 5-2 20 i
194 5-2 20 i

i i

12 3 4 5 6 7 8
Moles ammonium nitrate per mole hexamlne

Fig. 27. Effect of ammonium nitrate on relative yields of cyclonite and octogen,
according to Epstein and Winkler [72].

In explosive power octogen is somewhat less powerful than cyclonite, its lead
block expansion being 450 cm 3 whereas that of cyclonite is 500 cm 3 .

By virtue of these properties, the presence of octogen in cyclonite is not very
harmful. Octogen, however, is not used independently as an explosive, being em-
ployed solely as a substance accompanying cyclonite.

HOMOCYCLONITE

N0 2
I
N— CH 2

/ \

H 2 C N— N0 2

H 2 C

CH 2

s N /

N0 2
XXXIII

120 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Homocyclonite (Homohexogen) is an homologue of cyclonite with a 7-membered
ring. Wright and Myers [43] prepared it by the nitrolysis of compound XXXIV
(which arises as a result of the action of formaldehyde and ammonia on ethylene-
dinitramine):

N0 2 N0 2

I I

N— CH 2 CH 2 — N

/ \ / \

H 2 C N— CH 2 — N CH 2

I I I I

H 2 C CH 2 H 2 C CH 2

I I

N0 2 N0 2

XXXIV
m.p. 205°C (decomp.)

NITRO DERIVATIVES OF MELAMINE

Melamine is now a very important 'chemical product since it is a starting substance
for the manufacture of plastics. The presence in melamine of a triazine ring, as in
cyclonite, and of three amino groups induced attempts to nitrate this substance in
order to obtain an explosive nitramine.

Whitmore and Cason [73] examined the mechanism of the direct nitration of
melamine with nitric acid in the presence of acetic anhydride at 5°C and obtained
an explosive product (I). By nitrating a triacetyl derivative of melamine with nitric
acid in the presence of acetic anhydride at 20-25°C Cason prepared product (II).
This substance hydrolyses, losing N 2 0, which results in the formation of a stable
compound (III)

NH 2 NHN0 2

I I

N N HNO3 N N

H 2 N— C C— NH 2 (CH3CO)2 ° HO— C C— NHN0 2

I (decomp. at 228°C)

(CHjCOahO

NHCOCH3 NHN0 2 OH

N N HNO3 N N -N z O N N

H3CCOHN— C C— NHCOCH3 (CH3CO)2 ° HO— C C— OH HO— C C— OH

II III

(decomp. at 248°Q

HETEROCYCLIC NITRAMIN ES 121

Atkinson and Whitmore [74] elucidated the structure of these compounds. They
showed that compound (I) is N,N'-dinitroammeline, compound (II) nitroammelide
and compound (III) cyanuric acid. The authors also showed that fuming nitric acid
at 20-25°C transforms triacetylmelamine into N-nitro-N',N"-diacetyl melamine (IV)

NHN0 2

•°\

N N

I I'

CH3COHN— C C— NHCOCH3

(decomposition at 300°C)

All these substances decompose without melting. Only dinitroammeline (I)
has explicit explosive properties.

NITROSAMINES

Trimethylenetrinitrosamine (TMTN) 1,3,5-trinitrosohexahydro-sym-triazine, 1,3,5-
trinitroso-l,3,5-triazacyclohexane (I) and dinitrosopentamethylenetetramine (DNPT)
or l,5-endomethylene-3,7-dinitro-l,3,5,7-tetrazacyclo-octane or 3,5-dinitroso-l,3,5,7-
tetrazabicyclo [3,3,1] nonane (II)

I
/ N \ CH 2 N CH 2

H 2 C CH 2

I I

ON— N N— NO

ON— N CH 2 N— NO

\ c / CH 2 N CH 2

I II

m.p. 105-107°C m.p. 206-207°C

Trimethylenetrinitrosamine (I) was first described in 1881 by F. Mayer [75] who
suggested the structure (I). Duden and Scharff [76], Bachmann and Deno [77], and
Aubertein [78] gave detailed descriptions of its preparation and chemical properties.
Finally, Ficheroulle and Kovache [79] worked out the mechanism of its production
on a semi-commercial scale.

Dinitrosopentamethylenetetramine (II) was first obtained by Griess and Harrow
[80]. Formulae of these two substances (I) and (II) were proposed by Cambier and
Brochet [81] and Duden and Scharff [76].

They are formed by the action of nitric acid on hexamethylenetetramine at a
low temperature (max +8°C). Depending on the pH of the solution, the compounds
(I) or (II) are obtained, viz.: at pH = l-2, trimethylenetrinitrosamine (I) is formed
whereas at p H = 3-6, dinitrosopentamethylenetetramine (II) is formed.

122

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The yield of both reactions is 65-70%. According to Aubertein [78] substance
(I) may be prepared in a yield 84% of theoretical.

The explosive properties of this compound (I) may be of particular interest.

Trimethylenetrinitrosamine

Physical properties. The specific gravity of the substance is 1.508. It is spar-
ingly soluble in water. Simecek and Dolezel [82] report the following figures for its
solubility, expressed as the amount in grammes dissolved per 100 g of solvent (Table
22):

Table 22

20°C

40°C

60 °C

Water

0.2

0.3

0.6

Ethyl ether

0.8

1.2 at 34°

Toluene

1.4

2.3

4.4

Methyl alcohol

4.3

7.7

18.1

Acetone

68.5

139.7

254.5

According to Medard and Dutour [83] when molten the substance mixes with
trinitrotoluene and gives an eutectic consisting of 58% of trinitrotoluene and 42%
of trimethylenetrinitrosamine. The eutectic melts at about 55 °C.

The thermochemical properties of this compound are of great interest. As early
as 1896 Delepine [84] determined its heat of formation (—AH { ) and found it to be
negative. This observation was confirmed in later work by the same author and by
Badoche [85] as well as in more recent experiments by Medard and Thomas [86].
According to the latter the heat of combustion of the compound (I) —AH V = 557.17
kcal/mole, hence its heat of formation is —AH { =1\.\ kcal/mole, i.e. 408 kcal/kg.
The heat of detonation was found to be 850 kcal/kg.

Chemical properties. Trimethylenetrinitrosamine (I) decomposes explosively
under the influence of concentrated sulphuric acid at room temperature. At a low
temperature it is hydrolysed to form trimethylenetriamine sulphate (III)

/ N \
H 2 C CH 2

+ 6H 2 so 4

NH 2 HSO©
H 2 C CH 2

ON— N

N— NO

n© I
H 2 SO©H 2 N

i© +3NOHSO©

NH 2 HSO©

\ /
X CH 2

N CH 2

III

According to Simecek [87] trimethylenetriamine undergoes further decomposi-
tion to a Scruff's base:

HETEROCYCLIC NITRAMINES 123

NH 2

H?C CH 2 © ©

@ 2 | | @ 2 -> 3CH 2 -NH 2 ?± 3CH 2 =NH 2

H 2 N NH 2

\ /
X CH 2

Sodium hydroxide causes slow decomposition in the cold and rapid in the hot,
with evolution of formaldehyde, nitrogen and ammonia. Rapid decomposition
also occurs in boiling water and slow decomposition occurs in water at room
temperature.

Oxidation of (I) leads to cyclonite. According to the work of Brockmann, Down-
ing and Wright [12] oxidation with a solution of hydrogen peroxide (30%) in nitric
acid (99%) in the ratio of 1 mole (I) to 82 moles of nitric acid, 3 moles of H 2 2
and 3.7 moles of H 2 0, at -40°C, gives dinitro-nitrosamine (IV) as an intermediate:

y CH 2

/ \
ON— N N— NO

| |

H 2 2

^CH 2

2 N— N N— N0 2

1 1
H 2 C CH 2

\ N /

1
NO

CH 2

/ \

„ - 2 N— N N— N0 2

h 2 ° 2 > ! 1

H 2 C CH 2

\ N /

1
NO

HNO3

HNO3 H 2 C CH 2

\ N /

1
N0 2

The yield of cyclonite in this reaction is 74%.

Explosive properties. The apparent density od trimethylenetrinitrosamine (I) is
0.84, according to Medard and Dutour [83]. The same authors give the following
relationship between density and the compressing pressure:

pressure

density

kg/cm 2

170

1.10

340

1.23

680

1.37

1020

1.44

1700

1.525

2380

1.57

3000

1.59

Complete detonation is obtained:

at a density of 0.85 by 0.30 g of mercury fulminate
at a density of 1.20 by 0.40 g of mercury fulminate
at a density of 1.40 by 0.50 g of mercury fulminate
at a density of 1.57 by 2.5 g of mercury fulminate

According to these authors sensitiveness to impact is of the same order as that
of trinitrotoluene.

Simecek and Sramek [88] give the following table for sensitiveness to impact
in the drop test, using a 5 kg weight:

124 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Weight falling

from a height of

of explosions

100

The value of the heat of detonation was reported earlier (p. 120).
According to Medard and Dutour the lead block expansion is 125.5 (taking
picric acid as 100). The rate of detonation at a loading diameter of 30 mm is as
follows :

Density Rate

m/sec
0.85 5180

1.00 5760

1.20 6600

1.40 7330

1.50 7600

1.57 7800

Charges of molten and solidified material at a density of 1.42 give a rate of
detonation of between 7000 and 7300 m/sec.

In air the substance takes fire fairly easily and burns regularly.

The same authors examined the rate of detonation of a molten and solidified
eutectic comprising 58% of trinitrotoluene and 42% of substance (I), and obtained
a value of approximately 7000 m/sec.

Medard and Dutour [83] made a detailed investigation of the stability of the
substance (I). At room temperature test samples of the substance remained ap-
parently unaffected for 6 years. Marked decomposition occurred with rising temper-
ature, beginning at about 150°C; at 160°C nitric oxides are evolved. Rapid heating
causes immediate decomposition at 300°C and at 200°C decomposition occurs
after 2 minutes.

The substance is exceptionally sensitive to the action of acids. When mixed
with picric acid, for example, it undergoes violent decomposition after 2 hours'
heating at 60°C. At 100°C decomposition ensues in 10-15 minutes. A mixture
with trinitrotoluene is decomposed at 85°C.

The molten substance may react with such metals as iron, copper, aluminium.
Thus, despite the fact that substance (I) is a powerful explosive, only slightly sensi-
tive to impact, its low stability even in the presence of traces of substances with
an acid reaction gives little promise for its practical use.

Dinitrosopentamethylenetetramine

Dinitrosopentamethylenetetramine (II) is used as a gasifiable product for the
production of porous plastics and rubber (e.g. Unical ND, Vulkacel BN).

HETEROCYCLIC NITRAMINES 125

LITERATURE

1. F. Lenze, according to H. Kast, Spreng- u. Ziindstoffe, Vieweg & Sohn, Braunschweig, 1921.

2. G. F. Henning, Ger. Pat. 104280 (1899).

3. G. F. Henning, Ger. Pat. 298412, 298539, 299028 (1916).

4. G. C. V. Herz, Brit. Pat. 145793 (1921); U.S. Pat. 1402693 (1922).

5. G. C. Hale, /. Am. Chem. Soc. 47, 2754 (1925).

6. P. Terpstra, Z. Krist. 64, 150 (1924).

p 7. R. Hultgren, /. Chem. Phys. 4, 84 (1936).

8. G. Edwards, Trans. Faraday Soc. 49, 152 (1953).

9. A. Majrich, Mem. artill. franc. 14, 127 (1935).

10. T. UrbaNski and B. Kwiatkowski, Roczniki Chem. 13, 585 (1933).

11. T. UrbaNski and I. Rabek-Gawronska, Roczniki Chem. 14, 239 (1934).

12. F. J. Brockmann, D. C. Downing and G. F. Wright, Can. J. Research 27 B, 469 (1949).

13. M. DELEPINE, Bull. soc. chim. France [4], 9, 1025 (1911).

14. E. Vernazza, Atti reale accad. sci. Torino 70, 1, 404 (1934/35).

15. J. SlMECEK, Chem. Listy 51, 1323, 1699(1957); XVI Congres International de la Chimie Pure
et Appliquee II, p. 310, Paris, 1957.

16. F. SOMLO, Z. ges. Schiess- u. Sprengstoffw. 35, 175 (1940).

17. S. Epstein and C. A. Winkler, Can. J. Chem. 29, 731 (1951).

18. W. H. Jones, /. Am. Chem. Soc. 76, 829 (1954).

19. M. Avogadro, Mem. artill. franc. 10, 875 (1931).

20. T. UrbaNski and J. Pillich, Wiad. Techn. Uzbr. 43, 79 (1939).

21. M. Tonegutti, Chimica e industria 17, 517 (1935),

22. A. Haid, F. Becker and P. Dittmar, Z. ges. Schiess- u. Sprengstoffw. 30, 66, 105 (1935).

23. A. J. B. ROBERTSON, Trans. Faraday Soc. 45, 85 (1949).

24. T. UrbaNski and W. Krawczyk, Wiad. Techn. Uzbr. 45, 490 (1939).

25. F. TABOUIS, M. ORTIGUES and P. Aubertein, Mem. poudres 33, 59 (1951).

26. T. UrbaNski and W. Malendowicz, Roczniki Chem. 18, 856 (1938).

27. M. Tonegutti, Z. ges. Schiess- u. Sprengstoffw. 32, 93 (1947).

28. L. Medard, XXVII Congres Chimie Industr.,Brussel, Chimie et industrie, No. special, 81 (1954).

29. A. Ya. Apin and Yu. A. Lebedev, Dokl. Akad. Nauk SSSR 114, 819 (1957).

30. T. UrbaNski, Przemysl chem. 20, 117, 179 (1936); Z. ges. Schiess- u. Sprengstoffw. 33, 41, 62
(1938).

31. H. Kast, Angew. Chem. 35, 72 (1923).

32. T. UrbaNski and T. Galas, Compt. rend. 209, 558 (1939).

33. P. Laffitte and A. Parisot, Compt. rend. 203, 1516 (1936).

34. M. A. Cook, The Science of High Explosives, Reinhold, New York, 1958.

35. A. Izzo, Riv. artiglieria e genio 373 (1932).

36. K. K. Andreyev, Thermal Decomposition and Burning of Explosives (in Russian), Gosenergo-
izdat, Moskva-Leningrad, 1957.

37. Poudreries reunies de Belgique, Soc. An., Belgian Pat. 488943 (1949).

38. Technical Report P.B. 925, General Summary of Explosive Plants, D.A.G. Krummel, Duneberg,
Christianstadt, U.S. Dept. of Commerce, Washington, 1945; Technical Report P.B. 262, RDX
Manufacture in Germany, U.S. Dept. of Commerce, Washington, 1945; W. De C. Crater,
Ind. Eng. Chem. 40, 1627 (1948).

39. A. H. Lamberton, Quart. Rev. 5, 75 (1951).

40. G. F. Wright, Gilman's Organic Chemistry, Vol. IV, p. 983, J. Wiley, New York, 1953.

41. E. L. Hirst, a. Carruthers, W. J. Dunning, J. K. N. Jones, H. D. Springall et al., unpub-
lished, according to A. H. Lamberton [39].

126 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

42. A. H. Vroom and C. A. Winkler, Can. J. Research 28 B, 701 (1950).

43. G. S. Myers and G. F. Wright, Can. J. Research 27 B, 489 (1949).

44. W. J. Chute, D. C. Downing, A. F. McKay, G. S. Myers and G. F. Wright, Can. J. Re-
search 11 B, 218 (1949).

45. K. SINGH, J. Sci. Ind. Research (India) A 15, 450 (1956).

46. L. Berman, R. H. Meen and G. F. Wright, Can. J. Research 29 B, 767 (1951).

47. W. J. Chute, A. F. McKay, R. H. Meen, G. S. Myers and G. F. Wright, Can. J. Research
27 B, 503 (1949).

48. K. W. Dunning and W. J. Dunning, /. Chem. Soc. 1950, 2920, 2925, 2928.

49. P. P. Karpukhin and W. N. Chetyrkin, Trudy Kharkovsk. Khim.-Technol. Inst. Kirova
4, 143 (1944); Chem. Abstr. 42, 5918 (1948).

50. W. J. Dunning, B. Millard and C W. Nutt, /. Chem. Soc. 1952, 1264.

51. V. Gilpin and C* A. Winkler, Can. J. Research 30 B, 743 (1952).

52. M. Kirsch and C. A. Winkler, Can. J. Research 28 B, 715 (1950).

53. T. UrbaNski, unpublished (1936).

54. F. Grottanelli, Chimica e industria 20, 801 (1938).

55. W. H. Simmons, A. Forster and R. C. Bowden, Ind. Chemist 24, 530, 593 (1948).

56. T. UrbaSski and K. Szyc-LewaSska, Bull. Acad. Polon. Sci., ser. chim. 6, 165 (1958).

57. W. P. Binnie, H. L. Cohen and G. F. Wright, /. Am. Chem. Soc. 72, 4457 (1950).

58. R. W. Schiessler and J. H. Ross, Brit. Pat. 595354 (1947); U.S. Pat. 2434230 (1948).

59. E. Aristoff, J. A. Graham, R. H. Meen, G. S. Myers and G. F. Wright, Can. J. Research
27 B, 520 (1949).

60. A. Gillies, H. L. Williams and C. A. Winkler, Can. J. Chem. 29, 377 (1951).

61. Technical Report P.B. 4272, Hexogen Manufacture in Bobingen, U.S. Dept. of Commerce,
Washington, 1945; CIOS XXXII 8, G.m.b.H. zur Verwertung Chemischer Erzeugnisse, Fabrik
Bobingen.

62. W. E. Bachmann and J. C. Sheehan, J. Am. Chem. Soc. 71, 1842 (1949).

63. W. E. Bachmann, W. J. Horton, E. L. Jenner, N. W. MacNaughton and C. E. Maxwell,
J. Am. Chem. Soc. 72, 3132 (1950).

64. w. E. Bachmann, W. J. Horton, E. L. Jenner, N. W. MacNaughton and L. B. Scott, J. Am.
Chem. Soc. 73, 2769 (1951); W. E. Bachmann and E. L. Jenner, ibid. 73, 2773 (1951); W. E.
Bachmann, E. L. Jenner and L. B. Scott, ibid. 73, 2775 (1951).

65. R. C. Brian, F. Chapman, A. H. Lamberton, C. Lindley, P. G.Owston,J.C. Speakman
and D. WOODCOCK, Chemistry & Industry 223 (1949).

66. F. Chapman, /. Chem. Soc. 1949, 1631 ; F. Chapman, P. G. Owston and D. Woodcock, ibid.
1949, 1647.

67. e.g. E. E. Roberts, unpublished.

68. R. A. Marcus and C. A. Winkler, Can. J. Chem. 31, 602 (1953).

69. W. C. McCrone, Anal. Chem. 22, 1225 (1950).

70. M. Bedard, H. Huber, J. L. Meyers and G. F. Wright, Can. J. Chem. 40, 2278 (1962).

71. W. H. RlNKENBACH in Kirk & Othmer Encyclopedia of Chemical Technology 6, 41, Interscience,
New York, 1951.

72. S. Epstein and C. A. Winkler, Can. J. Chem. 30, 734 (1952).

73. J. CASON (and F. C. Whitmore), J. Am. Chem. Soc. 69, 495 (1947).

74. E. Atkinson and F. C. Whitmore, J. Am. Chem. Soc. 73, 4443 (1951).

75. F. Mayer, Ber. 21, 2883 (1888).

76. P. Duden and M. Scharff, Ann. 288, 218 (1895).

77. W. E. Bachmann and N. C. Deno, /. Am. Chem. Soc. 73, 2777 (1951).

78. P. Aubertein, Mem.poudres 33, 227 (1951).

79. H. Ficheroulle and A. Kovache, Mem. poudres 33, 241 (1951).

80. P. Griess and G. Harrow, Ber. 21, 2737 (1888).

HETEROCYCLIC NITRAMINES 127

81. R. Cambier and A. Brochet, Compt. rend. 120, 105 (1895).

82. J. SimeCek and Z. Dole2el, in T. Urbanski, Chemie a Technologie Vybusin, Vol. Ill, p. 98,
SNTL, Praha, 1959.

83. L. Medard and M. Dutour, Mem. poudres 37, 19 (1955).

84. M. Delepine, Bull. soc. chim. France [3], 15, 1202 (1896).

85. M. Delepine and M. Badoche, Compt. rend. 214, 777 (1942).

86. L. Medard and M. Thomas, Mem. poudres 31, 173 (1949).

87. J. Simecek, Chem. Listy 51, 1323, 1699 (1957).

88. J. SimeCek and Sramek, in T. Urbanski, Chemie a Technologie Vybusin, Vol. Ill, p. 101,
SNTL, Praha, 1959.

Part 2
PRIMARY EXPLOSIVES: INITIATORS

CHAPTER I

GENERAL INFORMATION

The earliest mention of explosives are to be found in the alchemical writings of the
first half of the seventeenth century. Basilius Valentinus [1] described "explosive
gold" which was a complex explosive salt formed by dissolving gold oxide in
ammonia. At that time it was widely known that this substance is easily exploded
by heat or direct contact with a flame.

According to Romocki [2], in 1630, the Dutchman, van Drebbel was the first
chemist to investigate mercury fulminate, and "explosive gold". The first description
of the laboratory preparation of "mercury fulminate" is given in Kunkel's book
Laboratorium Chymicum published in 1690 [3]. This substance was described again
by Howard in 1799-1800 [4]. No further discoveries of other primary explosives
were made until the development of modern chemistry.

The invention of percussion compositions for igniting powders is usually attri-
buted to Forsyth [5]. In 1805 he employed pellets composed of a mixture of potassium
chlorate and combustible materials, coated with wax to render them safer to handle,
but even so they were still dangerous since the mixture was sensitive to friction.
The first ignition caps were invented in the early nineteenth century. In these caps
the ignitable composition was enclosed in a casing of brass or copper. This inven-
tion cannot be traced with any certainty to any individual. The literature on the
subject names several chemists including Bellot and Egg in 1815 [5].

The first application of mercury fulminate in ignition caps is attributed to Wright
[6] in 1823.

Prior to 1831, straws filled with blackpowder, or fuses, which were cords satu-
rated with a powder mixture, were used for igniting high explosive charges. The rate
of burning of these powder timetrains was very irregular and lead to a great many
accidents due to premature explosions. In 1831, a considerable advance was made

1129]

130 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

by Bickford [7], who invented slow burning (safety) fuses, which comprised an
inner core, filled with blackpowder, surrounded by layers of plaited jute.

Around 1860 Nobel [8] began to use a fuse with a small cartridge of blackpowder
at the end which produced a more intense ignition, for detonating nitroglycerine.
Subsequently he replaced the cartridges by caps and finally invented the detonator
[9], by elongating the cap and considerably increasing the charge of mercury ful-
minate. Similar work was done by Andreyevskii in Russia [10].

Mercury fulminate is easy to produce, has been known since earliest times
and is still widely used. The scarcity of mercury has however led to many attempts
to replace this substance by something else, in particular by substances containing
a different metal. Some success was achieved as a result of work of Will and Lenze
[5] in 1892 on the application of heavy-metal azides as initiating agents.

Primary explosives are a group of substances which are highly sensitive to the
action of mechanical shock and are readily ignited by direct contact with flame
or electric sparks.

Special care should be taken, therefore, during their manufacture. The danger
becomes greatest when the initiating substance is being transferred to the drier
after it has been washed with water. From that time onwards all possible safety
measures must be taken and strictly observed.

All operations must be handled from an adjacent room or at least from behind
a strongly built wall or an armoured shield, progress being observed through
a sight-glass or by television. Weighing and pouring materials etc., should also be
carried out from behind a shield.

A plant used for drying, grinding, sieving, weighing and stirring primary explo-
sives must be designed to minimize friction, and power units should be located in a
separate building.

In such a plant, conditions are particularly favourable for the accumulation of
static electricity, due to friction between the crystals themselves, and between the
crystals and parts of the plant, even between the crystal and the air, during drying.
All the parts of the plant should therefore be well earthed. If they are manufactured
of non-conductive material, as for example, plastics, ebonite or leather, these materi-
als should contain conductive substances such as graphite or aluminium dust, to
help dissipate the static electricity generated.

If the rule that all the parts of the plant must be earthed is not observed, ac-
cumulation of static electricity may occur, and under these circumstances, initiators
only need an electric spark to start them burning. Some of them, for instance lead
styphnate, are particularly susceptible to ignition by sparks.

It is also desirable to cover the floors of any buildings in which initiators are
housed with a conductive material. In order to minimize friction, the flooring
material should be soft. Hence polyvinyl chloride or rubber containing aluminium
powder or graphite, laid on earthed metallic tapes, are used. A floor that conducts
electricity facilitates the escape of static electricity and also removes any danger
which may arise from the building up of static charges on operators. According

INITIATORS— GENERAL INFORMATION 131

to Freytag [11], a man walking on an insulated floor, covered, for example, with
a woollen rug, amasses a static charge of 14,000 volts. It is obvious that a person
so charged could, on approaching an earthed instrument, cause a spark capable
of igniting a sensitive initiating compound.

Since accumulation of static electricity is favoured by dryness, moderate humi-
dification of the atmosphere increases its conductivity, and thereby decreases the
danger in handling initiating materials.

Recent technical literature discusses new methods of minimizing or even prevent-
ing the build-up of static electricity by using salts of radioactive elements to ionize
the air introduced into the factory buildings.

It is also very important to destroy any explosive substances which may be
entrained in liquid wastes and wash-water.

They can be settled out and subsequently precipitated. They should be destroyed
by chemicals which decompose them. Mercury fulminate, for example, is decom-
posed by a solution of thiosulphate, and lead azide by dilute nitric acid plus sodium
nitrite.

LITERATURE

1. Basilius Valentinus, according to Feldhaus, Z. ges. Schiess- u. Sprengstoffw. 4, 258 (1909)

2. S. J. V. Romocki, Geschichte der Sprengstoffchemie,Bd. I, Oppsnhsim (Schmidt), Berlin
1895.

3. Kunkel, Laboratorium Chymicum, Hamburg, 1690.

4. Howard, Phil. Trans. 90, I, 204 (1800).

5. According to H. Kast, Spreng- u. Zundstoffe, Vieweg & Sohn, Braunschweig, 1921.

6. Wright, Phil. Mag. 62, 203 (1823).

7. Bickford, Brit. Pat. 1659 (1831).

8. A. Nobel, Brit. Pat. 2359 (1863); 1813 (1864).

9. A. Nobel, Brit. Pat. 1345 (1867).

10. According to A. G. Gorst, Porokha i vzryvchatye veshchestva, Oborongiz, Moskva, 1957.

11. H. Freytag, Raumexplosionen durch statische Elektrizitat, Verlag Chemie, Berlin, 1938.

CHAPTER II

FULMINIC ACID AND ITS SALTS

FULMINIC ACID

Fulminic acid Cfe=NOHis a gaseous, highly toxic substance with an odour re-
sembling that of hydrogen cyanide. It is isomeric with other acids cf the same
empirical formula HCNO. The chief of these is cyanic acid HCNO, which is
obtainable only in the form of its salts; free cyanic acid is unstable. The action
of inorganic acids on cyanates leads to the evolution of cyanic acid which hydro-
pses to form carbon dioxide and ammonia:

NaCNO + HC1 -> HCNO + NaCl
HCNO + H 2 -+ C0 2 + NH 3

The treatment of cyanates with organic acids (oxalic acid, for example) in a non-
aqueous medium, for instance by grinding the two together, does not result in the
decomposition of cyanic acid, but in the formation of a trimer, cyamelide, to which
the formula of the trioxymethylene derivative is attributed (I)

/ C \
o o

I I

HN=C C=NH

\o/

. I

When distilled, this substance undergoes depolymerization, and isocyanic acid,
a stable liquid with the structure (II), is formed at a temperature of 0°C:

NH=C=0
II

Its salts also exist in the form of a trimer, cyanuric acid (III), which is produced
on heating the salts of isocyanic acid with acetic acid. Esters of cyanuric acid undergo
isomerization when heated and are converted into esters of isocyanuric acid (IV):

[132]

FULMINIC ACID AND ITS SALTS 133

OH O

c! C

/ C \ / \

N N HN NH

II I ^ \ q

hc/V^oh </ViN

III IV

The free acid — a trimer of isocyanic acid — probably has the structure (IV). Cyan-
uric acids are of some importance as sources of initiating materials (cyanuric azide,
for example, p. 194).

The salts of fulminic acid differ basically from those of isocyanic acid and its
trimers.

The formula (V) attributed originally to fulminic acid, considers it to be an

oxime of carbon monoxide [7] ; a more recent interpretation suggests the formula

(Va):

C==NOH C±=NOH

V Va

The linear structure of the fulminate ion

e e e e

C==N— O or C^N— O

was recently confirmed by spectroscopic investigation of some salts of fulminic
acid.

Singh [1] has examined infra-red spectra of mercuric, silver and lead fulminates,
and Beck [2] those of sodium and potassium fulminates. The maxima 2147 and 1225
cm -1 were found to be characteristic of asymmetric and symmetric vibrations of the
0—N— C group, respectively. The maximum 1181 cm- 1 was assigned to the bending
frequency of the same group [1]. Beck also found that the transient formation of
an isomeric ion e N=C— O can occur on thermal decomposition of fulminates.

This kind of structure is also proved by the fact that chemically fulminic acid
behaves as an unsaturated compound, adding hydrogen chloride at a temperature
of 0°C, to form the crystalline chloroformoxime (VI):

H \

>C=NOH

CI
VI

A certain amount of crystalline isocyanylic acid is also produced. This substance
has no explosive properties and according to Wieland [3] has the structure of fur-
oxanedialdoxime (VII):

HON=CH CH=NOH

I !

C C

N N-»0

\o/

VII

134 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

When mercury fulminate is boiled with water, polymerization occurs with the
formation of salts of fulminic acid, which is probably a mixture of substances (VII)
and (VIII). This compound hydrolyses to produce formic acid and hydroxylamine :

H \

>C=NOH + 2H 2 -> HC1 + HCOOH + NH 2 OH

Thus, fulminic acid C=N— OH (or C==N— H) has been regarded as a substance
analogous to hydrogen cyanide C=N— H (or C=N— H). This analogy is also
borne out by the facility with which fulminic acid forms complex compounds with
for instance iron-sodium ferrofulminate, Na 4 Fe(CNO) 6 - 18H 2 0, analogous to
sodium ferrocyanide.

It must be emphasized that the analogy between fulminic acid and hydrogen
cyanide is very deep. Liquefied hydrogen cyanide (b.p. +26°C) has explosive pro-
perties. This is explicable by the fact that like acetylene it is a strongly endothermic
substance. The explosive decomposition of hydrogen cyanide proceeds theoretically
according to the equation:

HCN -> C + |N 2 + |H 2

The quantity of heat evolved amounts to 1017 kcal/kg, the volume of gases
evolved, V , is 830 l./kg, the temperature is 2250°C. This was first reported by Walker
and Eldred [4].

Wohler and Roth [5] proved that liquefied HCN can be exploded by means of
a blasting cap.

Wohler calls attention to the fact that some cyanides of heavy metals, e.g. mer-
curic cyanide, are highly sensitive to friction and impact and may initiate detona-
tion of liquefied hydrogen cyanide.

Liquefied or solid (m.p. — 14°C) hydrogen cyanide is also capable of polymeriza-
tion, a strongly exothermic reaction which may involve apparent explosion due to
local overheating. Local overheating may also induce a genuine explosion, particu-
larly in the presence of the cyanides of some heavy metals.

The unsaturated nature of fulminic acid accounts for its tendency to polymerize
and suggests that polymers constitute the brown impurity produced in the manufac-
ture of mercury fulminate. The ability of fulminic acid to polymerize is also proved
by the formation of cyamelide (I). It has also been established that an ether so-
lution of fulminic acid is converted into metafulminuric acid on standing :

HON NOH

L- h

i i

HC O

VIII

FULMINIC ACID AND ITS SALTS 135

MERCURY FULMINATE

Mercury fulminate (CfeNO) 2 Hg is an initiating material of the greatest impor-
tance. It is obtained very simply by treating a solution of mercuric nitrate with
alcohol in nitric acid. A method for preparing it was described in alchemical writings.
This reaction, together with its product has been studied by a number of chemists,
including Liebig [6], who gave an account of the elementary chemical composition
of fulminate in 1823. Nothing was known of its structure until Nef's suggestion [7]
in 1^94, that fulminic acid is an oxime of carbon monoxide. This structure was
subsequently supported for sodium fulminate by Wohler and Teodorowicz [8].
More recent investigations altered these views, as described above (p. 133). The
mechanism of reaction which results in the formation of mercury fulminate was
reported by Wieland [9]. Solonina [10] made a detailed examination of its properties
and its manufacturing technology.

According to Wieland, reactions between mercury nitrate, nitric acid and alcohol
leading to the formation of mercury fulminate proceed as follows:

(1) Oxidation of alcohol to acetic aldehyde

C 2 H 5 OH + HNO3 -> CH3CHO + HN0 2 + H 2

(2) Nitration of acetic aldehyde to nitrosoacetic aldehyde

CH3CHO + HN0 2 -> NO— CH 2 CHO + H 2

(3) Isomerization of nitrosoacetic aldehyde to isonitrosoacetic aldehyde

ON— CH2CHO -> HON=CH— CHO

(4) Oxidation of isonitrosoacetic aldehyde to the corresponding acid

HON=CH— CHO -> HON=CH— COOH

(5) Nitration of isonitrosoacetic acid to nitrolacetic acid

N0 2
HON=CH— COOH + HNO3 -+ HON=C— COOH + H 2

(6) Decarboxylation of nitrolacetic acid in methylnitrolic acid

N0 2
I
HON=C— COOH -> HON=CH— N0 2 + C0 2

(7) Decomposition of methylnitrolic acid into fulminic acid and nitrous acid

HON=CH— NO2 -> C=NOH + HN0 2

(8) Formation of mercury fulminate

2CfeNOH + Hg(N0 3 ) 2 -^ (G=NO) 2 Hg + 2HN0 3

This reaction gives the volatile by-products: ethyl nitrate, ethyl nitrite and
acetic acid, NO, N0 2 and C0 2 .

136 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

One hundred and forty-two parts of mercury fulmin ate ca n be obtained from
100 parts of mercury ., A considerable amount of oxalic acid, produced by oxidation,
remains in the solution, hence crude mercury fulminate is washed with distilled water.

Beside this basic method of manufacturing mercury fulminate, which is widely
practised, there are alternate processes. Angelico [11] recognized that mercury
fulminate is formed by treating a mercury solution in an excess of nitric acid with
a concentrated aqueous solution of malonic acid in the presence of a small amount
of sodium nitrate. The reaction results in a considerable rise of temperature, C0 2
evolution and the precipitation of the fulminate (L. W. Jones [12]).

Nef [7] showed that the mercuric salt of nitromethane (obtained by the action
of HgCl 2 on the sodium salt of nitromethane) decomposes when boiled with dilute
hydrochloric acid to produce mercury fulminate. In all probability the following
reaction takes place:

(CH 2 =NOO) 2 Hg -> (Ct=NO) 2 Hg + 2H 2

PHYSICAL PROPERTIES

Mercury fulminate consists of octahedral crystals, belonging to the orthorhombic
system with the axial relationship a : b : c=0.712 : 1 : 1.353 (Miles [13]). The pure
substance crystallizes into the form of white, silky needles.

Commercial mercury fulminate may be greyish, pale brown or white, the colour
depending on the method of its preparation. The reason for these different colours
has been the subject of many investigations. A white product is formed when a cer-
tain amount of hydrochloric acid or cupric nitrate or chloride is added to the
reaction mixture. A pale grey or a pale brown fulminate occurs on the application
of pure reagents without the above-mentioned admixtures. The grey fulminate
usually consists of very regular crystals (Fig. 28) whereas those of white fulminate
are less regular (Fig. 29) due to the presence of impurities of mercuric chloride, or
copper salts if copper has been added.

The grey or brown colour is usually uniform throughout the whole mass of
crystals, although there are cases when the colour occurs in certain places only,
forming stains or dyeing the edges of the crystals.

Kast [14] found that the grey mercury fulminate is the purest and contains 99.7-
99.9% of mercury fulminate, soluble in hydrochloric acid. The insoluble residue is
composed mainly of mercurous chloride which is probably derived from impurities
of the starting substances.

The white fulminate contains 99.3-99.4% of the pure substance. The insoluble
residue is again composed chiefly of mercurous chloride, but it also contains sub-
stances which turn dark under the influence of ammonia, hence the white product
is less pure than the grey one. The more hydrochloric acid is added to the reaction
mixture, the higher is the content of mercuric chloride.

It was believed at first that grey fulminate was contaminated with metallic mer-
cury, since it had been observed that when it is dissolved in certain solvents a resi-

FULMINIC ACID AND ITS SALTS

137

due of metallic mercury is obtained. Solonina [10] showed, however, that this residue
is produced from both kinds of fulminate and that the mercury which constitutes
an insoluble residue in ammonia and potassium cyanide solutions or in pyridine
does not occur in the fulminate crystals but is formed as a result of decomposition

Fig. 28. Crystals of "grey" mercury fulminate, according to Kast [17].

by reaction with the solvent. Thus the idea that the presence of mercury in the
crystals of grey fulminate causes the grey colour cannot be considered proven,
although a number of authors still believe in it (Wohler and Berthmann [15]). In
point of fact, a dark grey product, copiously contaminated with mercury, may be

Fig. 29.

Crystals of "white" mercury fulminate, according to Kast [17],

prepared by carrying out the reaction in dilute aqueous solution. The formation of
mercury fulminate is then accompanied by the reduction of mercuric nitrate to
metallic mercury. The same effect can be obtained by using an insufficient amount
o nitric acid or too low a temperature of reaction. Such contaminated mercury fulmi-
nate cannot be used since it amalgamates with the metal body of the caps, causing
err corrosion and may also form copper fulminate by reaction between the amalgam
m mercury fulminate. The course of the reaction itself may also influence the
°ht OUr ^ product - According to Wohler [16] the white coloured product may be
ootained if acetaldehyde is used in the reaction instead of alcohol.

J38 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

On the basis of all these experiments Kast [17] expressed the opinion that dif-
ferent colours are produced by differences in the size and shape of the crystals and
not by impurities. This however appears to be incorrect. There seems every reason
to believe that the grey and particularly the brown colour of mercury fulminate
are produced by the presence of organic impurities, i.e. resinous product of the
polymerization of fulminic acid (Marshall [18]). The white product, in Marshall's
opinion, contains the same impurities, but in disguised form.

Apart from the impurities that influence its colour, mercury fulminate may
contains a trace of mercuric oxalate, the presence of which was discovered by Shish-
kov [19] as early as 1856. Oxalic acid is always formed during the reaction as a by-
product resulting from the oxidation reaction and according to Solonina the amount
of oxalic acid formed is larger if hydrochloric acid is present in the reaction mixture.

Nicolardot and Boudet [20] found that mercuric nitrate may also be an impurity
of mercury fulminate.

According to various authors, the specific gravity of mercury fulminate is:

4.42 (Berthelot and Vieille [21])

4.394 (Solonina [10], product recrystallized from

an aqueous solution of sodium cyanide)
4.307 (Miles [13], product recrystallized from

an aqueous solution of ammonia)

According to Patry [22] the crystallized product has a lower specific gravity
(4.32) than the crude product (4.40).

The apparent density of loosely-poured fulminate depends to a great extent
upon the size and shape of the crystals. According to various authors it may range
from 1.22 to 1.60. A fine crystalline product has a low apparent density, a coarse
crystalline product a high one. These variations are of great importance when load-
ing caps with fulminate measured volumetrically. In detonators the density is
usually as high as 2.5.

The solubility of mercury fulminate in water is low. According to Holleman
[23], 100 ml of water dissolve:

at a temperature of g

12°C 0.07

49°C 0.176

100°C 8

Mercury fulminate crystallizes from water as a yellow coloured product contain-
ing \ H 2 (Shishkov [19]). It was believed at first that the yellow colour is due to
the presence of mercuric oxide resulting from the hydrolysis of the fulminate. It
is now considered that this colour should be ascribed to the formation of products not
fully defined and partly to the mercuric salts of metafulminuric acid (p. 134). Further-
more on boiling in water, the hydrolysis of mercury fulminate may be fairly extensive.

Mercury fulminate dissolves in alcohol rather more readily than in water. The
best solvent for fulminate is an aqueous solution of ammonia. At 30-35° a concen-

FULMINIC ACID AND ITS SALTS 139

trated aqueous solution of ammonia dissolves a fourfold amount of mercury ful-
minate, but at 60°C decomposition ensues with the formation of urea and guani-
dine. From an ammonia solution fulminate can be crystallized either by evaporating
off the ammonia, by diluting the solution with water or by acidifying in the cold
with acid (e.g. acetic acid). Mercury fulminate also dissolves in acetone saturated
with ammonia.

A good solvent for mercury fulminate is a mixture of a concentrated ammonia
solution with alcohol and water (Miles [13]). According to Singh [24] the best results
are obtainable by a mixture of the above components in the volume ratio 2:1 : 1 .

Mercury fulminate dissolves in an aqueous solution of potassium cyanide (Steiner
[25], Grigorovich [26]), to form a complex salt. According to Solonina the fulminate
is best precipitated from this solution by treatment with dilute nitric acid. Thus,
12 g of mercury fulminate, dissolved in a solution of 6 g of KCN and 30 ml of water
is diluted with water to 100 ml and treated carefully with 50 ml of nitric acid, i.e.
10 ml of acid, s.g. 1.40 diluted with water to 50 ml.

Pyridine is also a good solvent for mercury fulminate. 14.5 g of mercury fulmi-
nate may be dissolved in a 1 g of pyridine on moderate heating. The fulminate may
by recovered if the solution is poured into water. Large crystals, an addition
compound of mercury fulminate with pyridine, then separate. This compound
loses pyridine on drying.

Mercury fulminate also dissolves in many solutions of various salts, but in some
of them (e.g. potassium iodide, sodium thiosulphate) it undergoes rapid decomposi-
tion.

Majrich [27] established that ethanolamine or an aqueous solution of ammonia
with ethanolamine are good solvents for mercury fulminate. By dilution with water
or acidification of the solution, mercury fulminate is precipitated in a highly pure
form, suitable for further use.

Mercury fulminate crystals are not so hard as those of lead azide (Todd and
Parry [28]).

According to Yuill [29] the specific heat of mercury fulminate is :

at 110°C 0.119 cal/g

at 125°C 0.120 cal/g

Its thermal conductivity according to Belayev and Matyushko [30] is 0.00029.

CHEMICAL PROPERTIES

As previously stated, mercury fulminate is hydrolysed by heating in water;
in boiling water hydrolysis is very rapid. Farmer [31] noticed that on heating with
water under pressure, mercury fulminate undergoes decomposition to metallic
mercury. Marked decomposition also takes place on heating or standing for long
periods at room temperature in an aqueous solution of ammonia or potassium

140 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

cyanide, or in pyridine, i.e. solvents for fulminate. When purifying the fulminate
by crystallization, special care must therefore be taken to see that temperatures
during dissolution and precipitation are as low as possible and that the latter fol-
lows the former with the greatest possible speed.

After a 14 day immersion of mercury fulminate in an aqueous solution of potas-
sium cyanide, precipitation can be inhibited by the addition of nitric acid.

On boiling the solution takes on a violet colour. Dissolved in an aqueous solution
of ammonia, fulminate decomposes even after 12 hr. On boiling a pyridine solution
of fulminate, complete decomposition occurs.

Mercury fulminate is relatively resistant to the action of dilute acids, in particu-
lar to that of nitric acid, but concentrated acids cause decomposition. Thus, under
the influence of nitric acid decomposition occurs with evolution of NO, CO, acetic
acid and mercuric nitrate. Under the influence of concentrated hydrochloric acid
free fulminic acid is evolved (with an odour resembling that of hydrogen cyanide)
as well as the decomposition products: hydroxylamine hydrochloride, formic acid,
mercuric chloride (Carstanjen and Ehrenberg [32]; Scholl [33]). Mercury fulminate
explodes on direct contact with concentrated sulphuric acid.

Strong alkalis decompose mercury fulminate easily. Heating with aniline leads
to the formation of phenylurea, diphenylguanidine and metallic mercury (Steiner
[34]).

On treatment with phenylhydrazine, mercury fulminate undergoes reduction
to free mercury. The phenylhydrazine changes colour from olive, grey (at the mo-
ment when mercury is set free) to reddish-brown. Several hours after the addition
of alcohol and dilute sulphuric acid, a red-violet colour appears (Langhans [35]).
This reaction may be used for the qualitative detection of mercury fulminate.

Mercury fulminate undergoes rapid decomposition by the action of ammonium
sulphide to form mercuric sulphide. The fulminate dissolves in sodium thiosulphate,
according to the reaction:

(CNO) 2 Hg + 2Na 2 S 2 3 + 2H 2 0=HgS 4 6 + (CN) 2 + 4NaOH

This reaction may be used to determine fulminate quantitatively by back titra-
tion of the sodium hydroxide formed. It can also be used to destroy fulminate resi-
dues and waste material. The impurities in mercury fulminate (oxalate and nitrate)
are insoluble in thiosulphate.

Reactions with metals. When mercury fulminate is boiled with water containing
metallic suspensions, the majority of metals (e.g. aluminium, zinc, copper), form
their fulminates and mercury is precipitated. Reaction can also occur at room
temperature, except with nickel. Other metals may be ranged according to increasing
reactivity: silver, tin, bismuth, cadmium, iron, lead, copper, zinc, brass, aluminium.
With aluminium, the reaction takes only a few hours, yielding a large amount of
A1 2 3 .

A similar reaction was observed when mercury fulminate was kept in contact
with metals in a damp atmosphere. Aluminium gave a white bloom after only four

FULMINIC ACID AND ITS SALTS

141

days. Iron and brass became slightly corroded in six days and zinc and lead in
14 days. The remaining metals, i.e. copper, cadmium, tin and silver showed no
change after 28 days (Langhans [36]).

Chemical stability and behaviour at high temperatures. Mercury fulminate under-
goes marked thermal decomposition even at 50°C. Rathsburg [37] found that
a sample of the technical product stored at 50-60°C for 6 months in a dry atmo-

Com me*^2-

50 days

Fig. 30. Comparison of the rate of decomposition of mercury fulminate and other
primary explosives at 75°C, according to Wallbaum [38].

sphere lost 3.6% in weight, while in a damp atmosphere 7.6% was lost. A recrystal-
lized sample, however, showed a greater stability, under the same conditions only
losing 0.2-0.5% in weight.

Heating mercury fulminate at 75°C causes distinct decomposition. According
to Wallbaum [38], during the first 10 days the loss in weight is not significant (ca.
0.12%), but after that the rate of decomposition begins to increase. After 46 days
the loss in weight reaches 8%. The shapes of the decomposition curve (Fig. 30)
clearly shows the increasing rate of decomposition. The curve for mercury fulminate
compares unfavourably with that for other initiating materials, i.e. lead and silver
azides, lead styphnate and tetrazene.

Hess and Dietl [39] found that 0.5 g samples of fulminate at 90-95°C undergo
partial decomposition to form substances with reduced explosive properties after
35-40 hr; they also showed that after 75-100 hr a brown-yellow powder of low
inflammability is produced.

Langhans [40] examined the changes which occur in white and grey mercury
fulminates during heating at 90°C under reduced pressure. After 100 hr a brown-
yellow, non-explosive substance was formed which retained the original crystalline
form. This substance was named mercury pyrofulminate by the author. It differs
from the starting substance by containing more mercury (76.4%) the empirical
formula being Hg 4 C 5 5 N 7 and by being insoluble in an aqueous solution of
ammonia and in pyridine.

142 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Aqueous solutions of organic acids such as formic, acetic, and oxalic, decompose
mercury fulminate, forming the corresponding mercuric salts. On the other hand, the
action of dilute inorganic acids involves decomposition with formation of C0 2 .

The decomposition of fulminate heated at 60-100°C under a reduced pressure
(5 mm Hg) was investigated in detail by Farmer [31]. At 80°C the brown fulminate
began to decompose with the evolution of gas after the induction period was over,
i.e. after 80 hr, whereas decomposition of the white fulminate began after a much
longer induction period, lasting 140 to 190 hr, after which it then proceeded with
the evolution of gas which was almost exclusively carbon dioxide. The rate of de-
composition of the white fulminate was higher at this stage than that of the grey
fulminate. The grey fulminate was transformed into a non-explosive product after
about 200 hr from the beginning of decomposition.

Farmer quotes the following figures for the time required for the production
of 5 cm 3 of gas by heating mercury fulminate (this corresponds to the decomposi-
tion of 11% of substance):

Brown fulminate: at 60.0°C 1227 hr

at 89.6°C 39 hr

White fulminate: at 60.0°C 2010 hr

at 89.6°C 67 hr

The difference in the behaviour of the two modifications was due largely to the
differences in crystal size. Fine fulminate, ground under water decomposes more
rapidly than a coarse crystalline product.

Farmer's experiments were repeated and extended by Garner and Hailes [41].
They examined the behaviour of mercury fulminate at about 100°C and came to
the conclusion that during the initial induction period, decomposition is accom-
panied by a slow evolution of gas at a constant velocity (linear decomposition). At
the end of this phase the main decomposition period begins with an increased rate
of gas evolution. The authors noticed that if the fulminate is finely ground, rapid
evolution of gas begins at once, without any initial period.

Garner and Hailes believe that decomposition proceeds by a chain mechanism
with a constant coefficient of branching.

Grinding increases the number of centres at which the reaction originates. Since
the decomposition reaction passes from one grain to another at the points where
thin surfaces are in contact, grinding which increases the surface area, would be
expected to have the effect described above.

A number of later authors, e.g. Prout and Tompkins [42], Vaughan and Phil-
lips [43] have confirmed that the thermal decomposition of mercury fulminate is
a chain reaction.

In a later work Garner and Haycock [44], came to the conclusion that the first
10% of the substance undergoing the accelerated decomposition breaks down
according to the cubic equation (1):

P-Po=ki(t-h) i ( J )

FULMINIC ACID AND ITS SALTS

143

where : p is the pressure of gaseous decomposition products during time t, p and
t are the pressure of gases and the time at the moment of completion of linear
decomposition (initial period) i.e. at the beginning of accelerated decomposition,
and k 3 is the constant for the reaction rate.

On the basis of this equation the authors draw the conclusion that after com-
pletion of linear decomposition, i.e. at the point p , t , spherical nuclei of decom-
position are formed.

iog(t-t )

1.4 1.6 1.8 2.0

2.6
2.2
1.8
1.4
1.0
0.6
0.2 ,„..-.-

Time, (win)

Fig. 31. Influence of various methods of treatment on the thermal decomposition of

mercury fulminate, according to Bartlett, Tompkins and Young [45]. A— pre-irra-

diated, £— crushed, C— aged.

Bartlett, Tompkins and Young [45] suggested a modified (2) equation of Garner
and Haycock:

p-Po^kiit-toy+kit (2)

Here k x is the constant for the linear reaction.

These authors studied the influence of the various methods of treatment on the
thermal decomposition of mercury fulminate crystals. This is shown in the graph
(Fig. 31).

Curve A represents the decomposition of mercury fulminate irradiated with
ultra-violet rays, curve B the decomposition of ground mercury fulminate,
and curve C the decomposition of ordinary (freshly-prepared) mercury fulminate.

According to Garner's calculations [44, 46] the activation energy of the acceler-
ated decomposition period is about 32 kcal/mole. Vaughan and Phillips [43] gave
the figure 25.4 kcal/mole and log 5=11.05.

Bartlett et al. found for the activation energy the figure of 27 kcal/mole and
for linear decomposition the figure of about 5 kcal/mole.

Singh [24] noticed that when heated for a few minutes at a temperature nearing
that of immediate decomposition mercury fulminate crystals undergo decomposi-

144

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

tion first along crystallographic planes (010) and (100) on the surface of the crystals.
Thus crystals heated, for instance, for 7 min 36 sec at 160°C undergo the cracking
shown in Fig. 32 (b). (The same crystal before heating is shown in Fig. 32 (a)).

(a)

(b)

(c)

Fig. 32. Crystals of mercury fulminate: (a) before heating, (b) after heating at
160°C for 7 min 36 sec, (c) after heating at 80°C for 96 hr, according to Singh [24].

Similar behaviour in silver azide crystals was observed by Bowden et al. [47].

When crystals of mercury fulminate are heated at lower temperatures the decom-
position reaction is localized mainly around lattice defects such as growth marks
on the surface of crystals or points where dislocations emerge at the surface (Fig.
32(c)).

The nuclei formed in crystals of mercury fulminate are yellowish-brown. They
probably consist of Langhans' mercury pyrofulminate [40].

The admixture of various substances acts in different ways on the decomposi-
tion rate of fulminate: inorganic acids accelerate decomposition; organic acids

FULMINIC ACID AND ITS SALTS

145

exert no influence whatever while organic bases sometimes accelerate decomposi-
tion considerably. E.g., the addition of 10% of centralite leads to the explosion of
fulminate heated for 3 hr at 80°C.

According to various authors, the ignition temperature of mercury fulminate
is 187-190°C on rapid heating; on heating at a rate of 5°C/min it is 160-165°C.
A test sample, when thrown onto a metallic surface heated to 215°C, explodes

Fig. 33. Cracks produced in mercury fulminate crystals by exposure to ultra-violet
light (500 x), according to Tompkins et al. [45].

after 5 sec, while an immediate explosion takes place on throwing a test sample
on a surface heated to 277°C. On a surface heated to 139°C the explosion occurs
after 39 min; at lower temperatures there is no explosion at all (data according
to Laffitte and Patry [48]).

The behaviour of mercury fulminate at high temperatures depends on its
purity. The recrystallized substance explodes immediately on a surface heated to
287°C.

Ignition depends on the size of the test sample used. A 10 g sample explodes
after about 7 hr of heating at 115°C, while smaller ones decompose completely at
132°C without exploding.

Evans and Yuill [49] investigated the ignition of mercury fulminate by the
adiabatic compression of the atmosphere surrounding it. They calculated that
a 50% response corresponds to a temperature of 530° C.

146 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The action of light. Mercury fulminate is sensitive to sunlight. Farmer found
that on exposure to the sun's rays for 5 weeks in summer a test sample of fulminate
showed considerable decomposition with gas evolution.

Berchtold and Eggert [50] and Eggert [51] established that mercury fulminate
(like silver fulminate and other primary explosives) is exploded when strongly
irradiated. To explode mercury fulminate a light with an intensity of 1.65 J/cm 2
is required (to explode silver fulminate— 2.1 J/cm 2 ).

According to Patry [22] fulminate darkens under the influence of irradiation
by a mercury arc. After a month test samples become almost black and the mercury
content increases from 70.8 to 71.3. As on heating, the crystalline structure remains
unchanged but the optical properties of the crystals undergo alteration.

According to Tompkins et al. [45] crystals of mercury fulminate exposed to ultra-
violet light (wavelength 2537 A) develop cracks, roughly parallel and about 10/*
apart (Fig. 33). The cracks occur preferentially at points of weakness and high
reactivity, i.e. at points where sub-grain boundaries emerge at the surface; it is
at such sites that photolysis is most likely to occur. A similar phenomenon of crack-
ing under the action of high temperature was described by Singh [24] (p. 144,
Fig. 32).

According to Kaufman [52] strong y-radiation (on average 10 5 r per hour) can
decompose mercury fulminate. Mercury fulminate evolves large amounts of gas
during irradiation and eventually loses its explosive properties.

EXPLOSIVE PROPERTIES

The densities of loading for mercury fulminate obtainable by applying differ-
ent pressures are tabulated below:

re kg/cm 2

Density

200

3.0

660

3.6

1330

4.0

3330

4.3

Under a pressure greater than 1660 kg/cm 2 mercury fulminate becomes "dead
pressed" i.e. takes fire with difficulty and burns without detonation.

According to Gorst [53] mercury fulminate pressed at 500 kg/cm 2 gives 3 %
misfires following flame ignition, at 600-650 kg/cm 2 it gives 5% and at 3000 kg/cm 2
almost 100% misfires. In detonators, therefore the fulminate is compressed under
pressures of 250-350 kg/cm 2 .

The sensitiveness of mercury fulminate to penetration by a striker depends
upon the pressing pressure. Under pressures up to 750 kg/cm 2 the sensitiveness to
perforation rises with the increase of pressing pressure. For pressures between
700 and 750 kg/cm 2 an optimal sensitiveness to perforation is observed, and at
still higher pressure, the sensitiveness decreases and finally disappears at about
2000 kg/cm 2 .

FULMINIC ACID AND ITS SALTS

147

The most important explosive property of mercury fulminate is that its burning
with a moderate rate, started by ignition, impact or friction, is easily converted
into detonation. According to Patry [22], a charge of mercury fulminate (of density
1.25) in glass tubes of diameter 3-12 mm, when ignited at one end by the flame
of a blackpowder fuse, burns over a certain distance (up to 30 cm at a small diam-
eter of tube, and about 3 cm in one of larger diameter) with a moderate rate, that
varies from 10 m/sec in the narrower tube to 20 m/sec in the wider tube. A deto-
nation wave then arises which moves with a rate that ranges between 2250 and
2800 m/sec (the higher rate occurs in the larger diameter tubes). If the tube is very
small, with a diameter less than 3 mm, no detonation wave may be produced.

o 4o so now

Fig. 34. Rate of burning of pressed mercury fulminate against initial temperature,
according to Belayev and Belayeva [54].

Belayev and Belayeva [54] studied the influence of the initial temperature on the
rate of burning of mercury fulminate, pressed into pellets. The results are shown
in Fig. 34. A 100 °C temperature increase in the mercury fulminate accelerated the
rate of burning approximately 1.7 times. The authors also investigated the rate of
burning of mercury fulminate at various temperatures, and expressed this relationship
by the following equations :

at 20°C
at 90°C
at 105°C

= 0.47+1.05/>
= 0.65+1.44 />
= 0.71 + 1.60/7

where K=the linear rate in cm/sec
p=the pressure in kg/cm 2

The rate of detonation, as reported by various authors, ranges from 2250 to
6500 m/sec, depending on the density and the diameter of loading.

According to Kast and Haid [55] the rate of detonation against the density of
mercury fulminate is related as follows:

»ensity

Rate of detonation

m/sec

1.25

2300

1.66

2760

3.30

4480

At the Chemisch-Technische Reichsanstalt [56] the figures given on p. 148 were
found for the rate of the detonation of mercury fulminate, pressed into detonators:

148

CHEMISTRY AND

TECHNOLOGY OF EXPLOSIVES

Density

Rate of detonation

m/sec

3.07

3925

3.96

4740

A mixture of mercury fulminate containing 10% of KC10 3 detonates at a density
of 3.16 with a rate of 4090 m/sec.

Patry [22] quotes the following rates of detonation in glass tubes (Table 23).

Table 23
The rate of detonation of mercury fulminate (according to Patry [22])

Diameter of tube
in mm

7.5

Density of loading

Rate of detonation in m/sec

0.85
2270

1.25
2700

1.0
2500

1.35
3000

1.45
3300

1.45
2700

According to other data the rate of detonation is 3975 m/sec at a density of
3.0 and 5400 m/sec at a density of 4.2.

234 cm 3 of gas are formed from the detonation of 1 g of mercury fulminate
in an atmosphere of nitrogen. The gas consists of:

co 2

0.15%

65.7% ,

N 2

32.25%

H 2

1.9% (Berthelot and Vieille [21])

This conforms to the decomposition equation:

(CNO) 2 Hg = 2CO + N 2 + Hg

The calculated heat of decomposition is 114.5 kcal/mole (mercury as liquid) or
99.1 kcal/mole (mercury as vapour). When based on 1 kg of fulminate the corre-
sponding figures are : 403 kcal and 349 kcal.

Kast [17] reports the following physical constants for mercury fulminate:

heat of formation -221.5 kcal/kg (- AH ( =62.9 kcal/mole)

heat of explosion 357 kcal/kg

volume of gases (To) 316 l./kg

temperature of explosion (f) 4350°C(?)
specific pressure (f) 5530 m

According to the same author expansion in the lead block is 110 cm 3 .

The sensitiveness of mercury fulminate to impact and friction is high. The height
from which a weight must be dropped to cause explosion is however not the same
in reports by different authors. E.g. Stettbacher [57] states that fulminate was exploded
by a 2 kg weight from a drop of 4 cm and nitroglycerine by the same weight from
a drop of 6 cm. According to R. Robertson [58] fulminate exploded on impact
from a drop only 10% of that necessary to explode picric acid.

FULMINIC ACID AND ITS SALTS 149

Fulminate is desensitized by the addition of water. When it contains 5% of
water it only partially detonates on impact; at 10% water content it decomposes
without explosion and 30% of water prevents its decomposition. These results,
however, refer to small scale tests, on a large scale they may be rather different.
Substances such as oils, glycerine and paraffin act similarly as desensitizers. Ful-
minate containing 20% of high-melting paraffin wax was used for the manufacture
of a detonating fuse, employed in Austria. Fulminate so phlegmatized is insensitive
to moderate impact and friction, but is detonated by a blasting cap.

TOXICITY

Mercury fulminate has a sweetish "metallic" taste. When administered orally
it is as poisonous as the majority of mercury compounds. Since, however, it is very
sparingly soluble in water its toxicity through contact with the skin is insignificant.
Nor is it toxic to lower plants, e.g. moulds often form on the moist bags in which
mercury fulminate is stored.

Poisoning from mercury does occur, however, among workers employed in the
first stages of production, when handling mercury.

MERCURY FULMINATE MANUFACTURE

There are numerous specifications for the technical manufacture of mercury
fulminate. They may be divided into three groups :

(1) methods in which a cold solution of mercury in nitric acid is employed;

(2) methods in which a warm solution of mercury in nitric acid is employed;

(3) methods, in which substances to bleach the product are included in addi-
tion to the essential raw materials: a mercury solution in nitric acid and ethyl alcohol.

The most important safety consideration in manufacturing fulminate is to ensure
that only very small quantities (usually limited to 500 g of mercury) are produced
in each reactor. Since mercury and mercuric nitrate readily react with metals the
manufacture has to be carried out in glass reactors (retorts or flasks). A very pure
product is thereby obtained and safety is ensured by the absence of metallic parts,
against which friction and impact would be dangerous.

(1) One of the earliest descriptions ' of the manufacture of mercury fulminate
given by Chevalier [59] is as follows: 300 g of mercury are dissolved in 3000 g of
cold nitric acid (54% HN0 3 , s.g. 1.34) and the solution is poured into a flask con-
taining 1900 g of 90% alcohol. After few minutes a vigorous reaction begins and
crystals are precipitated. On completion of the reaction 238 g and 158 g of alcohol
are added in turn. The fulminate is filtered off through a cloth filter and carefully
washed free of acid with water. The yield is 118-128 parts of fulminate per 100
parts of mercury, i.e. 83-90% of theoretical.

Too large a quantity of cold alcohol added during the reaction may cause the
formation of contaminated fulminate.

150

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(2) Chandelon's method [60] is more widely used. Here 1 part by weight of mer-
cury is dissolved in 10 parts of nitric acid (65% HN0 3 , s.g. 1.40) by moderate
heating; the solution, heated to 55%, is poured into a flask or retort with a capa-
city at least 6 times that of the liquid, containing, apart from the above mentioned
solution, 8.3 parts of 87% alcohol. The reactor is connected by a vent pipe with
stoneware jars and a cooling tower in which the vapours evolved during the reaction
are condensed.

After approximately 15 min the reaction begins, as shown by the evolution of
gas. The liquid soon begins to boil, and the reactor fills with white vapour. The

^ l "

q a,n l)ii,(l;ll,;n,i>,aiM;(i,C,ltM

Fig. 35. Fume-cupboard with long necked
flask (i) and heaters (2) for dissolving mer-
cury in nitric acid, according to Budnikov
et al. [61].

Fig. 36. Reaction flask,

according to Budnikov

et al. [61].

violence of the reaction may be suppressed by the addition of cold alcohol, but too
large an addition may excessively inhibit the reaction and may lead to the formation
of free mercury, which contaminates the product.

The fulminate is precipitated in the form of greyish needles. When the reaction
is complete, the reactor is allowed to stand for approximately 30 min while the
contents are cooled. 1-2 1. of water are then poured in and the liquid is decanted
from above the precipitated crystals. The precipitate is transferred to a cloth filter
and washed with distilled water until completely free of acid. The product is then
screened on a silk sieve (approximately 100 mesh/cm 2 ) which retains the larger
crystals. The smaller crystals are collected for direct use. The large ones are ground
under water, passed through the same sieve and added to the previous batch. 125
parts of fulminate are obtainable from 100 parts of mercury, which corresponds
to a yield of 88%.

The condensate that collects in the jars and tower consists of ethyl nitrate and
nitrite, acetaldehyde and unreacted alcohol. The vapours of these substances are
noxious so care must be taken that the apparatus is tightly closed.

(3) Solonina [10] gives two methods for the manufacture of mercury fulminate:
the first (a) produces a white fulminate and the second (b) a grey one.

FULMINIC ACID AND ITS SALTS

151

(a) 500 g of mercury is dissolved in 4500 g of nitric acid (62% HN0 3 , s.g. 1.383);
5000 cm 3 of 92-95% alcohol to which 5 g of copper dissolved in 5 g of hydrochloric
acid (23% HC1, s.g. 1.115) has been previously added, is poured into the solution.

Tower with coke

Fig. 37. A lay-out of the manufacture of mercury fulminate, according to Kast [17].

(b) 400 g of mercury is dissolved in 4200 g of nitric acid (62%) and 4000 cm^
of 96% alcohol is then added.

In both methods the solution of mercury in nitric acid is heated cautiously
to 50-56 °C and the added alcohol to 40°C.

152

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(4) According to a method employed in German factories (Kast [17]) 150 g
of mercury is dissolved in 1072 cm^ (1500 g) of nitric acid (65% HN0 3 , s.g. 1.40)
and 1500 cm' of 79.5% alcohol is added.

The temperature of the mercury solution lies between 40 and 55 °C, that of the
alcohol between 20 and 35 °C. A grey product is formed. In order to obtain a white

Fig. 38. General view of plant for manufacturing mercury fulminate in Atlas Powder

Company, according to Davis [62]. On the left, conical flasks with mercury nitrate

in nitric acid. On the right, reaction flasks.

product, a little concentrated hydrochloric acid, s.g. 1.185 is added to the alcohol
before the reaction.

The series of figures explains various stages of the manufacture. Figure 35 shows
a fume cupboard with long-neck flask for dissolving mercury (and small amount

&=-- Water

Fig. 39. Removal of mercury fulminate from the reaction flask to a rubber bucket,
according to Budnikov et al. [61],

of copper) in nitric acid with the addition of a small amount of hydrochloric acid.
Figure 36 shows a thick walled flask in which mercury fulminate is produced. The
lay-out of reaction retorts, condensation jars and a cooling tower is given on Fig. 37,
and a general view of the reaction flasks on Fig. 38.

FULMINIC ACID AND ITS SALTS

153

Figure 39 shows a method of transferring mercury fulminate from the reaction
flask to a rubber bucket. After preliminary washing in a vacuum filter, mercury
fulminate can be transferred for final washing in the apparatus shown in Fig. 40

To a trap

-4

Water

Fig. 40. Washing apparatus, according to Budnikov et al. [61]: 1— glass funnel,
.2 —mercury fulminate, 5— filter plate, 4— wash-water, 5— over-flow of wash -water.

in which water passes upwards through the fulminate layer, until the product
is free of acid (this usually requires 40-60 min).

Storage and the further, processing of mercury fulminate

Mercury fulminate prepared by one of the methods outlined above is tested to
check the acid content and the content of other mercury compounds. When tested
with litmus paper, the moist product "should give a neutral reaction. A 5 g test sample
mixed with 2 g of sodium hydrogen carbonate should not give a black or bluish tint.

Fig. 41. Vacuum drier of mercury fulminate, according to Budnikov et al. [61].

154

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Mercury fulminate containing about 50% of water is stored in glass jars or in
linen or linen-rubberized bags placed in bakelite, bakelized cardboard of paraffinized
boxes loosely covered with a lid. In this condition it can also be transported.

Fig. 42. Vacuum drier. Courtesy H. Orth G.m.b.H., Ludwigshafen-Oggersheim.

After accurate determination of its moisture content the fulminate may be used
for the manufacture of mixtures by the so-called wet method. Alternatively it can
be dried before mixing.

/S^&c

Barrier

Radiators

Fig. 43. Diagrammatic view of a drier with natural draught for mercury fulminate,
according to Vennin, Burlot and Lecorche [66].

Drying mercury fulminate has always been a difficult problem. For a long time
cylindrical vacuum driers (Fig. 41) were used in which the suction effect of the
vacuum pressure (100-200 mm Hg), normally held the lids tightly in place.

FULMINIC ACID AND ITS SALTS

155

In the event of an explosion the lids were blown off, and this sometimes prevent-
ed the destruction of the drier. The method was not very safe, however, especially
when drying other primary explosives. The fulminate was dried at a temperature
between 35 and 45°C.

Another former method for drying mercury fulminate in a warm air stream
made use of a natural draught (Fig. 43). The fulminate is spread in thin layers over

Fig. 44. Diagrammatic view of a German brattice drier for primary explosives with
screening of the product [67].

frames wrapped with silk and arranged on shelves (7) (2) (3). Each frame contains
approximately 1.5 kg of fulminate.

Brattice driers, incorporating a device for removing the dried material (Fig. 44)
have been used in Germany. The moist material, spread in a thin layer over
cloth stretched on wooden frames (7), is dried in warm air supplied via the ducting
(2) at a rate of about 0.5 m/sec. Next to the frame on which the material is dried
there is a tin funnel (3) with a built-in sieve in its base. This funnel is connected with
the ventilating duct by a flexible tube. Each frame (7) contains about 1.2 kg
of fulminate (dry substance). To dry a batch of fulminate at 65-70°C takes
1-1.5 hr.

After drying the material is cooled in a current of cold air produced by intro-
ducing cold water into the system that normally produces warm air. Since the dried
material may become electrically charged even though the frame on which it was
dried is earthed, to permit the charge to escape, a certain amount of water is sprayed
over it from a water feed nozzle located near the ducting supplying the drying air.
These operations take approximately 0.5 hr.

After cooling the frames (7) are tilted so that dried material is allowed to run
through the funnel (3) onto a vibrating sieve. A special device is employed to tap

156 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

the inclined frame, to ensure that pouring proceeds smoothly. Both the tilting and
tapping of the frames is carried out by mechanisms controlled from behind a wall
or an adjacent room.

The funnel (3) through which the dried material is poured onto the sieves is
made of graphitized leather, and the sieves of stainless wire are suspended on belts
also made of graphitized leather.

The sifted material is poured into a cylindrical vessel of graphitized plastic,
approximately 18 cm high and 6 cm in diameter, loosely covered with a lid of soft,
black rubber.

The floors in the building are covered with polyvinyl chloride containing graphite
or aluminium powder to make them conduct electricity. The air must not be too dry,
since this favours the build-up of static charges so it is kept humid by hanging cloth
saturated with water on the walls and by moistening the floors from time to time
or by using air-conditioning equipment.

Treatment of waste

Waste substances from the manufacture of mercury fulminate are :

(1) the spent liquor decanted from above the product;

(2) the sediment removed from the mercury fulminate by washing ("slime");

(3) condensed vapours.

(1) The spent liquor contains about 3% of dissolved substances, comprising
90-96% of oxalic acid and 3-6% of mercury in the form of mercurous nitrate.
The recovery of mercury in the form of mercurous chloride or mercury proper is
usually profitable. This is achieved by adding nitric acid to the spent liquor (1 1.
of concentrated hydrochloric acid to 50 1. of liquid). The sediment which is precip-
itated is separated by decantation and dissolved in concentrated hydrochloric acid,
using 10 kg of hydrochloric acid for 10 kg of sediment. Next a solution of pieces
of tin (1 kg) in hydrochloric acid (6 kg) is added and pure mercurous chloride is
precipitated.

Mercury may also be recovered in a metallic form by neutralizing the liquor
with milk of lime and dissolving the precipitated sediment in hydrochloric acid,
recovering the mercury by electrolysis or by displacement with zinc.

(2) The "slime" is processed in a similar manner, viz. it is dissolved in hydro-
chloric acid and the mercury recovered from the solution by one of the methods
outlined above.

(3) The liquid in the jars and condenser tower may be distilled over sodium
hydroxide. The recovered alcohol can be recycled.

(4) Residues of mercury fulminate are destroyed either by dissolving them in
sodium thiosulphate or by covering them with quick lime and treating the mixture
with live steam.

FULMINIC ACID AND ITS SALTS

157

OTHER SALTS OF FULMINIC ACID

Among other fulminates, the silver salt, (CNO) 2 Ag, is of some importance.
It is prepared in a way similar to mercury fulminate, by the action of alcohol on
a silver solution in nitric acid. Silver fulminate, however, is of little value as an
explosive since silver is an expensive raw material. Detonators of silver fulminate
were employed only in the Italian Navy.

The other fulminates are of no practical value. They are prepared from mercury
fulminate either by reacting the metal amalgam with a suspension of mercury ful-
minate in water (this is applicable to the majority of metals, including the alkali
metals, or simply by the action of the metal itself (e.g. zinc or thallium) which dis-
places mercury from mercury fulminate (also in water). For example, chips of thal-
lium, zinc, or copper are allowed to stand for some time in a suspension of mercury
fulminate in water, the corresponding metal fulminate is gradually formed.

Table 24
Comparison of the properties of fulminates

Fulminate

Initiation temperature

Sensitiveness to impact,

°C

work in kgm/cm 2

Sodium

215

Potassium

225

Calcium

195

165

Strontium

205

170

Barium

220

175

Cadmium

210

110

Copper(II)

205

110

Manganese

215

150

Thallium

110

Silver

170

Rosenberg [63] investigated the properties of sodium, potassium, calcium,
strontium, barium, cadmium, cupric, copper, manganese, thallium and silver fulmi-
nates and compared them with mercury fulminate. Some of this results are shown
in Table 24.

Table 25

Smallest amount

which will cause

Initiated high explosive

detonation of

Tetryl

Picric acid

TNT

Mercury fulminate

0.29

0.30

0.36

Silver fulminate

0.02

0.05

0.095

Cadmium fulminate

0.08

0.05

0.11

Copper fulminate

0.025

0.08

0.15

Thallium fulminate

0.30

0.43

158

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Martin [64] examined the initiating properties of certain fulminates, and found
that silver, cadmium and copper fulminates have stronger initiating properties than
mercury fulminate. Table 25 and Fig. 45 show the figures obtained by Wohler and
Martin [65], expressed as the smallest amounts of the fulminate of different metals
necessary to produce detonation of various high explosives.

Tetryl

\PicridMnitrotoluene\7rinUro-JrinUrv-
\ acid | I anisote | xylene

Fig. 45. Initiating ability of various priming explosives, according to Wohler and

Martin [65].

These compounds, however, have not been used in practice (apart from silver
fulminate, as mentioned above) due to the high cost of preparing them.

Fulminic acid can be regarded as the simplest oxime. A number of compounds
with an oxime group C=NOH can form salts which possess initiating properties,
for example salts of nitroformoxime (methylnitrolic acid) (I) which can be obtained
in a known way by the action of nitrous acid on nitromethane. Salts of form-
hydroxamic acid (II), particularly the mercuric salt, also possess initiating properties
[66]. Formhydroxamic acid can be obtained by the action of hydroxylamine on formic
acid esters or by oxidation of methylamine.

/N0 2

HC \

^NOH

LITERATURE

1. K. Singh, /. Chem. Soc. 1959, 459.

2. W. Beck, Chem. Ber. 95, 341 (1962).

3. H. Wieland, Ann. 444, 20 (1925).

4. M. Walker and D. N. Eldred, Ind. Eng. Chem. 17, 1074 (1925).

5. L. WOhler and I. F. Roth, Chem. Ztg. 50, 761 (1926).

6. J. Liebig, Ann. Chim. 2, 294 (1823).

7. J. U. Nef, Ann. 280, 263, 305 (1894).

8. L. WOhler and K. Teodorowicz, Ber. 38, 1345 (1905).

FULMINIC ACID AND ITS SALTS 159

9. H. Wieland, Ber. 42, 821 (1909).

10. A. Solonina, Z. ges. Schiess- u. Sprengstoffw. 5, 41, 67 (1910).

11. F. Angelico, Atti reale accad. Linzei, Roma 10, 476 (1901).

12. L. W. Jones, Am. Chem. J. 20, 1 (1898).

13. F. D. Miles, J. Chem. Soc. 1931, 2532.

14. H. Kast, Jahresber. Mil. Vers. Amt. 13, 79 (1908).

15. L. Wohler and A. Berthmann, Angew. Chem. 43, 59 (1930).

16. L. WOHLER, Ber. 38, 1351 (1905).

17. H. KAST, Spreng- u. Ziindstoffe, Vieweg & Sohn, Braunschweig, 1921.

18. A. Marshall, Explosives, Vol. II, Churchill, London, 1917.

19. L. Shishkov, Ann. 97, 54 (1856).

20. P. Nicolardot and J. Boudet, Bull. soc. chim. France 25,119 (1919).

21. M. Berthelot and P. Vieille, Compt. rend. 90, 946 (1880).

22. M. Patry, Combustion et detonation, Paris, 1933.

23. A, F. .HOLLEMAN, Rec. trav. chim. 15, 159 (1896).

24. K. Singh, Trans. Faraday Soc. 52, 1623 (1956).

25. A. Steiner, Ber. 9, 779 (1876).

26. P. Grigorovich, Zh. Russ. Khim. Obshch. 37, 113 (1906).

27. A. Majrich, Z. ges. Schiess- u. Sprengstoffw. 31, 147 (1936).

28. G. Todd and E. Parry, Nature 181, 260 (1958).

29. A. M. Yuill, Ph. D. Thesis, Cambridge, 1953, according to F. P. Bowden and A. D. Yoffe,
Fast Reactions in Solids, Butterworths, London, 1958.

30. A. F. Belayev and N. Matyushko, Dokl. Akad. Nauk SSSR 30, 629 (1941).

31. R. C. Farmer, /. Chem. Soc. 121, 174 (1922).

32. E. Carstanjen and A. Ehrenberg, /. prakt. Chem. 25, 232 (1882).

33. R. Scholl, Ber. 27, 2916 (1894).

34. A. Steiner, Ber. 7, 1244 (1874); 8, 518, 1177 (1875).

35. A. Langhans, Z anal. Chem. 57, 401 (1917); Z. angew. Chem. 31, 1, 161 (1918); Z. ges. Schiess-
u. Sprengstoffw. 13, 345, 406 (1918); 14, 300, 399 (1919); 15, 7, 23, 89, 219 (1920).

36. A. Langhans, Z. anal. Chem. 60, 93 (1921); Z. ges. Schiess- u. Sprengstoffw. 16, 105 (1921).

37. H. Rathsburg, Ber. 54, 3185 (1922).

38. R. Wallbaum, Z. ges. Schiess- u. Sprengstoffw. 34, 124, 197 (1939).

39. Hess and Dietl, Mitt. Art. -Gen. Wesen 18, 405 (1887).

40. A. Langhans, Z. ges. Schiess- u. Sprengstoffw. 17, 9-28, 122-159 (1922).

41. W. E. Garner and H. R. Hailes, Proc. Roy. Soc. (London) A 139,576 (1933).

42. E. G. Prout and F. C. Tompkins, Trans. Faraday Soc. 40, 488 (1944).

43. J. Vaughan and L. Phillips, /. Chem. Soc. 1949, 2741.

44. W. E. Garner and E. W. HAYCOCK, Proc. Roy. Soc. (London) A 211, 335 (1952).

45. B. E. Bartlett, F. C. Tompkins and D. A. Young, /. Chem. Soc. 1956, 3323.

46. W. E. Garner, Science Progr. 33, 209 (1938-39).

47. F. P. Bowden et al., Nature 178, 409 (1956).

48. P. Laffitte and M. Patry, Compt. rend. 193, 173 (1931).

49. J. I. Evans and A. M. Yuill, Discussion Roy. Soc, Initiation and Growth of Explosion in
Solids, Proc. Roy. Soc. (London) A 246, 176 (1958).

50. J. Berchtold and J. Eggert, Naturwiss. 40, 55 (1953).

51. J. Eggert, Physik. Bl. 12, 549 (1954).

52. J. V. R. Kaufman, Discussion Roy. Soc, Initiation and Growth of Explosion in Solids, Proc.
Roy. Soc. (London) A 246, 219 (1958).

53. A. G. Gorst, Porokha i vzryvchatyye veshchestva, Oborongiz, Moskva, 1957.

54. A. F. Belayev and A. E. Belayeva, Dokl. Akad. Nauk SSSR, 33, 41 (1941); 52, 507 (1946);
54, 1381 (1946).

160

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

55. H. Kast and A. Haid, Angew. Chetn. 37, 973 (1924).

56. Jahresber. chem.-techn. Reichsanstalt 8, 122 (1929).

57. A. STETTBACHER, Nitrocellulose 8, 3 (1937).

58. R. ROBERTSON, J. Chetn. Soc. 119, 1 (1921).

59. Chevalier, /. des connaissances usuelles, 1836, according to Dinglers polyt. J. 61, 191 (1837).

60. CHANDELON, Mem. soc. roy. sci. Liige, 1848, according to Dinglers polyt. J. 108, 21 (1848).

61. M. A. Budnikov, N. A. Levkovich, I. V. Bystrov, V. F. Sirotinskii and B. I. Shekhter,
Vzryvchatyie veshchestva i porokha, Oborongiz, Moskva, 1955.

62. T. L. DAVIS, The Chemistry of Powder and Explosives, J. Wiley, New York, 1943.

63. Rosenberg, Dissertation, Darmstadt, 1913; Chem. Ztg. 37, 933 (1913).

64. F. MARTIN, fiber Azide und Fulminate, Darmstadt, 1913.

65. L. WOHLER and F. MARTIN, Angew. Chem. 30, 33 (1917); Ber. 50, 586 (1917).

66. L. Vennin, E. Burlot and H. Lecorche, Les Poudres et Explosifs, Librairie Polytechnique
Beranger, Paris et Liege, 1932.

67. BIOS Final Report No. 1074, The Manufacture of 22 Rimfire Ammunition, Dynamit A.G. at
Nurnberg and Stadeln.

CHAPTER III

HYDRAZOIC ACID AND ITS SALTS

HYDRAZOIC ACID

Hydrazoic acid is a colourless liquid, of sharp, irritating odour with a boiling
point of about 37°C and a freezing point of about -80°C. It is highly poisonous,
its toxicity being of the order of that of hydrogen cyanide. Even the earliest investi-
gators (Curtius [1]) reported that the vapours of hydrazoic acid irritate the respi-
ratory tract, particularly the nasal mucosa, and that its aqueous solution burns
the skin. Stern [2] describes a serious case of poisoning with hydrazoic acid. Accord-
ing to the studies of Pravdin and Shakhnovskaya [3] and Shakhnovskaya [4], hydrazoic
acid interferes with the oxidation-reduction processes in the human body. Concen-
trations in air within the range 0.0005-0.007 mg/1. evoke marked symptoms of
intoxication. The toxic effect may be delayed, symptoms appearing the day fol-
lowing exposure. The salts of hydrazoic acid, e.g. sodium azide, are also highly
poisonous.

Hydrazoic acid in liquid form is very dangerous to handle owing to the ease
with which it explodes.

Its chemical structure has been the subject of many investigations. At first, it
was assumed to possess either a ring structure as an acid radical — N 3 (I) (Curtius [1]),
or a chain structure (II) (Thiele [5], Franklin [6]):

" N <l.

— N=N=N
N

X-Ray investigations of the structure of azides by Hendricks and Pauling [7],
and Frevel [8] confirmed Thiele's formula (II) and indicated that the chain structure
(III) has the extreme forms a and b :

e e e e

— N=N=N — N— N=N

Ilia IHb

The interatomic distances are equal, amounting on average to 1.15 A.

162 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The more recent investigations by Llewellyn and Whitmore [9] with strontium
azide, for example, have confirmed the above findings. The distances between the
atoms of nitrogen are 1.12 A, and between the furthest atoms of nitrogen in the two
N 3 groups and the atom of strontium: 2.63 and 2.77 A respectively.

Knaggs [10] found that in the case of cyanuric triazide the distance between
the pairs of nitrogen atoms is not the same, being 1.26 and 1.11 A respectively.
Examination of the Raman spectrum of sodium azide solutions has confirmed the
chain structure of hydrazoic acid (Langseth, Nielsen and Sorensen [11]). The same
conclusion is drawn from investigations of the absorption spectrum in the infra-red
(Herzberg et al. [12]).

For the gaseous state the interatomic distances and angles

were determined by investigation of the rotational spectrum (Eyster [13]) and the
microwave spectrum (Amble and Dailey [14]).

The structure of methyl azide, the simplest organic derivative of hydrazoic acid,
is:

H 3 C 135 °

i.e. a characteristic bending of valency bond occurs at the end of the N 3 system.
The structure of the — N 3 group has also been elucidated by Clusius and
Weisser [15], by reacting phenylhydrazine with nitrous acid labelled with the heavy
isotope 15 N:

C 6 H 5 NHNH 2 + Hi5N0 2 -> C 6 H 5 N 2 15 N

The phenyl azide so obtained was reacted with phenylmagnesium bromide to
give diazoaminobenzene, which, in turn, was split into aniline and ammonia. De-
termination of the isotope content of the reaction products proved that the only
linear formula that fits is that in which the extreme nitrogen is the isotope 15 N:

e e e

C 6 H 5 N=N=15N + C 6 H 5 MgBr -> }C 6 H 5 N=N— 15NHC 6 H 5 + iC 6 H 5 NH— N=15NC 6 H 5

The reduction of both of these compounds would give, in all, 1 mole of aniline,
1 mole of labelled aniline (C 6 H 5 i5NH 2 ) and 1 mole of ordinary ammonia. This is
in agreement with the results of the experiment.

Apart from compound (I) an insignificant amount of product (II) is formed
in which the heavy nitrogen occupies a different position C 6 H 5 — N= 15 N=N.

On the basis of their observations, the authors drew up the following reaction
mechanism for the formation of phenyl azide:

HYDRAZOIC ACID AND ITS SALTS 163

"NO

h»no, -tfCgHsN— NH 2 -> C 6 H 5 N=15N=N

C 6 H 5 NHNH 2 C

X C 6 H 5 NH— NH— 15N0 2 -> C 6 H 5 N=N=15N

If the group N 3 had a ring structure, the equivalence of the nitrogen atoms (2)
and (3)

N(2)

— N

(D\
N(3)

would involve the addition of phenylmagnesium bromide at both positions (2) and
(3) with the same probability. The addition products would be the following labelled
diazoaminobenzenes :

iC 6 H 5 N=N— NHC 6 H 5 + iC 6 H 5 NH— N=NC 6 H 5

if. $

+ iC 6 H 5 N=N— NHC 6 H 5 + iC 6 H 5 NH— N=NC 6 H 5

The reduction of these compounds would give \ mole of labelled aniline,
\\ moles of ordinary aniline, \ mole of labelled ammonia and \ mole of ordi-
nary ammonia, which is incompatible with the experimental results actually
obtained.

Spectrographs analysis of the derivatives of hydrazoic acid

Electronic spectrum. The N 3 group is a chromophore, and may be classified
as a chromophore with two cumulated double bonds (Braude [16])

N=N=>N

Such chromophores are characterized by absorption bands of low intensity.

Ethyl azide and azidoacetic acid N 3 CH 2 COOH may serve as examples of com-
pounds in which N 3 is linked to an organic residue by a covalent bond. In an alcohol
solution they show the following bands:

285 m# of very low intensity {E=ca. 20) and

220 van of higher intensity (E=ca. 150) (V. Henri [17], W. Kuhn and Braun [18], Mohler [19],

Sheinker [20]).

Sheinker and Syrkin [21] discovered a difference between the absorption spectrum
of the ion Nf in inorganic and organic azides (salts). They concluded that the transi-
tion from- the azide ion to the azide linked with alkyl group by a covalent bond
involves a change in the symmetry of the N 3 group due to the change in length of
the bonds N — N.

According to Jacobs and Tompkins [22] in the reflection spectrum sodium azide
and barium azide give a band with the maximum in the vicinity of 240 mfi and
202 ny* respectively. Evans and Yoffe [23] found a value of 248 m/i for potassium

164

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

azide. Tompkins and Young [24] established that potassium azide gives new absorp-
tion bands on irradiation in the ultra-violet at a low temperature (— 196°C).

For other azides (in the solid state) the following bands were found in the reflec-
tion spectrum:

Table 26

Azide

Bands
max.

Temperature

Author

T1N 3

275 ran

room temperature

Evans, Yoffe [23]

320 mn

room temperature

Evans, Yoffe [23]

425 mn

room temperature

Evans, Yoffe [23]

AgN 3

359 mn

-175°C

McLaren, Rogers
[25]

Hg 2 (N 3 )2

390 mn

room temperature

Deb, Yoffe [26]

Pb(N 3 ) 2

400 mn

room temperature

McLaren [27]

Infra-red spectrum. The spectrum of hydrazoic acid has been examined by a
number of authors. Thus, two bands were found by Eyster [28] :

2141 cm -1 asymmetric stretching vibrations and
1269 cm" 1 symmetric stretching vibrations.

Dows and Pimentel [29] present the following table for the frequencies of vibra-
tions of HN 3 and DN 3

Table 27

Approximate description

Frequency, cm"'

HN 3

DN 3

H-

- N stretching

3336

2480

N-

-N — N asymmetric stretching

2140

2141

N-

-N — N symmetric stretching

1274

1183

H-

-N — N bending

1150

955

N-

-N — N bending

672

638

N-

-N — N bending

522

498

The asymmetric and symmetric vibrations of methyl azide have frequencies of
2141 cm- 1 and 1351 cm- 1 respectively (Eyster and Gillette [30]). For a number of
aliphatic and aromatic azides Lieber et al. [31] found the figures 2114-2083 cm" 1
for asymmetric vibrations and 1297-1256 cm -1 for symmetric ones. Among the
other authors who have examined organic azides the investigations of Boyer [32] and
Evans and Yoffe [33] are noteworthy.

The most extensive investigations of the infra-red and Raman spectra of metal
azides were made by Gray and Waddington [34]. The Raman spectrum of azides
was studied by: Kahovec et al. [35], Kohlrausch and Wagner [36], and Sheinker
and Syrkin [21],

HYDRAZOIC ACID AND ITS SALTS

165

The results of the investigations of Gray and Waddington are tabulated below.

Table 28

Fundamental vibration frequencies of N 3 ion
(Gray and Waddington [34])

Salt

Frequencies, cm~i

Raman | Infra-red

LiN 3

1369

1277

2092

635

NaN 3

1358

1267

2128

639

KN 3

1344

1273

2041

645

RbN 3

1339

1271

2024

642

CsN 3

1329

1267

2062

635

Ca(N 3 ) 2

1380.5

1267

2114

638

Sr(N 3 ) 2

1373

1273

2096

635

Ba(N 3 ) 2

1354

1278

2123
2083

650
637

NH 4 N 3

1345

2041

661
650

The infra-red spectra of the heavy metal azides, which are the most interesting
because of their explosive properties, were investigated by Garner and Gomm [37],
Lecomte et al. [38], and the Raman spectra have been studied by Kohlrausch and
Wagner'[36], and by Deb and Yoffe [26]. The results are given in Table 29.

Table 29

Salt

Frequencies, cm -

Author

Raman

Infra-red

T1N 3

1941

636

Lecomte et al. [38]

AgN 3

2173

644
680

Lecomte et al. [38]
Lecomte et al. [38]

CuN 3

1337

2110

615

Deb and Yoffe [26]

1268

642

Lecomte et al. [38]

Hg 2 (N 3 ) 2

1300

2080

1322, 1273,1
647, 592 J

Deb and Yoffe [26]

Hg(N 3 ) 2

1313

2070

675, 644, 6421
584 J
630

Deb and Yoffe [26]

Pb(N 3 ) 2

1352

Lecomte et al. [38]

1350

1254

2006
2080

628

Kohlrausch et al. [35, 36]
Garner and Gomm [37]

1352

2040

Summing up the infra-red and Raman spectra data of various authors, Bellamy
[39] quotes the following frequencies:

for asymmetric vibrations
for symmetric vibrations

2160-2120 cm-i
1340-1180 cm-i

166

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The first of these bands has a frequency very similar to that of a diazo group
and of the C=N group in isocyanates and other compounds.

Hydrazoic acid is a weak acid like acetic acid. Mendeleyev [40] noted its acidic
properties. It reacts with zinc, iron, magnesium and aluminium, to form azides with
evolution of hydrogen (Curtius and Rissom [41]; Curtius and Darapsky [42]. A
small amount of ammonia is also produced, due to the reduction of hydrazoic acid.
According to Sofianopoulos [43] heated Al dust reacts with HN 3 to form alu-
minium nitride

6A1 + 2HN 3 -> 6A1N + H 2

Hydrazoic acid also reacts with copper, silver and mercury but in a different
way : it forms azides without loss of hydrogen and a considerable amount of hydra-
zoic acid is reduced to ammonia or hydrazine and free nitrogen. The reaction with
copper recalls the action of nitric acid on this metal.

Cxi + 3HN 3 -> Cu(N 3 ) 2 + N 2 + NH 3 (3)

[3Cu + 8HN0 3 -> 3Cu(N0 3 ) 2 + 2NO + 4H 2 0]

Hydrazoic acid, like nitric acid, oxidizes hydrogen sulphide to form sulphur:

HN 3 + H 2 S -> S + N 2 + NH 3 (4)

[2HN0 3 + 3H 2 S -> 3S + 2NO + 4H 2 0]

When mixed with hydrochloric acid it forms a solution resembling that of nitro-
hydrochloric acid in its properties and capable of dissolving noble metals, e.g.
platinum:

Pt + 2HN 3 + 4NHC1 -+ PtCl 4 + 2N 2 + 2NH 3 (5)

[3Pt + 4HN0 3 + 12HC1 -> 3PtCl 4 + 4NO + 8H 2 0]

Hydrazoic acid reacts with potassium permanganate. They mutually reduce each
other to evolve a mixture of nitrogen and oxygen. Nitrous acid oxidizes hydrazoic
acid with the evolution of nitrogen.

Hydrazoic acid and its salts give a deep red colour with ferric chloride resembling
that of ferric chloride with the salts of thiocyanic acid. This colour fades under the
influence of hydrochloric acid.

Liquid hydrazoic acid explodes on heating to 100°C in a tube (Dennis and
Isham [44]). According to R. Meyer and Schumacher [45] an explosion may also
occur either on rapid cooling, on filtering the liquid acid under vacuum or by passing
compressed oxygen into a vessel containing liquid hydrazoic acid. These authors
examined the decomposition of gaseous hydrazoic acid at temperatures ranging
from 306 to 330°C under pressures between 30 and 200 mm Hg. Decomposition
proceeds quantitatively according to the equation

3HN 3 -> NH 3 + 4N 2

The reaction is monomolecular. The half-life at 330°C is 12min. In a dilute
aqueous solution hydrazoic acid is stable and not liable to decompose even on long
boiling (Curtius [46]). However, a 17% aqueous solution of hydrogen azide can
probably detonate [47].

HYDRAZOIC ACID AND ITS SALTS 167

Gaseous hydrazoic acid is liable to non-explosive decomposition at a temper-
ature above 250°C. At 33°C the half-life is 12 min.

Hydrazoic acid is decomposed by ultra-violet radiation. In all probability the
decomposition proceeds gradually with the formation of free radicals (Beckman and
Dickinson [48]) :

HN 3 + hv -> HN + N 2

HN + HN 3 -> H 2 + 2N 2

HN + HN 3 -> H 2 N 2 + N 2

H 2 N 2 + HN 3 -> NH 3 + 2N 2

According to Stewart [49] hydrazoic acid decomposes under the influence of
active nitrogen to form the NH radical.

A number of investigations have been devoted to the thermal decomposition of
hydrazoic acid or to decomposition produced by electric discharge. Thus Rice and
Freamo [50] established that its thermal decomposition at 77°K leads to the formation
of a blue-coloured sediment. At a higher temperature, 148°K, it changes colour,
forming a white substance which has been identified as ammonium azide. They
suggested that the blue colour is caused by the presence of the free imino radical NH.

Mador and Williams [51] and Dows, Pimentel and Whittle [52] continued investi-
gations on the subject. According to the former the blue substance contains the
radicals NH and NH 2 while the latter believe that the radicals NH and (NH)* are
present. Spectrographic analysis by Pannetier and Gaydon [53] has confirmed the
presence of the electronically excited NH radical.

Using Norrish's "flash photolysis method" [54], Thrush [55] examined the de-
composition of hydrogen azide in the presence of an excess of inert gas. The absorp-
tion spectra characteristic of the radicals NH and NH 2 were observed. He therefore
suggested an alternative scheme for the decomposition of hydrazoic acid, different
from that proposed by Beckmann and Dickinson [56]:

HN 3 + hv -> NH + N 2

NH + HN 3 -> NH 2 + N 3

NH 2 + HN 3 -> NH 3 + N 3

2N 3 -> 3N 2

He also observed an absorption band at 2700 A which he attributed to the N 3
radical. According to Gray and Waddington [57] the initial step in the decomposi-
tion of HN 3 is nearly thermoneutral :

HN 3 -> HN + N 2 , —AH^ 5 kcal/mole

Becker, Pimentel and Van Thiel's [58] equally interesting study of the photolysis
of solid hydrazoic acid led to the assumption that the radicals NH, NH 2 and N 3
are formed during decomposition.

Franklin, Herron, Bradt and Dibeler [59] studied the decomposition of hydrazoic
acid under a reduced pressure, on ignition by means of Tesla discharges. The reaction

168

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

products included hydrogen, nitrogen, ammonia and undecomposed hydrazoic acid.
Similar results were obtained by Foner and Hudson [60].

Irradiation of an aqueous solution of HN 3 leads to the formation of hydroxyl-
amine according to the reaction.

HN 3 + H 2 -> NH 2 OH + N 2 (Glen [61])

According to Alekseyev [62] explosion of a mixture of HN 3 and hydrogen gives
ammonium azide and the intermediate compound N 2 H 5 N 3 .

The salts of hydrazoic acid, the azides, have solubilities similar to those of the
corresponding chlorides. Sodium azide dissolves in water. Silver azide does not
dissolve in water or in nitric acid, but dissolves easily in an aqueous solution of
ammonia. Lead azide, like lead chloride, is sparingly soluble in cold water but more
soluble in hot water; it is also soluble in ammonium acetate.

There are two methods for the manufacture of hydrazoic acid or its salts. One
is derived from a number of investigations by Curtius [1] and is based on the action
of nitrous acid on hydrazine :

NH 2 — NH 2 + HONO -> HN 3 + 2H 2 (6)

An excess of nitrous acid decomposes hydrazoic acid in accordance with the
equation :

HN 3 + HONO -> N 2 + N z O + H 2

(Seel and Schwaebel [63], G. Stedman [64]).

The reaction is quantitative and is used in practice for destroying waste azides.

Reaction (6) is particularly well suited for use in the preparation of the organic
derivatives of hydrazoic acid from the corresponding derivatives of hydrazine. Nitrous
esters may be employed, instead of the acid, e.g. ethyl nitrite in the presence of
sodium hydroxide

NH 2 — NH 2 + C 2 H 5 ONO + NaOH -> NaN 3 + C 2 H 5 OH + 2H 2 (7)

This method is preferable to reaction (6) since the sodium salt is formed instead
of the volatile and highly toxic hydrazoic acid.

The other widely used method, that of Wislicenus [65], is based on the action of
nitrous oxide on sodium amide. The reaction takes place in the following sequence:

2NH 3 + 2Na
NaNH 2 + N 2
NaNH 2 + H 2

300°C

190-230°C

-* 2NaNH 2 + H 2

-+ NaN 3 + H 2
-+ NaOH + NH 3

(8)
(9a)
(9b)

Reactions (9a) and (9b) proceed concurrently: nitrous oxide first reacts with
molten sodium amide to form sodium azide and water vapour. The latter then
reacts with another molecule of sodium azide, hydrolysing it with the formation
of sodium hydroxide and ammonium (9b).

The mechanism of reaction (9a) has been examined by isotope techniques using
nitrous oxide labelled with 15N (Clusius et al. [66, 67]). The course of the reaction

HYDRAZOIC ACID AND ITS SALTS 169

may be summarized by the equations :

-NaOH,-NH 3 X Nal5NNN (22 %>

2NaNH 2 + OiSNN : -<

x NaNi5NN (78%)

Among other reactions leading to the production of hydrazoic acid, the following
are noteworthy for theoretical reasons :

The formation of hydrazoic acid and its derivatives together with ammonia
from diazo compounds under the influence of hydrazine or its derivatives was
explained by Thiele [68]. At an intermediate stage a diazohydrazine, e.g.
C 6 H 5 N 2 NHNH 2 , is formed which then undergoes decomposition according to two
parallel reactions :

C 6 H 5 N 2 NHNH 2 -> C 6 H 5 NH 2 + HN 3

^ (10)

QH5N3 + NH3

Hydrazoic acid may be formed from hydrazine under the influence of oxidizing
agents. This was first noticed by Sabaneyev [69] who acted hydrazine with nitric
acid. According to Jannasch and Jahn [70] chlorates act similarly in an acidic medium
while Turrentine and Olin [71] found that hydrazine can be oxidized by electrolysis
to hydrazoic acid. On the other hand, according to Tanatar [72] a number of other
oxidizing agents such as chromic acid, permanganates, and hydrogen peroxide give
hydrazoic acid only in the presence of hydroxylamine. The reaction probably occurs
in the following way:

N 2 H 4 + NH 2 OH + 20 -> HN 3 + 3H 2 (11)

The preparation of inorganic azides was reviewed by Audrieth [73] and more
recently by Evans, Yoffe and Gray [74].

LEAD AZIDE

Neutral azide

Apart from mercury fulminate, lead azide is the most important primary explo-
sive.

Lead azide can exist in two allotropic forms: the a-form is orthorhombic, the
^-form monoclinic. (Miles [75], Garner and Gomm [37]). The crystallographically
stable modification is the a-form. It is prepared by rapidly stirring a solution of
sodium azide with a solution of lead acetate or lead nitrate.

The /?-form is prepared by slow diffusion of sodium azide and lead nitrate solu-
tions. This form has a tendency to revert to the a-form, either on standing at an
elevated temperature {ca. 160°C), or when crystals of the jff-form are kept in water
containing a crystal of the a-form (Azdroff [76]) or on contact with a lead salt solu-
tion (Miles [75]).

170

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The properties of both those forms, as reported by Gray and Waddington [77],
together with new figures for the density, are tabulated below.

Table 30

Species

Density

Unit cell size

No. of mole-
cules in unit
cell

Heat of form-
ation — AH{
kcal/mole

orthorhombic
monoclinic

4.71
4.93

6.628
5.090

11.312
8.844

16.256
17.508

12
8

115.5
115.8

The apparent density of the a-form is 1.2.

The earlier opinion that the jS-form is the more sensitive to impact appears to
be incorrect. This problem will be discussed more fully in the section on the explosive
properties of lead azide.

Lead azide is insoluble in an aqueous solution of ammonia. Acetic acid causes
its decomposition but it is soluble in water and concentrated solutions of sodium
nitrate, sodium acetate or ammonium acetate. There are fairly big differences of
solubility, depending on temperature.

Solonina [78] quotes the following figures for the solubility of lead azide:

in 100 ml of water

in 100 ml of concentrated
solution of NaNC>3

in 100 ml of concentrated
solution of CH 3 COONa

at 18°C
at 70°C
at 18°C
at 80°C
at 18°C
at 80°C

0.023 g
0.090 g
0.125 g
0.487 g
1.542 g
2.020 g

Owing to the difference between its solubility in the cold and in the hot, lead
azide may be recrystallized from water or from the solutions mentioned above.
Under these conditions the crystals are obtained in the form of long, colourless
needles.

Majrich [79] reports that lead azide dissolves in ethanolamine, but it is not
practicable to precipitate a pure form, suitable for commercial use by dilution of
this solution.

According to Wohler and Krupko's [80] data, recrystallization of lead azide
from water or from aqueous solutions is not free from hazard, since the salt often
explodes during crystallization. For this reason lead azide is not recrystallized in
practice. The phenomenon will be discussed later on p. 173.

The specific heat of lead azide, as reported by Yuill [81], is :

at 100°C
at 200°C
at 250°C

0.100 cal/g
0.117 cal/g
0.116 cal/g

McLaren [82] determined the thermal conductivity of pressed pellets of azides
and obtained a value of 4xl0"4 (c.g.s. units) at 45°C, the density of the pellets
being 3.62 g/cm*.

▼^

HYDRAZOIC ACID AND ITS SALTS 171

The V.D.H. hardness values of the crystals, according to Yuill [81] are

for a-lead azide: 114 kg/mm 2 at a load of 50 g

103 kg/mm 2 at a load of 20 g

for j?-lead azide: 65 kg/mm 2 at a load of 5g

Todd and Parry [83] quote figures for the a-form which prove that lead azide is
much harder than mercury fulminate.

Lead azide, like hydrazoic acid, is liable to undergo oxidation and reduction
reactions. It is partially decomposed by atmospheric oxygen to form free hydrazoic
acid, nitrogen and ammonia. This reaction is promoted by the presence of
carbon dioxide in the air. When boiled in water, lead azide undergoes slow decom-
position with the evolution of hydrazoic acid.

Lead azide is completely decomposed by the action of dilute nitric or acetic
acid in which sodium nitrite has been dissolved, and the products pass to
the solution. This reaction may be used for the destruction of lead azide wastes
and residues, using 15% nitric acid and 8% sodium nitrite.

In an aqueous suspension, lead azide is oxidized by eerie sulphate to form nitro-
gen:

Pb(N 3 ) 2 + 2Ce(S0 4 ) 2 -> PbS0 4 + Ce 2 (S0 4 ) 3 + 3N 2

This reaction may be employed for the quantitative determination of azide.

Long experience in the storage of blasting caps filled with lead azide has shown
that this substance reacts with copper or brass to form cupric azide, which is highly
sensitive to friction and impact. For this reason lead azide is compressed only into
aluminium and zinc cases.

When exposed to light lead azide soon turns yellow on the irradiated side. The
layer of changed substance protects the deeper layers from further decomposition
and thus irradiation does not entail changes in the explosive properties of the sub-
stance. However, as Wohler and Krupko [80] have shown, if the lead azide is subject-
ed to stirring during irradiation, decomposition may proceed too far.

Slow decomposition of lead azide takes place under the influence of ultra-violet
irradiation, as demonstrated by the investigations of Garner and Maggs [84] and •
Tompkins et al. [22, 85], but if the irradiation is very intense, explosion may occur,
as was shown by Berchtold and Eggert [86] and Meerkamper [87]. In another paper
Eggert [88] reported that the light intensity required to cause the explosion of lead
azide is 2.0 J/cm2 when the electrical energy of the flash is 240 J, and the half-life
of the flash 0.8 msec.

As shown by the investigations of a number of authors, irradiation of lead
azide (and other azides) with a-particles, X-rays and y-rays does not cause explosion
(Haiissinsky and Walden [89]; Gunther, Lepin and Andreyev [90]). However, it
produces a slow decomposition of lead azide, according to Kaufman [91].

Groocock [92] noticed that lead azide irradiated with y-rays at an elevated tem-
perature decomposed more rapidly than that which has not been irradiated.

172 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The thermal stability of lead azide is very high. At 75°C it loses approximately
0.8% of its weight during the first 4 days, after which further heating involves a
loss of 0.03-0.05% per week (Wallbaum [93] cf. graph in Fig. 30). At 115°C, in the
dark, it undergoes no changes for 24 hr unless the temperature reaches 170°C when
a distinct loss of weight takes place during that time. At temperatures above 200°C
decomposition is quite rapid, ranging from a few hours to several minutes, depending
on the temperature, and the substance loses its explosive properties. On the other
hand, in the light, decomposition may be observed even at 50°C.

Bowden and Singh [94] and later Bowden and McAuslan [95] using the electron
microscope, observed that on heating at a temperature above 120°C the separate
crystals of lead azide (like those of cadmium or silver azides), break down into fine
particles, approximately 10-5 cm j n <jj a anc } decomposition reaction takes place
chiefly on the newly-formed surfaces. This makes it evident that the thermal decom-
position of azides cannot be regarded as a surface reaction or a process occurring
within large crystals only; the whole mass is involved, due to crystal breakdown.

According to various authors, the ignition temperature of lead azide ranges from
327 to 360°C. When a test sample is dropped onto a metal plate instant explosion
ensues if the temperature of the plate is 380°C or higher. The ignition temperature
of lead azide is the highest ignition temperature of an explosive ever to have been
observed.

According to Sudo [96] lead azide prepared by the action of sodium azide on
an aqueous solution of lead acetate has a lower ignition temperature (332-336°C)
than that obtained by the action on a solution of lead nitrate (339-359°C).

According to Garner and Gomm [37] the activation energy of the thermal decom-
position of lead azide is 38.0 kcal/mole, assuming that the reaction can be expressed
by an equation of the form p = kt.

Evans and Yuill [97] investigated the ignition of dextrinated lead azide by the
adiabatic compression of air surrounding it, and estimated that the 50% explosion
level corresponds to a temperature of 990°C.

Bryan and Noonan [98] carried out similar investigations using helium and
estimated the minimum energy required to ignite lead azide with a 3 msec delay to
be 0.087 cal/cm 2 , i.e. less than that for blasting gelatine, 0.15 cal/cm 2 , PETN,
0.25 cal/cm 2 and tetryl, 0.33 cal/cm 2 .

Lead azide detonates with a high rate, amounting to 4500 m/sec at a density of
3.8 and 5300 m/sec at a density of 4.6 (Kast and Haid [99]).

According to the Chemisch Technische Reichsanstalt [100] the rate of detona-
tion of lead azide at a density of 2.75 is 3620 m/sec and at a density of 3.65—4700 m/sec.

The rate of detonation of a thin film of lead azide (0.1-0.5 mm thick) is 2100 m/sec
(Bowden and Williams [101]). Lead azide is less sensitive to impact than mercury
fulminate, but drop test figures quoted by various authors differ widely. Some of
them report a negligible difference between the two, while others state it is consider-
able (e.g. that azide requires 2-3 times the height of drop necessary to explode
fulminate). On the other hand, when mixed with pulverized sand lead azide is more

HYDRAZOIC ACID AND ITS SALTS

173

sensitive than mercury fulminate, which makes it evident that lead azide is more
sensitive to friction than fulminate. Its high sensitiveness to friction has been confirm-
ed by a number of accidents. Nevertheless, lead azide does not necessarily explode
when rubbed in a porcelain mortar. On the other hand, numerous cases of spon-
taneous explosion of lead azide are known, e.g. during pouring, weighing, drying
and even when simply left standing or left for crystallizing. It was formerly believed

start

haze and crystals,
(no explosion)

quiescent period,
(explosion)

"mat "

formation,

(no explosion)

Fig. 46. The diffusion growth of /?-lead azide [103].

that the /?-form is particularly liable to explosion when crystallized in the form of
long needles. Hawkes and Winkler [102], however, prepared crystals of the /?-form
3-4 mm long which did not explode on being crushed or broken.

As stated above, Wohler and Krupko [80] noticed that lead azide may explode
during crystallization. Miles [75] confirmed this and stated that spontaneous explo-
sion could occur during the growth period of crystals of /?-lead azide, i.e. when the
two solutions forming the crystals diffused slowly. When isolated, however, the
crystals (some of which had reached a length of 4 cm) were not found to be particu-
larly sensitive.

Rogers and Harrison [103] tried to determine the conditions governing this
phenomenon, i.e. the explosion during the growth of )S-lead azide. Their experi-
ments, which are illustrated diagrammatically in Fig. 46, were carried out in a
test-tube. Three solutions were carefully introduced so that they did not mix. The
bottom layer consisted of 20% lead nitrate (2cm3). The middle layer was 20%
sodium nitrate (1 cm^). The top layer was 10% sodium azide (2 cm^). Crystals of
lead azide formed in the sodium nitrate layer after i-^hr. They appeared to
start from the walls and spread inwards. A major explosion generally occurred in
the system after the crystals had been growing for 6-12 hr. A series of very small
explosions accompanied by "clicks" sometimes preceded the major explosion.

Explosion occurred under a number of conditions. For example, if a solution
was made up by mixing 5% lead nitrate (0.3 cm'), 2% sodium azide (0.3 cm 3) and

174

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

water (2 cm 3 ) no immediate precipitate was formed. Crystals appeared after ca.
\ hr, and spontaneous explosions occurred in certain cases. In all the experiments
where crystals exploded the common feature was that the initial concentration of
azide in the solution around the crystals was only slightly above the saturation

Fig. 47. Lead azide for use in detonators (75 x ) precipitated in the presence of dextrin

according to Davis [104].

value. No explosions were observed when the azide was precipitated rapidly, using
very much stronger solutions. A small rise in temperature e.g. of 10°C appears to
increase the probability of explosion.

Spontaneous explosions of lead azide also take place during crystallization from
saturated solution in ammonium acetate. A detailed study of this phenomenon has.
been made by Taylor and A. T. Thomas [105]. When the concentration of the solu-
tions and the temperature and conditions of cooling were carefully controlled, they
were able to predict the time at which spontaneous explosions occur. E.g. :

with a 1% lead azide solution in 5% ammonium acetate the explosion may be

obtained after 40-45 min,

Fig. 48. Lead azide precipitated in different conditions, according to Sudo [96] : (a) —
from high concentration of lead acetate (10%) and low of sodium azide(4%); (b) —
from low concentration of lead acetate (4%) and high of sodium azide (10%); (c) —
from low concentrations of lead acetate (4%) and sodium azide (2%); (d) — from high
concentrations of lead acetate (25%) and sodium azide (10%) in the presence of
gelatine (5%); (e)— from high concentrations of lead nitrate (25%) and sodium azide
(10%) in the presence of gelatine (5%) (very small crystal form).

(c)

HYDRAZOIC ACID AND ITS SALTS

175

(b)

. aaate-.._ aMira '

IOXj!5*X pfi.l^

(d)

176 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

with a 0.6% and 0.5% lead azide solution, the time is 225-255 min and ca. 480 min

respectively.

Taylor and Thomas have shown that spontaneous explosions are not associated
with the large crystals of lead azide that are formed : they filtered off large crystals
ca. 30 min before the predicted time of explosion and at the predicted time the
mother liquor exploded while the filtered lead azide crystals remained intact.

They also found that by adding small amounts of dextrin, polyvinyl alcohol or
other hydrophilic polymers explosion could be prevented. It is known that these
compounds are able to alter the crystal habit of several substances, including lead
azide.

Clearly, explosion is associated with the very early stages of crystallization.

According to Kaufman [91] spontaneous explosion can also take place during
the growth of a-lead azide crystals, e.g. when a supersaturated solution of lead
azide in ammonium acetate is seeded with crystals of the a-form. Spontaneous ex-
plosions have also been observed with mercuric azide and in some cases with cad-
mium azide.

According to Sudo [96] spontaneous explosion can occur during the formation
of lead azide from sodium azide and lead acetate, when the concentration of reacting
solutions is high (10% or more).

When manufacturing lead azide, efforts should be made to precipitate small,
highly regular, free flowing crystals of a length not exceeding 0.1 mm.

Considerable progress in the manufacture and application of lead azide was
achieved by the addition of dextrin to the solution in which it was produced, as
mentioned above.

The presence of dextrin in the solution favours the precipitation of tiny, equal
sized, rounded crystals. In Fig. 47 magnified crystals of dextrinated lead azide are
shown and in Fig. 48 crystals of ordinary lead azide are shown for comparison.
According to Sudo [96] gelatine exerts an influence similar to that of dextrin. In
this author's opinion when precipitated from a solution containing gelatine lead
azide is less sensitive to friction than that from solutions without added colloids.

The sensitiveness of moist lead azide is not much lower than that of the dry
product. According to Wohler and Krupko [80] a 30% water content does not
render the lead azide insensitive.

Yuill [81] investigated the sensitiveness of lead azide to impact at room tem-
perature and at -190°C. He found that 10-15% more energy is required for the
initiation of explosion by impact at a low temperature than that at room temperature
(Table 31).

The disadvantage of lead azide lies in the difficulty with which it is ignited by
a flame. For this reason it is usually mixed with lead styphnate i.e. a substance
particularly easy to ignite, or the charge of lead azide in a detonator is covered
with a layer of lead styphnate. Such a layer not only facilitates the ignition of the lead
azide, but also protects it against the action of carbon dioxide. It is also difficult
to ignite lead azide by an electric spark (Brown, Kusler, Gibson [106]).

hydrazoic acid and its salts

Table 31

initation of explosion of lead azide pellets by impact

177

Weight of pellets

Specific energy
cal/g

Energy
difference

20°C -190°C

cal/g

0.35
0.42
0.51

119

104

134
117
107

15
13

Lead azide passes very rapidly from burning to detonation. When used in very
small amounts, it is therefore capable of initiating detonation in other explosives
hence it is very suitable for use in detonators though it cannot be employed in caps*

Wallbaum [93] determined the minimum charges of several primary explosives
necessary for initiating the explosion of a 0.4 g charge of PETN, loosely poured
or pressed. His results are tabulated below.

It is characteristic of lead azide that even under a pressure as high as 2000 kg/cm2
it cannot be "dead pressed". This is a great advantage. In practice a pressure of
500-600 kg/cm2 is used.

Other data concerning the initiating properties of lead azide, as compared with
the other primary explosives, are given in Table 32.

Table 32
Initiation effects of primary explosives

Pressure on PETN, kg/cm*

2000

Pressure on initiator, kg/cm 2

2000

Primary Explosive

Lead azide (technical)
Lead azide (crystallized)
Silver azide
Mercury fulminate
Tetrazene
Lead styphnate

500

1000

2000

Minimum initiating charge, g

0.040

0.170

0.050

0.040

0.015

0.100

0.010

0.005

0.110

0.005

0.300

0.330

dead pressed

0.160

0.250

dead pressed

0.550

1.000

no detonation

According to Garner [107] lead azide decomposes in the following way:

^Pb 2 ® + N© -» £Pb + N 2 + N — 51 kcal (12)

£Pb 2 ® + N 3 e -+ £Pb + 2N 2 + 157 kcal (13>

According to the equation (12) one atom of nitrogen is expelled from the N e ion to
form two molecules of nitrogen by reaction with another Nf ion. Reaction (13) is
highly exothermic hence the decomposition of one N 3 group may involve that of

178 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

2-3 neighbouring N 3 groups. If these groups decompose simultaneously, the decom-
position of 22 ions Nf may ensue etc. Thus, the rapid transition of lead azide to
detonation may be accounted for by the fact that the decomposition of a small
number of molecules of lead azide may induce explosion in a sufficiently large
number of N^ ions to cause the explosion of the whole mass.

The decomposition of other salts of hydrazoic acid takes a similar course.

Basic azides

Hydrolysis of lead azide with water or a solution of sodium hydroxide yields
basic salts. They are also formed by the treatment of a lead salt solution with sodium
azide plus sodium hydroxide solution.

Feitknecht and Sahli [108] found five different basic salts, each of which exists
in several crystalline modifications as established by X-ray analysis :

I Pb(N 3 ) 2 -PbO (3 crystalline modifications)

II 3Pb(N 3 ) 2 -5PbO (3 crystalline modifications)

III 2Pb(N 3 ) 2 -5PbO

IV 2Pb(N 3 ) 2 -7H 2

V Pb(N 3 ) 2 -«PbO, where
«=from 4 to 9.

LEAD AZIDE MANUFACTURE

Lead azide is manufactured on a technical scale by the action of sodium azide on
an aqueous solution of lead nitrate. According to a description of manufacture in
the Wolfratshausen factory in Germany [109], the reaction is conducted in an open
reactor of stainless steel, provided with a jacket warmed by hot water and a stirrer
which may be lifted out of the reactor (Fig. 49). The reactor is emptied by tilting. Its
upper edge is therefore fitted with a spout so that the contents pour easily. The size
of the reactor is such that 4.5 kg of lead nitrate in the form of a 9-10% solution
can be used in each batch. This solution is poured into the reactor, warmed to 50°C
and neutralized with sodium hydroxide to a pH of about 4.0 (in the presence of
methyl orange) and 150 g of dextrin mixed with a small amount of water, is added.
The suspension or solution of dextrin in water should be decanted before use to
separate mechanical impurities, such as sand.

Next, 1.5 kg of sodium azide is added as a 2.7-3.0% aqueous solution. The
solution should be alkaline (50 cm 3 of solution should require 8-10 cm 3 1.0 N H 2 S0 4
for neutralization using phenolphthalein as indicator). If the alkalinity is too low,
a calculated amount of NaOH should be added to the solution. The above-mention-
ed quantity of sodium azide is poured into the reactor for 1 hr, maintaining a tem-
perature of 50°C.

After the two solutions have been mixed, stirring is stopped. The lead azide
so produced should be allowed to settle after which the liquid above is decanted

HYDRAZOIC ACID AND ITS SALTS

179

and the solid is conveyed by a water stream to a cloth filter, stretched on a frame.
The azide is washed on the filter with a large amount of water, and then the filter
loaded with azide is placed in a bakelite vessel and transferred to the storehouse.

Fig. 49. Diagram of the design and operation of a reactor for the manufacture of lead
azide and other primary explosives (tetrazene, lead styphnate and lead picrate).

Drying and sieving is carried out as described in the chapter on mercury fulmi-
nate. About 1.2 kg of material is dried at a time on the frame, at a temperature
of 65-70°C.

The destruction of azide residues in solutions or suspensions is a matter of
great technical importance.

For this purpose the reaction of hydrazoic acid with nitrous acid is utilized
(p. 168).

THE CONTINUOUS METHOD OF LEAD AZIDE MANUFACTURE
(ACCORDING TO MEISSNER [110, 111])

The method involves introducing a continuous supply of lead nitrate and sodium
azide solutions in equivalent proportions into the upper part of the reaction column
(Fig. 50) from the bottom of which a suspension of lead azide is removed. The
reaction mixture flows down in countercurrent to air blown in.

Lead azide with a purity under 95%

Precipitation. In a dissolving vessel a solution of sodium azide is prepared in
distilled water and then aqueous sodium hydroxide is added. The quantity of sodium
hydroxide used depends on the amount of lead azide to be produced. The solution

180

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

thus prepared is conveyed, via a filter, into a receiver standing in the upper floor.
Here, it is heated up to the reaction temperature. (According to some information
[112] the optimum temperature is 36±1°C. Too high a temperature may cause
the precipitation of irregular, small crystals.) In another dissolving vessel a solution

Fig. 50. Column for continuous manufacture of heavy metal azides and lead styphnate,
according to Meissner [110]: 1,2— inflow of reacting solutions, 3,4— reaction column,
5— air nozzle with exit openings, 6 and 7 directed up- and downwards, respectively,

S— outlet, 9— overflow.

of lead nitrate in distilled water is prepared. The solution is filtered and transferred
into the receiver. The appropriate amount of dextrin dissolved in water is also added,
and the solution is heated up to the reaction temperature. Before the precipitation
reaction is begun the precipitation column should be filled with warm distilled water at
the desired precipitation temperature, and the air agitation must be switched on. Then
the flowmeters for feeding the lead nitrate solution and the sodium azide solutions
are opened in turn. After some minutes, the clip of the drain hose is adjusted in such
a way that the liquid level in the precipitation column remains at the same height,
i.e., the amount of mother liquor discharged with the crystals equals that of the
solutions run in. The liquid discharged from the column is led to a vacuum filter.
When sufficient lead azide is gathered on one of the suction filters, the liquid
from the column is discharged to the second vacuum filter. The lead azide is washed
with distilled water until it is no longer alkaline. As soon as sufficient lead azide has-
been collected on the second suction filter, the liquid discharged from the column

HYDRAZOIC ACID AND ITS SALTS

181

is directed once more onto the first suction filter. Each time, the lead azide accu-
mulated on the filter is washed until no alkalinity is shown.

The crystal structure of the lead azide is checked by examination under a micro-
scope at frequent intervals from the beginning of the precipitation. The lead azide
is precipitated as spherical crystals (Fig. 51).

Fig. 51. Lead azide precipitated by Meissner method [111]. Courtesy J. Meissner.

The mother liquor and washing water sucked off from the vacuum filters are
led into a destruction vat.

Cleaning and destruction. After the precipitation has been completed, the pre-
cipitation column and suction filters should be washed thoroughly with water.
For cleaning, the precipitation column is filled with water, nitric acid and sodium
nitrite are added, and the air agitation is switched on for half an hour. Then the
contents are discharged and the column is rinsed again with water to remove the
acid.

The mother liquor and washing waters from precipitation and cleaning carry
some lead azide. All the mother liquors and washing waters are therefore collected
in a vat, where nitric acid and sodium nitrite are added, and the liquid is stirred
for half an hour. After the lead azide has been destroyed, the acid waste waters —
which are no longer explosive - are discharged into a sump or the sewerage system.

Drying and sifting. The washed lead azide is removed from the filters while
still damp, and carried in a bucket to the separate drying house. There it is laid in

182 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

a thin layer on trays covered with conductive rubber. The warm air used for
drying is supplied from a separate room.

After the drying process, cold air is admitted for a short time into the drying
room to cool the lead azide to ambient temperature and bring its humidity to that
prevailing at room temperature.

Subsequently, the lead azide is transferred into composition boxes behind a pro-
tective wall. Sifting takes place in a separate bulding, on a special vibratory sieve.
The explosive is poured out of the composition boxes into the sieve funnel in a safe
place behind a protective wall. The sifted lead azide is collected in an empty compo-
sition box placed underneath.

Lead azide with a purity of more than 95%

Pure lead azide may be produced in the same equipment, but instead of lead
nitrate and dextrin lead acetate is employed. The precipitation temperature is lower
than in the manufacture of technical-grade lead azide (having a purity of less than
95%). All other operations are the same as already described.

SILVER AZIDE

Silver azide is slightly hygroscopic — at room temperature in a damp atmo-
sphere it picks up approximately 2% of water.

Silver azide is a very vigorous initiator, almost as efficient as lead azide (cf.
Table 32).

The researches of Wischin [113] and those of Garner and Maggs [84] have shown
that metallic nuclei are formed during the slow thermal decomposition of silver
azide. These researches were carried out by means of an optical microscope.

Sawkill [114] recently confirmed this observation using an electron microscope
and found that silver is evolved as the result of slow reactions. In the early stage
of decomposition intermediate compounds, richer in silver than azide, are formed.
The pure metal, which is evolved only in the final stage of decomposition has a mark-
edly oriented structure and a grain size of 0.1 x 0.1 x0.05 mfi.

Like lead azide, silver azide decomposes under the influence of ultra-violet irra-
diation. If the intensity of radiation is sufficiently high the crystals may explode
(cf. p. 171).

The explosion of silver azide under the influence of brief, intensive irradiation
was studied by Berchtold and Eggert [86] and Meerkamper [87]. The intensity of
light causing the explosion of silver azide is 2.6 J/cm 2 , the el ectrical energy necessary
to produce a flash with a half-life of 0.8 m/sec being 310 J.

The same problem was later investigated by Courtney- Pratt and Rogers [115].
They found that the energy required to cause silver azide to explode should be great-
er than 8 x 10 -4 cal per mm 2 of the crystal surface.

HYDRAZOIC ACID AND ITS SALTS 183

Eggert and Courtney- Pratt and Rogers state that the decomposition of silver azide
under the influence of irradiation has a thermal character, i.e. that light absorbed
by a thin surface layer of the crystal is degraded into heat in a very short time in-
terval (less than 1/50 //sec), whereupon explosion occurs by the normal thermal
mechanism.

According to Bowden and Yoffe [116] other possible mechanisms should be
considered, including direct photochemical decomposition. A number of experi-
ments have been carried out by members of the Bowden school.

Thus, according to Rogers [117] when the surface of crystals which have been
given a flash, but which have not exploded or broken down, is examined by an opti-
cal microscope, it can be seen that the crystal is much darkened on the irradiated
face, and contains many irregular but parallel cracks. The cracks are not visible on
the other side of the crystals suggesting that they penetrate only a short distance
into it.

McAuslan [118] and Rogers [117] attempted to measure the time that elapses
between the absorption of light by the silver azide and the ignition of the crystals.
This proved to be less than 20 /*sec.

If silver azide has been sensitized by the dyestuff erythrosin (Rogers [117]) it is
about 2.5 times more sensitive to photo-initiation than normal silver azide. Evans
[119] examined the sensitization of silver azide by the incorporation of gold powder.
He found that the critical light energy necessary for explosion is reduced by incor-
porating gold powder. The greatest effect was produced when the mixture contained
28% of gold by weight.

According to Bowden and Yoffe [116] these and other results suggest, that the
initiation of flash decomposition may be of a true photochemical character.

These authors suggest the following mechanism for the initiation of photochem-
ical decomposition of azides:

N© < > N 3 + e

i.e. an N 3 radical is formed.

The spread of the explosion from the decomposed surface layer however depends
on thermal factors, i.e. the heat liberated by the reaction is greater than that lost
by self heating, conduction etc. The heat liberated during decomposition is sufficient
to melt the surface of the azide and give rise to a reaction that will be self-support-
ing in the thermal sense.

The sensitizing action of gold can be interpreted by postulating that the gold
particles act as electron traps.

Gray and Waddington [57,120] examined the physico-chemical properties of
silver azide and state that its melting point is 300°C. On the basis of the latest
opinion that the explosive decomposition of azides results from processes involving
ions and electrons caused by imperfection and deficiencies in the crystal lattice
(Jacobs and Tompkins [22]), the authors incorporated silver cyanide, Ag 2 (CN) 2 ,

184 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

into silver azide. The (CN)| e ion has the same electron configuration as the Nf ion
and probably has the same linear structure and similar dimensions as Nf ion.

It is assumed that the presence of silver cyanide increases the sensitiveness of silver
azide at elevated temperatures; sensitized AgN 3 explodes at a lower temperature
than the ordinary compound and the induction period is shorter.

Ignition Minimum ignition

at 260°C after temperature
PureAgN 3 induction 280 min 340°C

AgN 3

precipitated from a solution

containing 10% of Ag 2 (CN 2 ) 2 induction 10 min
AgN 3

precipitated from a solution
containing more than 10% of
Ag 2 (CN 2 ) 2 induction 1 min 270°C

The same authors investigated the kinetics of decomposition

2AgN 3 -» 2Ag + 3N 2 + 148 kcal at 230-300°C

and found the approximate equation

d(N 2 )

— — =£(AgN 3 )2/3
at

The activation energy is 35 kcal/mole.

The electrical conductivity of AgN 3 at a temperature above 160°C

<7= 1.60 exp (- 10.7 kcal/ RT)
is therefore very high.

Thus migration of the cation is possible, and the authors presume that this is
the cause of the initiating property of silver azide.

Other authors quote the following values for the activation energy of the thermal
decomposition of silver azide :

40.0 kcal/mole (Audubert [121])

29.0 kcal/mole (Haycock [122])

44.0 kcal/mole (below 190°C) (Bartlett, Tompkins and Young [123D

31.0 kcal/mole (above 190°C) (Bartlett, Tompkins and Young [123])

The ignition temperature is 273°C and is thus much lower than that of lead
azide, although the sensitiveness of silver azide to impact is also lower than that
of lead azide. Taylor and Rinkenbach [124] report that with a 0.5 kg weight a 77.7 cm
drop is necessary to cause detonation of silver azide, whereas for mercury fulminate
a 12.7 cm drop is sufficient.

The rate of detonation of a thin (0.1-0.5 mm) film of silver azide is about
1550 m/sec; that of the same film in an enclosed space about 1700 m/sec (Bowden
and Williams [101]).

Silver azide, AgN 3 , is manufactured in the same way as lead azide, in aqueous
solution, by action of sodium azide on silver nitrate.

HYDRAZOIC ACID AND ITS SALTS 185

CUPRIC AZIDE

Cupric azide, Cu(N 3 ) 2 , is of great practical significance since it can be formed
in addition to cuprous azide by long term action of lead azide on copper or its alloys.

Curtius and Rissom [41] prepared cupric azide by the action of an aqueous
solution of sodium azide on an aqueous solution of cupric sulphate, obtaining the
salt in a hydrated form. The anhydrous salt was prepared by Straumanis and Ciru-
lis [125] in the form of dark brown, reddish sediment by reaction of lithium azide
on cupric nitrate in an alcohol solution. Another method described by Curtius
consists of reacting hydrazoic acid with metallic copper in an aqueous medium.

Green cupric azide has also been described (Dennis and Isham [44]). It is formed
by the action of hydrazoic acid on cupric hydroxide or (according to Straumanis
and Cirulis) on cupric oxide. It is sometimes grey in colour.

Cupric azide is insoluble in water, but is soluble in dilute acids and in acetic
acid. It is decomposed by concentrated sulphuric acid, evolving nitrogen. It dis-
solves in an aqueous solution of ammonia and aliphatic amines to form a complex
compound.

Boiling in water (Wohler and Krupko [80]) leads to hydrolysis with the forma-
tion of basic cupric azide. Long-continued boiling causes complete hydrolysis to
cupric oxide and free acid. Black cupric azide, Cu(N 3 ) 2 , when exposed to the action of
air for 2 months, is completely converted into a yellow basic salt. This is discussed later.

The ignition temperature of cupric azide is 202-205°C. The dry substance is
exceptionally sensitive to friction, especially the green modification, and is often
exploded by contact. It is also very sensitive to impact; the green modification is
exploded by a 2 kg weight falling from a height of less than 1 cm, the black one from
a drop of about 1 cm.

Its rate of detonation ranges between 5000 and 5500 m/sec.

Straumanis and Cirulis [125] emphasize its exceptionally strong initiating
properties, viz 0.0004 g only of the substance is sufficient to detonate penthrite.

Basic cupric azide, Cu(OH)N 3 , prepared by Wohler and Krupko, is yellow col-
oured. According to Straumanis and Cirulis it is less sensitive to friction and impact.
It is exploded by a 1 kg weight falling from a height of 7-8 cm. Its ignition temper-
ature is the same as that of the neutral salt (203-205°C).

Complex salts of cupric azide are also explosive. The salt Cu(NH 3 ) 4 (N 3 )2
is much less sensitive to impact than cupric azide itself. The complex lithium-cupric
salt Li 4 [Cu(N 3 ) 6 ] has exceptionally strong initiating properties.

OTHER METAL AZIDES

Martin [126] prepared nickel, cobalt, zinc and manganese azides by the action
of an ether solution of hydrazoic acid on the dry metal carbonate. Cuprous azide,
CuN 3 , was obtained in the form of a light grey sediment by the reaction of sodium
azide with a solution of cuprous sulphate.

186

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The preparation of ferric azide failed, however, due to the formation of basic
ferric azide. A complex compound was obtained by Martin from chromium hydro-
xide and hydrazoic acid. It contained no chromium or Nf ions and has the probable
formula of a complex compound [Cr(N 3 ) 4 ]H.

Martin carried out extensive research into the explosive properties of the azides
of various metals (Table 33). The high sensitiveness of cuprous azide to impact is
noteworthy.

Table 33
Explosive properties of azides (according to Martin [126]

Azide

Ignition

temperature

°C

Sensitiveness

to impact

(work in kgm/cm 2 )

Minimum
initial charge, g

trinitrotolu-
ene

tetryl

Silver

273

13.97

i
0.07 0.02

Lead

327

4.76

0.09 0.025

Mercurous

281

4.76

0.145 1 0.045

Cadmium

291

18.54

0.04 ! 0.01

Zinc

289

17.53

Cuprous

174

2.66

0.095 0.025

Nickel

200

5.46

Cobalt

148

5.88

Manganese

203

6.30

Barium

152

7.70

Strontium

169

9.10

Calcium

158

10.14

Lithium
Thallium

245
320

(no explosion)
16.18

0.115 0.07

Mercuric azide, Hg(N 3 ) 2 , occurs in two allotropic modifications: a (orthorhom-
bic) and p (monoclinic), like lead azide (Miles [75]; Garner and Gomm [37]). The
latter is obtained like /Mead azide, by slow diffusion of the solutions. During crystal-
lization spontaneous explosion may occur. Mercuric azide also explodes during
crystallization from a hot, aqueous solution.

The physical properties of mercurous azide — Hg 2 (N 3 ) 2 — were examined by
Evans and Yoffe [33] and its photochemical decomposition by Deb and Yoffe [26].
The activation energy was found to be 8.4 kcal/mole.

Cadmium azide may also detonate spontaneously, but under different experi-
mental conditions, i.e. when a rod of metallic cadmium is immersed in hydrazoic acid
and cadmium azide is formed on the surface of the rod (A. T. Thomas [127]).

Gray and Waddington [128] have produced a graph showing changes in the
rate of detonation in relation to the heat of formation of various azides (Fig. 52).

The same authors [57] determined the heats of formation and decomposition of
a number of azides (Table 34).

hydrazoic acid and its salts

Table 34

187

Salt

Heat of formation

kcal/mole

HN 3 gas

-71.66

HN 3 liquid

-64.37

LiN 3

- 2.58

NaN 3

- 5.08

KN 3

- 0.33

RbN 3

+ 0.07

CsN 3

+ 2.37

NH4N3

-26.79

CaN 6

-11.03

SiN 6

- 1.72

BaN 6

+ 5.32

CuN 3

-67.23

CuN 6

- 140.4

AgN 3

-74.17

Hg 2 N 6

-141.5

T1N 3

-55.78

PbN 6

-115.5

CdN 6

ca. -108

The heat of decomposition to metal and nitrogen, AH, has of course, the oppo-
site value AH= -AH f .

With regard to their sensitiveness to ignition by light, the azides of monovalent
metals may be ranged as follows:

KN 3 < TIN3 < AgN 3 < CuN 3

According to Deb [129] the corresponding values are: for T1N 3 - 92; AgN 3 - 39;
CuN 3 - 12 J.

The activation energies of azides not mentioned above are shown in Table 35.

Table 35

Salt

Activation energy

Author

NaN 3

34.0 kcal/mole

Garner and Marke [130]

KN 3

36.0 kcal/mole

Garner and Marke [130]

Ba(N 3 ) 2

21.0-27.0 kcal/mole

Jacobs and Tompkins [22]

Sr(N 3 ) 2

20.0 kcal/mole

Maggs [131]

Ca(N 3 ) 2

18.0 kcal/mole

Marke [132]

Cu(N 3 )

26.5 kcal/mole

Singh [133]

The properties of thallous azide have been examined in detail by Gray and
Waddington[120].

188

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Thallous azide is a yellow, crystalline substance, m.p. 334°C. The crystal struc-
ture as determined by X-ray analysis points to the isomorphism of T1N 3 with KN 3
and RbN 3 azides.

Thallous azide is sensitive to the action of light and decomposes under its in-
fluence to evolve metallic thallium.

The ignition temperature, as determined by throwing a test sample on a heated
metallic surface, is 490°C. The activation energy of its pre-explosive state as calcu-
lated by the authors is said to be approximately 40 kcal/mole.

The introduction of a thallium salt with a bivalent anion, e.g. thallous sulphide,
T1 2 S, produces defects in the crystal lattice. According to a recent view (Jacobs

3000

§2000

1
-8

1000

1
°Pb(N 3 ) 2

/°AgN 3
'TIN 3

^°Ca(N

»h

50 100 150
Heat of formation, kcal

200

Fig. 52. The variation of the rate of detonation with heat of formation of azides,
according to Gray and Waddington [128].

and Tompkins [22]), the explosive decomposition of azides is caused by imper-
fection and deficiencies of the crystal lattice. Due to defects produced artificially with
thallium sulphide, the ignition temperature is reduced, e.g. the ignition temperature
of thallium azide containing 18% of T1 2 S is reduced to 420°C.

Deb and Yoffe [134] examined the decomposition of thallous azide under the
action of ultra-violet light in the wavelength region 3200-3800 A. Two exciton
bands 3415 and 3348 A have been observed in thallous azide by low-temperature
spectroscopy (Nikitine and Gross's method). The refractive index has been measured
by the Brewster angle method, the electron energy levels have been estimated and
the results of the photochemical decomposition have been related to the electron
energy level and to measurement of photoconductivity [33].

The overall reaction is

2T1N 3 -*2T1 + 3N 2 ,

where nitrogen gas is formed by the reaction at the surface

2N 3

3N 2

The activation energy was found to be 3.2 ±0.1 kcal/mole. The suggested mecha-
nism of photochemical decomposition is as follows :

HYDRAZOIC ACID AND ITS SALTS 189

N3 + h v -> N 3 (exciton formation)

*
N3 ->- N3 + q (reversion of the exciton to the ground

state)
N 3 -> N 3 + e (thermal dissociation of the optically

formed exciton)
N3+e -» N3 (recombination of electron and positive

hole)
2N 3 -> 3N 2 + Q (combination of two positive holes)

Tl„+Tl©+e -> Tl„ +1 (electron trapping and metal formation)

The electrical conductivity of T1N 3 is very high: 6-5 x 10-s/275°C (i.e. a million
times higher than that of KN 3 ).

The explosive properties of sodium, calcium, strontium and barium azides
have been investigated at the Chemisch-Technische Reichsanstalt [135]. These azides
differ markedly from lead, silver and cupric azides in that they show none of the
properties of primary explosives. All three may be ignited by a spark, a glowing
wire or the flame of blackpowder. Calcium azide burns most rapidly and has dis-
tinctly marked explosive properties. Larger quantities of it may explode when
ignited in a closed tin, while strontium and barium merely burn violently. Calcium
azide detonates under the influence of a detonating cap. The sodium azide does
not decompose in these conditions. The other azides show weak decomposition
under the influence of a standard (No. 3) detonator. Their most important pro-
perties are tabulated below.

Table 36

Azide

Ignition
temperature °C

Lead block

expansion

cm 3

Heat of de-
composition
kcal/mole

Sodium

Calcium

Strontium

Barium

no ignition

up to 300

171-176

190-202

190-200

120
30-90
25-30

13-14

Tompkins et al. [22, 85] studied the photochemical decomposition of potassium
and barium azide. Originally they found that the rate of photolysis was proportion-
al to the square of the intensity of the radiation.

In more recent studies, Jacobs, Tompkins and Young [136] examined the rate
of evolution of nitrogen from barium azide as a criterion of the rate of photolysis,
and have shown the reaction to be more complex than was previously indicated.
A mechanism for the photolysis involving the production and reaction of both exci-
tons and positive holes has been formulated.

According to Ficheroulle and Kovache [137] barium azide has a low sensitive-
ness to impact (a 2 kg weight falling from 100 cm causes 14% of explosions) but it
is very sensitive to friction. It does not possess the properties of a primary explosive,

190 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

but in large quantities may burn very violently. The toxicity of BaN 6 is negligible,
but in the presence of strong acids it decomposes to evolve strongly poisonous
hydrazoic acid.

Strontium azide possesses similar properties, but is highly sensitive to the action
of even weak acids (e.g. with C0 2 it forms SrC0 3 ) and easily is hydrolysed.

Barium and strontium azides are used in the manufacture of valves in radio
technology.

Ficheroulle and Kovache recommend two methods for the manufacture of
barium azide. One of them is based on the reaction of ethyl nitrite with hydrazine
in the presence of barium hydroxide: 19.5 kg of ethyl nitrite is cooled with brine
to -15°C when a mixture of 31 kg of Ba(OH) 2 in 100.1. of the 10% hydrazine
hydrate previously cooled to — 15°C is added at a rate of 4 1. per hr. The whole is
strirred for 10 hr while room temperature is attained gradually. A stream of C0 2
is allowed to pass through the liquid to precipitate excess barium. Barium carbo-
nate is then removed by nitration. The filtrate is greatly concentrated and BaN 6
precipitated by the addition of alcohoL Thus 12.6 kg of BaN 6 is obtained i.e. 44%
of the theoretical yield. A yield up to 55% may sometimes be obtained. The other
method is based on the double decomposition of Ba(C10 4 ) 2 and KN 3 . Sparingly
soluble KC10 4 is precipitated while BaN 6 passes into solution. This method gives
a product which is less pure and not suitable for valve manufacture.

Sodium azide (see above, p. 189 and Table 36) can be decomposed on heating
but it is of low sensitiveness to impact or friction and is not listed as an explosive
in transport regulations. According to Gunther et al. [138] rubidium azide is much
more sensitive to impact and friction than sodium azide. Gunther believes this to
be due to the fact that the radius of the orbit of nitrogen atoms in rubidium azide
is much shorter than that in sodium azide.

Curtius [1] who prepared ammonium azide did not notice its explosive proper-
ties. They were reported by Berthelot [139] who found ammonium azide to be an
endothermic substance with a heat of formation — AH { of —19.0 kcal.

Berthelot and Vieille [140] reported that the explosive decomposition of ammo-
nium azide proceeds according to the following equation :

2NH 4 N 3 -» 3N 2 + H 2 + 2NH 3 + 502 kcal/kg
Volume of gases (Vo) is 1148 l./kg
Temperature of explosion t, 1350-1400°C
Specific pressure (/), 7290 m

A low explosion temperature together with a great amount of gaseous products
and a high specific pressure suggested the used of ammonium azide as a propellent
explosive. In practice the use of the substance, however, is prevented by its high
volatility.

Azides of complex salts with ammonia ("ammines") are described below (p. 231).

Among other inorganic azides those prepared by Wiberg et al. [141] i.e. boron,
B(N 3 ) 3 , and silicon, Si(N 3 ) 4 , azides are of interest. Grundman and Ratz [142] ob-

HYDRAZOIC ACID AND ITS SALTS 191

tained highly explosive "phosphorous azide", P 3 N 21 . As early as 1915 Curtius
and Schmidt [143] described the preparation of S0 2 (N 3 ) 2 from S0 2 C1 2 and NaN 3 .
By the action of S0 3 on KN 3 Lehman and Holznagel [144] obtained an addition
compound, KN 3 -2S0 3 , which is transformed on heating into disulphuryl azide,
S 2 5 (N 3 ) 2 , a highly explosive substance. Sundermeyer [145] described recently
silily-azides, e.g. (CH 3 ) 2 Si(N 3 ) 2 .

Halogen azides, e.g. IN 3 , BrN 3 , C1N 3 , FN 3 are also known. They are, how-
ever, highly unstable.

An extensive review of physics and chemistry of inorganic azides is given
by Evans, Yoffe and Gray [74].

ORGANIC AZIDES

In the search for powerful explosives, attempts have been undertaken to introduce
the — N 3 group into organic molecules. Generally this so enhances their sensitive-
ness to friction and impact that they cannot be used. Moreover, the increase in
explosive power is not always commensurate with the rise in the manufacturing
costs of the substance.

Triazoethanol nitrate N 3 • CH 2 CH 2 • ON0 2 (II) is an example of an organic
aliphatic azide and nitric ester. It was prepared by T. Urbanski and Rusiecki [146]
by the following steps :

CH 2 C1 NaNj CH 2 N 3 CH 2 N 3

CH 2 OH CH 2 OH CH 2 ON0 2

I II

Compound (I) was obtained by Forster et al. [147]. Substance (II) is a liquid
explosive resembling nitroglycerine in its properties: it gives the same lead block
expansion and a rate of detonation ranging from 2000 to 6550 m/sec (in a lead pipe
17/21 mm diameter). Mixtures with ammonium nitrate give an even larger expansion
and a higher rate of detonation than analogous mixtures with nitroglycerine.
Triazoethanol nitrate has a higher sensitiveness to impact than has nitroglycerine.
The ignition temperature is 190°C in a closed vessel. According to Urbanski and
Rusiecki the alcohol itself, triazoethanol (I), is an explosive substance and gives
a lead block expansion of 130 cm3.

Azides containing the — CON 3 group have received little attention, and have
poor prospects for practical use. They are formed either by the action of nitrous
acid on hydrazides or by the action of sodium azide on acid chlorides:

HONO
R— CONHNH 2 > R— CON 3

R— COC1

IM* 41 -

192 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Some azides are capable of the so-called Curtius rearrangement, which results
in the formation of isocyanates with loss of nitrogen.

Organic azides with initiating properties may be exemplified by the following
substances (Curtius [148]): oxamidoazide (III) and triazide of trimesic acid (IV):

CON 3

' 0N3 A

conh > N30C-1J1-CON3

m iv

The first of these is obtained by the action of nitrous acid on oxamidohydrazide.
It is unstable and decomposes explosively at 115°C. The second is formed by the
action of nitrous acid on the trihydrazide of trimesic acid. The explosive properties
of neither of these azides have been reported.

Diazide of carbonic acid (V) is exceptionally sensitive to friction: it explodes on
contact with a glass rod (Curtius and Heidenreich [149]) Succinyl azide [150] and other
acyl azides [151] behave similarly.

N 3

/
CO

N 3

Organic azides in which the N 3 group is combined with an aromatic radical may
be prepared by a general method based on the action of hydroxylamine on a diazo
compound (Mai [1 52]) :

Ar— N 2 C1 + NH 2 OH -> (Ar— N=N— NHOH) -> ArN 3

\
ArNH 2

The initial amine is formed again, together with azide.

Tetra-azido quinone (VI) was obtained by Sorm [153], by the action of sodium
azide on chloranil. This explosive is powerful but of no practical use due to its
inadequate stability; its ignition temperature is low: 91°C.

The organic derivatives of hydrazoic acid which contain an aromatic ring with
nitro groups comprise an important group of initiators. Picryl azide (VII), m.p.
89-90°C is a typical example. It has been prepared both by the action of nitrous
acid on trinitrophenylhydrazine (Purgotti [154], Schrader [155] and Korczyiiski [156])
and by the action of sodium azide on picryl chloride. Rathsburg [157] suggested

HYDRAZOIC ACID AND ITS SALTS 193

the use of picryl azide as an initiator but its initiating properties proved to be too
weak and the compound has not found practical application.

N 3

2 N— f\~ N0 2

no 2

VII

Trinitrobenzoyl azide (VIII) was prepared by Vasilevskii. Blokhshtein and
Kustria [158] by the action of sodium azide on trinitrobenzoyl chloride

CON 3
I
2 N— /\— NO,

N0 2
VIII

Trinitrotriazidobenzene (IX) is the only representative of organic azides posses-
sing properties of primary explosives which has some prospect of practical use.
Turek [159] prepared it by the action of sodium azide on sym-trichlorotrinitrobenzene
(Vol. I. p. 469) and on the basis of its properties which he himself determined he
suggested its use as an initiator.

N0 2

o 2 nk y \no 2

Trinitrotriazidobenzene is insoluble in water, easily soluble in acetone and
moderately soluble in chloroform and alcohol. It is not hygroscopic and is moist-
ure-resistant. In the presence of moisture it has no effect on iron, steel, copper
or brass. At its melting point, 131°C, it undergoes decomposition to evolve nitrogen
and to form benzotrifuroxane ("hexanitrosobenzene") also an explosive substance
(Vol. I, p. 603).

N0 2

N 3

-> 3N 2 +

N3 \,/ /N3

o 2 n/~V \no 2

N 3
IX

194 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The same reaction also occurs at a lower temperature. 0.665% of the substance
decomposed at 20°C to form benzotrifuroxane in 3 years; 2.43% at 35°C in one
year; 0.65% at 50°C in 10 days; and at 100 C C the substance underwent complete
change in 14 hr. This decomposition is not, however, autocatalytic. This reaction —
the formation of furoxane derivatives from aromatic azides with nitro group in the
ortho-position — is of a general character (Boyer qt al. [160]). Despite the ease
with which it decomposes trinitrotriazidobenzene has not been rejected for use as an
initiator. In some countries large scale experiments are in progress to examine the
possibilities of developing its practical application.

Trinitrotriazidobenzene is less sensitive to impact and friction than mercury
fulminate.

As an explosive the substance is very powerful. Its expansion in the lead block
is about 500 cm 3 , i.e. its explosive strength is midway between that of tetryl and
penthrite. At a density of 1.54 it detonates at a rate of 7500 m/sec.

The specific gravity of trinitrotriazidobenzene is 1.8054. Under a pressure of
3000 kg/cm 2 it gives a density of 1.751 and under 5000 kg/cm 2 the density achieved
is 1.7526. A pressure higher than 300 kg/cm 2 may make it "dead pressed". 0.02 g
of the substance compressed under the pressure of 300 kg/cm 2 detonates trinitro-
toluene, and 0.01 g detonates tetryl. It is therefore one of the most vigorous initia-
tors.

As a starting material for the preparation of sym-trinitrotriazidobenzene, 1,3,5-
trichlorobenzene is used. It is obtained by the chlorination of aniline and the removal
of the amino group. Nitration to the trinitro derivative is described in Vol. I. The
final reaction is simple: powdered l,3,5,-trichloro-2,4,6-trinitrobenzene is added to
an aqueous alcohol solution of sodium azide. The precipitated product is washed
with alcohol and water and dried at a moderate temperature. The product so obtain-
ed may be purified by crystallization from chloroform.

A powerful initiator in which azido groups are combined with a heterocyclic
ring is cyanuric triazide (XI) '

N 3
I

N N
N 3 — C C— N 3

XI
m.p. 94°C

It was obtained by Ott and Ohse [161] in the following way:
HCN — > CNCl + HC1

3NaN 3
(CNC1) 3 * (CN) 3 (N 3 ) 3

XII XI

HYDRAZOIC ACID AND ITS SALTS 195

On reaction with chlorine, hydrogen cyanide gives cyanogen chlorides, forming
a trimer— cyanuric chloride (XII). The latter is a liquid with a melting point of
146°C and a boiling point of 196 G C. Next, cyanuric triazide (XI) is obtained by the ac-
tion of sodium azide in an aqueous solution, at room temperature on compound (XII).

Cyanuric triazide is insoluble in water, sparingly soluble in cold alcohol and
readily soluble in hot alcohol, acetone, benzene, chloroform, ether, and molten
trinitrotoluene, It is slightly hygroscopic and slightly volatile. It irritates the skin
causing dermatitis. Its heat of formation — AH t is 219 kcal/mole (H. Muraour [162]).

Much attention has been devoted to this substance as an initiator since it was
found that its initiating properties are stronger than those of mercury fulminate.
However it proved to be highly dangerous to handle and sensitive to impact and
friction; it has been known to explode during manufacture, e.g. during drying.
Large crystals, which explode even under the pressure of a rubber cork, are particu-
larly dangerous. Attempts to press it into capsules often resulted in explosion.
This accounts for the fact that the substance has found no practical application.
It is nevertheless very interesting from the theoretical point of view. The following
data are characteristic of its properties.

The ignition temperature, when heated at the rate of 20°C/min, is 205-208°C
i.e. higher than that of fulminate, but decomposition becomes evident on heating
at a temperature slightly exceeding 100°C. The substance is exploded by a drop
three times less than that of mercury fulminate.

The rate of detonation at a density of 1.15 is 5545 m/sec (Kast and Haid [99]),
it is therefore an initiator that detonates extremely fast.

The initiating properties of the substance are characterized by the following

figures (Taylor and Rinkenbach [163]) which indicate the amounts of primary

explosive required to initiate different high explosives compressed under a pressure

of about 14 kg/cm 2 .

j. 10 g of cyanuric triazide
Trinitrotoluene {^ g of mercury {u]minate

(0.05
J0.21

IS:

i g of cyanuric triazide

Picric acid { n „ , g of mercury fu i minate

14 g of cyanuric triazide
e ry '0.24 g of mercury fulminate

A new type of initiator containing the azido group has recently been described
by Glowiak [164]. These are salts of heavy metals (e.g. lead), phenols of phenolic
acids containing nitro groups and azido group, e.g. the plumbous salt of dinitro-
azidophenol (XIII) and 5-nitro-3-azidosalicylic acid (XIV) :

OH COOH

2 N V j^ /N 3 1 .OH

OzN—.

^ \N 3

XIV

196 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Plumbous salts of these compounds are very sensitive to mechanical and thermal
impulses but have no distinctly marked ability to initiate secondary explosives and
are similar to lead styphnate in their properties.

Interesting material with which the mechanism of decomposition of organic
azides was investigated is triphenylmethyl azide

(C 6 H 5 ) 3 CN 3

Saunders and Ware [165] examined its thermal decomposition in the molten
state at 170-200°C and found it gives a benzophenone anil (C 6 H 5 ) 2 C = NC 6 H 5 .
They suggest the following mechanism for the decomposition

— C— N=N=N -> — C — N=N=N -> — C=N— <^ ^> + N 2
I II

The transition state is represented as the resonance hybrid of (I) and (II).

The same product was obtained by Deb and Yoffe [26] as the result of photo-
chemical decomposition of the substance. The activation energy of this reaction was
found to be 8.82 kcal/mole. See also [33].

The chemistry of organic azides has been the subject of many theoretical and
practical investigations. Apart from their application as explosives some organic
azides possess interesting pharmacological and bacteriostatic properties.

A new group of azides containing silicon of general formulae

R 3 SiN 3

R 2 Si(N 3 ) 2

RSi(N 3 ) 3 (R=C 6 H 5 orCH 3 )

were recently prepared by Reichle [166]. He also described similar compounds with
other group IV elements: Ge, Sn and Pb, e.g. (C 6 H 5 ) 3 PbN 3 .

The compounds are remarkably stable.

An extensive review of the chemistry of aliphatic and aromatic azides is given
by Boyer and Canter [167] and Gray [168]. Organic azides are subject to various
reactions such as the Bergmann degradation and the synthesis of peptides, the well
known Curtius rearrangement, the Darapsky synthesis of a-aminoacids [169], for
synthesis of triazoles [170], tetrazoles ("Schmidt reaction") [169] and [171] etc.
These reactions lie beyond the scope of the present book.

LITERATURE

1. T. Curtius, Ber. 23, 3023 (1890); 24, 3341 (1891); 29, 759 (1896); J. prakt. Chem. [2], 43,
207 (1891); 48, 273 (1893); 91, 1, 415 (1915).

2. R. Stern, Klin. Wochenschr. 6, 304 (1927).

3. PRAVDIN and Shakhnovskaya. Farmakologya i toksikologya 8, 50 (1948).

HYDRAZOIC ACID AND ITS SALTS 197

4. Shakhnovskaya, Farmakologya i toksikologya 13, 3, 4 (1950).

5. J. Thiele, Ber. 44, 2524, 3336 (1911).

6. E. C. Franklin, /. Am. Chem. Soc. 46, 214 (1924); 56, 568 (1934).

7. L. B. Hendricks and L. Pauling, /. Am. Chem. Soc. 47, 2917 (1925).

8. L. K. Frevel, J. Am. Chem. Soc. 58, 779 (1936).

9. F. J. Llewellyn and F. E. Whitmore, J. Chem. Soc. 1947, 881.

10. I. E. KNAGGS, Proc. Roy. Soc. {London) A 150, 5776 (1935).

11. A. Langseth, J. R. Nielsen and J. U. SOrensen, Z. physik. Chem. (B) 27, 100 (1934).

12. G. Herzberg, F. Patat and H. Verleger, Z. Electrochem. 41, 522 (1935).

13. E. H. Eyster, /. Chem. Phys. 8, 135 (1940).

14. E. Amble and B. P. Dailey, /. Chem. Phys. 18, 1442 (1950).

15. K. Clusius and H. R. Weisser, Helv. Chim. Acta 35, 1548 (1952).

16. E. A. BRAUDE, Ann. Rep. Progr. Chem. 42, 105 (1945).

17. V. HENRI, International Critical Tables 5, 359, McGraw-Hill, New York, 1929.

18. W. Kuhn and E. Braun, Z. physik. Chem. 138, 281 (1930).

19. H. MOHLER, Helv. Chim. Acta 26, 121 (1943).

20. Yu. N. Sheinker, Dokl. Akad. Nauk SSSR 77, 1043 (1951).

21. Yu. N. Sheinker and Ya. K. Syrkin, Izv. Akad. Nauk SSSR, Ser. Fiz. 14, 478 (1950).

22. P. W. M. Jacobs and F. C. Tompkins, Proc. Roy. Soc. {London) A 215, 254 (1952).

23. B. L. Evans and A. D. Yoffe, Proc. Roy. Soc. {London) A 250, 346 (1959); Atoi/re 183, 1241
(1959).

24. F. C. Tompkins and D. A. Young, Proc. Roy. Soc. {London) A 236, 10 (1956).

25. A. C. McLaren and G. F. Rogers, Proc. Roy. Soc. {London) A 240, 484 (1957).

26. S. K. Deb and A. D. Yoffe, Proc. Roy. Soc. {London) A 256, 528 (1960), and unpublished
work.

27. A. C. McLaren, Ph.D. Thesis, Cambridge, 1957.

28. E. H. Eyster, /. Chem. Phys. 8, 135 (1940).

29. D. A. Dows and G. C. Pimentel. /. Chem. Phys. 23, 1258 (1955).

30. E. H. Eyster and R. H. Gillette, J. Chem. Phys. 8, 369 (1940).

31. E. Lieber, D. R. Levering and L. J. Patterson, Anal. Chem. 23, 1594 (1951); E. Lieber,
C. N. Ramachandra Rao, T. S. Chao and C. W. W. Hoffman, Anal. Chem. 29, 916 (1957). ,

32. J. H. Boyer, J. Am. Chem. Soc. 73, 5248 (1951).

33. B. L. Evans and A. D. Yoffe, J. Chem. Phys. 30, 1437 (1959).

34. P. Gray and T. C. Waddington, Trans. Faraday Soc. 53, 901 (1957).

35. L. Kahovec, K. W. F. Kohlrausch, A. W. Reitz and J. Wagner, Z. physik. Chem. (B)
39, 431 (1938).

36. K. W. F. Kohlrausch and J. Wagner, Monatsh. Chem. 77, 180 (1947).

37. W. E. Garner and A. S. Gomm, /. Chem. Soc. 1931, 2133.

38. A. Delay, C. Duval and J. Lecomte, Bull. soc. chim. France 12, 581 (1945).

39. L. J. Bellamy, The Infra-red Spectra of Complex Molecules, Methuen, London, 1958.

40. D. Mendeleyev, Ber. 23, 3470 (1890).

41. T. Curtius and J. Rissom, /. prakt. Chem. [2], 58, 263 (1898).

42. T. Curtius and A. Darapsky, /. prakt. Chem. [2], 61, 409 (1900).

43. A. J. Sofianopoulos, Bull. soc. chim. France [4], 5, 614 (1909).

44. L. M. Dennis and H. Isham, /. Am. Chem. Soc. 29, 18, 216 (1907); Ber. 40, 462 (1907).

45. R. Meyer and H. J. Schumacher, Z. physik. Chem. (A) 170, 33 (1934).

46. T. Curtius, Ber. 23, 3023 (1890).

47. A. R. Joyner, according to J. Taylor, Detonation in Condensed Explosives, p. 158, Oxford
University Press, London, 1952.

48. A. O. Beckman and R. G. Dickinson, J. Am. Chem. Soc. 50, 1870 (1928); 52, 124 (1930).

49. K. Stewart, Trans. Faraday Soc. 41, 663 (1945).

198 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

50. F. O. Rice and M. Freamo, /. Am. Chem. Soc. 73, 5529 (1951); 75, 548 (1953).

51. J. L. Mador and M. C. Williams, /. Chem. Phys. 22, 1627 (1954).

52. D. A. Dows, G. C. Pimentel and E. Whittle, /. Chem. Phys. 23, 1606 (1955).

53. G. Pannetier and A. G. Gaydon, J. chim. phys. 48, 331 (1951).

54. R. G. W. Norrish, G. Porter and B. A. Thrush, Proc. Roy. Soc. (London) A 216, 1 56 (1 953).

55. B. A. Thrush, Proc. Roy. Soc. (London) A 235, 143 (1956).

56. A. O. BECKMAN and R. G. Dickinson, /. Am. Chem. Soc. 50, 1870 (1928); 52, 124 (1930).

57. P. GRAY and T. C. Waddington, Proc. Roy. Soc. (London) A 235, 106, 481 (1956).

58. E. D. Becker, G. C. Pimentel and M. van Thiel, /. Chem. Phys. 26, 145 (1957).

59. J. L. Franklin, J. T. Herron, P. Bradt and V. H. Dibeler, J. Am. Chem. Soc. 80, 6188
(1958).

60. S. N. Foner and R. L. Hudson, /. Chem. Phys. 28, 719 (1958).

61. K. GLEN, Ber. 61, 702 (1928).

62. D. Alekseyev, Zh. fiz. chim. 2, 535 (1931).

63. F. Seel and R. Schwaebel, Z. anorg. Chem. 274, 169 (1953).

64. G. Stedman, /. Chem. Soc. 1959, 2943, 2949.

65. W. Wislicenus, Ger. Pat. 66813 (1892); Ber. 25, 2084 (1892).

66. H. Clusius and E. EFFENBERGER, Helv. Chim. Acta 38, 1834, 1843 (1955).

67. H. Clusius and H. Knoff, Helv. Chim. Acta 39, 681 (1956).
• 68. J. TfflELE, Ann. 270, 10 (1892).

69. A. Sabaneyev, Z. anorg. Chem. 20, 28 (1899).

70. P. Jannasch and A. Jahn, Ber. 38, 1576 (1905).

71. J. W. Turrentine and J. M. Olin, J. Am. Chem. Soc. 37, 1114 (1905).

72. S. TANATAR, Ber. 35, 1810 (1902).

73. L. F. AUDRIETH, Chem. Rev. 15, 169 (1934).

74. B. L. Evans, A. D. Yoffe and P. Gray, Chem. Rev. 59, 515 (1959).

75. F. D. Miles, J. Chem. Soc. 1931, 2532.

76. L. V. Azaroff, Z. Krist. 107, 362 (1956).

77. P. Gray and T. C. Waddington, Nature 176, 653 (1955).

78. A. Solonina, Z. ges. Schiess- u. Sprengstoffw. 5, 41, 67 (1910).

79. A. Majrich, Z. ges. Schiess- u. Sprengstoffw. 31, 147 (1936).

80. L. WOhler and W. Krupko, Ber. 46, 2056 (1913).

81. A. M. Yuill, Ph.D. Thesis, Cambridge, 1953, according to F. P. BowDENand A.D.YOFFE
[116].

82. A. C. McLaren, unpublished work (1955), according to F. P. Bowden and A. D. Yoffe
[116].

83. G. Todd and E. Parry, Nature 181, 260 (1958).

84. W. E. Garner and J. Maggs, Proc. Roy. Soc. (London) A 172, 299 (1939).

85. J. G. N. Thomas and F. C. Tompkins, Proc. Roy. Soc. (London) A 209, 550 (1951).

86. J. Berchtold and J. Eggert, Naturwiss. 40, 55 (1953).

87. B. Meerkamper, Z. Elektrochem. 58, 387 (1954).

88. J. Eggert, Physik. Bl. 12, 549 (1954).

89. M. HaIssinsky and R. J. Walden, Compt. rend. 208, 2067 (1939).

90. P. L. Gunther, L. K. Lepin and K. K. Andreyev, Z. Elektrochem. 36, 218 (1930).

91. J. V. R. Kaufman, Discussion Roy. Soc, Initiation and Growth of Explosion in Solids, Proc.
Roy. Soc. (London) A 246, 219 (1958).

92. J. M. Groocock, Discussion Roy. Soc, Initiation and Growth of Explosion in Solids, Proc.
Roy. Soc. (London) A 246, 225 (1958).

93. R. Wallbaum. Z. ges. Schiess- u. Sprengstoffw. 34, 126, 197 (1939).

94. F. P. Bowden and K. Singh, Proc. Roy. Soc. (London) A 227, 22 (1954).

95. F. P. Bowden and J. McAuslan, Nature 178, 408 (1956).

HYDRAZOIC ACID AND ITS SALTS 199

96. H. Sudo, Kristalle des Bleiazids, Rep. Fac. Eng., Chuo University, Tokyo, February 1958.

97. J. I. Evans and A. M. Yuill, Discussions Roy. Soc, Initiation and Growth of Explosion in
Solids, Proc. Roy. Soc. {London) A 246, 176 (1958).

98. G. J. Bryan and E. C. Noonan, Discussion Roy. Soc, Initiation and Growth of Explosion
in Solids, Proc. Roy. Soc. (London) A 246, 167 (1958).

99. H. Kast and A. Haid, Angew. Chem. 37, 973 (1924).

100. Jahresber. Chem.-Techn. Reichsanstalt 8, 122 (1929).

101. F. P. BOWDEN and H. T. Williams, Proc. Roy. Soc. (London) A 208, 176 (1951).

102. A. S. Hawkes and C. A. Winkler, Can. J. Research 25, 548 (1947).

103. G. T. Rogers and P. L. Harrison (1955), according to F. P. Bowden and A. D. Yoffe [1 16].

104. T. L. Davis, The Chemistry of Powder and Explosives, J. Wiley, New York, 1943.

105. G. W. C. Taylor and A. T. Thomas (1957), according to F. P. Bowden and A D Yoffe
[116].

106. F. W. Brown, D. J. Kusler and F. C. Gibson, U.S. Bureau of Mines, Techn. Paper 3852,
Washington, 1946.

107. W. E. Garner, Trans. Faraday Soc. Symposium of Solid Phase Reactions (1938).

108. W. Feitknecht and M. Sahli, Helv. Chim. Acta 37, 423 (1954).

109. CIOS-XXVII 38, Manufacture of Initiating Explosives.

110. J. Meissner, Brit. Pat. 500711 (1939).

111. J. Meissner, private information (1964).

112. Private information.

113. A. WlSCfflN, Proc. Roy. Soc. (London) A 172, 314 (1939).

114. J. SAWKILL, Proc. Roy. Soc. (London) A 172, 135 (1955).

115. J. S. Courtney-Pratt and G. T. Rogers, Nature 175, 652 (1955).

116. F. P. Bowden and A. D. Yoffe, Fast Reactions in Solids, p. 88, Butterworths, London, 1958.

117. G. T. Rogers, Ph.D. Thesis. Cambridge, 1955, according to [116].

118. J. H. McAuslan, Ph.D. Thesis, Cambridge, 1957, according to [116].

119. B. L. Evans, Ph.D. Thesis, Cambridge, 1957, according to [116].

120. P. Gray and T. C. Waddington, Chem. Ind. 1255 (1955).

121. R. Audubert, Trans. Faraday Soc. 35, 197 (1939); /. Chem. Phys. 49, 275 (1952).

1 22. E. W. Haycock - see Chemistry of the Solid State by W. E. Garner (Ed.), p. 238, Butterworths,
London, 1955.

123. B. E. Bartlett, F. C. Tompkins and D. A. Young, Discussion Roy. Soc, Initiation and
Growth of Explosion in Solids, Proc. Roy. Soc. (London) A 246, 206 (1958).

124. C. Taylor and W. H. Rinkenbach, Army Ordnance 5, 114 (1925); Z.ges. Schiess- u. Spreng-
stoffw. 20, 114(1925).

125. M. STRAUMANlsand A. Cirulis, Z. anorg. Chem. 251, 315, 332, 335 (1943); 252, 121 (1943).

126. F. Martin, Vber Azide und Fulminate, Darmstadt, 1913.

127. A. T. Thomas, unpublished work (1957), according to F. P. Bowden and A.D. Yoffe[116].

128. P. Gray and T. C. Waddington, 27th Intern. Congres Chimie Industrielle, Brussels, 1957.

129. S. K. Deb, unpublished work (1957), according to F. B. Bowden and A. D. Yoffe [116].

130. W. E. GARNER and D. J. B. Marke, /. Chem. Soc. 1936, 657.

131. J. Maggs, Trans. Faraday Soc. 35, 433 (1939).

132. D. J. B. Marke, Trans. Faraday Soc. 33, 770 (1937).

133. K. Singh, Trans. Faraday Soc. 55, 124 (1959).

134. S. K. Deb and A. D. YOFFE, Proc. Roy. Soc. (London) A 256, 514 (1960).

135. Jahresber. Chem.-Techn. Reichsanstalt 6, 80 (1927); 8, 102 (1929).

136. P. W. M. Jacobs, F. C. Tompkins and D. A. Young, Disc. Faraday Soc. 28, 234 (1959).

137. H. FlCHEROULLE and A. Kovache, Mem. poudres 33, 7 (1956).

138. P. L. GOnther, J. Porger and P. Rosbaud, Z. physik. Chem. (B) 6, 459 (1930).

139. M. Berthelot, Ann. chim. [6], 28, 138 (1893).

200 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

140. M. BERTHELOT and P. VlEILLE, Bull. soc. chim. France [3], 11, 744 (1894).

141. E. WlBERG, Z. Naturforsch. 9b, 495 (1954).

142. C. Grundman and R. RAtz, Z. Naturforsch. 10b, 116 (1955).

143. T. Curtius and F. Schmidt, Ber. 55, 1571 (1922).

144. H. A. Lehman and W. Holznagel, Z. anorg. Chem. 293, 314 (1958).

145. W. Sundermeyer, Chem. Ber. 96, 1293 (1963).

146. T. UrbaSski and A. Rusiecki, Wiad. Techn. Uzbr. 26, 442 (1934).

147. M. O. Forster and H. E. Fierz, /. Chem. Soc. 93, 1867 (1908); M. O. Forster and S. H.
Newman, J. Chem. Soc. 97, 2573 (1910).

148. T. Curtius, J. prakt. Chem. 91, 1, 415 (1915).

149. T. Curtius and K. Heidenreich, Ber. 27, 2684 (1894); J. prakt. Chem. [2], 52, 472 (1895).

150. A. D. G. France and R. A. Jeffreys, Chemistry & Industry 1962, 2065.

151. P. A. S. Smith, The Curtius Reaction in R. ADAMS (Ed.) Organic Reactions,\ol 3, p. 337,
J. Wiley, New York, 1946.

152. J. MAI, Ber. 24, 3418 (1891); 25, 372, 1685 (1892); 26, 1271 (1893).

153. F. Sorm, Chem. Obzor 14, 37 (1939).

154. A. PURGOTTI, Gazz. chim. ital. 24, 1, 554 (1894).

155. F. SCHRADER, Ber. 50, 777 (1917).

156. A. KORCZYNSKI and S. NAMYSLOWSKI, Bull. soc. chim. France [4], 35, 1186 (1924).

157. H. Rathsburg, Ger. Pat. 341961 (1919).

158. V. V. Vasilevskii, F. I. Blokhshtein and B. D. Kustria, Zh. obshch. khim. 5, 1652 (1935).

159. O. TUREK, Chimie et industrie 26, 781 (1931).

160. J. H. BOYER, R. S. Buriks and U. Toggweiler, J.Am. Chem. Soc. 82,2213 (1960); J. H.
Boyer and R. S. Buriks, ibid. 82, 2216 (1960).

161. E. Ott and E. Ohse, Ber. 54, 179 (1921); Ger. Pat. 350564 (1919); Chem. Zentr. 1922 II, 1194;
Ger. Pat. 355926 (1920); Chem. Abstr. 17, 1242 (1923); U.S. Pat. 1390387 (1921); Brit. Pat.
170359 (1921); Chem. Zentr. 1922 II, 1118.

162. H. MURAOUR, Bull. soc. chim. France 51, 1152 (1932).

163. C. TAYLOR and W. H. Rinkenbach, J. Franklin Inst. 19b, 551 (1923).

164. B. Glowiak, Bull. Acad. Polon. Sci., sir. chim. 8, 5 (1960).

165. W. H. Saunders and J. C. Ware, Jr., /. Am. Chem. Soc. 80, 3328 (1958).

166. W. T. Reichle, Inorg. Chem. 3, 402 (1964).

167. J. H. Boyer and F. C. Canter, Chem. Rev. 54, 1 (1954).

168. P. Gray, Quart. Rev. 17, 441 (1963).

169. J. E. Gowan andT. S. Wheeler, Name Index of Organic Reactions, Longmans, London, 1960.

170. O. Dimroth, Ber. 35, 4041 (1902); Ann. 399, 91 (1913).

171. J. H. BOYER et al., J. Am. Chem. Soc. 81, 4671 (1959); /. Org. Chem. 25, 286, 458 (1960).

CHAPTER IV

OTHER INITIATING EXPLOSIVES

DIAZO COMPOUNDS

Berthelot and Vieille [1] examined the explosive properties of diazobenzene nitrate
(C 6 H 5 N = N)®NOf . Later Wohler and Matter [2] demonstrated that it is unsuitable
for use as an initiator due to its very weak initiating properties and very high sensi-
tiveness to friction and impact. Herz [3] suggested the use of m-nitrodiazobenzene
perchlorate. This, however, is hygroscopic and insufficiently stable (it explodes at
154°C).

DINITROBENZENEDIAZO-OXIDE
(DINITRODIAZOPHENOL)

The only diazo compound of practical value is dinitrobenzenediazo-oxide (dini-
trodiazo-oxide, or less correctly dinitrodiazophenol). In technical literature it may
be denoted as DDNP or Dinol.

The following formulae have been ascribed to this substance: cyclic (I) (Bamberger
[4]), diazonium (II) (Hantzsch and Davidson [5], Klemenc [6]) or quinonoid (III)
(Wolff [7]).

II N ill

\/\c/ Y^o

N0 2 N0 2

I II

In modern transcription formulae (II) and (III) take the forms (Ha) or (Ilia)
(Hodgson and Marsden [8], Anderson and Le Fevre [9]):

OzN^^N^N 2 N.

Ilia
[201]

202 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Recently from a comparison of the infra-red absorption spectrum of this com-
pound with that of o-benzoquinone Glowiak [10] came to the conclusion that dinitro-
benzenediazo-oxide has a quinonoid structure. Both substances show the presence
of the strong absorption band of the carbonyl group: 1666 cm -1 for dinitrobenzene-
diazo-oxide and 1680 cm -1 for o-benzoquinone. In addition dinitrobenzenediazo-
oxide gives a band with a frequency of 2190 cm -1 , characteristic of a double bond
between nitrogen atoms. (Some derivatives of this compound may also have the
diazo structure (Ha), which is discussed later on.)

This substance was the first diazo compound to be discovered. It was prepared
by Griess [11] by diazotizing picramic acid. Its explosive properties attracted the
attention of Lenze [12] who found it to be as valuable as mercury fulminate in spite
of its higher sensitiveness to impact. This compound is also of interest as being the
first initiator containing no heavy metals. It has now been utilized in the United
States of America and Japan as a component of initiating charges in detonators
and caps.

Physical properties

Dinitrodiazophenol has a specific gravity of 1.63, and occurs as yellow needles
which decompose without melting on heating to 188°C.

Its crystalline form is of great importance from the practical point of view.

The needle-like shape of the crystals prevents their being easily poured (into
the capsule) making them liable to felt so the aim is to produce short crystals either
by a suitable selection of conditions for the reactions of diazotization and precipita-
tion of the product (D. Smolehski and Plucinski [13]) or by the addition of certain
substances to the solution from which the product is to be precipitated; Garfield
[14], for instance, suggests for this purpose the addition of triphenylmethane dye-
stuffs to the solution.

The physical and explosive properties of dinitrodiazophenol were investigated
by Clark [15] and by D. Smolenski and Plucinski [13]. The solubility of dinitro-
diazophenol at 50°C (in 100 g of solvent) is: 2.45 g in ethyl acetate, 1.25 g in methyl
alcohol, 2.43 g in ethyl alcohol, 0.23 g in benzene, 0.11 g in chloroform.

The substance is also soluble in concentrated hydrochloric acid, acetone, acetic
acid, nitrobenzene, aniline, pyridine, and nitroglycerine, at room temperature. In
water its solubility is only 0.08% at 25°C.

Chemical and explosive properties

Dinitrodiazophenol is not decomposed by concentrated acids at room tem-
perature, but on the other hand a dilute (e.g. 0.5%) solution of sodium hydroxide
causes its decomposition with the evolution of nitrogen even at room temperature.
This property finds application in the destruction of residues.

OTHER INITIATING EXPLOSIVES 203

Dinitrodiazophenol is more stable than mercury fulminate. It may be stored
without change at 50°C, in dry condition, for 30 months (under these conditions
fulminate is stable only for 9 months) and under water for 12 months.

Vaughan and Phillips [16] investigated the decomposition of dinitrodiazophenol
at temperatures between 111 and 120°C in vacuo. The gaseous products of decom-
position contain: 61.5% of N 2 , 3% of NO, 4.0% of N0 2 , 2.5% of CO, 28% of
C0 2 . This is evidence of the fact that decomposition consists not only in the loss
of diazo group nitrogen, but also in the decomposition of the benzene ring.

Smolenski and Plucinski [13] examined the effect of sunlight and found that signs
of decomposition are perceptible after only 40 hr of irradiation. A sample so irradi-
ated shows a somewhat lower ignition temperature.

Kaufman [17] found that ^-radiation produces gas evolution from dinitrodiazo-
phenol. Partial decomposition occurred after 45 days of irradiation (on average
10 5 r per hour) and the explosive power of the irradiated substance was reduced
and irregular.

Dinitrodiazophenol explodes [13] on a metal plate at 180°C after 10 sec; at
185°C after 5 sec; at 190°C after 2.5 sec; at 200°C after 1 sec.

In spite of its high specific gravity the apparent density of the needle-shaped
crystals, according to Clark [15], is only 0.27; under a pressure of 240 kg/cm 2 it
is 0.86. The crystals may be obtained in the form of pellets whose apparent density
is about 0.8. The effect of the conditions of preparation have been thoroughly exam-
ined by Smolenski and Plucinski [13]. They found that at a diazotization temperature
as recommended by Clark, i.e. 15°C, the product pours with difficulty. Conversely,
diazotization at a higher temperature (25-45°C) results in formation of a product
with a density of about 0.82.

Smolenski and Plucinski prepared dinitrodiazophenol in the form of free-flowing
crystals by applying the following reaction conditions:

A solution of 320 g of sodium nitrite in 2 1. of water is added to a suspension
of 1000 g of the sodium salt of picramic acid in 8 1. of water. Next, 6 1. of 5.5%
hydrochloric acid is added dropwise for 2 hr, stirring continuously. The initial tem-
perature of 20°C rises to 25°C. Completion of the reaction is determined by means
of starch-iodide paper. The product is filtered off, washed with cold water and dried
at 35-40°C. Its yield amounts to 80% of the theoretical.

T. Urbahski, Szyc-Lewanska et al. [18] have recently found that dinitrobenzene-
diazo-oxide can be prepared by oxidation of picramic acid with chromium trioxide
in the presence of sulphuric acid at 55-60°C. One part of picramic acid is fully
oxidized by chromic acid to yield gaseous products: CO, C0 2 , N0 2 , NH 3 and
H 2 0. Nitrogen dioxide acts further as a diazotizing agent on undecomposed picramic
acid to yield the diazo compound. The yield of this reaction does not exceed 31%
of theoretical calculated on the picramic acid used.

Clark confirmed that dinitrodiazophenol does not become "dead pressed" even
under a pressure of 9140 kg/cm 2 which is a great advantage in an initiating material.
Smolenski and Plucinski quote the following figures which are characteristic of its

204 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

explosive properties. The substance is endothermic. Its heat of formation is about
365 kcal/kg. The reaction proceeds according to the following equation :

C28.57H9.52O23.8N19.0 = 20CO+ 1 .2C0 2 + 1 .4H 2 + 1 .5H 2 + 3.65C+ 3.72HCN + 7.64N 2

+955.3 kcal/kg
The volume of gases (Vo) 876 l./kg

The temperature of explosion 3700 G C.

Owing to the great volume of gases, high heat of formation and high explosion
temperature this substance is a much stronger explosive than those initiators which
contain metal in the molecule.

According to Clark [15] 1 g of dinitrodiazophenol on being pressed at a pressure
of 240 kg/cm 2 into a copper capsule gives an expansion of 25 cm 3 in a small lead
block (mercury fulminate 8 cm 3 , lead azide 7 cm 3 ). Using 0.75 g of the substance,
Smolehski and Plucihski obtained a lead block expansion of 17-23 cm 3 .

Clark found dinitrodiazophenol less sensitive to impact than mercury fulminate,
lead azide or lead styphnate; it is exploded by a drop of 375 g weight from a height
of 22.5 cm, whereas mercury fulminate is exploded by a drop of 15 cm.

Smolehski and Plucihski also disclosed that a fine crystalline product is more
sensitive (15 cm drop) and coarse crystalline less sensitive (30 cm drop).

A. Belayev and A. Belayeva [19] found its linear rate of burning to be 2.15 cm/sec.

The initiating properties of dinitrodiazophenol were investigated by Clark [15]
and Smolehski and Plucihski [13]. According to Clark the initiating power of the
substance is approximately twice as great as that of mercury fulminate, but a little
less than that of lead azide.

Thus to initiate picric acid the following quantities of priming explosives are
required :

0.115 g of dinitrodiazophenol
0.225 g of mercury fulminate
0.12 g of lead azide

For trinitrotoluene the corresponding figures are as follows :

0.163 g, 0.240 g, 0.16 g
and for tetryl :

0.075 g, 0.165 g, 0.03 g.

In Smolehski and Plucihski's opinion dinitrodiazophenol alone is not suitable
as an initiating material for detonators since it requires too long a path for burning
to change into detonation, hence it is necessary to add another initiating substance
e.g. lead azide. Nevertheless it is suitable for filling caps.

The properties of benzenediazo-oxides

Vaughan and Phillips [16] studied the decomposition of 4-diazo-l -oxide (III) and
nitro derivatives of this compound (IV, V) and of the nitro derivatives of 2-diazo-l-
oxide (VI, VII, VIII).

OTHER INITIATING EXPLOSIVES 205

Nr^N N=>N N=;N

III IV V

Their experiments showed that the nitro derivatives of 4-diazo-l -oxide are more
stable than the corresponding derivatives of 2-diazo-l -oxide.

The introduction of a nitro group at the or/Ao-position to the oxygen atom
in 4-diazo-l-oxide (IV) increases the stability. Great stability is also demonstrated
by the o- and ^-substituted nitro derivatives of 2-diazo-l -oxide (VI and VII). On
the other hand, the m-substituted compound (VIII) has a lower stability than com-
pounds (VI) and (VII). Dinitro substituted derivatives, ortho-ortho (V) and ortho-
para (Ha), are distinguished by a higher stability than the mononitro derivatives
of the same oxides (IV) or (VI) and (VII).

Glowiak [20] examined the properties of the diazotization products of numerous
nitro derivatives of o-aminophenol, viz. :

2,6-dinitro-4-amino-m-cresol
4,6-dinitro-2-amino resorcinol
5-nitro-3-aminosalicylic acid
4,6-dinitro-2-amino-m-hydroxybenzoic acid.

From them he prepared the diazo compounds (IX), (X), (XI), and (XII).

From their infra-red absorption spectra he ascribes a quinonoid structure to the
first two compounds and a diazo structure to the last two and to their plumbous
salts :

CH 3 O COOH COOH

A/ N ° 2 ° 2N \A^ N2 /\/° e

B ^o V\h o 2 n/ / v / \n©

N 2 N0 2

IX X XI

The compounds with a quinonoid structure differ from those with a diazo struc-
ture by their darker colour and lower chemical stability. They are for example
easily decomposed by light and concentrated acids; they are less resistant to heat
and show a higher sensitiveness to impact, friction and flame than compounds
with a diazo structure.

Phenyldiazonium nitroformate. Nitroform derivative (XIII) is of particular interest
among derivative diazonium salts. It was prepared by Ponzio [21] who reacted the
potassium salt of trinitromethane (nitroform) with an aqueous solution of phenyldia-
zonium acetate:

(N0 2 ) 2 C=NOOK + CH3COO N 2 C 6 H 5 -> (NC>2) 2 C=NOO N 2 C 6 H,

XIII

206 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Quilico [22] questioned the formula (XII), assuming the substance to be an azo
compound with the structure (N0 2 ) 3 C— N=NAr, but in a later paper Ponzio [23]
insists that in principle his formula is correct, while modifying it into (N0 2 ) 3 CN 2 C 6 H 5
(XHIa).

On the basis of its infra-red absorption spectrum Glowiak [20] deduced an ionic
structure (XHIb), confirming, in principle, the formula (XHIa) :

C(N0 3 ) e

XHIb

According to Glowiak's researches this substance has the following pro perties
Its ignition temperature with an induction period of 1 sec is 124°C. To initiate
0.5 g of tetryl, compressed under a pressure of 500 kg/cm 2 , the minimum charge
of the phenyl diazotate of nitroform, loosely poured or compressed under a pressure
of 100 kg/cm 2 , is 0.3 g. When compressed at higher pressure (300 kg/cm 2 ) the charge
must be increased to 0.5 g.

Dinitrobenzene diazo-oxide reacts with hydrazine hydrate to yield 2,4-dinitro-
6-[tetrazene-(l)]-phenolhydrazine salt (XIV)

OH(NH 2 NH 2 )

I
2 N-A-N=N-NHNH 2

N0 2

XIV

The product (XIV) can form metal salts. Some of them (e.g. potassium salt)
possess initiating properties [24].

THE DERIVATIVES OF AMINOGUANIDINE

TETRAZENE

N— K

II ^>C— N=N— NH— NH— C— NH 2 -H-,0

N-NH i H

Tetrazene or tetrazolylguanyltetrazene hydrate was first prepared by Hoffmann
and Roth [25], by the action of a neutral solution of sodium nitrite on aminoguani-
dine salts (without an excess of inorganic acid). According to these authors the
reaction proceeds as follows :

HN X JH 1 r Hi. NH

C— NH— N

+ HOJ

— NiO

HzN' Net ! H

!>N-C<(

NH-

H I -> tetrazene + 3H 2

+ HO NO

OTHER INITIATING EXPLOSIVES 207

The course of the reaction in the presence of inorganic acids is different from
that in the presence of acetic acid.

ye— NH— NH— N=N— cf
H 2N X \NH— NH— NO

Hoffmann etal. [26] suggested the structural formula (la) for tetrazene, i.e.
l-guanyl-4-nitrosoaminoguanyltetrazene. The correctness of this formula was later
questioned by Patinkin, Horwitz and Lieber [27]. The synthesis of tetrazene by the
action of tetrazolediazonium hydroxide (II) on aminoguanidine salts (III) at 0°C,
suggested that tetrazene has the structure of l-(5-tetrazolyl)-4-guanyltetrazene hyd-
rate (I) :

J )>C-N 2 OH + NH 2 -NH-C-NH 2 -> II ~~ ^CN=N-NH-NH-C-NH 2 -H 2
N-NH NH N-NH

II III

The formation of guanyl azide (IV) at the first stage may account for the forma-
tion of tetrazene by the action of nitrous acid on aminoguanidine, i.e. by Hoffmann
synthesis.

^ T „ HONO

NH 2 — C— NH— NH 2 > NH 2 — C— N 3

NH (IV)

v ,N N

NH 2 -C<f ||

\NH— N

(V)
Guanyl azide is then isomerized to aminotetrazole (V) which undergoes diazo-
tization and couples with aminoguanidine as stated above.

Rathsburg [28] suggested the use of tetrazene in explosive technology.
Tetrazene is a light, crystalline substance, s.g. 0.45, colourless or pale yellow,
practically insoluble in water and in the majority of organic solvents. It is only
slightly hygroscopic (it absorbs 0.77% of moisture, at 30°C, in an atmosphere with
a relative humidity of 90%). It has basic properties and is soluble in concentrated
hydrochloric acid. Tetrazene hydrochloride may be precipitated from such a solu-
tion with ether. Free tetrazene is evolved from the hydrochloride by reaction with
sodium acetate or ammonia. With an excess of silver nitrate, tetrazene gives a precipi-
tate of the double salt C 2 H 7 N 10 OAgAgNO 3 -3H 2 O.

It is stable at ordinary temperatures, whether wet or dry, but hydrolyses on
being boiled in water with the evolution of 2 N 2 per molecule.

Under the influence of sodium hydroxide, tetrazene undergoes decomposition
with the evolution of ammonia, cyanamide and triazidonitrosoaminoguanidine (VI) :
N— K

/>C-N=N— NH— NH— C— NH 2 H 2 -> N 3 — C— NH— NH— NO + NH 3 + NH 2 CN

N-NH II II

NH NH

208 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Substance (VI) reacts in an enol form (Via) :

N 3 — C— NH— N=N— OH

Via

On the addition of copper acetate to a solution of compound (VI) the copper
salt (VII) is formed which on treatment with acid decomposes to form 5-azido-
tetrazole (VIII):

.NH— N=N /NH— N

N 3 — C< | -> N 3 — C\ || +CuO

N— Cu— O ^N N

VII VIII

Tetrazene is stable at temperatures up to 75°C. At 100°C it undergoes marked
decomposition. The ignition temperature of tetrazene is lower than that of mercury
fulminate. On a metal plate, heated to 160°C, it explodes after 5 sec (mercury ful-
minate behaves in the same way at 190°C). According to Wallbaum [29] tetrazene
explodes at 140°C on being heated at the rate of 20°C/min.

The explosion heat of tetrazene is rather low, i.e. 663 kcal/kg. This is charac-
teristic of explosive substances containing a guanyl group in the molecule.

According to some authors tetrazene is rather more sensitive to impact than
mercury fulminate. Others consider it to be equally sensitive.

The ease with which tetrazene is detonated by ignition depends to an exception-
ally great extent on its density. It has been shown that tetrazene detonates most
easily when it is poured freely into the capsule ; when pressed it gives a much weaker
detonation. Rinkenbach and Burton [30] obtained the following data in a sand
test, using a 0.4 g charge of tetrazene :

Table 37

Tetrazene charge

pressed under

Crushed sand,

a pressure of

kg/cm 2

13.1

16.7

9.2

7.5

200

2.0

Thus, at a pressure of 200 kg/cm 2 the substance nears the condition of being
"dead pressed". In spite of the fact that burning under this condition passes to
detonation with difficulty, when greatly compressed the material maintains its ability
to be detonated by a cap. Thus, 0.4 g of tetrazene, pressed under a pressure of
200 kg/cm 2 , develops its maximum power, i.e. 21.1 g of sand crushed, when initiated
with 0.4 g of mercury fulminate. The difficulty in passing from burning to detonation
makes tetrazene unsuitable for detonators and its application is thus limited to

OTHER INITIATING EXPLOSIVES 209

ignition caps, where even 2% in the composition results in improved uniformity of
percussion and friction sensitiveness and makes it suitable as a sensitizer for friction
compositions.

The explosive properties of tetrazene perchlorate are also of interest.

Tetrazene manufacture

In the Wolfratshausen factory [31] in Germany, an 8% solution of sodium
nitrite and a 12.5% solution of aminoguanidine sulphate, slightly acidified with
acetic acid in the presence of litmus, were used for the reaction.

Into a reactor of the type used for the manufacture of lead azide (cf. Fig 49)
50 1. of a solution of sodium nitrite (4 kg of NaN0 2 ) were introduced and heated
to a temperature of 50-55°C. To the warm solution 401. of a solution of amino-
guanidine sulphate, containing 5 kg of the dry substance, was added during a period
of 1-2 hr. The rate at which the solution was introduced influences the dimensions
of the crystals formed. If the solution was introduced rapidly, small crystals resulted,
if it was introduced slowly, the crystals were large.

In some factories (e.g. at Stadeln [32]) dextrin was added to the reacting solu-
tions to obtain more uniform crystals. After a solution of aminoguanidine sulphate
has been added, the contents of the reactor were stirred for a further 30 min, when
the stirrer was stopped. The precipitate of tetrazene settled on the bottom, the
liquid from above the precipitate was decanted, the precipitate itself was covered
with water, agitated, allowed to stand, decanted and finally transferred by a powerful
stream of water onto a cloth filter. The product was washed first on the filter with
water and finally with alcohol. The tetrazene, containing alcohol, was then trans-
ferred in a bakelite vessel together with the filter. Washing with alcohol is necessary
for uniform drying of the product. If tetrazene containing only water is dried, the
crystals are liable to stick together due to the low solubility of tetrazene in water.
The presence of alcohol completely prevents this. Tetrazene was dried in the usual
way (cf. p. 155) at a temperature from 45 to 55°C.

From the above-quoted amounts of raw material 2.6-2.7 kg of product is
obtained.

The waste tetrazene, collected in settling tanks (vessels), is destroyed by injection
of live steam.

Other reactions of aminoguanidine with nitrous acid

During the reaction of tetrazene preparation care should be taken that the
solution contains neither free inorganic acid nor an excess of acetic acid.

When guanidine sulphate is treated with sodium nitrite, a molecule of sulphuric
acid combined with guanidine is sufficient for the liberation of nitrous acid from
sodium nitrite in an amount necessary for the reaction.

210 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In the presence of an excess of inorganic acid, the reaction proceeds differently
and the formation of azidoguanidine (IV) as a final product takes place

NH 2 — C— NH— NH 2 + HONO -> NH 2 — C— N 3 + 2H z O

II II

NH NH

This substance forms salts with acids. The nitrate, perchlorate and picrate of (IV)
have explosive but not initiating properties.

Azidoguanidine is not decomposed by boiling in water. It is, however, hydrolysed
on treatment with concentrated sodium hydroxide to form sodium azide. Under
the influence of a dilute solution of sodium hydroxide or of dilute acids it is isomerized

to 5-aminotetrazole (V):

,N N

NH2-C-N3 -> NH 2 -C<f ||

|| \NH— N

Attempts have been made to employ aminotetrazole as a constituent of smokeless
and fiashless propellants.

The reaction of aminoguanidine with nitrous acid in the presence of an excess
of acetic acid gives also 1 ,3-ditetrazyltriazine (IX). It is possible that 5-aminotetr-
azole first arises, which then undergoes diazotization and the diazo compound
thus formed couples with the remaining aminotetrazole:

N— N^ mono N— N. e

II \N— NH 2 > I >C— N 2 CH 3 COO ->

II / CH3COOH II /

N— NH N— NH

N-N JS N

|| X C __ N==N _ NH — C<f ||

II / ^NH — N

N— NH

NITROSOGUANIDINE

NH— NO NH 2

C=NH or C=N— NO

^•NHz ^NH 2

Davis and Rosenquist [33] suggested the use of nitrosoguanidine as a weak
initiator. This compound occurs in the form of pale yellow crystals which explode
on being heated to 165°C.

According to these authors, nitrosoguanidine is prepared by the reduction of
nitroguanidine with zinc dust in a neutral medium, in the presence of ammonium
chloride at room temperature (below 20-25°C). The product is filtered off together
with a precipitate of zinc oxide and zinc salt, from which it is then extracted with
hot (65°C) water. Nitrosoguanidine crystallizes when the solution is cooled to 0°C.
The yield is about 50%.

OTHER INITIATING EXPLOSIVES 211

Nitrosoguanidine decomposes explosively on contact with sulphuric acid. When
dry, it is very stable, but it decomposes in the presence of water and in a moist
atmosphere. Its lack of stability prevents its practical application.

CYANAMIDE SALTS

Silver cyanamide

NAg 2

Chretien and Woringer [34] described the preparation of silver cyanamide from
calcium cyanamide by the action of silver nitrate and also described its explosive
properties. Montagu-Pollock [35] described a method for growing large crystals of
the salt from its aqueous solution in the presence of ammonium nitrate, ammonia
and a surface active agent. Bowden and Montagu-Pollock [36] and Montagu-Pol-
lock [35] studied the slow decomposition of the crystals when heated at temper-
atures from 150 to 360°C. The course of decomposition was studied by electron
microscope.

The main conclusions from this work were:

(1) Nucleation by metallic decomposition products was observed only in the
special cases involving the initiation of holes of crystallographic shape.

(2) In general, decomposition was found occurring everywhere on the crystal
surface.

(3) The silver produced by the decomposition was very mobile on the crystal
surface.

. (4) Boundaries appeared separating areas of greater and lesser decomposition.

NITROCYANAMIDE SALTS
NH— N0 2

The explosive properties of the potassium salt of nitrocyanamide first attracted
McKay's [37] attention. He separated this potassium salt as a by-product from the
preparation of aliphatic diazo compounds by the hydrolysis of N-alkyl-N-nitroso-
N'-nitroguanidines with an aqueous solution of sodium hydroxide at temperatures
from 0°C to room temperature:
^Alkyl

, ' N ° one /N /NH-N0 2

C=NH > Alkylene< II + &i

^NH— N0 2

212

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The same author et al. (McKay, Hatton, G. W. Taylor [38]) prepared a number
of nitrocyanamide salts by the action of the chlorides of various metals on the silver
salt in suitable solvents or the by reaction of the carbonates of various metals
with a solution of nitrocyanamide in acetonitrile, and found that some of these
salts have initiating properties. The majority, however, cannot be recommended
for practical use, their sensitiveness to impact being too low (they are not exploded
by a 5 kg weight falling from a height of 300 cm).

Harris [39] reports that only the following nitrocyanamide salts possess greater
sensitiveness and may be considered to be of practical use: potassium, tin, lead,
barium and silver nitrocyanamides.

Comparing with the experiments of Grant and Tiffany [40] Harris states that
silver and barium nitrocyanamides show an initiation capacity the same as that
of an 80 : 20 mixture of mercury fulminate and potassium chlorate, but weaker
than that of a mixture of lead azide and lead styphnate. This can be seen from Table
38 quoted by Harris:

Table 38

Minimum initiating charge

Primary explosive

necessary for explosion

Sand crushed

of 1.25 gof tetryl

per g

Lead azide-lead styphnate

(80:20)

0.15

110

Diazodinitrophenol-potassium

chlorate (75: 25)

0.25

103

Mercury fulminate-potassium

chlorate (80: 20)

0.40

Silver nitrocyanamide

0.45

Barium nitrocyanamide

0.50

In spite of their fairly promising initiating properties, nitrocyanamide salts are
of no practical use due to their high hygroscopicity.

For instance, due to its hygroscopicity, the air-dried tetrahydrated lead salt
is not exploded by a 5 kg weight falling 325 cm. Only after being dried over magnesi-
um perchlorate it is exploded with 100% probability by a 15 cm drop, whereas
the anhydrous lead salt is exploded by a 10 cm drop.

NITROPHENOL SALTS

LEAD PICRATE

A number of salts of picric acid have been described already (Vol. I). Some
salts of polynitrophenols and of heavy metals have initiating properties. One of the
earliest known substances of this kind is lead picrate. Its high sensitiveness to the
action of mechanical impact, however, raised difficulties in its practical utilization.

OTHER INITIATING EXPLOSIVES 213

T. Urbanski and Kruszynska [41] made a comparative study of the sensitiveness
to impact of lead picrate and other initiating explosives. They found lead picrate
to be more sensitive than any other substance. They also examined the decomposi-
tion of lead picrate on hot metal plates. On contact with a metal surface heated to
341 °C it explodes after 3 sec, and on one heated to 370°C after 1 sec.

From their results the authors calculated the activation energy of the thermal
decomposition which leads to explosion of this substance, and obtained a value of
55.6 kcal/mole.

During World War II the Germans employed, on a small scale, cap compositions
containing lead picrate for the manufacture of electric fuses giving few gaseous
products.

The lead picrate for this purpose was produced in the following way [42]. Into
a stainless-steel reactor equipped with a stirrer of the type used for the manufacture
of lead azide and other initiators (cf. Fig. 49) 8 1. of a solution containing 1.44 kg
of lead nitrate and 15 1. of ice water were poured. Fifteen litres of a solution contain-
ing 1.5 kg of picric acid were then added. During the reaction the temperature
should be maintained between 6 and 10°C. Since the temperature rises with the
precipitation of lead picrate, 7-8 more litres of ice water must be poured into the
reactor, usually a few minutes after the picrate has begun to precipitate. After 4 hr
the liquid was decanted from above the precipitate; the latter was transferred to
a cloth filter and washed with alcohol (10 1.) to which an aqueous solution of lead
nitrate (500 ml of a 30% solution) has been added to avoid the dissolution of lead
picrate during washing. 2.2 kg of product was obtainable from one batch.

The very fine crystalline percipitate of lead picrate which with water takes the
form of a paste, was then dried for 4 days, first at 40°C and finally at 60°C. The pro-
duct so dried was sieved through a silk screen of 600 mesh per cm2. The lead picrate
was then mixed with silicon and lead chromate.

LEAD STYPHNATE

-29
2 N ^

NO,

29
Pb-H 2

Lead trinitroresorcinate or lead styphnate is also known under the names:
Bleitrizinat, Trizinat in Germany and Teneres in the U.S.S.R.

It is usually prepared by adding a solution of lead nitrate to one of magnesium
styphnate. The latter is an easily soluble and weakly basic salt of trinitroresorcinate.
The use of suitable salts and conditions (pH, temperature, rate and sequence in
adding the raw materials) is of great importance, since unsuitable salts and condi-

214

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

tions may easily lead to the formation of basic lead styphnate, which has consider-
ably weaker initiating properties.

The basic salt is formed by the reaction of lead acetate with sodium or magnesium
styphnate. The chemical composition of the basic salts formed depends upon the
reaction conditions. The basic salt is usually credited with the structural formula
(II), corresponding to the dibasic salt:

OPbOH

2 N-/\-N0 2

I L-OPbOH

2 N

2 N-

N0 2

PbPb(OH) 2

2©

PbPb0 1.5H 2

III

OPbOH

2 N

IV V

Griess [43], who possibly dealt with the same salt, ascribed to it the structure III.

According to Zingaro [44] the dibasic salt (II) may be prepared by slowly adding
a solution of styphnate in a 2% solution of sodium hydroxide at 65-70°C to an
aqueous solution of lead nitrate. Zingaro also reports that structure (IV) is pos-
sible for this salt.

Finally there is the possibility of the formation of monobasic salt, probably
with the structure (V). Basic salts may be converted into neutral ones by the care-
fully controlled action of nitric acid.

The neutral salt has a characteristic reddish-brown colour, whereas the basic
salts are yellow.

Lead styphnate (neutral salt I) is practically insoluble in water (0.04 g in 100 ml
of water, at 15°C) and in the majority of organic solvents. It is very stable at room and
elevated temperatures (e.g. 75°C) and is not hygroscopic. In a moist atmosphere,
at room temperature, it absorbs only 0.05% of water. Its specific gravity is 3.1, its
apparent density, 1.0-1.6.

A method of preparation of the neutral salt (I) has been given by Herz [45].
He claims that the anhydrous salt crystallizes from an aqueous solution. In the
light of other authors' works (e.g. Zingaro), it is doubtful whether in such condi-
tions, an anhydrous salt can really be formed. The dehydration of neutral lead
styphnate (I) was investigated by Zingaro who found that complete dehydration
may be effected by heating the substance at 115°C for 16 hr. At higher temperatures
(135-145°C) dehydration takes place more quickly (Fig. 53). Stettbacher [46] repor-
ted that in a moist atmosphere anhydrous lead styphnate absorbs water to reform

OTHER INITIATING EXPLOSIVES

215

the hydrate. This observation was confirmed by Zingaro [44]. The hydration curve
at 30°C, according to Zingaro, is represented in Fig. 53.

Zingaro has also shown that the neutral lead styphnate, which is orange in
colour, reacts with pyridine at 50°C, to form pale yellow needles after approximately
1 hr. This is a molecular addition product of \ pyridine molecule to 1 mol. of basic
lead styphnate :

C 6 H 5 (N0 2 ) 3 OH(OPbOH) -iC 5 H 5 N

The neutral salt crystallizes with one molecule of water (according to some authors,
e.g. Rinkenbach [47]), or with half a molecule of water, which exerts a favourable
influence on the substance's sensitiveness to impact. This sensitiveness is relatively

Time, hr

FIg. 53. Loss of water of crystallization by the neutral salt of lead styphnate at various
temperatures: 7-145°C,2-135°C,3-115°C, ^-hydration curve at 30°C, according

to Zingaro [44].

small. Wallbaum [29] reports a drop test figure of 23 cm for a 1 kg weight, whereas
mercury fulminate is exploded by an 8-10 cm drop and the corresponding figure
for lead azide is 23 cm.

Research by T. Urbanski and Kruszynska [41] showed that lead styphnate
(neutral salt) is exploded in the drop test by a 2 kg weight performing work of
5.0 kgm/cm2, whereas lead picrate is exploded by work of 0.04 kgm/cm2.

The ignition temperature of lead styphnate is 267-268°C.

A number of authors have investigated the thermal decomposition of lead styph-
nate: Hailes [48] examined the decomposition of this substance within the temper-
ature range 200 to 228°C, and Garner, Gomm and Hailes [49] derived the following
equation for the decomposition curve:

p=ct m where

pressure developed by the decomposition products,

— constant,

— temperature,
exponent with values ranging from 0.88 to 4.43.

P
c
t
m

216

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Tompkins and Young [50] investigated the decomposition of the substance at
temperatures between 195 and 229°C. Explosive decomposition begins at a temper-
ature above 235°C. Decomposition curves as a function of time at various temper-
atures are reproduced in Fig. 54, the kinetic equation taking the form:

p=k(r-r ) 2 where

p — pressure developed by the decomposition products,

k — the rate constant,

t — time,

t — time of about 10 min, denoting the end of the process.

40 60 80
Time, (min)

Fig. 54. The decomposition of lead styphnate monohydrate. Pressure developed with

time at various temperatures: A— at 229.3°C,B— at 224.4°C, C— at 216.9°C,D— at

214.5°C, E— at 194.9°C, according to Tompkins and Young [50].

The activation energy of decomposition, according to Hailes, is 46.7 kcal/mole,
according to Tompkins and Young 33 kcal/mole and according to T. Urbanski
and Kruszynska [41] 42.2 kcal/mole.

Tompkins and Young confirmed Zingaro's [44] figure for the activation energy
of the dehydration of lead styphnate as 13 kcal/mole.

The inflexion in curves (A) and (B) occurs after a decomposition of about 30%
of substance. The plots of the acceleratory period are approximately parabolic. The
mechanism of the decomposition probably consists of nucleation of sub-grains at
the edges and progression of the reaction into the grains with a non-coherent inter-
face.

Evans and Yuill [51] showed that lead styphnate may be ignited by adiabatic
compression of air, the calculated temperature amounting to 660°C, whereas Bryan

OTHER INITIATING EXPLOSIVES

217

and Noonan [52] using helium as the gas surrounding the styphnate, found that
the energy to obtain the necessary compression for igniting this substance is 0.046 cal/
/cm 2 .

McAuslan [53] examined the ignition of lead styphnate by irradiation with the
intense light of a spark from an electric discharge. The light energy required to ini-
tiate lead styphnate to explosion was 29 J. This author also studied the relationship
between the light energy required to initiate lead styphnate and the temperature

50 100 150 200 250 300 350 400 450 500
Temperature ,°C

Fig. 55. Variation of the minimum light energy for ignition of lead styphnate with
ambient temperature, according to McAuslan [53]. x— ignition; o— no ignition.

of the substance. He obtained a straight line graph (Fig. 55). Extrapolation gives
a temperature of 480°C for zero light energy. This value is higher than the ignition
temperature of non irradiated lead styphnate.

Kaufman [17] found lead styphnate to be exceptionally resistant towards nuclear
radiation. After 90 days strong y-radiation from 198 Au (an average irradiation
of 10 5 r per hour) produced practically no change in the substance. The volume of
gas evolved after 90 days was only 0.12 ml/g.

High ignitability by direct action of a flame or electric spark is a characteristic
feature of lead styphnate. The salts enormous sensitiveness to the discharge of static
electricity was first disclosed in 1938 by Barcikowski, Dobrzynski and Kielczewski
[54].

This property has since been confirmed by many authors (Hartmann, Nagy
and Brown [55] ; Morris [56] ; and Taylor and Hall [57]). It became clear that numerous
accidents due to the ignition of lead styphnate during drying, pouring, moving,
stirring etc. had been caused by the discharge of accumulated static electricity.
Attempts to reduce its sensitiveness by the addition of graphite have been unsuc-
cessful, and lead styphnate continues to be very dangerous to handle.

Special attention is now paid to the careful earthing of all parts of the plant in
which dry lead styphnate is handled. Floors in factory buildings should be made

218

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

of asphalt or of a soft material (plastic) containing graphite or aluminium to make
them conduct electricity. When manual work with lead styphnate is unavoidable,
the operator should stand on a felt-cloth sheet saturated with a solution of calcium
chloride.

To prevent excessive dryness of the working atmosphere, which favours the
build up of static electricity, wet cloths should be hung about the building or air-
humidifying equipment used.

Fig. 56. Blasting caps: (a)— charged with a single explosive (mercury fulminate or

lead azide); (6)— charged with a primary explosive and a high explosive; (c)— charged

with three layers: a primary explosive and two high explosives.

A more modern procedure is to introduce by means of radioactive salts ionized
air into the premises.

Lead styphnate is a fairly weak explosive because of its high metal content
(44.25%).

The rate of detonation is :

at a density of 0.93 2100 m/sec
at a density of 2.6 4900 m/sec
at a density of 2.9 5200 m/sec

It is also very weak as a primary explosive. Even 1.0 g of it fails to initiate penth-
rite compressed under-a pressure of 2000 kg/cm 2 . It is therefore not used for filling
detonators, and its applications are limited to use in

(a) non-corrosive ignition caps (so called "Sinoxyd");

(b) addition to lead azide in detonators to facilitate ignition;

and (c) as a covering layer to protect lead azide against carbon dioxide and to
facilitate ignition (Fig. 56).

Lead styphnate manufacture

In the method employed at Wolfratshausen [31] a solution of magnesium styph-
nate was first prepared, adding to 20 kg of magnesium oxide a suspension (partly
solution) of 120 kg of trinitroresorcinol in 350 1. of water. The temperature rose

OTHER INITIATING EXPLOSIVES 219

spontaneously due to the reaction, but should be raised further by heating to 60°C.
The solution so obtained was filtered through a cloth filter, diluted with water to s.g.
1.043 (6°Be) and poured into a vat, in which it was allowed to stand for 10 hr. The
temperature then fell to 25-30°C.

From the magnesium styphnate solution so prepared, 86.4 1. of liquid was de-
canted, leaving the lower layer in which the sediment was collected. This solution
was heated to 60°C, while stirring, and 22.7 1. of 34% solution of lead nitrate, s.g.
1.274 (31°Be) was then poured into it during a period of 20-30 min, while stirring
continued and the temperature was maintained at 60°C. When the solutions were
mixed, the contents of the reactor were cooled as quickly as possible to 25°C; when
this temperature has been reached the stirrer was stopped and the precipitated sedi-
ment of lead styphnate was allowed to settle. The liquid from above the sediment
was then decanted, and the latter was first washed out of the reactor by a stream of
water, and transferred onto a cloth filter, where it was washed again as is the custom
with other primary explosives. From the above mentioned amounts of raw material
about 8 kg of lead azide was obtained.

The product was dried in a drier (as described for mercury fulminate) at temper-
atures from 65 to 70°C in batches of 1.2 kg at a time; it was then sieved as de-
scribed above. A sieve analysis of the product showed, for example, the following
sizes of crystals:

on sieves with a clearance of 0.1 mm 8 % of the substance was retained
on sieves with a clearance of 0.075 mm 33% of the substance was retained
on sieves with a clearance of 0.060 mm 32% of the substance was retained
on sieves with a clearance of 0.040 mm 18% of the substance was retained
and 9% of the substance passed through.

Waste lead styphnate was destroyed by adding an excess of sodium carbonate;
most of the lead was then precipitated as a carbonate and a solution of sodium
styphnate was formed. This solution was then treated with iron filings and acidified
with sulphuric acid, reduction of the nitro groups took place, and the substance
ceased to be dangerous.

The continuous method for the manufacture of lead styphnate (according to Meissner

[58, 59])

The continuous preparation of lead styphnate can be carried out in the same
equipment as for the manufacture of lead azide (Fig. 50) after previous cleaning of
the apparatus and exchanging the flowmeters.

For this production, the starting materials are needed: trinitroresorcinol and
magnesium oxide to form magnesium styphnate and lead nitrate, all disssolved in
distilled water.

220

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 57. Lead styphnate precipitated by Meissner method [59]. Courtesy J. Meissner.

The precipitating temperature is, in this production, somewhat higher than in the
manufacture of technical-grade lead azide.

All other operations are similar to those already described before.

Lead styphnate crystals obtained by this method are sphere-shaped (Fig. 57).

OTHER STYPHNATES

Tompkins and Young [50] examined the thermal decomposition of barium
styphnate and found it to be similar to that of lead styphnate hydrate (Fig. 54).

The point of inflexion of the decomposition curve occurs after 50% of the sub-
stance has decomposed. The energy of decomposition is 36.5 kcal/mole.

According to T. Urbanski and Kruszynska [41] thallium styphnate has similar
properties to those of lead styphnate. The former, however, is much more sensi-
tive to impact than the latter. Their sensitiveness to temperature is similar: contact
with a metal surface heated to 351°C causes explosion after 1 sec. The activation
«nergy of thermal decomposition is nearly 80 kcal/mole.

LEAD DINITRORESORCINATE

_ N0 2

2®

OTHER INITIATING EXPLOSIVES 221

This substance is prepared by the action of a hot solution of lead nitrate on a so-
lution of sodium dinitroresorcinate. The lead salt is precipitated as the solution
cools. Lead dinitroresorcinate ignites from a direct flame readily and burns with
great velocity. Its initiating action is weaker than that of lead styphnate and it is
less sensitive to impact and friction than styphnate. It has therefore recently found
application as a component of cap compositions. It is valuable because it is safer to
handle than styphnate. The preparation of dinitroresorcinol is discussed in Vol. I
p. 536.

NITROSOPHENOL SALTS

Nitrosophenols are formed very easily by the action of nitrous acid on phenols.
Some salts of heavy metals have weak initiating properties. These are: lead dini-
trosophenate and lead trinitrosophloroglucinate.

Due to the readiness with which it ignites under the direct influence of flame,
lead dinitrosophenate has found a certain application in the manufacture of cap
compositions for ignition by spark or flame. Its disadvantage lies in its relatively
low stability; heating to 120°C causes explosion after 2 hr, and after 10 days of
heating at 80 C C marked decomposition occurs.

NITRAMINE SALTS

* T. Urbanski, Piskorz and Mazur [60] prepared a number of methylenedinitra-

mine salts

,NH— N0 2
H 2 C<

X NH— N0 2

and found that the silver and lead salts have initiating properties. Their ignition
temperatures and sensitiveness to impact are shown in the following table:

Table 39

Salt

Ignition
temperature

50% of explosions by
2 kg weight dropped from

Silver
Plumbous

195°C

213°C

10 cm
12 cm

"ISONITRAMINE" (NITROSOHYDROXYLAMINE) SALTS

T. Urbanski, Zacharewicz and Pietrzyk [61] suggested the use of some methyl-
enedi-isonitramine salts (CH 2 ) (N 2 2 H) 2 as primary explosives.

The sodium salt of methylenedi-isonitramine was prepared by Traube [62] from
acetone and nitric oxide in the presence of sodium alcoholate according to the
following chain of reactions :

CH3COCH3 + 4NO + 2C 2 H 5 ONa -> CH 3 COCH(N 2 2 Na) 2 + 2C 2 H 5 OH (14)

222 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Product (I) so produced undergoes hydrolysis in water:

CH 3 COCH(N 2 2 Na) 2 + H 2 -» CH3COOH + CH2(N 2 2 Na)2 (15)

The sodium salt of methylenedi-isonitramine (II) is formed. In an aqueous solu-
tion this precipitates with the salts of heavy metals. Some of these salts (e.g. the
thallium salt) have the properties of weak initiators.

The structure of these compounds remained obscure for a long time. Traube
originally assigned the structure — N— N— OH to the isonitramino group, but in
further researches he found that isonitramines and the derivatives of nitrosohydro-
xylamine which he prepared by the action of nitrous acid on ^-derivatives of hydro-
xylamine were identical. He did not, however, draw from this any definite conclu-
sions as to the structure of isonitramines in spite of the fact that by the synthesis
of isonitraminoisobutyric acid, Gomberg [63] had confirmed the nitrosohydro-
xylamine structure of the isonitramino group.

Hantzsch [64] described isonitramines as compounds which have either the
structure reported by Traube (I) or that of nitrosohydroxylamine (Ha):

R— N— N— OH R— N— OH

V I

1 Ha

for which the tautomeric formula (lib) is also possible:

R— N-K)

N— OH

lib

Hantzsch expressed the opinion that the active hydrogen in isonitramines is
always combined with the oxygen atom whereas in nitramines it may be combined,
in different tautomeric modifications, either with the nitrogen or the oxygen atom.

In the light of Angeli's [65] investigations, which led to the conclusion that the
azoxy group has an unsymmetrical structure (— N=N— ), and not as formerly

assumed the symmetrical one of (— N=N— ), it did not seem possible to accept a

symmetrical structure for the isonitramine group (I).

Hantzsch and Strasser [66] came to this conclusion and assigned the structure
of nitrobenzyl-N-nitrosohydroxylamine to esters of nitrobenzylisonitramine :

2 N-C 6 H 4 CH 2 N(OR)NO (thus Ha)
or 2 N • C 6 H 4 CH 2 (NO)NOR (thus lib)

However, on investigating the ultra-violet absorption spectra of methylenedi-
isonitramine, R. N. Jones and Thorn [67] failed to detect a band characteristic of

OTHER INITIATING EXPLOSIVES 223

the nitroso group. This was accepted as evidence against the nitrosohydroxylamine
structure of these compounds. On the other hand, the methyl ester of methylenedi-
isonitramine has a spectrum typical of a nitro compound. Carmack and Leavitt
[68] confirmed the absence of the NO group band in the ultra-violet spectrum of
nitrosohydroxylamine derivative prepared by Cason and Prout [69] and ascribed
this to the existence of a hydrogen bond :

R— N H

X N=0"'''

III

This bond cannot, of course, exist in an ester, in which a band characteristic of
the nitroso group does occur.

Urbanski and Piskorz [60] found that the properties of methylenedi-isonitramine
salt differ markedly from those of methylenedinitramine salt. They also established
that methylenedi-isonitramine is not a tautomeric modification of methylenedinitr-
amine. In their most recent work [70] they have confirmed the nitrosohydroxylamine
structure of isonitramines by examining the infra-red absorption spectra.

Under the influence of aqueous solutions of the water soluble salts of heavy
metals the sodium salt of methylenedi-isonitramine gives precipitates of the salts
of these metals. T. Urbanski, Zacharewicz and Pietrzyk [61] suggested the use of
certain heavy metal salts as initiators. Particularly interesting properties were de-
monstrated by the thallous salt CH 2 (N 2 2 T1) 2 .

In another investigation, T. Urbanski and Wesolowski [71] studied the salts of
nitr omefhylisonitramine :

/ N0 2
CH 2
\n 2 o 2 h

According to Traube [62] the sodium salt of the aci-modification of this compound
is obtained by the action of nitric oxide on the sodium salt of nitromethane in
the presence of sodium alcoholate:

-NOONa
CH 2 =NOONa + 2NO + C 2 H 5 ONa -> CH + C 2 H 5 OH (16)

\sr 2 2 Na

Some of the heavy metal salts of nitromethylisonitramine appear to have initi-
ating properties, which are however considerably weakened by the presence of a
nitro group. The salts of type III are therefore weaker initiators than the corre-
sponding metal salts of methylenedi-isonitramine.

224 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

SALTS OF METAZONIC ACID

Metazonic acid is a nitroacetaldoxime:

CH 2 N0 2
CH=NOH

The aci-modification has the form

CH=NOOH

I
CH=NOH

It is produced by the action of sodium hydroxide on nitromethane at an elevated
temperature (Friese [72]; Steinkopf and Kirchhoff [73]). Urbariski and Kowalczyk
[74] found that some heavy metal salts of this compound have weak initiating
properties and that metazonic acid itself is a fairly weak explosive. Its expansion
in the lead block is 240 cm 3 .

SALTS OF OXALIC ACID

As early as 1883 Berthelot [75] noticed that some salts of oxalic acid (e.g. mercuric
or silver oxalates) have the properties of primary explosives.

This group of initiators has no practical application. Nevertheless it is interesting
from the theoretical point of view, due chiefly to the fact that the general equation
for the decomposition of oxalates is :

A solid -> B solid + C gas

Thus it is similar to the decomposition of azides. There have been several papers
on silver oxalate — Ag 2 C 2 4 . Macdonald and Hinshelwood [76] confirmed the
Berthelot equation, according to which the only products of decomposition of silver
oxalate are metallic silver and C0 2 .

Benton and Cunningham [77] found that the rate of thermal decomposition of
silver oxalate may be increased by previously exposing it to ultra-violet radiation.

During the thermal decomposition of silver oxalate, fragments of metallic silver
are formed. This has been confirmed by conductivity measurements (Macdonald
and Sandison [78]) or by X-ray examination (Griffith [79]).

Tompkins [80] investigated the thermal decomposition of silver oxalate at 110—
130°C. Its decomposition, in his opinion, is similar to that of barium azide.

Mercuric oxalate appears to undergo decomposition by a somewhat different
mechanism with the formation of mercury and mercurous oxalate as intermediate
products before full decomposition occurs (Prout and Tompkins [81]).

OTHER INITIATING EXPLOSIVES 225

PEROXIDES

The explosive properties of peroxides have attracted attention for a long time,
mainly because of their initiating properties: namely, in a confined space burning'
readily passes into detonation. In spite of this, virtually none of the peroxides has
found practical application. Some are rather unstable, others are very volatile and
all are highly sensitive to friction and impact: e.g. acetone peroxide, very easily
prepared by the action of potassium persulphate on acetone in the presence of
sulphuric acid (Baeyer and Villiger [82]) possesses, according to T. Urbaiiski's [83]

h 3 ci .0— a .ch 3

>c< >c/
H3C/ \o— <y \ch 3

studies, a very high vapour pressure and is highly volatile at room temperature.
In an open vessel a thin layer of the substance loses half its weight in approximately
3 months. The Chemisch-Technische Reichsanstalt [84] determined the rate of
detonation of acetone p|jpxide in a tube of 6.3 mm diameter: the rate was found
to be 5190 m/sec when the s.g. was 1.6 and 5290 m/sec when the s.g. was 1.2.

A peroxide which might havtf practical importance and which has been the
subject of fairly extensive studies, is hexamethylenediamine peroxide, so-called
HMTD, a substance with a m.p. 145°C with the probable formula:

/CH 2 - O— O— CH 2 .
N— CH 2 — O— O— CH 2 — N
\CH 2 — O— O— CH 2 /
I

?-CH 2 /CH 2 -^0

I >N— CH 2 -O0— CH 2 — N< I

O-CH/ \CH 2 -0

I, according to Baeyer and Villiger [82] or II, according to Girsewald and Sie-
gens [85].

Legler [86] was the first to prepare this substance by the action of hydrogen
peroxide on ammonium salts in the presence of formaldehyde. Later, Girsewald
[87] obtained it by the treatment of hexamethylenetetramine with hydrogen peroxide.

At present, it is rather difficult to decide between these two formulae, but it
seems certain that the substance contains methyleneamino groups N-CH 2 and
peroxy groups, —O—O—. This view is supported by the observation of T. Urban-
ski and Szyc-Lewanska [88] that the action of nitric acid on hexamethylenediamine
in the presence of ammonium nitrate leads to the formation of cyclonite and formic
acid.

Hexamethylenediamine peroxide is prepared by dissolving 14 g of hexamethylene-
tetramine in 45 g of 30% hydrogen peroxide, in a beaker chilled with a mixture of ice

226 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

and common salt. 21 g of powdered citric acid is then poured into the solution while
maintaining the temperature below 0°C. When all the citric acid has been added,
the whole is agitated for 3 hr at 0°C and allowed to stand for 2 hr at room tem-
perature. The white, crystalline product is filtered off and washed with water and
alcohol to aid drying at room temperature.

The same compound was prepared by Leulier [89] by the action of hydrogen
peroxide on hexamethylenetetramine in the presence of nitric acid, but in a lower
yield than that in the Girsewald method. Leulier ascribed an incorrect formula
to the substance as pointed out by Girsewald and Siegens [85].

Hexamethylenediamine peroxide is practically insoluble in water and in the
majority of organic solvents. According to Taylor and Rinkenbach [90] it is volatile
at a temperature higher than room temperature and at 75°C it decomposes markedly
losing methylamine. At 100°C it is totally decomposed after 24 hr. When boiled
in water it decomposes, and passes into solution with evolution of oxygen, the
aqueous solution contains ammonia, formaldehyde, ethylene glycol, formic acid
and hexamethylenetetramine.

When thrown onto a metallic surface heated to 200°C it explodes instantly, or
on a surface at 149°C after 3 sec. It is a very powerful explosive. Its rate of detonation
at a density of 0.88 in a pipe 5.5 mm in diameter is 4510 m/sec.

It is less sensitive to impact than mercury fulminate (a 3 cm drop is necessary
to cause explosion, with a 2 kg weight, whereas for mercury fulminate a 2.5 cm drop
is sufficient), but as an explosive it is much powerful than the latter.

As an initiator it is also much more powerful than mercury fulminate. Thus,
trinitrotoluene is detonated by as little as 0.08 g of hexamethylenediamine peroxide
compressed under a pressure of 67 kg/cm 2 , whereas with mercury fulminate 0.26 g
is required. For picric acid and tetryl 0.05 g of peroxide is sufficient to produce
detonation as compared with 0.21-0.24 g of fulminate.

Its specific gravity is 1.57, but its apparent density only 0.66. A density of 0.91 can
be obtained under a pressure of 170 kg/cm 2 . It is not liable to become dead pressed
even under a pressure of 730 kg/cm 2 .

In spite of its initiating qualities, hexamethylenediamine peroxide is of no prac-
tical use due to its doubtful stability.

Recently Lefevre and Baranger [91] recommended it as a chemotherapeutic
agent against cancer. They obtained positive results from oral treatment of cancer
of the prostate gland.

Another peroxide, i.e. tetramethylenediperoxidodicarbamide has similar pro-
perties. It probably has the following structure:

,NH— CH 2 — O— O— CH 2 — NH.
CO CO

\nh— ch 2 — o— o— ch 2 — nh/

This compound was obtained by Girsewald and Siegens [92] by the action of
hydrogen peroxide and nitric acid on an aqueous solution of urea and formaldehyde.

OTHER INITIATING EXPLOSIVES 227

Spaeth [93] suggested its application as an initiator, but without practical success
for the same reasons as with other peroxides, i.e. insufficient stability and a very
high sensitiveness to mechanical stimulus.

Peroxides and ozonides are a wide group of compounds which are receiving a
growing amount of attention for both theoretical and practical reasons. Particularly
in the polymer industry peroxides have found wide application as catalysts and
intermediates.

This subject is outside the scope of this book. Those interested should consult
monographs and review articles and books by Rieche [94], Davies [95] and others
[96-98].

ACETYLENE AND ITS SALTS (ACETYLIDES)

It has long been known that acetylene explodes under the influence of com-
pression. Experiments by Rimarski and Metz [99] showed that at a temperature
below 500°C acetylene does not explode if the pressure is lower than 3 kg/cm 2 .
An explosion may occur at 510°C under a pressure of 2.05 kg/cm 2 . At room tem-
perature acetylene may explode provided it is compressed adiabatically with a
pressure of 170 kg/cm 2 .

Acetylene is an endothermic compound, its heat of formation (-AH t ) being
-54.9 kcal/mole. Its heat of explosion is therefore very great, viz. 1870 kcal/kg, al-
though the explosion is not connected with an oxidation reaction:

volume of gases (V ) is 60 l./kg

temperature of explosion approximately 2700°C

Solid acetylene (m.p. about -83°C) is insensitive to impact, but at a density
of 0.64 it can be detonated by a number 8 detonator, showing a rate of detonation of
2500 m/sec and a lead block expansion of 300 cm^.

Gaseous, compressed acetylene also detonates. According to Penny [100] when
compressed to 8 atm in a pipe i-in.in diameter, it detonates with a rate of 1817 + 7
m/sec and in a pipe 1 in. dia. with a rate of 1870 ±22 m/sec. According to Mayes
[101] when compressed to 3-6 atm in a pipe 1£ in. dia., it detonates with a rate of
1848 m/sec.

The heavy metal salts of acetylene have the properties of primary explosives,
but only cuprous acetylide was found to be satisfactory for practical use.

CUPROUS ACETYLIDE

Cuprous acetylide was prepared by Berthelot [102] as early as 1866 by the action
of acetylene on an ammonia solution of cuprous chloride. Cuprous acetylide takes
the form of a russet or reddish-brown powder, insoluble in water and in the majority
of organic solvents.

228 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Much research was necessary in order to establish the chemical composition of
cuprous acetylide. Blochmann [103] and Scheiber and Reckleben [104] showed that
the freshly precipitated and dried product has the approximate composition
Cu 2 C 2 H 2 0. According to Keiser [105] drying this salt over sulphuric acid or
calcium chloride (Scheiber and Reckleben [104]) gives an anhydrous product.

Kuspert [106] drew attention to the fact that cuprous acetylide may form a
colloidal solution. The colloidal state is favoured by the use of diluted ammonia
solutions of cuprous salts.

The substance is stable at ordinary temperatures and up to 100°C. Like cupric
acetylide it decomposes on being heated in hydrochloric acid (Berthelot [102],
Sabaneyev [107]). A solution of potassium cyanide also causes decomposition with
the loss of acetylene. Makowka [108] showed that aldehyde-like compounds are
formed from cuprous acetylide on reaction with a 30% solution of hydrogen perox-
ide.

Cuprous acetylide explodes in air at 120-123°C, but in an acetylene atmosphere,
under a pressure of 5 atm it decomposes without explosion at 250°C. According to
Morgan [109] it is very easily exploded by an electric spark.

Apart from cuprous acetylide, with the formula Cu 2 C 2 , there are complex
cuprous salts prepared by the action of acetylene on certain cuprous salts in a neutral
or slightly acidic medium; e.g. Bhaduri [110] obtained a cuprous acetylide contain-
ing a thiosulphate group by the action of acetylene on cuprous thiosulphate,
and in the presence of potassium iodide Scheiber and Reckleben [104] precipitated
an acetylide containing iodine.

Care should be taken that cuprous acetylide is not contaminated with cupric
acetylide which may occur if the cuprous chloride used for the reaction contains
cupric salt. This is of importance since cupric acetylide is unstable and explodes
on heating even between 50 and 70°C. It is also more sensitive to impact and friction
than cuprous acetylide. The pure cupric acetylide is black or brown.

Many authors recommend the precipitation of cuprous acetylide in the presence
of reducing substances such as hydroxylamine (Ilosvay [111]), S0 2 (Rupe[112]),
hydrazine sulphate (Cattelain [113]), so as to avoid contamination with cupric
acetylide.

The precipitation of cuprous acetylide was introduced into analytical chemistry
for the quantitative determination of copper. Since cupric acetylide was dangerous
to handle, Makowka [108] worked out a method in which cupric salts are previ-
ously reduced, e.g. with hydroxylamine, to cuprous salts, when the acetylide is pre-
cipitated. Cuprous salts in a solution of hydroxylamine are employed as reagents
for acetylene (e. g. Pietsch and Kotowski [1 14]).

Cuprous acetylide is used as the chief component of match heads in electric
fuses, being particularly susceptible to ignition by sparks or a glowing wire to give
a sharp, hot flame.

OTHER INITIATING EXPLOSIVES 229

SILVER ACETYLIDE

Silver acetylide, Ag 2 C 2 , is a white powder formed when acetylene is passed
through an ammoniacal solution of silver chloride. It has even stronger explosive
properties than cuprous acetylide due to its exceptionally large negative heat of
formation {-AH f = -87.15 kcal/mole). Its ignition temperature is 200°C. It is of
no practical value.

VARIOUS INITIATORS
NITROGEN SULPHIDE

Nitrogen sulphide, N 4 S 4) m.p. 178°C, was prepared by Soubeiran [115] by the
action of ammonia on sulphur chloride dissolved in benzene

6SC1 2 + I6NH3 -> N4S4 + 2S + I2NH4CI

It is prepared by dissolving 1 volume of sulphur chloride in 8-10 volumes of
carbon disulphide, cooling and passing in dry ammonia until the dark brown precipi
tote first formed, has redissolved, producing an orange-yellow solution containing
flocks of ammonium chloride. The latter is filtered off and the filtrate is evaporated
to dryness. The dry residue is extracted with boiling carbon disulphide to remove
the sulphur. The undissolved material is crude nitrogen sulphide. On cooling the hot
extract deposits a further quantity of the substance. The combined crude product is
recrystallized from carbon disulphide.

Nitrogen sulphide is insoluble in water, slightly soluble in alcohol and ether
somewhat soluble in carbon disulphide and benzene. At room temperature it is
nydrolysed to some extent by water to form free sulphur, sulphur dioxide and am-
monia. Its specific gravity is 2.22.

Nitrogen sulphide is less sensitive to friction and impact than mercury fulminate
It is a weaker initiator than mercury fulminate, its rate of acceleration being con-
adorably less than that of the latter. Its ignition temperature is 207°C. It is excep-
Uonally strongly endothermic, its heat of formation {-AH t ) being - 138.8 kcal/mole
(Berthelot and Vieille [116]).

In many patents nitrogen sulphide is recommended as a filling for fuses, primers
etc. but in spite of this it has not been used in practice.

NITROGEN SELENIDE

Nitrogen selenide, N 4 Se 4 , was prepared by Espenschied [117] by the action of
ammonia on selenium chloride. It is an orange-red, amorphous, explosive powder.

lne explosive properties of this substance were studied by Verneuil [118], and Ber-
thelot and Vieille [1 16]. Its ignition temperature is 230°C. It is very sensitive to friction
and impact. Its heat of formation (-AH t )is strongly endothermic: - 169.2 kcal/mole

230 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

SALTS OF THIOCYANIC ACID

At the beginning of the twentieth century the thiocyanates (rhodanates) of
certain metals (e.g. mercury, copper) were recommended as components of cap
compositions with potassium chlorate. The rhodanates were intended as a substi-
tute for mercury fulminate, but only lead rhodanate acquired any practical signifi-
cance.

Lead rhodanate, Pb(CNS) 2 is formed, when a solution of lead nitrate, slightly
acidified with nitric acid, is treated with a moderately concentrated solution of the
rhodanate of an alkali metal.

When exposed to long day-light or ultraviolet irradiation lead rhodanate becomes
yellow. When boiled in water it is converted into a basic salt Pb(CNS) 2 • Pb(OH) 2
which behaves as a weak initiator. It is used as a constituent in some cap compositions.

COMPLEX SALTS

The complex salts of precious metals, formed by the action of ammonia either
on aqueous solutions of silver, gold and platinum salts or on silver oxide were the
first substances tp reveal the ability to explode violently on heating, on direct contact
with flame or by friction or impact ("fulminating" silver and gold).

Later it was found that a number of other metals which can give typical complex
salts (Cd, Co, Cr, Cu, Fe, Hg, Mn, Ni and Zn) can form explosive co-ordination
compounds.

These substances have a variable composition depending upon the reaction
conditions, chiefly the concentration of the reagents used. Salvadori [119] called
attention to the interesting explosive properties of the chlorates and perchlorates
of complex "ammines" of the type :

[M n (NH 3 ) 4 ](C10 3 ) 2
[M n (NH 3 )4](C10 4 )2
[M ln (NH 3 )6](C10 3 )2
[M m (NH 3 ) 6 ](C10 4 ) 2

The explosive properties of these salts were partially studied by Ephraim and
Jahnsen [120] and were later investigated in detail by Friedrich and Vervoorst [121].
The latter also investigated the analogous combinations described by Franzen and
Mayer [122], in which ammonia was replaced by hydrazine*

Friedrich and Vervoorst found that the substance [Cu(NH 3 )4.](C10 3 )2 has
initiating properties. It is however of no practical use, losing ammonia fairly rapidly
in air. In damp air it is easily hydrolysed. Nickel ammino perchlorate [Ni(NH 3 ) 6 ]
(C10 4 ) 2 is a fairly strong explosive with the rate of detonation 5300 m/sec at the
density 1.39.

OTHER INITIATING EXPLOSIVES 231

Friedrich and Vervoorst also prepared chlorates of hydrazinometals :

[Cd(N 2 H 4 )2](C103)2

[Ni(N 2 Hj) 4 ](C10 3 )2

and similar perchlorates, which proved to be basic salts.

The chlorates are strong initiators with an initiating power exceeding that of
mercury fulminate and even that of lead azide. The ignition temperature of the
cadmium salt is 125°C, and that of the nickel salt 170°C. Basic perchlorates of cad-
mium and nickel are also initiators, weaker than the chlorates, but somewhat strong-
er than mercury fulminate.

The sensitiveness of all these salts to impact is very high, in many cases higher
than that of mercury fulminate.

The corresponding nitrates (N0 3 instead of G10 3 or C10 4 ) do not possess ini-
tiating properties.

Another interesting group of explosive co-ordination compounds is formed by
the azides of "ammines" of general formulae

[M 11 (NH 3 ) 2 ](N 3 )2
[M n (NH 3 )4](N 3 )2
[M m (NH 3 ) 6 ](N 3 ) 2
[M ra (NH 3 ) 6 ](N 3 ) 3

The salts of Cd, Cr, Cu 11 , Ni were first prepared by Strecker and Schwinn [123].
They do not seem to possess sufficient stability to be of practical use.

Another type of complex cobalt ammine including azide groups in the co-ordi-
nation were obtained by Linhard and Flygare [124]:

[Co(NH 3 ) 5 N 3 ]X 2

where X=C1, Br, I, N0 3 , N 3

and [Co(NH 3 ) 5 N 3 ]Y

where Y=Cr0 4 , S2O3, S 2 6

Only the perchlorate (X=C10 4 ) shows marked sensitiveness and possibly ini-
tiating properties. The other compounds do not seem to possess properties interesting
from the practical point of view.

Another kind of complex salts was obtained by Hodgkinson and Hoare [125]
by the action of an ammoniacal solution of cupric, nickelous-nickelic or silver
oxide on an alcoholic solution of tetranitromethane. Precipitates are then formed
which are insoluble in water, explosive, but unstable, decomposing on boiling in
water.

There are many other explosive complex salts but at present they are of no prac-
tical use.

An extensive review of various explosive complex compounds has been published
by Fedoroff et al. [126].

232

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES
SILVER PERCHLORATE

There are some indications that silver perchlorate has initiating properties since
in some unexplained cases, the large crystals detonate on slight friction in a way
similar to the detonation of crystals of lead azide or silver azide (Hein [127]).

These properties have not been examined in detail.

INITIATING COMPOSITIONS

Initiators are usually compositions from which the desired results are obtained
by the interrelation of the components. This is of particular importance in the manu-
facture of percussion caps which are fired by striker pins to give a hot flame capable
of igniting propellants and should therefore have ability to detonate.

Blasting caps, i.e. detonators, (Fig. 56, p. 218) were originally filled with mercury
fulminate alone or with a 80:20 mixture of mercury fulminate and potassium chlorate.
The chlorate is added to facilitate pressing since mercury fulminate alone cannot
be pressed conveniently. In addition, potassium chlorate increases the ignitability
of mercury fulminate, thus permitting a greater pressing pressure to be used which
in turn improves coherence.

Mercury fulminate, or its mixture with potassium chlorate, is usually pressed
under a pressure of 250-300 kg/cm 2 . Pressed at pressures of 600 kg/cm 2 it ignites
with difficulty and misfires may occur.

Detonators for mining explosives are manufactured in various sizes and numbered
according to the amount of fulminate which they contain. The charges of fulminate
for each size are shown in the following table :

Table 40

No. of detonator

7 | 8

Charge of mercury
fulminate, g

0.3

0.4

0.54

0.65

0.8

1.0

1.5

2.0

2.3

3.0

In mining, a No. 8 detonator is used for ammonium nitrate explosives, No. 6
and No. 3 for nitroglycerine explosives.

Modern detonators have a double filling, i.e. a charge of high explosive such
as tetryl or penthrite at the bottom, initiated by a layer of mercury fulminate or
fulminate-chlorate placed on top (Fig. 56b). In this way more powerful detonators
have been produced for mining purposes, containing the following charges :

Table 41

No. of detonator

Tetryl, g

Mercury fulminate, g

0.3
0.3

0.4
0.4

0.75
0.5

0.9
0.5

OTHER INITIATING EXPLOSIVES

233

Detonators containing lead azide are loaded into aluminium capsules. Lead
azide may be used alone or in conjunction with lead styphnate or tetrazene, the
use of which facilitates ignition of azide. The layer of azide may also be covered

so so ioo no """ w
Weight of priming charge, centigrams

Fig. 58. Minimum initiating charge of lead azide, lead styphnate and their mixtures
for 1.25 g of tetryl, according to Grant and Tiffany [40].

with a layer of lead styphnate. Lead azide or its mixtures may be pressed at a very
high pressure (e.g. 1000 kg/cm 2).

A general idea of the change in the minimum initiating charge of lead azide-lead
styphnate mixtures with their composition is given in Fig. 58 by Grant and Tiffany
[40]. The low priming ability of lead styphnate alone is also indicated.

The minimum charge required to detonate the base explosive in the detonator
was determined by the sand test of Snelling [127a].

Aluminium detonators with lead azide and other explosives were used in the
mining industry for some time, e.g. a No. 8 detonator, contained 1 g of tetryl and
0.3 g of a mixture of lead azide and lead styphnate. These were more powerful
than those with a fulminate-tetryl charge, but the use of detonators with aluminium
sheathing was soon forbidden in coal-mines due to the danger created by the burn-
ing of the aluminium.

The modern detonator TAT-1, used in U.S.S.R. is filled in three layers (Fig.
56c) as follows:

bottom — tetryl 0. 1 2 g

middle— lead azide 0.21 g
top— lead styphnate 0.06 g

Modern artillery primers also include a charge of penthrite and lead azide sensi-
tized to flame by the addition of lead styphnate or tetrazene.

234

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(1) The bottom layer is loaded with 0.35 g of penthrite (pressed at 1800 kg/cm 2 ),
the middle with 0.35 g of penthrite (not pressed) and the top with 0.30 g of a mixture
of 92.5% of lead azide and 7.5% of tetrazene (pressed at 1100-1800 kg/cm 2 ).

Thus a layer of bursting charge adj^ent to the initiator is compressed more
weakly and in consequence complete explosion occurs with greater ease.

(2) The bottom layer is loaded with 0.2 g of penthrite (pressed under a pressure
of 500 kg/cm 2 ), the middle with 0.2 g of penthrite (not pressed) and the top with
0.4 g of a mixture of 80% of lead azide and 20% of lead styphnate (pressed under
a pressure of 500 kg/cm 2 ).

Mixtures of mercury fulminate, potassium chlorate (as an oxidizing agent),
antimony sulphide and ground glass were widely utilized for many years in percus-
sion caps. The content of mercury fulminate was small so that the mixture had no
explosive properties. For the same reasons significant amounts of potassium chlo-
rate were used as an oxidizing agent, thus diluting the fulminate to some extent.
Antimony sulphide is a combustible component which gives a hot flame. Ground
glass was added in order to increase the internal friction and make it more sensitive
to percussion.

Some compositions also contained an adhesive, e.g. shellac, gum etc.

A German composition of 1883 contained, for example:

Mercury fulminate 27%

Potassium chlorate 37%

Antimony sulphide 29%

Ground glass 7%

To 100 parts of this mixture 0.6 part of shellac were added.

The composition of other caps according to Gorst [128] is given in Table 42.

Table 42

Cap

Mercury
fulminate

Potassium
chlorate

Antimony
sulphide

Rifle and pistol

Fuse

16.5

55.5

37.5
25

28.0

37.5
25

The English caps contained a little blackpowder which elongates the flame pro-
duced by the explosion of

Mercury fulminate
Potassium chlorate
Antimony sulphide
Sulphur
Blackpowder

Mixtures containing mercury fulminate, potassium chlorate, and antimony
sulphide tend to destroy the inside of firearm barrels, since on decomposition the
mercury fulminate evolves free mercury which causes erosion of the barrel at the

15%

35%

45%

2.5%

OTHER INITIATING EXPLOSIVES

235

high temperatures created inside the bore. Decomposition of the potassium chlorate
gives potassium chloride which remains in the bore and strongly corrodes the steel.
Sulphur dioxide formed by the combustion of antimony sulphide also helps to destroy
the barrel. For a long time therefore, the use of compositions not containing mercury
fulminate and potassium chlorate were advocated, but a satisfactory formulation
for non corrosive mixtures was found ("Sinoxyd") only when lead styphnate was
introduced as their chief component. Since styphnate is hard to ignite by impact
it was sensitized by an admixture of tetrazene. At the same time potassium chlorate
was replaced by barium nitrate.

German compositions for rifle and pistol caps are tabulated below [31, 32].

Table 43

Components

Composition No. 30/40
for rifle and pistol caps

Composition for rifle

caps manufactured at

Stadeln

Lead styphnate

30-35

Tetrazene

2-3

Barium nitrate

4(M5

Lead dioxide

5-8

Calcium silicide

6-12

Antimony sulphide

6-9

ie Czechoslovak mixta

]
]

re, Oxyd, for pistol cartridges

Lead styphnate 45%
retrazene 5%
Barium niiSkte 33%

has a similar compos

\ntimony sulphide 2

viz.:

Aluminium-magnesium alloy 5%

The German mixture No. 30 for the friction fuses of hand grenades has the
following composition:

Lead styphnate 25%

Barium nitrate 25%

Lead dioxide 24%

Silicon 15%

Ground glass 3%

THE PREPARATION OF PRIMER COMPOSITIONS

Formerly cap compositions were prepared by mixing the well-damped compo-
nents on a glass-topped table, while adding water continuously to the mixture
(especially at the edges which are liable to more rapid drying). A wooden spade
was used for mixing. The mixture in the form of a paste, was made into pellets which
were then dried and pressed either into the sheaths of blasting caps or into the
capsules of percussion caps.

236

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

This method was widely employed before World War I and for some time after-
wards, but it is now discontinued due to the changes which occur during storage
to caps so manufactured. These changes reduced the caps efficiency possibly due

Fig. 59. Silk bag for mixing an initiating composition: 1— rubber frame, 2— leather

lug for suspending the bag, 3— strings, 4— rubber rings, 5— mixing string, according

to Vermin, Burlot and Lecorchd [132].

Fig. 60. Mixing of primary explosive charge on a rubber tray at Hercules Powder
Company, according to Davis [133].

OTHER INITIATING EXPLOSIVES

237

Fig. 61. A device for mixing and simultaneously drying an initiating composition:

/-ebonite drum, 2— tilted revolving drum plate, i-warm air conduit, according

to Vennin, Burlot and Lecorche" [132].

^/V////////////////////^^^^

Fig. 62. Automated arrangement for mixing initiating compositions at Stadeln [130]:
/—papier mach<5 drum {B— in a horizontal position when mixed, B'— when loaded, B"— when un-
loaded), 2— papier macb.6 spoon, 3— paper funnel.

238

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

to traces of moisture in the cap composition which are particularly persistent if
binders such as gum arabic etc. have been incorporated. The presence of moisture also
enhances the reaction of mercury fulminate with the metal into which the composi-
tion is pressed.

For this reason the components are now not mixed wet. There are several me-
thods of dry mixing. One of them (the earlier "jelly-bag" method) is to place the

Fig. 63. Detonator manufacture— a line of compartments in which the filling and press-
ing of detonators is carried out automatically. Courtesy Imperial Chemical Indus-
tries Ltd., Nobsl Division.

dried and sieved components of the composition into a conical silk bag (Fig. 59)
which is fastened at its base to a rubber frame. There are strings within the bag
fitted with rubber rings. Another string fastened to the top of the cone passes over
a pulley fixed to the ceiling. This string leads outside the building or protective wall.
By manipulating the strings the bag is put into motion and its contents are stirred.
The bag is emptied by turning it upside down by a pull on the string.

The same type of arrangement for mixing primary explosive charge used by the
Hercules Powder Company, according to Davis [129], is shown in Fig. 60. It consists
of a triangular rubber tray. The composition is mixed by lifting and lowering the
corners of the tray in turn. The lifting and lowering is accomplished behind a con-
crete safety barriers.

OTHER INITIATING EXPLOSIVES

239

In a more modern method, the mixing may be combined with drying. A diagram
illustrating such a device is presented in Fig. 61. Here the composition is placed in
an open ebonite drum which is laid on a tilted, revolving plate. The moist compo-
nents are placed in the drum (weighed and containing a known amount of moisture),

Fig. 64. Detonator manufacture— the plating section of the automatic plain detonator
unit. Courtesy Imperial Chemical Industries Ltd., Nobel Division.

and the latter is then aerated with a warm air stream which dries the contents when
stirred. This system can also be used for drying initiators, e.g. mercury fulminate,
lead azide etc.

In the German plant at Stadeln [130] drums of papier mache about 20 cm in dia-
meter and 45 cm long are used, mounted on a horizontal, revolving axis, and provided
with an aperture on the drums' cylindrical surface. Before being set in motion this
opening is closed with a rubber cork. Subsequently it is opened by remote control
by means of hooks and strings. After being stirred, the contents of the drum are
poured into boxes previously ranged in readiness. Each drum is sited in a chamber,
protected by a concrete partition wall. It is desirable to cover the floor with a soft
material that conducts electricity. With compositions which are very dangerous to
handle, containing lead styphnate, stirring proceeds stepwise so that the preparation
of the final, dangerous composition lasts for as short a time as possible. Thus a com-
position consisting of barium nitrate, calcium silicide, antimony sulphide and lead

240 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

dioxide is first prepared. This composition is manufactured in large amounts in
ordinary, wooden drums. A weighed amount of the composition is then mixed
automatically with tetrazene and lead styphnate (Fig. 62). The papier mache drums
are in the shape of a truncated cone with bases of about 8-30 cm in diameter. Such
a drum will hold about 400 g of material. Above the drum there is a papier mache
spoon into which a worker pours a batch of^the mixture. The worker then leaves
the building and sets the driving mechanism in motion from a distance. This engine
first rotates the spoon so that its contents run into the drum. Then the drum is moved
into a horizontal position, in which it revolves for 7 min with a velocity of 60 r.p.m.,
from which it is tilted downwards, pouring its contents into a paper funnel. After
being emptied the drum is returned to the initial position and the drive is stopped.
The whole mechanism is so arranged that all these movements are carried out auto-
matically.

An idea of a modern plant lay-out for the manufacture of priming compositions
and detonators is given by Figs. 63 and 64 showing installations at Imperial Chemical
Industries, Nobel Division, in Great Britain.

COMPOSITIONS FOR EXPLOSIVE RIVETS

In 1937 rivets containing a small explosive charge in their shanks were constructed
(Aircraft Factory Heinkel and Rheinisch-Westphalische Sprengstoff A. G. [131]).
The composition of the explosive mixture is so selected that a slight explosion occurs
on heating the head of the rivet with a hot iron. The explosion causes an expan-
sion of the shank thus fixing the rivet in place. Explosive rivets have found wide
application primarily for riveting aircraft components in which rivets are not
accessible from both sides of the riveted surface.

Initially only aluminium alloy rivets (duralumin) were employed. Later, during
World War II, steel explosive rivets were also used. The principle is limited to rivets
of small size. For filling these rivets, charges of chlorate explosive mixtures with
fairly low ignition temperatures (e.g. 180°C) were originally used. Their composition
was similar to that of priming caps. During World War II the following mixture
was used for aluminium rivets:

Nitromannite 15%
Tetrazene 10%
Aluminium 70%
Adhesive 5%

LITERATURE

1. M. Berthelot and P. Vieille, Mem. poudres 1, 99 (1882-83).

2. L. WOhler and O. Matter, Z. ges. Schiess- u. Sprengstoffw. 2, 181 (1907).

3. E. Herz, Ger. Pat. 258679 (1911).

4. E. Bamberger, Ber. 28, 444 (1895).

5. A. Hantzsch and W. B. Davidson, Ber. 27, 1522 (1896).

OTHER INITIATING EXPLOSIVES

241

6. A. KLEMENC, Ber. 47, 1407 (1914).

7. L. Wolff, Ann. 312, 119 (1900).

8. H. H. HODGSON and E. Marsden, /. Soc. Dyers Colourists 59, 271 (1943).

9. J. D. C. Anderson and R. J. W. Le Fevre, Nature 162, 449 (1948).

10. B. Glowiak, Bull. Acad. Polon. ScL, ser. chim. 8, 1 (1960).

11. P. GRIESS, Ann. 106, 123 (1858).

12. W. Lenze (1892), according to H. Kast, Spreng- u. Ziindstoffe, p. 433, Vieweg & Sohn, Braun-
schweig, 1921.

13. D. Smolenski and J. Plucinski, Biul. WAT, 4-5, 22 (1953).

14. F. M. Garfield, U.S. Pat. 24080059 (1946).

15. L. V. CLARK, lnd. Eng. Chem. 25, 663, 1385 (1933).

16. J. Vaughan and L. Phillips, /. Chem. Soc. 1947, 1560.

17. J. V. R. Kaufman, Discussion Roy. Soc, Initiation and Growth of Explosion in Solids Proc
Roy. Soc. (London) klA6, 219 (1958).

18. T. Urbanski, K. Szyc-LewaNska, M. Bednarczyk and J. Ejsmund, Bull. Acad. Polon
ScL, ser. chim. 8, 587 (1960).

19. A. F. Belayev and A. E. Belayeva, Dokl. Acad. Nauk SSSR 52, 507 (1946)- Zh fiz khim
20, 1381 (1946).

20. B. Glowiak, Bull. Acad. Polon. Sci., ser. chim. 8, 9 (1960).

21. G. Ponzio, Gazz. chim. ital. 45, II, 12 (1915).

22. A. QuiLICO, Gazz. chim. ital. 62, 503, 912 (1932).

23. G. Ponzio, Gazz. chim. ital. 63, 471 (1933).

24. J. F. Kenney, (Remington Arms Co. Inc.) U.S. Pat. 2728760 (1955); Chem. Abstr 50 7462
(1956); Chem. Zentr. 1957, 1082.

25. K. A. Hoffmann and R. Roth, Ber. 43, 682 (1910).

26. K. A. Hoffmann, H. Hock and R. Roth, Ber. 43, 1087 (1910); K. A. Hoffmann and
H. Hock. Ber. 43, 1866 (1910); 44, 2946 (1911); K. A. Hoffmann, H. Hock and H. Kirm-
reuther, Ann. 380, 131 (1911).

27. S. H. Patinkin, J. P. Horwitz and E. Lieber, /. Am. Chem. Soc. 77, 562 (1955).

28. H. Rathsburg, Ger. Pat. 362433, 400814 (1921).

29. R. Wallbaum, Z. ges. Schiess- u. Sprengstoffw. 34, 126, 197 (1939).

30. W. H. RiNKENBACH and O. E. Burton, Army Ordnance 12, 120 (1931).

31. CIOS Report XXVII-38, Manufacture of Initiating Explosives.

32. CIOS Report XXXIII-48, Report on the Visit to D.A.G. Small Arms Factory, Stadeln.

33. T. L. Davis and E. N. Rosenquist, /. Am. Chem. Soc. 59, 2114 (1937).

34. A. Chretien and B. Woringer, Compt. rend. 232, 1115 (1951).

35. H. M. Montagu -Pollock, Proc. Roy. Soc. (London) A 269, 219 (1962).

36. F. P. Bowden and H. M. Montagu -Pollock, Nature 191, 556 (1961).

37. A. F. McKay, Can. J. Research 23 B, 683 (1950).

38. A. F. McKay, W. G. Hatton and G. W. Taylor, unpublished work, according to S. R.
Harris [39].

39. S. R. Harris, /. Am. Chem. Soc. 80, 2302 (1958).

40. R. L. Grant and J. E. Tiffany, lnd. Eng. Chem. 37, 661 (1945).

41. T. Urbanski and K. Kruszynska (Szyc-Lewanska), unpublished work (1937).

42. BIOS Final Report No. 833.

43. P. Grtess, Ber. 7, 1224 (1874).

44. R. A. Zingaro, /. Am. Chem. Soc. 76, 816 (1954).

45. E. Herz, U.S. Pat. 1443328 (1919).

46. A. Stettbacher, Nitrocellulose 8, 3 (1937).

47. W. H.RTNKENBACH in R. E. Kirk and D. F. Othmer (Eds.), Encyclopedia of Chemical Tech-
nology, Vol. 6, Interscience, New York, 1951.

242 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

48. H. R. HAILES, Trans. Faraday Soc. 29, 544 (1933).

49. W. E. Garner, A. S. Gomm and H. R. Hailes, /. Chem. Soc. 1933, 1393.

50. F. C. Tompkins and D. A. Young, J. Chem. Soc. 1956, 3331.

51. J. I. Evans and A. M. Yuill, Discussion Roy. Soc, Initiation and Growth of Explosion in
Solids, Proc. Roy. Soc. (London) A 246, 176 (1958).

52. G. J. Bryan and E. C. Noonan, Discussion Roy. Soc, Initiation and Growth of Explosion
in Solids, Proc. Roy. Soc. (London) A 246, 167 (1958).

53. J. H. McAuslan, Ph.D. Thesis, Cambridge, 1957, according to F. P. Bowden and A. D. Yoffe,
Fast Reactions in Solids, Butterworths, London, 1958.

54. J. Barcikowski, Dobrzynski and Kielczewski, unpublished work (1938).

55. I. Hartmann, J. Nagy and H. R.Brown, U.S. Bureau of Mines Techn. Paper 3722 (1943).

56. G. Morris, Engineering 164, 49, 73 (1947).

57. J. Taylor and R. H. Hall, according to Morris [56].

58. J. Meissner, Brit. Pat. 500711 (1939).

59. J. Meissner, private information (1964).

60. T. Urbanski, M. Piskorz and J. Mazur, Biul. WAT 8, No. 84, 112 (1959).

61. T. Urbanski and J. Zacharewicz, Wiad. Techn. Uzbr. 18, 16 (1932); T. Urbanski and
C. Pietrzyk, Wiad. Techn. Uzbr. IV, 10 (1935).

62. W. Traube, Ber. 27, 1507, 3291 (1854); Ann. 300, 81 (1898).

63. M. Gomberg, Ann. 300, 59 (1898).

64. A. Hantzsch, Ber. 31, 177 (1898).

65. A. Angeli, Ber. 29, 1885 (1896); 59, 1401 (1926); Gazz. chim. ital. 46, II, 67 (1916).

66. A. Hantzsch and E. Strasser, Ber. 64, 656 (1931).

67. R. N. Jones and G. D. Thorn, Can. J. Research TIB, 828 (1949).
€8. M. Carmack and J. J. Leavitt, /. Am. Chem. Soc. 71, 1221 (1949).

69. J. Cason and F. S. Prout, /. Am. Chem. Soc. 71, 1218 (1949).

70. M. Piskorz and T. Urbanski, Bull. Acad. Polon. Sci., sir. chim. 11, 597 (1963).

71. T. Urbanski and T. Wesolowski, Wiad. Techn. Uzbr. 18, 28 (1932).

72. P. Friese, Ber. 9, 394 (1876).

73. W. Steinkopf and C. Kirchhoff, Ber. 42, 2030 (1909).

74. T. Urbanski and M. Kowalczyk, Wiad. Techn. Uzbr. IV, 22 (1935).

75. M. Berthelot, Sur la force des matures explosives, Paris, 1883.

76. J. Y. Macdonald and C. H. Hinshelwood, /. Chem. Soc. 127, 2764 (1925).

77. A. F. Benton and G. L. Cunningham, /. Am. Chem. Soc. 57, 2227 (1935).

78. J. Y. Macdonald and R. Sandison, Trans. Faraday Soc. 34, 589 (1938).

79. R. L. Griffith, /. Chem. Phys. 11, 499 (1943).

80. F. C. Tompkins, Trans. Faraday Soc. 44, 206 (1948).

81. E. G. Prout and F. C. Tompkins, Trans. Faraday Soc. 43, 148 (1947).

82. A. Baeyer and V. Villiger, Ber. 33, 2479 (1900).

83. T. Urbanski, unpublished work (1930).

84. Jahresber. Chem.-Techn. Reichsanstalt 5, 124 (1926).

85. C. Girsewald and S. Siegens, Ber. 54, 490 (1921).

86. L. Legler, Ber. 18, 3343 (1885).

87. C. Girsewald, Ber. 45, 2571 (1912); 47, 2464 (1914); Ger. Pat. 274522 (1912).

88. K. Szyc-Lewanska and T. Urbanski, Bull. Acad. Polon. Sci., sir. chim. 6, 165 (1958).

89. A. Leulier, /. pharm. chim. [7], 15, 222 (1917).

90. C. Taylor and W. H. Rinkenbach, Army Ordnance 5, 463 (1924).

91. R. Lefevre and P. Baranger, Acta Union Intern, contre Cancer 16, 887 (1960).

92. C. Girsewald and S. Siegens, Ber. 47, 2464 (1911).

93. C. P. Spaeth, U.S. Pat. 1984846 (1935).

94. A. Rieche, Alkylperoxyde (und Ozonide), Steinkopff, Dresden-Leipzig, 1931.

OTHER INITIATING EXPLOSIVES 243

95. A. G. Davies, Organic Peroxides, Butterworth, London, 1961.

96. A. V. Tobolsky and R. B. Mesrobian, Organic Peroxides, Interscience, New York, 1954

97. R. Criegee, Methoden der Organischen Chemie, Houben- Weyl, E. MOixer (Ed ) Thieme
Stuttgart, 1957.

98. V. Karnojitzki, Les Peroxydes Organiques, Hermann, Paris, 1958.

99. W. Rimarski and L. Metz, Autogene Metallbearbeitung 26, 341 (1933).

100. E. Penny, Disc. Faraday Soc. 22, 157 (1956).

101. H. A. Mayes, Disc. Faraday Soc. 22, 213 (1956).

102. M. Berthelot, Ann. chim. [4], 9, 393 (1866).

103. R. Blochmann, Ann. 173, 167 (1874).

104. I. Scheiber and H. Reckleben, Z. anal. Chem. 48, 529 (1909); Ber. 44, 210 (1911)

105. E. H. Keiser, Am. Chem. J. 14, 285 (1892).

106. F. KOspert, Z. anorg. Chem. 34, 453 (1903).

107. A. Sabaneyev, Ann. 109 (1875).

108. O. Makowka, Ber. 41, 824 (1908).

109. J. D. MORGAN, Phil. Mag. 45, 968 (1923).

110. K. Bhaduri, Z. anorg. Chem. 76, 419 (1912); 79, 355 (1913).

111. L. ILOSVAY, Ber. 32, 2697 (1899); Z. anal. Chem. 40, 123 (1901).

112. H. RUPE, /. prakt. Chem. [2], 88, 79 (1913).

113. E. Cattelain, /. pharm. chim. [8], 3, 321 (1926).

114. E. Pietsch and A. Kotowski, Z. angew. Chem. 44, 309 (1931).

115. Soubeiran, Ann. Chim. [2], 67, 71 (1838).

116. M. Berthelot and P. Vieille, Mem. poudres 2, 3 (1884-89).

117. R. Espenschied, Ann. 113, 101 (1860).

118. Verneuil, Bull. soc. chim. France [2], 338, 548 (1882).

119. R. Salvadori, Gazz. chim. ital. 40, II, 9 (1910); 42, I, 458 (1910).

120. F. Ephraim and A. Jahnsen, Ber. 48, 41 (1915).

121. W. Friedrich and P. Vervoorst, Z. ges. Schiess- u. Sprengstoffw. 21, 49 (1926).

122. H. Franzen and O. V. Mayer, Z. anorg. Chem. 60, 247 (1908); 70, 145 (1911).

123. W. Strecker and E. Schwinn, /. prakt. Chem. 152, 205 (1939).

124. M. Linhard and H. Flygare, Z. anorg. Chem. 262, 233, 245, 338 (1950).

125. W. R. HODGKINSON and F. R. I. Hoare, according to Chem. Zentr. 1914 II, 435.

126. B. T. Fedoroff, H. A. Aaronson, E. F. Reese, O. E. Shefield, G. D. Clift et al,
Encyclopedia of Explosives and Related Items, Vol. I, Picatinny Arsenal, Dover, N. J., 1960.

127. F. Hein, Chem. Technik 9, 97 (1957); Angew. Chem. 69, 274 (1957).

127a. According to C. E. Munroe and J. E. Tiffany, Physical Testing of Explosives, Bull.
346, U.S. Dept. of Commerce, Bureau of Mines, Washington, 1931.

128. A. G. Gorst, Porokha i vzryvchatyye veshchestva, Oborongiz, Moskva, 1957.

129. T. L. Davis, The Chemistry of Powder and Explosives, J. Wiley, New York, 1943.

130. BIOS Final Report No. 1074, The Manufacture of 22 Rimfire Ammunition Dynamit A.G.
at Nurnberg and Stadeln.

131. E. Heinkel Flugzeugwerke, Rostock und Rheinisch-Westphalische Sprengstoff A.G., Nurn-
berg, Ger. Pat. 648842, 655669 (1937); W. Teske, Aluminium 19, 523 (1937); 21, 655 (1939);
Der Aluminium Praktiker 1939, 655.

132. L. Vennin, E.Burlot and H.Lecorche, Les poudres et explosifs, Librairie Polytechnique
Beranger, Paris-Liege, 1932.

133. T. L. Davis, The Chemistry of Powder and Explosives, J. Wiley, New York, 1943.

Part 3
COMPOSITE EXPLOSIVES

GENERAL INFORMATION

Most modern explosives used for practical purposes are not single chemical sub-
stances, but composite mixtures, their ingredients being selected to obtain the re-
quired properties.

Thus, if there is a need to reduce the melting point of a nitro compound, e.g.
picric acid, other nitro compounds are added to form an eutectic mixture; to decrease
the sensitiveness of picric acid to mechanical impact it is mixed with paraffin.

Smokeless propellants may be taken as another example of composite explo-
sives. These may be either mixtures of nitrocellulose of differing degrees of nitration,
partly in a colloidal and partly in a fibrous state with an admixture of the remain-
ing solvent and a stabilizer or a solution of nitrocelluloses in carbamite (centralite)
and nitroglycerine with an admixture of components such as aromatic nitro com-
pounds, nitroguanidine, graphite etc.

Blackpowder is yet another type of composite explosive being an intimate mix-
ture of potassium nitrate, charcoal and sulphur. A distinguishing feature of this
composition is that none of its components is explosive. Blackpowder and similar
compositions are examples of a large group of explosives which contain an oxi-
dizing agent as a chief component. Salts of nitric acid, e.g. ammonium, sodium and
potassium nitrates and oxy-salts of chlorine such as ammonium and potassium per-
chlorates or sodium and potassium chlorates are used as oxidizing agents. Other
constituents of such mixtures are combustible substances. These are often explosive
substances such as aromatic nitro compounds. Explosive mixtures are frequently
formulated so as to fully utilize the oxygen present in the oxidizing agents.

The amount of high melting material present, e.g. ammonium nitrate, regulates the
fusibility of these mixtures so that they may be fusible, semi-fusible or infusible, free-
pouring. There are also explosives with a plastic consistency. Usually these consist of
a plastic substratum containing a polymeric component. Again, there are liquid
explosives, containing liquid oxygen or a liquid oxidizing agent, e.g. nitrogen dioxide.

Finally, mention should be made of a group of explosives in which the compo-
nents are selected on another principle. These mixtures contain combustibles which

[245]

246 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

increase considerably the heat of explosion, for instance aluminium, ferrosilicon etc.
They burn by utilizing the oxygen contained in the explosive.

Composite explosives will be considered, according to their uses, in three main
categories: the high explosives, the low explosives or propellants and the primary
explosives or initiators.

High explosives will be classified primarily according to their consistency,
which may be fusible, semi-fusible or infusible, plastic or liquid. They will then be
arranged according to their most characteristic component.

Propellants will be grouped on the basis of their composition, into blackpowder
and similar mixtures, smokeless powders and rocket liquid propellants.

Because of the importance of rocket liquid propellants a separate chapter will
be dedicated to these mixtures. It will also include liquid high explosives.

Initiating compositions were reviewed together with primary explosives.

CHAPTER I

HIGH EXPLOSIVES

FUSIBLE EXPLOSIVES

MIXTURES OF NITRO COMPOUNDS

When picric acid was first used for filling shells by pouring it, in the molten state,
difficulties arose due to its high melting point. The necessity of using superheated
steam for melting was an added complication introducing the danger associated
with the prolonged heating of the explosive to a high temperature.

Data from the Griesheim factory [1] show that this may be avoided by adding
other aromatic nitro compounds, e.g. TNT, to the picric acid. The addition of even
a small amount (5-10%) of such a substance facilitates melting without seriously
decreasing the explosive power of the picric acid.

Easily fusible mixtures containing picric acid as the chief component were
very widely employed in Russia and France during World War I and in the period
immediately afterwards. A mixture of 51.5 or 80% picric acid with 48.5 or 20%
dinitronaphthalene was used in the U.S.S.R. for filling aerial bombs and manufac-
turing demolition charges.

The composition of mixtures used in France is given in Table 44.

Table 44
Composition of fusible explosive mixtures employed in France

Composition, %

Name

Nitro-

Dinitro-

of mixture

Picric
acid

Dinitro-
phenol

TNT

Trinitro-
cresol

naphtha-
lene

MDN

MNN

—

MTTC

—

MDPC

—

Cresylite No. 2

—

(60/40)

248 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Attempts were made in Britain to use mixtures of picric acid with dinitrobenzene.
Maxim [2] suggested mixing picric acid with dinitronaphthalene or nitronaphtha-

lene.

Kast [3] examined a number of readily fusible mixtures:

(1) 50 parts of picric acid | m p 80 _ 83 ° c
50 parts of trinitrocresol J

(2) 55 parts of TNT 1 42 _44<>c
45 parts of DNT J

(3) 10 parts of TNT I begins to melt at 39°C
90 parts of DNT J completely melts at 65°C

The use of a mixture of 35 parts of TNT and 65 parts of hexanitrodiphenylamine
(hexyl) for filling torpedoes was introduced by the Germans in 1912. Hexyl, which
is only very slightly soluble in molten TNT, on being heated to 80°C forms a suspen-
sion in the molten mixture; the resulting semi-fluid mass may be poured into the
shells. During World War I this mixture was widely used for filling torpedoes, mines
and aerial bombs. During World War II metallic aluminium was added (p. 266).
As the raw materials during World War I grew more difficult to obtain, this materi-
al, when intended for aerial bombs, was mixed with ammonium nitrate to obtain a
semi-fluid mass of the ammonal type. Where there was a lack of hexyl it was some-
times replaced by hexanitrodiphenyl sulphide and trinitroanisole was substituted
for TNT. This led to the following mixtures:

(1) 50% of TNT

50% of hexanitrodiphenyl sulphide

(2) 50% of trinitroanisole

50% of hexanitrodiphenylamine

(3) 50% of trinitroanisole

1 50% of hexanitrodiphenyl sulphide

These mixtures were poured into the shells as a molten mass usually containing
a suspension of unmelted hexanitro compounds.

When toluene was in short supply TNT was partly replaced either by dinitro-
benzene or by trinitronaphthalene. Thus the following mixtures were derived:

(1) DiFp consisting of TNT and dinitrobenzene in various proportions

(2) 65% of TNT

35% of trinitronaphthalene

During World War I, the Italians employed the following fusible mixtures:

(1) MAT-60% of picric acid 1 begins to melt at 55-56°C

40% of TNT } completely melts at 85°C

(2) MBT-60% of picric acid

40% of dinitrophenol

A more modern solution of the problem of fusible materials was given by
Lehman [4] in the U.S.S.R. who developed "L-alloy", an explosive consisting
of 95% TNT and 5% trinitroxylene melting at 74°C. L-Alloy has an explosive power

HIGH EXPLOSIVES 249

similar to that of TNT although it differs from the latter in detonating more readily
due to the fact that during the cooling down process very fine crystals of TNT
are formed in L-alloy.

During World War II the Germans partly replaced TNT either by trinitroxylene
or by a mixture of trinitroxylene with TNT. The following mixture was the one
most frequently used :

20% of trinitroxylene 1 ,. , ... „„

80% of TNT j melts cloudtly at 77°C

This mixture was prepared by the nitration of a mixture of nitrotoluenes with
nitroxylenes.

Another mixture used as a substitute for TNT had the following composition:

45% of trinitroxylene
5% of TNT
50% of tetryl,

It was prepared in a similar manner by nitrating a mixture of nitroxylenes with
dinitromethylaniline and mixing it with TNT. This mixture melts at 80°C and is
a more powerful explosive than TNT, but it requires a stronger detonator.

In the United States a fusible mixture of 70% tetryl and 30% TNT, Tetrytol,
was used for demolition charges and land mines, since this mixture has a higher
brisance than TNT and detonates more readily. The melting point of the mixture is
68°C. Cast mixture (solidified) has a density of 1.61-1.65, i.e. greater than that of
TNT, thus making possible the use of charges stronger than those of TNT. Its rate
of detonation is 7350 m/sec and in the ballistic pendulum gives a value of 120 (taking
100 for TNT). Its sensitiveness to impact by rifle bullet is a little higher than that
of TNT. Its chemical stability at temperatures of 100-1 20°C is somewhat lower
than that of tetryl, although specimens were successfully stored at 65°C for 2 years.

Mixtures with trinitroanisole have been employed in Japan; e.g. the 98 H 2
explosive, containing 60% of trinitroanisole and 40% of hexyl, which was used for
filling bombs, torpedoes and depth charges.

During World War II cyclonite was used by all the combatants to increase the
power of composite explosives. Fusible mixtures of TNT with cyclonite were prepared,
mainly with an admixture of aluminium, and mixtures of TNT with hexyl and cyclo-
nite also with admixture of aluminium (pp. 271-272).

In the United States a mixture of 60% cyclonite and 40% TNT was used under
the name of Cyclitol as a filling for aerial bombs. The density of the cast explosive
was 1.65-1.70, its rate of detonation 7800 m/sec, and in the ballistic pendulum it
gave a value of 130 (100 for TNT). In the United States this mixture is considered
to be only a little more sensitive to impact than TNT with a stability similar to that
of cyclonite. This does not agree with T. Urbanski's [5] investigations according
to which the sensitiveness to impact of such a mixture, in a powdered form, and its
ignition temperature (225°C) approximate to the corresponding values for tetryl.

250

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Table 45 and the graphs in Figs. 65 and 66 summarize T. Urbariski's findings
on the rate of detonation, the lead block expansion, the sensitiveness to impact and
the ignition temperatures of powdered mixtures of cyclonite with TNT, of various
composition.

t'C

320

300

>?p — /

s: / „ / -

280

. v>L>^

260

240

7 /

220

kgm/cm 2

4.0

3.5

3.0

2.5

2.0

1J5

1.0

0.5

20 40 60 80 100
Cyclonite, %
100 80 60 40 20
TNT, %

Fig. 65. The relation between the rate of de-
tonation (density 1 .04) and lead block ex-
pansion of mixtures of cyclonite and TNT,
and their composition (according to T. Ur-
banski [6]).

100 80
Cyclonite, %

40
60

60
40

80 100
TNT, %
20

Fig. 66. The relation between the ignition
temperature and sensitiveness to impact of
mixtures of cyclonite and TNT, and their
composition (according to T. Urbanski [6]).

Table 45
Explosive properties of mixtures of cyclonite with TNT

Content of TNT

Rate of detonation

(at a density

of 1.04)

m/sec

Lead block

expansion

cm 3

Sensitiveness

to impact

(50% of explosions)

kg/cm 2

Ignition

temperature

(10°/min)

°C

6590

480

0.22

225

6710

465

0.16

225

6620

445

0.21

224

6460

430

0.42

225

6335

390

0.95*

225

6260

365

2.40

226

6035

345

2.50**

231

5770

315

2.60

295

5570

310

2.80

298

5260

300

3.35

302

100

5230

290

4.10

328

* Equals the sensitiveness of tetryl.

** Equals the sensitiveness of picric acid.

HIGH EXPLOSIVES

251

In Japan, the 94 M mixture consisting of 60% trinitroanisole and 40% cyclonite
was employed for filling torpedoes and armour-piercing shells with hollow charges.
Such a mixture solidifies with a density of 1.64, its rate of detonation being 7700 m/sec.
In the United States a mixture of 75% HMX (Octogen) and 25% TNT was in use
under the name of Octol.

During World War II PETN was also used, although to a lesser extent than cyclo-
nite, since the former is more sensitive to impact and has a lower chemical stability.
Mixtures of various compositions were employed according to their intended use.
The most widely used mixture comprised 50% PETN and 50% TNT. This was
employed in the molten state for filling hand and anti-tank grenades, and powdered
and compressed, for filling detonators.

m/sec

20 40 60 80 100
PETN, %
100 80 60 40 20
TNT, %

100 80
PETN, %

100
TNT, %
20

Fig. 67. The relation between the rate
of detonation (density 1.04) and lead block
expansion of mixtures of PETN and TNT,
and their composition (according to T. Ur-
banski [6]).

Fig. 68. The relation between the ignition
temperature and sensitiveness to impact of
mixtures of PETN and TNT, and their
composition (according to T. Urbanski [6]).

Mixtures of PETN with TNT are known as pentrolit or pentolit. Pentolit 50/50
has a density of 1.63-1.67; its rate of detonation is 7450 m/sec. It is highly sensi-
tive to impact and its stability is lower than that of PETN alone (T. Urbanski,
Kwiatkowski and Miladowski [7]).

The rate of detonation, the lead block expansion, the sensitiveness to impact
and the ignition temperatures of mixtures of PETN and TNT, according to T. Ur-
banski [6, 8] and the rate of detonation according to Laffitte and Parisot [9] are pre-
sented in Table 46 and in the graphs shown in Figs. 67 and 68.

252

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Williamson [10] examined the cast structure of various fusible or semi-fusible
mixtures containing TNT as one component, the other being chosen from the fol-
lowing substances: ammonium nitrate, PETN, cyclonite and tetryl. The author also
prepared casts of TNT alone. When dealing with mixtures of TNT and PETN or
cyclonite, the author discovered that PETN and cyclonite recrystallize in suspension
in molten TNT. This phenomenon leads to an increase in the size of the crystals of
the suspended component (e.g. cyclonite) and, in consequence, reduces the fluidity
of the semi-molten mass on heating. Sometimes, however, a TNT mixture contain-
ing cyclonite in suspension may become more fluid on heating. This occurs when
cyclonite is present in the form of very small irregularly-shaped crystals. This is
also caused by recrystallization which in a given case leads to the formation of
a small amount of relatively large crystals without lowering the fluidity of the molten
mass.

In the solidification of a TNT-tetryl mixture the author did not establish the
presence of an addition compound of the two components in the solidified mass.

Viscosity — another important parameter of molten mixtures of TNT with
cyclonite containing mainly suspended particles of RDX — was examined by a number
of authors [11, 12]. The viscosity increases with increasing RDX content, viz.:

Table 46
Explosive properties of mixtures of PETN with TNT

Rate of detonation

Content of
TNT

Rate of detonation
(at a density of 1.04,

Lead
block

Sensitiveness
to impact

Ignition
temper-

according to Laf-
fltte and Parisot [9]

and 30 mm dia.)

expansion

(50% of explosions)

ature
( 10°/min)

°C

(at a density of 1.0

m/sec

cm 3

kg/cm 2

and 8 mm dia.)

m/sec

6005

515

0.20

200

5200

5870

480

0.15

201

5000

5785

440

0.22

262

4790

5675

425

0.42

267

4600

5510

390

0.80

275

4550

—

4680

5490

370

0.99*

277

4650

—

4950

5385

350

1.25

284

4400

5245

340

1.60

295

4100

5260

315

-1.65

306

3720

—

3800

5050

295

1.75

317

3650

—

2050-3000

100

4865

290

4.10

328

* Equab the Kuitiveneu of tetryl.

HIGH EXPLOSIVES 253

RDX, %

in 100 cm 3 of Viscosity

molten TNT cP

11.526

10 14.42

20 19.16

30 29.07

40 44.02

50 126.70

The Theological properties of the suspensions of RDX in molten TNT were
studied by Koch and Freiwald [13].

MIXTURES WITH AMMONIUM NITRATE

There is another type of fusible mixture with ammonium nitrate as the chief
constituent. These mixtures include substances which act as ammonium nitrate
fluxes, lowering its melting point. Most of them are nitrates of various metals and
various organic bases. Thus, Girard [14] by mixing equal amounts of guanidine
nitrate and ammonium nitrate obtained a mixture, m.p. 140°C. In numerous later
patents the following substances are mentioned as additives to ammonium nitrate:
10-20% of sodium nitrate and approximately 5% of other substances such as inor-
ganic chlorides, urea, acetates or dicyandiamide.

Between 1914 and 1918 the Germans used the following mixtures:

(1) No. 16 (2) No. 20

60-65% of ammonium nitrate 65-67% of ammonium nitrate

10% of sodium nitrate 10-12% of sodium nitrate

5% of dicyandiamide 3% of sodium acetate

25-20% of TNT 20% of TNT

These mixtures melted at 105-1 10°C.

The Dynamit A.G. factory [15] patented a number of mixtures consisting of
ammonium nitrate and nitrates of aliphatic amines, e.g. :

(1) 55% of ammonium nitrate
45% of ethylenediamine nitrate

(2) 60% of ammonium nitrate
40% of methylamine nitrate

(3) 55% of ammonium nitrate
40% of ethylenediamine nitrate

5% of methylamine nitrate

A different method for lowering the melting point of ammonium nitrate was
suggested a few years before World War II. It consists of the addition of hydrated
magnesium nitrate (i.e. containing water of crystallization) Mg(N0 3 ) 2 .6H 2 to

254

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ammonium nitrate. Due to the presence of the water of crystallization such mixtures
melt at temperatures below 100°C (Boyd [16]).

Fusible mixtures with ammonium nitrate as the chief constituent were also manu-
factured and utilized in Germany during World War II under the name of S explo-
sives.

In all explosives of this kind the fusible constituent of the mixture was ammonium
nitrate with other nitrates, e.g. of sodium, potassium or calcium (with water of crys-
tallization), of organic bases such as methylamine, ethylenediamine, guanidine etc.
and urea. Since on solidification these mixtures become very dense and detonate
with great difficulty, PETN, cyclonite or TNT were usually added. The composition
of some of these mixtures is given in Table 47.

Table 47
Composition of ammonium nitrate mixtures

~~ -^____^ Mixtures

Components — — __^^

S-19

H-5

—

43C

Ammonium nitrate

73.3

Sodium nitrate

17.4

Potassium nitrate

Calcium nitrate Ca(N0 3 ) 2 -4H 2

Ethylenediamine dinitrate

—

Guanidine nitrate

Urea

9.3

Cyclonite

—

TNT

1
1

The composition of similar mixtures, containing metallic aluminium, is given
below in Table 56 (p. 271).

Manueli and Bernardini [17] proposed an easily fusible mixture consisting of
ammonium nitrate, guanidine nitrate and nitroguanidine. According to them such
mixtures may be melted at a temperature below 130°C. The explosive Albit, based
on Manueli and Bernardini's patent, consisting of six parts of ammonium nitrate,
two parts of guanidine nitrate and two parts of nitroguanidine, has been used in
Italy.

Urbanski and Skrzynecki [18] found that a eutectic formed by these three compo-
nents solidifies at 113.2°C and contains:

60% of ammonium nitrate
22.5% of guanidine nitrate
17.5% of nitroguanidine

Le Roux [19] proposed the use of fusible mixtures of ammonium nitrate with
tetramethylammonium nitrate (Vol. II, p. 466).

HIGH EXPLOSIVES 255

It was also suggested recently that fusible explosives should be used instead of
semi-fusible ones. For this purpose TNT is replaced as a component by relatively
low-melting nitric esters.

Medard [20], for instance, recommends a mixture of 62% ammonium nitrate
and38%trimethylolpropane trinitrate. Such a mixture has on oxygen balance cor-
responding to complete combustion. A charge with a diameter of 30 mm and a den-
sity of 1.50 detonates with a rate of 6150 m/sec, and a charge with a density of 1 10
detonates with a rate of 4230 m/sec. Its lead block expansion is 127.5 (taking picric
acid as 100).

Its sensitiveness to impact is very low, lower than that of TNT.

A mixture of 40% ammonium nitrate and 60% trimethylolpropane trinitrate may
be melted at 60°C and used as a poured filling for shells. The substance solidifies
into a mass with a density of 1.36, giving a rate of detonation of 5200 m/sec.

MANUFACTURE AND SELECTION OF FUSIBLE MIXTURES
As a rule fusible mixtures are prepared in metallic kettles heated with steam or
water jackets and fitted with stirrers, which are emptied either by tilting or through
a valve placed in the bottom of the vessel.

In France mixtures containing picric acid were prepared in wooden tubs to
avoid the formation of picrates. They were heated with live steam injected through
an ebonite nozzle. After the mass has been mixed the steam supply was turned off and
water was decanted from above the layer of nitro compounds by a glass syphon.
The mixture was drawn off with wooden buckets in which it was cooled while being
stirred continuously. The granular mass was then poured onto wooden trays where
it was cooled further. Stirring was still continued during cooling, so as to produce
granules approximately 10 mm dia.

It must be borne in mind that generally aromatic nitro compounds are not
highly sensitive to impact and friction, but become more sensitive at elevated tem-
peratures as they melt (changes in the sensitiveness of TNT are discussed in
Vol. I, p. 319). If, therefore, a mechanical device is used for the preparation of
mixtures by melting, its construction should be such as to exclude any possibility of
friction or impact. It is probably best to use a converter heated with steam or water
jackets and fitted with a stirrer that can be lifted out by a special arrangement.
After the stirrer has been removed, the contents are poured out by tilting the vessel.
Fusible mixtures, with a composition having a suitable melting point, are selected
by the thermal analysis of the two- and three-component system. Thermal analysis
determines either the beginning and end of solidification or the beginning and end
of the melting of the mixture.

To examine the possibility of using a given mixture for filling shells the changes
in these temperatures are determined with mixtures of varied composition and it is
established whether the components are eutectic mixtures or molecular compounds.
The analysis also shows whether, in liquid phase, the components form a homoge-
neous system, solid solutions, etc.

256' CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In fact, the presence of solid solutions considerably effects the homogeneity of
the mixture, preventing the formation of inner cracks and fissures. If the solidifying
point of the eutectic is too low (i.e. below 60°C) the explosive may exude from the
shell, whereas too high a solidifying point hinders melting, and in turn complicates
the process of filling the shell.

The literature on thermal analysis of mixtures either deals with purely theoretical
studies of the formation of molecular compounds, solid solutions, etc., e.g. Kremann's
[21, 22] work, or describes the search for materials of immediate practical value.

Research relating to practical matters has two aims:

(1) To determine the contents of various isomers in nitration products. Such
studies include the work of Giua [23] on systems consisting of DNT and TNT iso-
mers, those of Pascal [24] on systems consisting of dinitro- and trinitronaphthalene
isomers, and those of Andrews [25] and of Wyler [26] on systems consisting of
dinitrobenzene isomers. The graphs obtained by these authors may be helpful
when studying the composition of nitration products.

(2) To decide whether a given mixture is suitable for filling shells or for pro-
ducing non-freezing dynamites. Thermal analysis of the components of mixtures
in practical use, e.g. nitroglycerine and centralite, fulfils a similar, practical aim by
explaining the interaction of these substances, in particular whether they form
simple eutectics, molecular compounds or solid solutions.

Studies of this kind which include a number of papers mentioned in corresponding
sections of this book, are recommended for reference. They are concerned with the
following substances :

(a) Mixtures containing aromatic trinitro compounds:

Yefremov et al. [27, 28] (systems containing TNT, picric acid, trinitrocresol,
trinitroresorcinol, tetryl, trinitroxylene etc.) ; Wogrinz and Vari [29] (systems contain-
ing TNT and picric acid); C. A. Taylor and Rinkenbach [30] (systems containing
TNT, picric acid and tetryl); Jovinet [31] (the system: picric acid-nitronaphthalene);
Hrynakowski and Kapuscinski [32] (systems containing TNT and picric acid);
T. Urbanski and Kwiatkowski [33] (systems containing picric acid and dinitro-
naphthalenes).

(b) Mixtures containing nitroglycerine: Tamburrini [34], Kurita and Hagui
[35], Medard [36], and Hackel [37].

(d) Mixtures containing cyclonite: T. Urbanski and Rabek-Gawrohska [40].
Khaibashev and Bogush [41] examined ternary systems containing cyclonite.

They discovered eutectic mixtures of the followin compositions:

(1) 82% of m-dinitrobenzene

9% of cyclonite m.p. 80.5°C

9% of trinitroxylene

(2) 74.5% of m-dinitrobenzene

7% of cyclonite m.p. 74.5°C

18.5% of 1,8-dinitronaphthalene

HIGH EXPLOSIVES 257

(e) Mixtures of guanidine nitrate, nitroguanidine and ammonium nitrate:
T. Urbanski and Skrzynecki [18].

(f) Three-component mixtures of inorganic nitrates: T. Urbanski and Koto-
dziejczyk [42] established the existence of a eutectic of the following composition :

66.5% of NH4NO3

21.0% of NaNOj f.p. 118.5°C

12.5% of KNO3

THE PHLEGMATIZATION OF FUSIBLE MIXTURES

Even by the end of the nineteenth century it had become clear that some explo-
sives safe to handle in principle (e.g. picric acid), are nevertheless too sensitive to
the impact that occurs when the projectile containing them strikes against heavy
armour plate. Attempts were made therefore to decrease this sensitiveness by de-
sensitization, or "phlegmatization" of the explosive.

In France, picric acid intended for armour-piercing shells was phlegmatized with
paraffin. MP (melinite paraffine) mixture, containing 88% of picric acid and 12% of
paraffin, was chosen for this purpose. Paraffin desensitizes picric acid, making
it less sensitive than TNT. Since paraffin is insoluble in molten picric acid, the mix-
ture is prepared by melting the paraffin in an aluminium kettle, heating it to 100°C
and adding powdered picric acid while stirring with a wooden paddle. The granules
so formed are transferred to a table, rolled out and mixed by rubbing several times
through a sieve.

Paraffin, however, is not a good phlegmatizing agent. It was discovered that
higher fractions of crude oil, with a waxy consistency, and composed of more polar
molecules, are considerably more effective so that desensitization may be achieved
by using a smaller amount of phlegmatizing substance. This is decidedly preferable
since too large an admixture of a phlegmatizing substance weakens the explosive.

Montan wax, widely used in Germany, is a better phlegmatizing agent than
paraffin. Hence before and during World War I German naval armour-piercing
shells were filled with TNT, desensitized by 6% of montan wax.

During World War II PETN, desensitized by 10% (sometimes 5 or 15%) of
montan wax was employed for small shells (e.g. 20 mm) in Germany.

The mixing of TNT or PETN with montan wax is usually carried out under
water at a temperature above the melting point of wax (70°C). With TNT the tem-
perature should be maintained below 80°C; with PETN the whole is heated to
95°C and vigorously agitated so that the phlegmatizing substance is dispersed. The
whole is then cooled to 30-35°C, while still being agitated. The granules and crystals
of the explosive, coated with a layer of wax, are filtered off and dried at 60-65°C.
A detailed description of PETN phlegmatization is given in Vol. II, p. 189.

The inclusion of wax-type phlegmatizing substances in the mixture facilitates
the pressing of the explosive as it decreases the friction between the crystals. Thus
a mixture with a higher density can be obtained by applying a lower pressure.

258 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Non-explosive substances which, when molten, act as solvents for explosives,
are another type of desensitizing agents. For instance, pentaerythritol tetra-acetate
was suggested by Bombrini-Parodi Delfino [43] as a phlegmatizer of PETN.

The phlegmatization of PETN with nitronaphthalene recommended by T.
Urbanski [6] is based on the same principle of the partial dissolution of PETN.
A mixture of PETN plus 20% nitronaphthalene has approximately the same sensi-
tiveness to impact as tetryl, and with 40% nitronaphthalene the same sensitiveness
to impact as picric acid.

Chlorofluoropolymers of the Kel and Exon types are now used in the U.S.A.
as modern phlegmatizers of cyclonite, e.g.

90/10 RDX/Kel F 3700 in granules 0.8-1.4 mm dia.

94/6 RDX/Exon 461 in granules ca. 0.3 mm dia.

Nitrocellulose dissolved ("gelatinized") with phosphoric acid esters can also
be used to produce a granulated and partly desensitized nitramine explosives, e.g.
a mixture:

94/3/3 HMX/NC/tris(j8-chloroethyl)phosphate with 0.1% diphenylamine as
stabilizer, in granules 1-7 mm dia. [44].

Phlegmatizing compounds in chlorate, and perchlorate explosive mixtures play
a special role since they are essential as a combustible constituent. They are discus-
sed in more detail on pp. 274-280.

SEMI-FUSIBLE AND INFUSIBLE EXPLOSIVES

The molecular composition of picric acid was established in the early ninteenth
century when it was the only highly-nitrated aromatic compound then known. It
was evident that its oxygen content was insufficient for complete combustion (to C0 2
and H 2 0). This was considered to be an adequate foundation for the erroneous
theory that, because of its insufficient oxygen content, the substance has no explosive
properties. It was believed that explosive properties are achieved only by mixing
picric acid with oxidizing agents such as chlorates, or sodium or potassium nitrates.

These mixtures found no practical application since picric acid gradually reacts
with salts to form picrates with the evolution of free acid. The picrates so formed
are highly sensitive to friction and impact, and the free acid acts corrosively. Mix-
tures with chlorates showed a particular sensitiveness to friction and impact, hence
doubt was expressed as to their practical value. Nevertheless, the idea of completing
the defective oxygen balance in aromatic nitro compounds by the addition of
such oxidizing agents as nitrates was carried out in such a way as to produce
mixtures useful for various practical purposes.

Cheltsov [45] obtained (1886) a stable explosive, suitable for storage, called
Gromoboi or Ma'izit, by mixing ammonium picrate with ammonium nitrate in the
ratio of 72.5 : 27.5.

HIGH EXPLOSIVES 259

To achieve the necessary stability he used its ammonium salt, which has no
acidic properties instead of picric acid.

The same principle has been employed in some countries (e.g. U.S.A.) to produce
the explosives in which the chief component was ammonium picrate together with
ammonium nitrate.

There have also been interesting and promising attempts to use ammonium
nitrate as a component of high explosives. Particularly noteworthy are the experi-
ments in which potassium nitrate in blackpowder is completely or partially replaced
by ammonium nitrate to improve brisance. This led to the ammonium nitrate mix-
tures, Ammonpulver M90 and M96, 15/85 mm. They were employed in Austria
for filling shells at the end of the nineteenth century but were not used for long,
due to the difficulty of detonating them.

Ohlsson and Norrbin [46] suggested another type of ammonium nitrate explo-
sive for mining purposes (see p. 395).

With the development of the organic chemical industry, aromatic nitro com-
pounds of the TNT type were introduced as ingredients of composite explosives. TNT
is preferable to picric acid since it has no acidic properties and hence is much less
reactive. Mixtures with TNT and similar nitro compounds showed an excellent
chemical stability.

High explosive mixtures with potassium and sodium chlorates or potassium
and ammonium perchlorate belong to a separate group; a mixture of these salts
with any combustible used in a suitable quantity will act as an explosive.

MIXTURES WITH NITRATES -MAINLY WITH AMMONIUM NITRATE

Mixtures of aromatic nitro compounds with ammonium nitrate were widely
used during World War I, when the enormous demand for high explosives could
not be met by the output of TNT, trinitronaphthalene, picric acid, trinitroanisole,
trinitrophenetole, dinitrobenzene, hexyl etc.

Ammonium nitrate, which was then being manufactured from atmospheric
nitrogen for the first time, is the most readily available explosive ingredient, and
all the more valuable since on decomposition it leaves no solid residue and ensures
a great volume of gaseous explosion products. Its great disadvantage lies in its high
hygroscopicity but this is unimportant if the explosive charge is tightly packaged.

The explosive properties of mixtures with ammonium nitrate depend on the
quantitative relationship between the oxidizing agent and the explosive or combustible
substance. According to Parisot and Laffitte's [9, 47] investigations the explosive
properties of mixtures of aromatic nitro compounds with ammonium nitrate vary
with the change in composition of the system in an almost rectilinear manner. The
graph in Fig. 69 shows how the rate of detonation depends on the composition
of mixtures of tetryl or picric acid with ammonium nitrate. T. Urbariski et al. [48]
also obtained a rectilinear relationship for nitrostarch mixtures with ammonium or
sodium nitrate (Fig. 71, p. 265).

260

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The sensitiveness to impact of mixtures of nitro compounds with ammonium
nitrate is higher than that of pure nitro compounds due to the friction produced
by the hard crystals of the ammonium nitrate. This was established for TNT mixtures
with ammonium nitrate by Hackel [49] (c.f. pp. 262-263).

6000

1000

NH4NO3 - tetryl
NH4NO3 - picric acid.

1400

10 20 30 40 50 60 70 80 90 100

NH4NO3, %

Fig. 69. The relation between the rate of detonation and the composition of tetryl or
picric acid mixtures with ammonium nitrate (Parisot and Laffitte [9]).

One of the most widely used mixtures comprised ammonium nitrate with
dinitronaphthalene in a ratio giving complete decomposition with a zero oxygen
balance :

87.5% of ammonium nitrate
12.5% of dinitronaphthalene

This mixture was used extensively during World War I for filling artillery shells
in France (under the name of Schneiderite), Italy and Russia. The spelling "sznajderyt"
was adopted in Poland and "shnaiderit" in the U.S.S.R.

Schneiderite may detonate at a density of 1 .30 with a rate ranging from 3815-5840
m/sec, depending on the way the ingredients are mixed. It is used either as a powder
pressed into shells with an initiating charge of TNT (called SCPT-explosive in France)
or as granules obtained by breaking up the pressed cake. The granules are poured
into the shells and the spaces between filled with molten TNT. In France this kind
of explosive bore the symbol STF.

This explosive used in France and Russia had the composition

78% of ammonium nitrate
22% of dinitronaphthalene

HIGH EXPLOSIVES

261

During World War II these substance were not widely employed, being rela-
tively weak explosives; they are not so valuable in this respect as similar TNT mix-
tures.

TNT mixtures with ammonium nitrate were in common use during the World
Wars I and II in Germany and Great Britain. The mixture consisting of:

40% of ammonium nitrate
60% of TNT

acquired considerable importance. It was known in Germany as Fp 60/40 (Fiill-
pulver 60/40) and in Britain as Amatol 40/60. It was heated to a temperature above
the melting point of TNT and then poured into shells as a semi-molten mass. Due
to the shortage of TNT in Germany during World War I originally a part and
then the whole of the TNT in this mixture was replaced by dinitrobenzene ' This
resulted in the mixture DiFp 60/40 with the following compositions:

40% of ammonium nitrate
60% of dinitrobenzene or a TNT mixture
with dinitrobenzene

The Germans also used a similar mixture containing trinitroanisole instead
of TNT for fining aerial bombs. In Britain other Amatols were also employed, i.e.
ammonium nitrate mixtures with TNT in various ratios (Table 48).

Table 48
Composition and properties of Amatols

Composition

Consistency

Properties

Name

NH 4 N0 3

TNT

Density

Rate of
detonation

Lead block

expansion

cm 3

m/sec

(TNT
=290 cm3)

Amatol 40/60
Amatol 45/55
Amatol 50/50

40
45
50

55
50

Semi-molten
when hot
Can be cast

1.54-1.59

1.56
1.60

6470-7440

7020
5850*

320-350
.340-360

Amatol 60/40

1.50

6060**

Amatol 80/20

Loose powder,
plastic when
hot ^

1.60

1.46-1.50

1.60

5600*

5080-5920*
5200*

350-370
370-400

* According to Evans [50].
** In a steel tube 17 mm dia. according to Copp and Ubbelohde [51].

Amatol 80/20 is a mixture approaching the proportion in which all the oxygen
contained in ammonium nitrate is used (the oxygen balance is +1.2% in Amatol

262

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

80/20). This is a free-flowing mixture when cold. It is usually stemmed or extruded
into the shells by means of a screw wheel (worm wheel).

Amatols 50/50 and 60/40 were loaded into shells as substitutes for Amatol
40/60, in a semi-molten state obtained by heating to above the melting point of

TNT.

Semi-fusible mixtures which contain 40-60% of nitro compounds tend to separate
out after the nitro compound has been melted. During World War II these mix-

40 60 80 100
NH4NO3, %

100 80 60 40 20
TNT %

Fig. 70. The sensitiveness to impact of TNT mixtures with ammonium nitrate. Curves
of probability of 10 and 50% explosions (Hackel [49]).

tures were used for filling large aerial bombs. To prevent stratification of the ingre-
dients (the separation of crystals of ammonium nitrate) emulsifying substances
were added to the mixtures.

TNT mixtures with ammonium nitrate are more sensitive to impact than TNT
itself. As shown by Hackel [49] (Table 49 and Fig. 70) mixtures containing 30-60%
of ammonium nitrate are equally as sensitive as picric acid. Mixtures of this kind
should not, therefore, be used for filling high initial velocity heavy calibre shells,
e.g. armour-piercing shells.

Amatols in the form of cast charges detonate more readily than TNT in the
same form. Initially, it was believed that the size of the ammonium nitrate crystals
influences the ease of detonation, viz. that finely ground ammonium nitrate
facilitates detonation.

Hackel [52], however, made it clear that in Amatols the ease of detonation also
depends on the size of the crystals of the solidified TNT and increases in proportion

HIGH EXPLOSIVES

Table 49

263

Sensitiveness to impact of TNT mixtures with ammonium nitrate (according to

Hacked

Composition of mixture

Sensitiveness to impact.

TNT

Ammonium

Impact of a 5 kg weight

nitrate

causing 10% of explo-

sions, kgm/cm 2

100

1.37

0.85

0.74

0.71

0.64

0.57

0.48

0.53

0.81

1.27

to the decrease of size of the crystals (i.e. in proportion to the rate of the cooling
of the explosive).

During World War II Amatols of increased explosive power, in which a part
of the ammonium nitrate was replaced by cyclonite, were also used. Thus German
Ammonals were evolved from Amatol 50/50. Dinitrobenzene was used in Amatols
39a and 40 (Table 50) as a substitute for TNT.

Table 50
Composition of Amatols with cyclonite

Ingredients

Amatol 39

Amatol 39a

Amatol 40

TNT

Dinitrobenzene

—

Cyclonite

5-10

Ammonium nitrate

40-45

The mining explosive Donarit was first used for filling hand grenades and
facturing demolition charges during World War I. It consisted of

manu-

80% of ammonium nitrate
12% of TNT

4% of nitroglycerine (gelatinized with
collodion cotton)

4% of wood flour

Due to the presence of nitroglycerine, this material was found to be too sensitive
to mechanical impulses (it exploded when struck by a 2 kg weight falling 30 cm).

264 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Kast [3] therefore proposed, as an alternative, an explosive called Perdit which

consisted of:

72% of ammonium nitrate
10% of potassium perchlorate
15% of liquid DNT
3% of wood flour

This material gives a lead block expansion of 380 cm^. It was employed widely up
to the end of the World War I not only for hand grenades and demolition char-
ges, but also for filling shells with a low muzzle velocity.

Apart from composite explosives with ammonium nitrate, mixtures with other
nitrates also acquired temporary significance, e.g. a mixture used in Russia during
World War I for filling hand grenades had the following composition:

60% of TNT
35% of potassium nitrate
4% of ammonium nitrate

and Baratol, employed in Britain during both World Wars, consisting of TNT
and barium nitrate in the ratio of 40 : 60.

An explosive called Macarit with a composition recommended by Macar [53] :

28% of TNT
72% of lead nitrate

also achieved some importance before World War I.

A similar substance under the name of Piombitto was used in Italy at the same
time for filling artillery shells.

During World War I, in the United States Trojan Explosive was widely employed
for filling hand grenades and mortar bombs. Its explosive constituent was nitrostarch:

23-27% of nitrostarch
31-25% of ammonium nitrate
36-40% of sodium nitrate
1.5-2.5% of charcoal
0.5-1.5% of lubricating oil

0.5-1.5% of calcium carbonate or magnesium oxide
0.2-0.4% of diphenylamine
0-1.2% of water

Since this explosive is infusible it was loaded by ramming down with a wooden
plug. Grenite was another granulated nitrostarch material for filling hand grenades,
consisting of 97% nitrostarch with lubricating oil (1.5%) and gum arabic (1.5%)
as granulating materials.

The explosive properties of nitrostarch mixtures (containing 12.7% of nitrogen)
with ammonium or sodium nitrates have been studied by T. Urbanski et al. [48].
Values for rate of detonation of mixtures with a density of 1.0 are given in Fig. 71.
They conform to Laffitte's rule that the variation in rate of detonation with the
composition of mixtures with ammonium nitrate is almost rectilinear.

HIOH EXPLOSIVES f265

Other examples of infusible explosives are the Ammonites with the composition
shown in Table 51.

A shortage of ammonium nitrate in Germany (from 1944) led to the use of a TNT
mixture with 40% of sodium nitrate and later even with 50-60% of sodium chloride
for shell filling as cast charges. Sodium chloride was used merely to keep the right
charge weight.

Sometimes sodium and potassium silicates were used instead of sodium chloride
Thus the consumption of TNT did not need to be increased while the consistency

m/sec

5400
5000

4000

3000
2800
2600
2400
2200
2000

10 20 30 40 50 60 70 80 30 100

inn nn an .,„ NltrOStdrCh, %

100 90 80 70 60 50 40 30 20 10
NH4NO3 or NaN0 3 , %

Fig. 71. The rate of detonation of nitrostarch mixtures with ammonium or sodium
nitrates (according to T. Urbanski et al. [48]).

and density of Amatols were maintained in mixtures which in turn averted the
necessity for the alteration of the design of shells and the method of filling. Such
mixtures were used to the same extent as Amatols for filling shells, bombs and
hand grenades. It is obvious that they were of little value.

Table 51
The composition of Ammonites

Ingredients

H-l

H-8

Ammonium nitrate

Potassium nitrate

Calcium nitrate (4H 2 0)

Cyclonite

PETN

In Japan explosive consisting of 60% of trinitroanisole and 40% of ammonium
nitrate was manufactured and used for filling bombs.

266 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

During World War II a mixture of TNT (48%) with ammonium picrate (52%)
known as Picratol was employed to replace pure ammonium picrate (Explosive D)
in the United States. This was a semi-fusible mixture cast into shells at a temperature
above the melting point of TNT. After solidification the density of the mixture
was 1.62. In accordance with the opinion prevailing in that country that ammonium
picrate is an explosive with a low sensitiveness to impact, it was used for filling
armour-piercing shells.

Various loose powder mixtures of ammonium picrate and aluminium were
also used by the Japanese during World War II. They will be dealt below, as mixtures
with aluminium.

The preparation of infusible (loose powder) ammonium nitrate mixtures is
usually carried out in the same way as the manufacture of composite mining explo-
sives. Since the military explosives are not very sensitive to mechanical stimulants,
and it is important to obtain a high density, mixing is usually performed by edge
runner mills. /

The mixing of semi-fusible explosives is commonly carried out in kneaders,
first heating the ammonium nitrate to a temperature of 85-90°C and then pouring
in the molten TNT. After stirring, the material is cooled down in the kneader to
produce a mixed crystalline mass.

MIXTURES WITH ALUMINIUM AND OTHER METALS

The addition to an explosive of combustibles which burn with very great evolution
of heat is advantageous in spite of the fact that the oxygen balance is impaired. The
heat of explosion so obtained is very great and the temperature of the explosion
products is very high.

The following metals have been suggested for this purpose: magnesium, alu-
minium, zinc and also silicon; sometimes ferro-silicon, alumino-silicon and calcium
silicide are also employed. Deissler [54] was the first (1897) to recommend aluminium
as a component of explosives. He was followed by Goldschmidt [55], Escales [56],
von Dahmen [57] and Roth [58]. In later years Kast [59] investigated military ex-
plosives which contained aluminium.

Magnesium and zinc are readily oxidized, and are liable to undergo oxidation
during the storage of mixtures containing them, hence they have not been utilized
for military purposes. Apart from this, magnesium is a valuable component of
various pyrotechnic mixtures such as those used in signals or for illumination, for
which it is hard to find a substitute. With the exception of calcium silicide the silicon
alloys burn with more difficulty and are less efficient. For this reason aluminium and
calcium silicides are the most widely used.

Originally aluminium was employed in the form of fine powder ("aluminium
bronze"). It appeared later that such a high degree of sub-division is unnecessary
and good results may also be achieved with aluminium filings, shavings and especially

HIGH EXPLOSIVES 267

flakes. The latter form is particularly advantageous as the smaller total surface of alu-
minium present with mixture slows down the rate of oxidation during storage.
In explosives in which there is no surplus of oxygen, aluminium reacts initially
with carbon dioxide formed as an explosion product, according to the equation:

3C0 2 + 2Al=3CO + AI2O3 + 180.7 kcal (1)

The heat effect of this reaction is very great and it makes a big contribution to the
general heat balance. t

Originally the addition of aluminium was limited to explosives with a positive
oxygen balance, i.e. mixtures containing a considerable amount of an oxygen car-
rier. However, during World War II, the Germans extended the use of aluminium
by adding it to nitro compounds, for example to a mixture of TNT with hexyl.

It might seem paradoxical to add aluminium to such explosives, as it would
result in a large reduction of volume of the gases:

3CO + Al ->- 3C + A1 2 3 (2)

However, under the very high pressures prevailing during detonation, the car-
bon monoxide decomposes in part into carbon dioxide and carbon. Aluminium
reacts with carbon dioxide according to eqn. (1) and with water according to eqn.

(3):

3H 2 + 2A1 -> 3H 2 + AI2O3 (3)

This would develop a considerable heat. Thus, replacing 15% of TNT-hexyl
with aluminium would increase the heat of detonation from ca. 1000 kcal/kg to
ca. 1400 kcal/kg [60].

In military ammonium nitrate explosives containing aluminium even those in
which the oxygen balance is negative, the main reaction is assumed to be that of
oxidation of the aluminium by reaction with ammonium nitrate:

2A1 + NH4NO3 -> AI2O3 + 2H 2 + 2350 kcal/kg

The gas volume is 502 l./kg and calculated explosion temperature 5400°C. It is
also possible that aluminium reacts with nitrogen to form aluminium nitride (A1 2 N 2 ).
The reaction is also exothermic (- J#=80 kcal). (The work was done at the Chem-
isch-Technische Reichsanstalt in Germany in the period between the two World Wars
[61].)

The mixtures containing the oxidizing agent, e.g. ammonium nitrate and alu-
minium were termed Ammonals.

The system: ammonium nitrate, aluminium, nitro compound (e.g. TNT) would
be expected to undergo gradual decomposition, e.g. that in stored shells and bombs
filled with such mixtures changes would occur, leading to the oxidation of the alumi-
nium. Obviously, a mixture containing oxidized aluminium has lower explosive
power than the same mixture containing metallic aluminium. It was therefore very
important to determine the mechanism of the oxidation of aluminium. It became
apparent that this is caused by impurities in the ammonium nitrate, not by the

268 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ammonium nitrate itself. Chaylan's [62] experiments showed that mixtures of pure
ammonium nitrate, TNT and aluminium are stable within the temperature range of
60-95°C, and at 100°C, after 1000 hr of heating, only a few cm^ of gas are evolved.
Experience has shown that traces of chlorides in ammonium nitrate should be
avoided as they can induce aluminium to react even at room temperature. Aluminium
also undergoes a corrosion reaction which is caused by the metallic walls of the
shell, and which has not yet been fully explained. In all probability its origin is
electrochemical. Other data show that aluminium reacts with copper alloys.

It is well known that ammonium nitrate evolves ammonia on storage, particularly
in the presence of moisture. It was found that the amount of ammonia evolved in
the presence of aluminium is much greater [63]. The evolution of ammonia is particu-
larly undesirable in the mixtures containing TNT, as TNT reacts with ammonia
to yield readily ignitable compounds (Vol. I, p. 304; see also [64]).

According to Aubertein [65] aluminium particles used as a component in explo-
sives may be coated with 2-8% of paraffin, mineral oil or resin to prevent premature
reaction with other ingredients.

Ammonals were originally intended for use in mines. In Austria [66] for instance,
ammonals with the following composition were used in gas-free mines:

80-90% of ammonium nitrate
4-6 % of charcoal
4-18% of aluminium

Soon afterwards, however, the use of ammonals in collieries was forbidden
and their utilization restricted to above ground operations, e.g. quarrying. During
World War I ammonals were extensively used for this purpose.

Ammonal was used to a certain extent in Russia and Great Britain. Its composi-
tion was as follows :

65% of ammonium nitrate
15% of TNT
17% of aluminium
3% of charcoal

The presence of charcoal, however, proved to be unnecessary; during World War I
Russian mines were loaded with ammonal made without it:

7.65% and 68% of ammonium nitrate
16.0% and 15% of TNT
7.5% and 17% of aluminium

Detonators for initiating a Schneiderite charge in Russian shells were also
filled with an aluminium mixture:

65% of ammonium nitrate
15% of TNT
20% of aluminium

HIGH EXPLOSIVES

269

According to Forg [66] even before 1914 T- Ammonal manufactured by the
G. Roth factory at Felixdorf was employed for filling artillery shells in Austria.
The composition of this explosive was as follows:

45% of ammonium nitrate
30% of TNT
23% of aluminium
2% of "red" charcoal (p. 345)

Kast [3] gives for this substance the following figures as characteristic of its ex-
plosive power:

Heat of explosion 1465 kcal/kg

Gas volume (F ) 605 l./kg

Temperature of explosion 4050°C

Specific pressure (/) 9900 m

Rate of detonation in an iron pipe at a density of

1.62 5650 m/sec
Lead block expansion 470 cm 3

In Germany "Deutsche Ammonal" was used for filling large calibre shells.
Various ammonals were employed to an appreciable extent for filling bombs, land
mines in Italy and to a lesser extent in France. The composition of these explosives
is represented in Table 52.

Table 52
Composition of ammonals

Italian ammonals

Belgian
Sabulite

Ingredients

Nitramit

Echo

French

British

German

Ammonium nitrate

TNT

Nitrocellulose

5.5

—

Paraffin

—

Animal grease

7.5

—

Charcoal

—

Aluminium.

—

Calcium silicide

—

Ammonals containing a little or no TNT detonated with difficulty and were
therefore not pressed but rammed down to a relatively low density. For the same
reasons their use was restricted to the shells with a low muzzle velocity. Shells with
a high muzzle velocity were loaded by pressing with ammonals rich in TNT (T- or
German ammonals). Ammonals were also used when a high brisance was required,
e.g. for filling torpedoes.

A few other simple German ammonals containing ammonium nitrate were
used for shell filling during World War II. Their composition is given in Table 53.

270

chemistry and technology of explosives

Table 53
Composition of German ammonals

Ingredients

Fp* 19 ,.

Fp* 13-113

Fp*110

Ammonium nitrate

TNT

Naphthalene

—

Wood meal

—

2.5

Aluminium

7.5

* Fp stands for Fiillpulver.

The compositions of American ammonals are collected in Table 54.

Table 54
Composition of U.S. ammonals

Ingredients

Ammonium nitrate

22.0

72.0

65.0

TNT

67.0

12.0

added 15.0

Aluminium (fine

powder)

11.0

16.0

Aluminium (coarse grain)

—

16.0

Charcoal

3.0

During World War I another type of explosive — "Mischpulver"— was used
by the Germans for filling hand grenades. This was blackpowder mixed with potas-
sium perchlorate and aluminium. The mixture consisted of:

83% of blackpowder
12% of potassium perchlorate
5 % of aluminium

The presence of perchlorate and aluminium conferred high explosive properties on
the mixture.

Perchlorate ammonal was also used in Britain for filling land mines. It consisted
of:

78% of ammonium perchlorate
16% of paraffin
6% of aluminium

After World War I the use of ammonals was restricted to quarrying: in coal
mines they were banned since their high flame temperature (due to presence of
metallic aluminium) makes them inherently dangerous there.

In Spain during the Spanish Civil War, General Franco's forces used Ammonal I
composed of [63]:

92.4% of ammonium nitrate
6.6% of charcoal
1.0% of aluminium

HIGH EXPLOSIVES

271

During World War II the use of aluminium in military explosives was resumed
on a wide scale. Thus sea mines and German torpedoes were loaded with a fusible
mixture of hexyl and TNT supplemented by 10-25% of metallic aluminium (Trial).

TNT was partly replaced by dinitronaphthalene in the KMA mixture. Alterna-
tively, hexyl was replaced by dinitrophenylamine to form fusible S-6 mixtures.

Table 55
Composition of KMA and S-6 mixtures

Ingredients

KMA

S-6

S-6
modified

TNT

Dinitronaphthalene

—

Hexyl

30-35

Dinitrodiphenylamine

15-20

Aluminium

15-25

Haid in the Chemisch-Technische Reichsanstalt in Berlin suggested [67] substitute
fusible mixtures of the S-type which were introduced towards the end of World
War II owing to the shortage of nitro compounds. Their composition is given in
Table 56.

Table 56

Composition

OF S-TYPE SUBSTITUTE MIXTURES

Ingredients

S-16

S-19
(modified)

S-22
(with cyclonite)

S-22 S-26
(with hexyl)

Ammonium nitrate

Sodium nitrate

6-8

Potassium nitrate

0-2

4.2

Ethylenediamine dinitrate (PH-Salz)

Urea

—

1.8

—

1.8

Hexyl

—

Cyclonite

Aluminium

For filling V2 missiles Trialen, a mixture of TNT, cyclonite and aluminium, was
used in Germany.

Another German explosive, Hexal, consisted of:

71% of cyclonite
4% of montan wax
25% of aluminium

The moulded slabs of this explosive were cast into the shell and the spaces between
them was filled with a fusible mixture, e.g. KMA.

In Great Britain a cast explosive Torpex was developed during World War II.
It was composed of 41% of RDX, 41% of TNT and 18% of aluminium. It was
used for filling torpedoes and bombs.

272

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Due to the sensitiveness of Torpex a few modifications were developed, as for
instance the cast explosives DBX and HBX [68] (Table 57).

Table 57

Ingredients

DBX

HBX

RDX

39.6

TNT

37.8

Ammonium nitrate

—

Aluminium

17.1

Desensitizer

—

5.0

Calcium chlorate

0.5

Density

1.68

In the U.S.A. Tritonal, a cast mixture of 80% TNT and 20% aluminium was
used. It gave a density of 1.73 [68].

In Japan a cast mixture composed of 60% of trinitroanisole and 40% of alumi-
nium was used for filling high explosive incendiary shells. On solidification this
mixture has a density of 1.90.

E-4, an infusible ammonium nitrate explosive was composed of:

44% of ammonium nitrate
10% of sodium nitrate
2% of urea
14%ofhexyl
30% of aluminium

During World War II the Japanese employed mixtures with aluminium or
ferro-silicon. Some examples of these are tabulated below (Table 58).

Recently The Dow Chemical Company (Midland, Michigan) developed a power-
ful new plastic explosive MS-80 containing 20% aluminium. No details on the
composition are available [69].

Table 58
Mixtures with aluminium and ferro-silicon

Type 4, Mk 5,
Ingredients Type 88 Ko | „-

! i K ->

Type 1, Mk 1,
PI

Type 1, Mk6,
P6

Ammonium nitrate
Ammonium perchlorate
Ammonium picrate
Aluminium
Ferro-silicon
Wood meal
Lubricating oil

75
55

10
5
1

81
16

2
1

11
2
1

Density

Rate of detonation Cm/sec)

1.05
3800

1.05

1.16
4280

1.13
4620

HIGH EXPLOSIVES

273

In addition to aluminium, calcium silicide or ferro-silicon, the use of silicon
was also recommended. The comparative figures illustrating the effect of the addi-
tion of these compounds on the strength of the explosive have been given by Sar-
torius [70] (Table 59). The power is expressed in terms of expansion in the lead
block (taking picric acid as 100).

Table 59
Influence of the ingredients on the explosive power of mixtures

Ingredients

Single
explosive

Mixtures with addition of

10% of
silicon

10% of
ferro-silicon

10% of
calcium silicide

PTN

Cyclonite
Picric acid
TNT

146

135

100

152

143

108

148
141
102
94.5

146

140

106

It was also observed that the increase in lead block expansion caused by the
addition of silicon is a half or two thirds smaller than that caused by the addition
of aluminium.

Zinc appeared to be too reactive and has not found any practical applica-
tion.

The preparation of S-type mixtures. The following description of the prepara-
tion of S-16 mixture is provided as an example. In a stainless steel kettle of 500 kg
capacity equipped with a stirrer and steam-heated to 118°C, 250 kg of the mixture
is prepared by pouring in the ingredients, in the following sequence:

50 kg of ethylenediamine dinitrate
30 kg of sodium nitrate
10 kg of potassium nitrate
160 kg of ammonium nitrate

The ingredients are added slowly, in batches, so that the mixture heats up before
the next batch is added. It is necessary to prevent the temperature from dropping
below 90°C otherwise melting is prolonged.

After all the ingredients have been introduced into the kettle the temperature
rises to 113-115°C. At this temperature the mixture forms a clear melt in 20 min.
If the mixture remains turbid, further heating and stirring is necessary.

When melting is complete the mixture is allowed to stand without stirring for
5 min, after which the contents are emptied through the bottom valve into another
kettle situated below. 200 kg of aluminium powder and 50 kg of cyclonite are added.
The temperature falls but must not be allowed to drop below 100°C. The contents
of the kettle are then heated to 113°C while stirring for 10 min.

After stirring the mass is cast straight into the shells or onto trays where it solidi-
fies, forming a layer 2 cm thick. The mass is then broken into pieces which are

274 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

loaded into the shells and the spaces between filled either with fresh material or TNT
or an alloy of TNT with cyclonite.

Since the mixture is highly hygroscopic it should be protected against atmos-
pheric moisture. During melting it is also necessary to avoid contact between the
mixture and brass or other copper alloys since the latter are liable to react with
metallic aluminium. Kettles should be washed with water every 24 hr to remove all
residues of the mixture.

MIXTURES WITH CHLORATES AND PERCHLORATES
Mixtures with potassium and sodium chlorates

As early as 1818 Berthollet [71] suggested the replacement of potassium nitrate
in blackpowder by potassium chlorate, which he had prepared for the first time in
1786. Thus, "white powder", which is a mixture of potassium chlorate with sugar,
was evolved. However, this mixture proved to be particularly sensitive to friction
and impact, hence its preparation and use was extremely dangerous. Further ex-
periments showed that pure potassium chlorate has no explosive properties but
when used in admixture with combustibles such as sugar, starch or aluminium it
gives rise to an explosive highly sensitive to mechanical action, especially to friction.
A disastrous explosion occurred in the potassium chlorate factory at St. Helens,
near Liverpool, in 1899. The storehouse in which barrels containing 150 tons of
potassium chlorate were kept was enveloped in flames. At first the contents of
the storehouse burned, but after a time an explosion took place. In all probability,
the explosion was caused by the fact that some of the potassium chlorate was melted
in the fire and the molten salt, together with the wood of the barrels, formed a
mixture which exploded under the influence of high temperature.

The manufacture and use of some mixtures containing potassium and sodium
chlorate were forbidden in several countries on account of their high sensitiveness
to friction. In Great Britain, for instance, the production of a mixture of potassium
chlorate and sulphur was prohibited by law in 1894. Mixtures of potassium chlorate
with phosphorus should also be avoided owing to the great dangers involved with
their preparation and handling. This is why the heads of safety matches are
made from mixtures of potassium chlorate with sulphur and with manganese dioxide
and potassium dichromate bound with glue.

The addition to potassium chlorate of vegetable oils, fats or mineral oils in which
aromatic nitro compounds are dissolved, as recommended by Street [72], proved
to be a milestone in the development of chlorate explosives. The application of an
admixture of castor oil was particularly useful. The presence of oils and fats in the
explosives reduced their sensitiveness to friction and impact, and the oily ingredient
conferred a slightly plastic consistency.

The effect of adding various organic substances to potassium chlorate on the
detonating capacity of the resultant mixtures, and on their power was studied system

HIGH EXPLOSIVES

275

-2

2 4 6 8 10 12 14
Paraffin content, %

4 o,-S

-5

- 7

§r

E 6

10g Picric acid

required
^Afo detonation

A? — *TTTT>

2 4 6 8 10 12 14
Vaseline content, %

1 &

2 §

4 Cn£

fl-S 3
o a, t>

7 t "
o

10g Picric acid

required
'•No detonation

Fig. 72. Explosive properties of the mixtures
of potassium chlorate with kerosene (according
to Chemisch-Technische Reichsanstalt [73]).

Fig. 73. Explosive properties of the mixtures
of potassium chlorate with vaseline (accord-
ing to Chemisch-Technische Reichsanstalt
[73]).

atically at the Chemisch-Technische Reichsanstalt [73]. The results that refer to
mixtures of potassium chlorate with kerosene, vaseline and nitrobenzene are shown
in Figs. 72, 73 and 74.

As was to be expected, mixtures with nitrobenzene are the most powerful and
the most readily detonated. Mixtures with kerosene or vaseline are of equal strength,

3
% a

2 4 6 8 10 12 14 16 18 20
Nitrobenzene content, %

10g Picric acid

required
"•No detonation

Fig. 74. Explosive properties of the mixtures of potassium chlorate with nitrobenzene
(according to Chemisch-Technische Reichsanstalt [74]).

those with a small kerosene content (e.g. 2-4%) detonate more easily than those
containing the same amount of vaseline while the mixtures which contain a larger
amount of kerosene or vaseline (e.g. 8-10%) detonate with more or less equal
ease.

The graphs for mixtures with paraffin oil lie close to those with kerosene while
mixtures with paraffin (m.p. 52°C) are similar to those with vaseline.

276 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Other studies of the effect of the structure of phlegmatizing substances on the
sensitiveness to impact of potassium chlorate mixtures were carried out by Blinov
[74]. He investigated mixtures consisting of 92% potassium chlorate and 8% liquid
phlegmatizing compound or of 85% potassium chlorate and 15% solid phlegma-
tizing compound. As far as phlegmatizing liquids are concerned it turned out that
the impact sensitiveness of the resultant mixtures is independent of the viscosity
of the liquid. Blinov proved that the phlegmatizing effect of such compounds in-
creases in proportion to the shortness of the carbon chain. This can be seen from the
examples given in Table 60. Moreover, increase in the plasticity of a solid substance
brings about a decrease in its sensitiveness ; this is shown by comparing the phlegma-
tizing effects of vaseline and paraffin.

Table 60
The phlegmatizing effect of various compounds

Phlegmatizing
compound

Sensitiveness

to impact,

Butyric acid
Valeric acid
Stearic acid
Oleic acid

50
45
20

Vaseline
Paraffin

55
26

Potassium chlorate can be replaced by sodium chlorate which is cheaper and
more widely available although being somewhat hygroscopic it is much less fre-
quently used as a constituent of explosives.

Chlorate explosive mixtures have the disadvantage that they cake and set solid
during storage. Some ingredients such as vaseline or paraffin tend to aggravate
this trouble. To counteract this tendency ingredients are added which have a loosen-
ing effect on the explosive composition, giving a relatively low density (e.g. wood
meal or cork dust). In the U.S.S.R. extensive work on this subject was carried
out by Shpitalskii and Krause [75].

The most effective method of preventing caking in chlorate explosives is to
manufacture them in granular form. For this purpose various resins (e.g. colophony)
or waxes (e.g. Carnauba) are added to the mixture. The moist mass is then rubbed
through a sieve and dried. The grains so formed are sifted through screens to separate
out the dust.

Explosives of this type were manufactured at Chedde in France, whence they
derived their name of Cheddite. They were recommended originally for use in
mines, but were gradually withdrawn as unsafe to use in the presence of methane
and coal-dust.

For military purposes Cheddites were usually produced with the following
composition given in Table 61.

high explosives

Table 61
The composition of Cheddites

277

Name

Constituents

Explosif O

No. 1

Type 41

Explosif O

No. 1

Type 60

bis

Explosif O

No. 2

modifie

TypeO
No. 6B

Explosif
S

Potassium chlorate

Sodium chlorate

Dinitrotoluene

Nitronaphthalene

Castor oil

Vaseline

Paraffin

2
13

15
1

3
7

The first three types of Cheddites were used in France for filling hand grenades
and shells with a low muzzle velocity and for manufacturing demolition charges
during World War I. When there was a shortage of nitro compounds in France,
Cheddite type O No. 6B was employed as a substitute for the first two. This was
also manufactured with sodium chlorate instead of potassium chlorate (Explosif S).
Kast [3] gives the following figures as characteristic of Cheddite type O No. 2,
modifie :

Apparent density 1.15

Heat of explosion 1185 kcal/kg

Gas volume (V„) 337 l./kg

Temperature of explosion ca. 4500°C

Specific pressure (f) 6090 m

Rate of detonation (at a density of 1.3) 3000 m/sec
(at a density of 1.5) 4000 m/sec
Lead block expansion 255 cm 3

Sensitiveness to impact (2 kg) 30 cm

The rate of detonation of Cheddite type O No. 6B appears to differ slightly
from that of the above composition although it contains no TNT; at a density of
1.4 it is 3500 m/sec.

The, chlorate explosives can be easily compressed to a density of 1.9 although
at such a high density they are more difficult to detonate (see also Fig. 75, p. 279).

Chlorate explosives yield only a small amount of gaseous products since the
major product of explosion is potassium chloride. The specific pressure /is therefore
relatively low (the high temperatures do not compensate for the small volume of
gases), hence the lead block expansion is low, but Cheddites without nitro com-
pounds—type O No. 6 and S — give an even smaller lead block expansion : 1 80-200 cm 3 .
Some chlorate explosives, when detonated in the open, do not transmit detonation
from cartridge to cartridge, differing in this respect from dynamites and ammonium
nitrate explosives. In a confined space, however, they behave differently. Here the

278 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

transmission of detonation over a distance is superior to that of ammonium nitrate
explosives. T. Urbanski [76] has reported that cartridges of Miedziankit (100 g,
30 mm dia.) can transmit detonation in the bore hole of the mortar of an experimen-
tal gallery for a distance of over 40 cm.

The main disadvantages of chlorate explosives such as sensitiveness to friction
and impact and caking in storage, were overcome by Sprengel [77], who introduced the
technique of mixing the ingredients immediately before use. To facilitate mixing, the
combustible component of the explosive had a liquid consistency. This led to the
development of the explosive "Rack-a-rock", consisting of potassium chlorate and
nitrobenzene. It played an important part in many engineering achievements at
the turn of the nineteenth century especially, in Russia and in the U.S.A.

A considerable advance was made by Laszczynski [78], when he worked out
the composition and method of preparation of the explosive Miedziankit consisting of :

1 90% of potassium chlorate it
\W/ of kerosene |

Cartridges containing only potassium chlorate were transported in safety to
the site, where they were dipped for a definite time into kerosene just before
use. Miedziankit was also manufactured by soaking potassium chlorate cartridges
with kerosene in the explosive factory. Kerosene with an ignition temperature
above 30°C was employed, to render the product safe for rail transport. Ac-
cording to T. Urbanski [76] the rate of detonation of Miedziankit is 3000 m/sec
in an iron pipe at a density of 1.7.

Miedziankit was one of the most widely used non-military explosives in Germany,
Russia and Poland before, during and immediately after World War I, when there
was a need to economize in the nitrates, including ammonium nitrate, used for
military purposes.

It is a feature of chlorate explosives that ammonium salts (e.g. ammonium
nitrate) must not be added to chlorate compositions (Vol. II).

The converse obviously applies to ammonium nitrate explosives, which must
not contain any chlorates, since during storage a double exchange reaction may
occur resulting in the formation of ammonium chlorate (p. 476, Vol. II), an un-
stable substance which decomposes spontaneously. A number of patents were taken
out between 1880 and 1895, for explosives based on the use of ammonium chlorate
or mixtures of ammonium nitrate with potassium or sodium chlorate. Many acci-
dents which occurred through the spontaneous decomposition of these explosives
proved the impossibility of using mixtures containing both chlorates and ammonium
salts (Hantke [79]).

MIXTURES WITH POTASSIUM AND AMMONIUM PERCHLORATES

Potassium perchlorate in the pure state, like potassium chlorate, is not an explo-
sive, indeed the decomposition of the former is endothermic:

KCIO4 = KQ + 20 2 - 7.8 kcal

HIGH EXPLOSIVES

279

When mixed with non-explosive combustibles, potassium perchlorate produces
compositions relatively difficult to detonate; e.g. the mixture of potassium perchlor-
ate with paraffin in a ratio of 85 : 15 gives a lead block expansion of only 60 cm*,
whereas a similar mixture with potassium chlorate gives approximately 200 cm 3 !
On the other hand, the higher content of oxygen in potassium perchlorate and the

8000

2.00

^6000

- & 1.50

•S

?»

%4000

- 1 1.00

■S

■«

•S

-to
c:

§2000

- %0.50

Density

Fig. 75. The relation between the density and the rate of detonation, and sensitiveness
to initiation by mercury fulminate of chlorate explosive [80].

smaller amount of potassium chloride produced during decomposition makes it
possible to form mixtures which are 10-15% stronger than chlorate mixtures. The
sensitiveness of perchlorate mixtures to friction and impact is lower than that of
chlorate mixtures, and their ignition temperature is higher.

In spite of their advantages, mixtures with potassium perchlorate as the chief
constituent are not very often used since the latter is too expensive. Other disadvan-
tages of such mixtures lie in their rather high sensitiveness to mechanical impulses,
the relatively great difficulty of detonating them, and their fairly high ability for
deflagration. Potassium perchlorate is therefore often employed simply as an addi-
tive to ammonium nitrate explosives (p. 264).

Another constituent of perchlorate explosives, ammonium perchlorate, unlike
ammonium chlorate, is stable. It is also dissimilar to potassium perchlorate in
being an explosive in the pure state, like ammonium nitrate. The greater specific
gravity of ammonium perchlorate gives to explosives with which it is mixed a greater
power than that of similar ammonium nitrate explosives. The former are also more
sensitive than chlorate explosives to friction and impact and to thermal ignition.

280

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

On explosive decomposition they may produce a certain amount of hydrogen chlor-
ide which is highly undesirable in mining explosives.

Perchlorate explosives, like chlorate explosives, can be compressed to a high
density, but the detonation at a high density is difficult. This is illustrated by a
graph published by the French Commission on Explosives (Commission des Sub-
stances Explosives) (Fig. 75) [80] which shows how the rate of detonation and the
amount of mercury fulminate required for detonation varies with density.

As early as 1865 Nisser [81] recommended the use of potassium perchlorate
instead of potassium chlorate. Ammonium perchlorate as a constituent of explosives
was proposed by Nobel [82] in 1888, and explosives with ammonium perchlorate
as a chief component were proposed by Alvisi [83] in 1895 and Carlson [84] in 1897.
Explosives containing ammonium perchlorate are used in Japan in the mining
industry (p. 474). Yonckites, developed by Yonck [85] (p. 447) were extensively
used in the Belgian mining industry. Cheddites with ammonium perchlorate instead
of potassium chlorate were manufactured in France.

Perchlorate Cheddites were employed in France and Italy for filling shells with a
low muzzle velocity during World War I (trench mortar shells, aerial bombs etc.).
Originally these were the explosives which imitated Cheddites containing DNT
(types B and C). Later, chiefly owing to a shortage of nitro compounds, paraffin
was utilized as a combustible ingredient. Thus emerged the explosives 86/14, 90/10
and E (Table 62).

Table 62
Composition of Cheddites

Constituents

TypeB

TypeC

TypeP

or 86/14

90/10

Ammonium perchlorate

61.5

Sodium nitrate

DNT

Castor oil

Paraffin

8.5

Perchlorate explosives for mining, and technical methods of manufacturing
chlorate explosives, will be discussed later (pp. 520-521). Mixtures of potassium
perchlorate or ammonium perchlorate with plastics or elastomers have re-
cently come into extensive use for jet propulsion (e.g. methyl polymethacrylate,
ester resins, and thiokol-rubber).

In this connexion a number of investigations have been undertaken to examine
the physico-chemical and explosive properties of these mixtures.

Gordon and Campbell [86], for instance, examined the exothermic decomposi-
tion of potassium perchlorate mixtures with carbon within the temperature range
300-360°C, while Grodzinski [87] studied the thermal decomposition of mixtures
of various combustibles with potassium perchlorate.

High explosives 281

PLASTIC EXPLOSIVES

Plastic explosives, such as dynamites, are explosives rich in a liquid con-
stituent, e.g. nitroglycerine, usually with dissolved high viscosity polymers. Guhr
dynamite (no more in use) composed of 75% nitroglycerine and 25% kieselguhr,
the first explosive to have a plastic consistency, owed this property to the high
proportion of liquid it contained.

Blasting gelatine is a plastic explosive consisting of 92-94% nitroglycerine and
6-8% collodion cotton (of high viscosity). Blasting gelatine is markedly elastic.
It loses its elasticity, becoming plastic, only on heating to a temperature of 40°C or
higher, depending on the type and amount of collodion cotton present. Recently
a new kind of dynamite was developed in the U.S.S.R. Its plasticity was achieved
by dissolving methyl polymethacrylate in nitroglycerine, in the proportions of
twenty to forty parts of nitroglycerine to one part of the polymer.

Blasting gelatine and dynamites are now used only for civil purposes. Originally,
they were suggested as military explosives, particularly for filling shells with a low
muzzle velocity, until it was shown that fillings made with nitroglycerine explosives
may cause premature explosions inside the barrel.

The Russian air force used bombs filled with a material resembling blasting
gelatine during World War I.

Dynamites were retained for military purposes for some time for use in
demolition charges. The disadvantage of these explosives lies in their limited chem-
ical stability. Ultimately, therefore they were replaced by explosives that remain un-
changed during storage (aromatic nitro compounds such as TNT, picric acid, and
more recently TNT with cyclonite or PETN).

The plasticity in an explosive can be put to practical use, e.g. for dem-
olition purposes. Thus, to sever an iron bar or to blow up a wall or a rock the
easiest way is to use a plastic explosive, moulded to fit the shape of the object to be
destroyed. Since blasting gelatine is not entirely safe to handle, being elastic, i.e.
difficult to shape as required, attention was directed to the development of plastic
materials based on other constituents. Thus, mixtures comprising a solution of
collodion cotton in liquid aromatic compounds as plasticizers were suggested. Such
were Plastrotyl, recommended by Bichel [88] with a composition:

86% of TNT

10% of liquid DNT (m.p. 20-25°C)
0.3% of collodion cotton
3.7% of turpentine

and the Swedish perchlorate plastic explosive Territ suggested by Nauckhoff [89] :

43% of ammonium perchlorate
28% of sodium nitrate
27.8% of TNT and DNT
1.2% of collodion cotton

282 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

According to Kast [3] the explosive properties of Territ are :

Density 1.67 (max. 2.15)
Rate of detonation 4700 m/sec
Lead block expansion 340 cm 3
Sensitiveness to impact (2 kg) 20 cm

Plastrotyl and Territ were rather difficult to detonate chiefly due to their great
density. To make these mixtures detonate more readily, nitroglycerine was added
but this increased their sensitiveness to impact.

In 1929, Stettbacher [90] suggested the use of a mixture of PETN with nitro-
glycerine, under the name of Penthrinite. Such a mixture may be plastic provided
suitable amounts of nitroglycerine and PETN with crystals of a proper size are
used and both constituents are thoroughly mixed. Stettbacher recommended a
composition ranging with the limits :

10-70% of PETN
90-30% of nitroglycerine

If a mixture rich in nitroglycerine is employed it may be converted into a plastic
one by the addition of collodion cotton.

In his later work Stettbacher developed the idea of using penthrinites as substi-
tutes for dynamite, with the following composition :

50% of PETN
46% of nitroglycerine
4% of collodion cotton

Stettbacher's proposals were subjected to sharp criticism by Naoum [91] who con-
sidered the substitution of such mixtures for dynamite to be inexpedient, since :

(a) they are considerably more expensive than nitroglycerine and,

(b) they cannot replace dynamite as mining explosives.

The Chemisch-Technische Reichsanstalt [73] examined the properties of PETN
mixtures with nitroglycerine and found that the latter easily exudes from mixtures
in which more than 20% is present, especially at an elevated temperature, e.g.
50°C. However, Stettbacher's observation that penthrinite can be compressed to
high density more easily than PETN itself was confirmed. Hand tamping of PETN
gives a density of 0.9, whereas a mixture of 80% PETN with 20% nitroglycerine
gives 1.3. The following densities of these mixtures were obtained under higher
pressures :

Pressure, kg/cm 2 Density

250 1.60

1000 1.62

2000 1.66

3000 1.67

The rate of detonation of a mixture compressed to a density of 1.67 is 7600 m/sec
whereas PETN of equal density detonates with a rate of 8400 m/sec.

HIGH EXPLOSIVES 283

Plastic explosives with cyclonite as the chief constituent were used extensively
during World War II. Cyclonite is preferable as it detonates easily even when strong-
ly phlegmatized (desensitized to impact) and as its rate of detonation is very high.
Originally a mixture of 88% cyclonite with 12% lubricating oil was employed.
This mass however was not sufficiently plastic and lost its plasticity readily by ex-
uding the oil at a high temperature. It was therefore replaced by a mixture named
Composition C-3 of 77% cyclonite and 23% gel made out of liquid nitro com-
pounds (e.g. liquid DNT) and nitrocellulose or of butyl phthalate and nitrocellulose.
The composition C-3 was later improved and designated Composition C-4.
It contains :

91.0% of cyclonite
2.1% of polyisobutylene
1.6% of motor oil
5.3% of di-(2-ethylhexyl)sebacate

It is less volatile than C-3 and has less tendency to harden at low temperature.
It has a density of 1.59, does not become hard even at -55°C (-70°F), and does
not exude at +77°C (170°F).

For some types of explosive working of metals, plastic sheet explosives EL-506
were developed by E. I. du Pont de Nemours and Co., Inc.

One of the representatives of this group of explosives, EL-506 A, consists of
PETN combined with plasticizers to form flexible sheets of 10x20 in., fabricated
in a number of thicknesses [68].

The German plastic material Hexoplast 75 was composed of:

75% of cyclonite
3.6-3.8% of TNT

20% of DNT (liquid)
1.2-1.4% of nitrocellulose

The freezing point of a TNT and DNT mixture was - 20°C. The ingredients were
mixed in a Werner-Pfleiderer kneader at 90°C, the cyclonite with the nitrocellulose
being added first and nitro compounds added after thorough mixing. This prevented
the formation of lumps of swollen nitrocellulose.

INCOMPATIBILITY IN EXPLOSIVE MIXTURES

It is well known that some ingredients of explosive mixtures should not be
brought together, as their mutual reaction produces undesirable changes of the pro-
perties of the explosive mixture.

Although a considerable amount of work was done on the compatibility of
various components of explosive mixtures no systematic study of this particular
problem was made until in 1938 T. Urbanski [6] began a series of investigations
which comprised examination of mixtures :

284

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(1) By thermal analysis [38, 40].

(2) By determining their chemical stability and ignition temperature, e.g. the
stability of nitroglycerine- or PETN-, or nitrocellulose-nitro compound mixtures
(Vol. IT, pp. 49, 181 and Vol. Ill, p. 566, respectively) [7].

The influence of various non-explosive substances, such as sulphur, was also
examined. It was found that the latter considerably reduces the temperature of

Fig. 76. The influence of sulphur on ignition temperature of nitro compounds and
cyclonite (according to T. Urbanski and Pillich [92]).

ignition of TNT and other high-nitrated aromatic compounds (Vol. I, p. 305)—
Fig. 76 [92].

(3) By determining the sensitiveness of mixtures to impact and friction. Foreign
crystals were found to increase the sensitiveness of nitro compounds when present
in relatively small quantities (e.g. a few percent) [5, 92] — Figs. 66, 68 and 70 (pp.
250, 251 and 262, respectively). On the contrary an addition of soft crystals of wax-
like substances is well known to desensitize explosives.

(4) By determining the explosive properties in mixtures [6]; this is discussed
on pp. 250-252.

The problem of incompatibility is discussed in various places in this book (e.g.
incompatibility of chlorates with ammonium salts, p. 476, Vol. II and p. 278, Vol.
III).

The action on TNT of various substances, which may occur in the explosive
(e.g. ferric oxide) is described on pp. 304-305, Vol. I.

HIGH EXPLOSIVES 285

The action of some inorganic substances on unsymmetrical isomers of trinitro-
toluene is mentioned on p. 331, Vol. I.

Another important practical problem is the action of tetranitromethane on
TNT (p. 339, Vol. I).

One of the important problems of compatibility in explosive mixtures is whether
ammonium nitrate can react with nitro compounds, such as TNT. This was discussed
by Lang and Boileau [64]. These authors concluded that no reaction can occur
between TNT and ammonium nitrate when they are pure. The evolution of ammonia
from ammonium nitrate on storage at room temperature does not suffice to produce
any reaction with TNT.

However any alkaline reaction which can be developed by some impurities may
produce an evolution of ammonia sufficient to form dark-coloured, readily ignitable
products.

Recently Rogers [93] pointed out that two types of incompatibility should be
distinguished, otherwise confusion may result. Incompatibility of the first type may
be caused by secondary chemical reactions or mobility of residual solvents, gases,
or plasticizers, leading to unexpected modifications of mechanical, physical, or
electrical properties.

Incompatibility of the second type appears as an unexpected increase in sensi-
tivenes or decrease in thermal stability, and may be caused by any of the foregoing
phenomena. Rogers found that zinc reacts readily with ammonium nitrate.
He also found that adding 20% urea to RDX reduces the thermal stability of the
latter.

LITERATURE

1. Chemische Fabrik Griesheim, Ger. Pat. 69897 (1892).

2. H. Maxim, U.S. Pat. 792716 (1905).

3. H. Kast, Spreng- u. Ziindstoffe, Vieweg & Sohn, Braunschweig, 1921.

4. Ya. I. Leitman, according to A. G. Gorst, Porokha i vzryvchatyye veshchestva, Oborongiz, Mos-
kva, 1949.

5. T. URBAflSKI, Przemysl chem. 20, 117, 179 (1936); Z. ges. Schiess- u. Sprengstoffw. 33, 41, 62
(1938).

6. T. URBAttSKI, Przemysl chem. 22, 521 (1938); Mem. artill. franc. 20, 237 (1946).

7. T. Urbanski, B. Kwiatkowski and W. Miladowski, Przemysl chem. 19, 225 (1935); Z. ges.
Schiess- u. Sprengstoffw. 32, 1, 57, 85 (1937).

8. T. Urbanski, Roczniki Chem. 17, 588 (1937).

9. P. Laffitte and A. Parisot, Compt. rend. 203, 1516 (1936).

10. W. O. Williamson, Research 11, 387 (1958).

11. A. W. Koch and H. Freiwald, Explosivstoffe 6, 279 (1958).

12. L. Andrussow, Explosivstoffe 7, 89 (1959).

13. A. W. Koch and H. Freiwald, Explosivstoffe 9, 221, 247 (1961).

14. C. Girard, Fr. Pat. -350371 (1904).

15. Dynamit A.G., Ger. Pat. 568000 (1932); Brit. Pat. 742312 (1932).

16. N. C. Boyd, Hercules Powder Co., U.S. Pat. 1840431, 1877520 (1932).

286 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

17. C. Manueli and Bernardini, Brit. Pat. 138371 (1920).

18. T. Urbanski and J. Skrzynecki, Roczniki Chem. 16, 353 (1936).

19. A. LE ROUX, Mem.poudres 35, 121 (1953).

20. L. MEDARD, Mem. poudres 36, 87 (1954).

21. R. KREMANN, Monatsh. 25, 1215, 1271 (1904); 26, 143 (1905); 27, 627 (1906).

22. R. Kremann et al., Monatsh. 27, 125 (1906); 29, 863 (1908); 32, 609 (1911); 37, 723 (1916);
38, 385 (1917); 39, 505 (1918); 40, 189 (1919); 41, 499 (1920); 42, 167, 181, 199 (1921); 44, 163
(1923).

23. M. GlUA, Ber. 47, 1718 (1914); Gazz. chim. ital. 45, I, 339, 557; II, 32, 348 (1915); 46, II, 272
(1916); 47, 1, 57 (1917); Lezioni di Esplosivi, Torino, 1932.

24. P. PASCAL, Bull. soc. chim. France [4], 27, 388 (1920).

25. D. H. ANDREWS, /. Phys. Chem. 29, 1041 (1925).

26. O. Wyler, Helv. Chim. Acta 15, 23 (1932).

27. N. N. Yefremov and A. M. Tikhomirova, Izv. Inst. Fiz.-Khim. Analiza 3, 269 (1926); 4, 65,
92 (1928).

28. N. N. Yefremov and O. K. Khaibashev, Izv. Sektora Fiz.-Khim. Analiza Akad. Nauk 11, 130
(1949).

29. A. Wogrinz and P. VAri, Z. ges. Schiess- u. Sprengstoffw. 14, 249 (1919).

30. C. A. Taylor and W. H. Rinkenbach, Ind. Eng. Chem. 15, 73, 795, 1070 (1923).

31. JoviNET, Mem.poudres 23, 36 (1928).

32. K. Hrynakowski and Z. Kapuscinski, Roczniki Chem. 14, 115 (1934).

33. T. Urbanski and B. Kwiatkowski, Roczniki Chem. 14, 941 (1934).

34. V. Tamburrini, Ann. chim. appl. 17, 275 (1927).

35. M. KURITA and J. HAGUI, /. Chem. Soc. Japan 50, 155 (1929).

36. L. Medard, Mem, poudres 24, 174 (1930-1931).

37. J. Hackel, Roczniki Chem. 16, 266, 323 (1936).

38. T. Urbanski, Roczniki Chem. 13, 399 (1933); 14, 925 (1934); 15, 191 (1935).

39. G. Desseigne, Mem. poudres 30, 127 (1948).

40. T. URBANSKI and I. R abek-Gawronska, Roczniki Chem. 14, 239 (1934).

41. O. K. Khaibashev and O. F. Bogush, Nauchno-issl. rabotyKhim. Inst. Akad. Nauk 1941-1943,
78 (1945).

42. T. Urbanski and S. Kolodziejczyk, Roczniki Chem. 16, 119 (1936).

43. Bombrini-Parodi Delfino, Ger. Pat. 651780 (1937).

44. M. J. Urizar, E. D Loughran and L. C. Smith, Explosivstoffe 10, 55 (1962).

45. I. M. Cheltsov, according to A. G. GORST, Porokha i vzryvchatyye veshchestva, Oborongiz,
Moskva, 1949.

46. C. J. Ohlsson and J. H. Norrbin, Swedish Pat. of 31 May 1867.

47. A. Parisot, Theses, Nancy, 1939.

48. T. Urbanski, J. Hackel and B. Kwiatkowski, Przeglqd Techniczny (1938).

49. J. Hackel, Wiad. Techn. Uzbr. 38, 519 (1937).

50. W. M. Evans, Proc. Roy. Soc. (London) 204 A, 14 (1950).

51. J. L. Copp and A. R. Ubbelohde, Trans. Faraday Soc. 44, 464 (1948).

52. J. Hackel, unpublished work (1946).

53. Macar, Belgian Pat. 142506 (1899).

54. Deissler, Belgian Pat. (1897).

55. Goldschmidt, /. Soc. Chem. Ind. (London) 543 (1898).

56. R. Escales, Brit. Pat. 24377 (1899); Belgian Pat. 145599 (1899).

57. von Dahmen, according to [63].

58. G. Roth, Ger. Pat. 172327 (1900).

59. H. Kast, Jahresber. Mil. Vers. Amt. 8, 48 (1902); 9, 18 (1903); 10, 4 (1905); 13, 17 (1908).

60. H. Muraour, Protar (Solothurri) 9, 62 (1943).

HIGH EXPLOSIVES

287

61. According to A. Stettbacher, Protar (Solothurn) 9, 212, 233 (1943).

62. E. Chaylan, Mem. poudres 30, 139 (1948).

63. B. T. Fedoroff et al., Encyclopedia of Explosives and Related Items, Vol I Picatinnv Arsenal
Dover, U.S.A., 1960. '

64. F. M. Lang and J. Boileau, Mem. poudres 34, 181 (1952).

65. P. Aubertein, Fr. Pat. 1180530 (1957-1959).

66. R. FORG, Ammonal, Otto Klemm, Wien-Leipzig, 1917.

67. H. Koenen and P. Dittmar, Explosivstoffe 11, 54 (1963).

68. According to J. S. Rinehart and J. Pearson, Explosive Working of Metals, Pergamon Press
Oxford, 1963. '

69. The Dow Chemical Company, Dow MS-80, Midland, Michigan.

70. R. Sartorius, Mem. poudres 34, 205 (1952).

71. C. Berthollet, Ann. Mm. 9, 22 (1818).

72. Street, Ger.Pat. 100522, 100523 (1897); 117051, 118102 (1898); Brit Pat 9970 13724 (1897V
12760, 12761, 24468 (1898); Fr. Pat. 267407 (1897). '

15. Jahresber. Chem.-Techn. Reichsanstalt 8, 115 (1929).

74. I. F. Blinov, Zh. prikl. khim. 8, 52 (1935).

75. E. Shpitalskii and E. Krause, Z. ges. Schiess- u. Sprengstoffw. 20, 103 (1925).

76. T. Urbanski, O/Schlesischer Berg- u. Hiittenman. Vereins Z. 65, 217 (1926).

77. Sprengel, Brit. Pat. 921, 2424, 2642 (1871); /. Chem. Soc. 26, 796 (1873).

78. Laszczynski, Ger. Pat. 215202 (1909); 237225 (1910).

79. Hantke, Angew. Chem. 50, 473 (1937).

80. H. Dautriche, Mem. poudres 14, 206 (1906-1907).

81. Nisser, Brit. Pat. 1934 (1865); 1375 (1868).

82. A. Nobel, Brit. Pat. 1471 (1880).

83. U. Alvtsi, Brit. Pat. 9190, 25838 (1898); Gazz. chim. ital. 29, 1, 121, 399; II 64 478 (1899V
31, 1, 221 (1901). '

84. O. F. Carlson, Brit. Pat. 10362 (1897); Swedish Pat. 8487 (1897).

85. Yonck, Belgian Pat. 143499, 143656 (1899); Brit. Pat. 24511 (1903).

86. S. Gordon and C. Campbell, Vth Symposium on Combustion,?. Ill, Reinhold, New York, 1955.

87. J. Grodzinski, /. Appl. Chem. 8, 523 (1958).

88. C. E. Bichel, Ger. Pat. 193213 (1906).

89. S. Nauckhoff, Swedish Pat. 30408 (1909).

90. A. Stettbacher, Z. ges. Schiess- u. Sprengstoffw. 24, 229 (1929).

91. Ph. Naovjm, Z. ges. Schiess- u. Sprengstoffw. 25, 376, 442 (1930).

92. T. Urbanski and J. Pillich, Wiad. Techn. Uzbr. 43, 79 (1939).

93. R. N. ROGERS, Ind. Eng. Chem., Products Res. and Dev. 1, 169 (1962).

CHAPTER II

LIQUID EXPLOSIVES

HISTORICAL

The use of explosives obtained from two non-explosive constituents, at least one
of which is in the liquid phase, was suggested by Sprengel [1] in 1871. The consti-
tuents were mixed just before the explosive was used to avoid the dangers of transport
and handling. Nitric acid was one of the liquid constituents, together with liquid or
solid aromatic nitro compounds. Such explosives were not, however, successful
apart from the Sprengel mixtures in which the oxidizing agent, e. g. potassium chlorate,
is a solid constituent; these substances are reviewed on p. 278.

The explosives proposed by Turpin [2], under the name of Panclastites, in which
nitrogen dioxide was the oxidizing agent, were more successful.

After the liquefaction of air had been achieved by Olszewski and Wroblewski
[3] and on a large scale by Linde [4] the use of liquid oxygen as an oxidizing agent
became possible in composite explosives called Oxyliquits (Linde [5]).

Liquid explosives came into extensive use during World War I when nitro com-
pounds and ammonium nitrate became scarce: panclastites were most commonly
used for military purposes and oxyliquits in the mining industry. During the World
War II the Germans employed liquid mixtures for jet propulsion including a new-
comer in this field— a mixture of concentrated (80-85%) hydrogen peroxide with
hydrazine for the propulsion of V2 rockets.

MIXTURES WITH NITROGEN DIOXIDE, NITRIC ACID AND
TETRANITROMETHANE

Nitrogen dioxide was used in mixtures with such combustibles as paraffin (with-
out aromatic compounds), carbon disulphide or nitrobenzene. These substances
were used in the proportions necessary to give complete decomposition into C0 2 ,
H 2 and N 2 , thus permitting full utilization of the oxygen present in the nitrogen
dioxide. To prevent the solidification of nitrobenzene at low temperatures a binary
combustible constituent, e.g. a mixture of nitrobenzene with carbon disulphide,
was used.

. #- CMS] ^

LIQUID EXPLOSIVES

289

Kast and Giinther [6] examined the explosive properties of these mixtures and
found them to be similar to nitroglycerine. They also had its advantages (great
power) and disadvantages (high sensitiveness to mechanical stimulants). They differ
from nitroglycerine in their lower specific gravity, which obviously contributes to

Fig. 77. Aerial bomb filled with liquid nitrogen dioxide (lower compartment) and
combustible liquid, e.g. petrol (according to Pascal [7]).

a somewhat lower brisance. A mixture of nitrogen dioxide with 21% by weight
of petrol (35% by volume) gives a rate of detonation of 7100 m/sec. A mixture
with 35.5% by weight of nitrobenzene (40% by volume) detonates with a rate
of 7650 m/sec.

During World War I the French used aerial bombs constructed in such a fashion
that the two constituents of the explosive filling were mixed after the bomb had
been released. The risk of handling and transporting such a dangerous explosive
was thus avoided (Fig. 77). The shell of this bomb was divided into two compart-
ments by a thin partition wall. The lower compartment, fitted with a percussion
fuse (7) and detonator (2) was filled with liquid nitrogen dioxide. A spring hammer (3)
held back by a hook was located in the upper compartment of the bomb which
was filled with petrol immediately before the aircraft had started. On release of the
bomb the hammer was unhooked and pierced the partition wall, and at the same
time the bomb turned upside down so that the compartment fixed with vanes was
uppermost and its contents— nitrogen dioxide (m.p.-10.2°C; b.p. + 22°C; density
at 0°C 1.4903)— the heavier constituent easily flowed down to mix with the other
constituent, i.e. petrol.

290 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The explosives recommended by Hellhoff [8] are also noteworthy. They consist
of concentrated nitric acid and dinitrobenzene or a mixture of nitrobenzene with
dinitrobenzene. They have not been used very extensively for practical purposes,
because their nitric acid content makes them extremely corrosive, but under war-time
conditions they were considered to be the cheapest explosives and the simplest to
prepare. During World War II they were suggested in Great Britain under the
name of Dithekite. According to Cook, Pack and Gay [9] Dithekite-13 or D-13
has the following composition :

24.4% of nitrobenzene

62.6% of nitric acid
13% of water

In liquid mixtures tetranitromethane may be used as an oxidizing agent. A fully
oxygen-balanced mixture consisting of 86.5% tetranitromethane and 13.5% toluene
has a density of 1.45. It is a powerful explosive (p. 591, Vol. I).

According to Medard and Sartorius [10] the solution of 33.5% dinitrotoluene,
50.0% nitric acid and 16.5% water in a glass tube 30 mm dia. can detonate at the
rate of 6700 m/sec. The solution of 47.0% TNT, 50% nitric acid and 3.0% water in
an aluminium tube 40/50 mm dia. detonated at the rate of 7500 m/sec.

Andrussow [11] described Nisalit, a mixture of 79.5% nitric acid (99%) with
20.5% acetonitrile. It develops the heat of detonation of 1670 kcal/1., gas volume
V = 708.0 l./kg and calculated explosion temperature is 4200°K. Its rate of detona-
tion (density 1.27 at 15°C) is 6250 m/sec and its lead block expansion 450 cm*.

The author suggested another mixture— Disalit:

22% of dimethyl ethe-
77% of HN0 3
1 % of water

He also suggested similar mixtures containing perchloric instead of nitric acid:
Niperchlorit and Diperchlorit, respectively.

MIXTURES WITH HYDROGEN PEROXIDE

The Germans used hydrogen peroxide of 80-85% concentration, alone or in
mixtures with combustibles, as a fuel for the big V2 rockets during World War II.
The utilization of hydrogen peroxide for rocket propulsion and the explosive proper-
ties of hydrogen peroxide and its mixtures will be discussed in later sections (pp.
299, 307).

MIXTURES WITH LIQUID OXYGEN (OXYLIQUITS)

These substances are reviewed together with mining explosives. Their use for
rocket propulsion will be considered in the section on p. 309.

LIQUID EXPLOSIVES 291

LIQUID ROCKET PROPELLANTS-PROPERGOLS

♦'

The liquid explosives of the type outlined above may serve not only as high
explosives but also as propellants for rocket propulsion. Liquids which are not
explosives in the strict sense, but which undergo violent decomposition under certain
conditions, with heat emission and gas evolution, may also be used for this purpose.
The liquids employed for rocket propulsion are called propergols.

The Germans were the first to use them during World War II. The use of con-
centrated (80-85%) hydrogen peroxide alone, or to a lesser extent, in a mixture
with such oxidizing agents as nitric acid, nitrogen dioxide, tetranitromethane, or
liquid oxygen, was an innovation.

According to the classification suggested by R. Levy [12] the following types
of propergols may be distinguished :

(1) Catergols, i.e. liquids which are decomposable by the action of catalysts,
e.g. hydrogen peroxide, decomposed by permanganates.

(2) Hypergols, i.e. systems composed of several (two, at least) liquids which
when mixed react spontaneously, usually after a certain induction period (e.g.
a mixture of petrol with an admixture of aromatic amines which reacts spontaneously
with nitric acid). For practical purposes the induction period of hypergols should
be as short as possible— at any rate shorter than 0.1 sec.

Propergols may also be classified according to their homogeneity:

(1) Monergols, i.e. monophase systems composed of at least two components,
one of which is an oxidant, the other a fuel, e.g. a solution of methyl nitrate (oxidant)
with methyl alcohol (combustible).

(2) Lithergols, i.e. polyphase, at least biphase systems, one of the phases being
liquid, another solid, e.g. carbon and liquid oxygen.

MIXTURES WITH NITROGEN DIOXIDE

Nitrogen dioxide can be used together with a combustible substance as a liquid
propellant (propergol) for rockets. A mixture of hydrocarbons e. g. petrol or paraffin,
may serve as the fuel.

On explosive decomposition the stoichiometric mixture of N 2 4 with paraffin
gives a considerable amount of heat: approximately 1560 kcal/kg.

Since the course of interaction of nitrogen dioxide and paraffin may often be
too slow, to facilitate and speed it up a substance that readily reacts with nitrogen
dioxide, e.g. aniline, should be added to the paraffin.

Nitrogen dioxide is noteworthy as an oxidant rich in oxygen (it contains 69.5%
by weight of oxygen and 1.01 kg of oxygen per 1 1. of substance). However, the
physico-chemical properties of nitrogen dioxide, such as its relatively high freezing
point (— 10.2°C) and a low boiling point ( + 22°C) limit its direct use as an oxidant.

292 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

On the other hand, nitrogen dioxide has proved to be a valuable component
of propergolic mixtures in which nitric acid is an essential oxidant. Nitric acid
containing approximately 20% N 2 4 is a particularly valuable oxidant as explained
below.

MIXTURES WITH NITRIC ACID

Mixtures of concentrated nitric acid with combustible substances have recently
found wide application as liquid propellants for rockets (propergols).

The first attempts to use them were made by the Germans during World
War II.

Nitric acid has many advantages as a component of these mixtures, being readily
available from large-scale manufacture.

The physico-chemical properties of the chemically pure substance are described
in Vol. I, p. 6. v

The high specific gravity of nitric acid, its high oxygen content (76% by weight),
low freezing and negative heat of formation all are very advantageous for its use as
an oxidant in propergols. Its disadvantage lies in its corrosive action.

For practical purposes commercial nitric acid, containing 2-4% of water, is
employed. Its freezing point is lower than that of the pure form (a 10% content of
water in HN0 3 lowers the freezing point to-68.5°C; a higher content of water
lowers this temperature to a lesser extent). Propergols containing nitric acid belong
to the hypergol group, i.e. the mixtures which react spontaneously.

Since the reaction of nitric acid alone with a fuel such as petrol or paraffin occurs
fairly slowly with a long induction period (longer than 0.1 sec) various substances
are added to the nitric acid or the hydrocarbon component to speed up the reaction
or induce its spontaneous initiation.

Since nitric acid, especially "red" fuming nitric acid "RFNA" which contains
a small amount of nitrogen oxides, reacts vigorously with aromatic amines, during
World War II the Germans employed solutions of these amines (e.g. aniline or
phenylenediamine) in benzene or xylene as the combustible component. They added
a small amount of ferric chloride as a reaction catalyst to the nitric acid. It was
also shown that the addition of vinyl ethers to amine solutions reduces the induc-
tion period.

To improve the properties of nitric acid as a component of propergols the follow-
ing admixtures were used, or recommended:

(1) Nitrogen dioxide. This substance alone may be used as an oxidant, though
it has some disadvantages, as outlined above. The addition of nitrogen dioxide to
nitric acid facilitates enormously the reaction of the latter with many organic com-
pounds, including amines. Moreover it also lowers the freezing point of the nitric
acid. The lowest freezing point ( — 73°C) is attained by a solution composed of
82% HN0 3 and 18% N 2 4 . Nitric acid containing 20% N 2 4 is employed most
frequently.

LIQUID EXPLOSIVES 293

According to Canright [13] nitric acid becomes "stabilized" by the presence of
nitrogen dioxide and water. Such nitric acid remains unchanged during prolonged
storage at an elevated temperature. ,

(2) Mineral acids. Corrosion by nitric acid can be reduced by the addition of
sulphuric acid (up to 10%).

It was found (according to Canright [13]) that the addition of small amounts
of hydrogen fluoride to nitric acid considerably reduces its corrosive action on alu-
minium and stainless steel since these metals become coated with a layer of fluorides.

(3) Mineral salts accelerating the combustion reactions. Apart from ferric
chloride, which was discussed above, the use of other salts has also been suggested.
Grollier-Baron and Wessels [14], for instance, suggest the addition of 4% potassium
dichromate to nitric acid in non-hypergolic propergols formed from nitric acid
and petrol. According to these authors, at a temperature of 670°C and when the
reagents are injected into the combustion chamber at a rate of 10.3 m/sec, a mixture
of nitric acid with petrol is ignited after 23 millisec; at an injection rate of 6.6 m/sec
ignition occurs after 33 millisec. At 920°C and an injection rate of 6.6 m/sec the
induction period is 31-35 millisec.

If nitric acid with 4% potassium dichromate is used at 650°C and the injection
rate is 10.3 m/sec, the induction period is 5.5 millisec and at an injection rate of
5..3 m/sec it is 3.6 millisec. At 850°C and a rate of 6.6 m/sec the induction period
is shorter than 1 millisec.

In mixtures containing nitric acid various fuels may be employed. Fuels used
for hypergolic ones, i.e. those which autoignite on mixing, differ essentially from
those used in other mixtures. In hypergolic mixtures, fuels are used which reac.
violently with nitric acid, e.g. aliphatic or aromatic amines, furfuryl alcohol, mercapt
tans, hydrazine etc. It is also advisable to add surface-active substances to the mixture-

Fuels for mixtures with nitric acid

The following fuels are already in use or recommended for employment in
mixtures with nitric acid.

Aliphatic hydrocarbons: petrol, paraffin. In the U.S.A. several types of combus-
tibles for liquid propellant jet aircraft are used. One of them, i.e. JP-4, is employed
for rocket propulsion, with nitric acid as an oxidizing agent (it can also be used
with hydrogen peroxide or liquid oxygen). The specification of JP-4, is as follows:

Specific gravity at 15.5°C

0.764

Vapour pressure at 38°C

134-160 mm Hg

Fractionation :

10% distills up to

84-102°C

50% distills up to

142-147°C

90% distills up to

209-227°C

294 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Content of aromatic hydrocarbons 10-12%
Bromine number 1.4

Aniline point 58°C

Ignition point -13.8°C

Freezing point — 60°C

Heat of combustion 10,400 kcal/kg

Aliphatic hydrocarbons are seldom used separately due to their relatively slug-
gish reaction with nitric acid. The ignition capacity of hydrocarbons may be in-
creased by dissolving in them aromatic amines or, as the Germans did in earlier
experiments, vinyl ethers.

The effect of aromatic amines, however, has been the subject of much contro-
versy. Thus, according to Qrollier-Baron and Wessels [14] the addition of 10%
aniline to petrol has no obvious influence on the length of the induction period.
E.g. if nitric acid containing 4% potassium dichromate is used with petrol contain-
ing 10% aniline at 620°C, and a rate of injection is 10.3 m/sec, the induction pe-
riod is 3.3 millisec; at a rate of injection of 5.3 m/sec it is 24 millisec.

Amines. Amines were the first ingredients to be used in hypergols. A mixture
of 50% xylidine with 50% triethylamine is the most widely used.

According to Fedosyev and Sinyaryev [15] this mixture, with 98% nitric acid
in stoichiometric proportions has the following physico-chemical constants:

Specific gravity 1.32

Heat of reaction 1500 kcal/kg or

1800 kcal/kg
Gas volume 784 l./kg

Explosion temperature approximately 2710°C

Amines with furfuryl alcohol. The commonest of these mixtures is that with
80% aniline and 20% furfuryl alcohol. Fedosyev and Sinyaryev quote the following
values characteristic of this mixture with nitric acid:

Specific gravity 1.39

Heat of reaction 1520 kcal/kg or

1900 kcal/kg
Gas volume 756 l./kg

Explosion temperature approximately 2780°C

" Barrere and Moutet [16] suggest the following mixtures which form hypergols
with nitric acid:

(1) 25% aniline and 75% furfuryl alcohol

(2) 25% dimethylaniline and 75% furfuryl alcohol

(3) 25% toluidine and 75% furfuryl alcohol

(4) 25-50% xylidine and 75-50% furfuryl alcohol

(5) 25% diethylamine and 75% furfuryl alcohol

(6) 25% triethylamine and 75% furfuryl alcohol

Mixtures (2) and (4) react most easily. They have the shortest induction periods,
of 18 and 20 millisec, respectively.

LIQUID EXPLOSIVES 295

The least reactive are mixtures (5), (6) and particularly (2), with induction periods
of 55, 61 and 96 millisec, respectively.

The same authors studied a mixture of nitric acid with furfuryl alcohol. In reac-
tivity it occupies a position midway between the above mentioned groups with an
induction period of 33 millisec. Similar results were obtained by Kilpatrick and
Baker [17] when studying the reaction of furfuryl alcohol with colourless nitric acid.

Mercaptans. McCullough and Jenkins [18] investigated the possibility of the
use of mercaptans, by-products of petroleum refining. The mixture of mercaptans
contains as chief ingredients :

Propyl mercaptan 27.8% mol.
Butyl mercaptan 65.3% mol.
Amyl mercaptan 6.6% mol.

and traces of hexyl mercaptan.

They employed 96.5% nitric acid or an acid containing 22% of N 2 4 and 1%
of water. The experiments indicated that mercaptans could be used as the combusti-
ble component in nitric acid hypergols.

Hydrazine. Concentrated hydrazine (96%) reacts spontaneously with nitric
acid. According to Kilpatrick and Baker [17] reaction with 96% colourless nitric
acid occurs with a delay of 5.0 ± 1.7 millisec, whereas with fuming nitric acid (con-
taining 24% N 2 4 ) there is a delay of 3.1 ± 1.4 millisec. Hydrazine of 71.5% con-
centration gives a delay of about 37 millisec with either acid.

In another series of experiments on using a molar ratio =2.87 and a rate

N 2 H 4

of injection of 9.2 m/sec these authors obtained an induction period of 0.2 millisec.

In addition to the foregoing tests they examined the possibility of using liquid

ammonia with an admixture of 9 or 14% hydrazine as a combustible component

in a mixture with nitric acid containing 24% N 2 4 . The induction period was 14

or 6-10 millisec, respectively.

The properties of hydrazine including its explosive properties will be discussed
further.

Ammonia. Lewis, Pease and H. S. Taylor [19] and Altman and Penner [20]
showed that the system liquid ammonia-nitric acid may be transformed into
a hypergolic system by the addition of an alkali metal, e.g. lithium, to the
ammonia.

Surface-active substances. Bernard's [21] investigations showed that the addition
of surface-active substances (wetting agents) to the fuel may reduce the induction
period prior to the ignition of hypergols.

The author quotes the following figures which illustrate the effect of the addition
of sodium alkyl sulphate on the induction period of a mixture of furfural with
98% nitric acid:

296 CHEMISTRY AND TECHNOLOGY^ EXPLOSIVES

Mixture without additions 29.9 millisec
Addition of 0.5% sodium

alkyl sulphate to furfuryl

alcohol 14.4 millisec

Addition of 0.5% sodium

alkyl sulphate to nitric

acid 22.7 millisec

Table 63

Characteristic properties of fuels with nitric acid according to Bellinger,

Friedman, Bauer, Eastes and Bull [22]

Composition

Content of
combus-
tibles

°/

Heat of
reaction

kcal/kg

Density

Pressure
in combus-
tion

chamber
atm

Flash

point

°C

Specific

impulse

sec*

Nitric acid (60%) + nitro-
gen dioxide (40%) + aniline
Nitric acid + aniline
Nitric acid + furfuryl alco-
hol
Nitric acid + hydrazine

26.6

34.5
38.5

1535
1440

1.45

1.37
1.28

20
21

21
21

2707
2760

2620

225
218

214
243

« The specific impulse h is measured in kg (or lb) of pressure exerted by kg (or lb) of fuel per sec. Hence
specific impulse is expressed in seconds.

Nitroparaffins. Nitroparaffins, such as nitromethane and tetranitromethane, may
also act as constituents of propergols, although there is obviously an essential dif-
ference in employing these two substances, resulting from their chemical and explo-
sive properties.

Nitromethane (Vol. I, p. 579) may be used as a monergol propellant. However its
negative oxygen balance may be reduced by the addition of liquid oxidants, e.g.
of tetranitromethane to form a bipropellant. In practice, however, nitromethane
decomposes too slowly, and it is difficult therefore to obtain hypergolic mixtures
from it. To facilitate and accelerate the decomposition of mixtures with nitromethane
it is necessary to add a catalyst, such as a salt of chromic acid.

Higher nitroparaffins such as nitroethane, and 2-nitropropane may be used as
constituents of rocket propellants but rather in the capacity of a fuel, e.g. in mixtures
with nitric acid, hydrogen peroxide, or liquid oxygen. Research by Tait, A. E. Whit-
taker and Williams [23] showed that the combustion of stoichiometric mixture of
2-nitropropane with 98% nitric acid in a closed bomb is maintained spontaneously
under pressure above 10 atm. The rate of burning depends on the pressure. For
pressures in the ranges 14-70 and 70-140 atm (Fig. 78) the increase in rate of burning
as a function of pressure differs significantly. This indicates a difference in the
mechanism of the combustion reaction at these two pressure ranges.

Among other nitroparaffins isomeric dinitroethanes are also recommended
(Wood [24]):

asymmetric

LIQUID EXPLOSIVES

CH(N0 2 ) 2

297

CH 3

and symmetric CH2NO2

(liquid, b.p. 185°C,
density 1.35)

I (m.p. 40°C, b.p. 135°C,

CH 2 N0 2 density 1.46)

These substances have explosive properties and may be employed as monopro-
pellent propergols. However, symmetric dinitroethane does not seem to be stable
enough (Vol. I, p. 394).

Tetranitromethane is of a different nature, being largely an oxidant.

35 70 140
Pressure, atm

350

Fig. 78. The relation between the rate of burning of the stoichiometric mixture of
2-nitropropane with nitric acid and the pressure (according to Tait, A. E. Whittaker

and Williams [23]).

During World War II the Germans experimented with liquid mixtures consisting
of tetranitromethane and combustibles as a liquid fuel for the propulsion of the
big V2 rockets.

The greatest advantage of tetranitromethane as an oxidant lies in its high density
(1.64 at 20°C). With its high content of oxygen (65.3% by weight) and high density,
one litre of tetranitromethane contains 1.07 kg of oxygen, i.e. slightly less than
liquid oxygen at a temperature of -183°C (1.14 kg of oxygen per litre). Consi-
dering that a molecule of tetranitromethane itself contains a certain amount of
combustible material in the form of a carbon atom, it is possible to calculate the
oxygen content in the oxidizing part of the molecule only, i.e. in N0 2 groups. Ac-
cording to the calculations of Tschinkel [25] the density of nitro groups in tetrani-

298

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

tromethane is 1.484 at a boiling point of 126°C, and the oxygen content 1.16 kg per
one litre of nitro groups, i.e. higher than liquid oxygen alone.

Owing to these properties a mixture of tetranitromethane with paraffin is prefer-
able to mixtures containing other oxidizing agents. The explosive decomposition of
the tetranitromethane mixtures with petroleum hydrocarbons in stoichiometric

la*

1
/

\//

J
? 0.25

J *

<K.

0.5

1.0

Stoichiometric ratio

Fig. 79. The rate of burning of mixtures of tetranitromethane with hydrocarbons:
/— cyclohexane, //— iso-octane and ///— n-heptane (according to Behrens [27]).

proportions gives a high heat effect, viz. 1620 kcal/kg (according to Fedosyev and
Sinyaryev [15]).

The disadvantage of tetranitromethane lies in its high freezing point (+13.8°C).
The Germans suggested lowering this temperature by the addition of nitrogen dio-

0.5

2.5

1.0 15 2.0

Stoichiometric ratio
Fig. 80. The rate of burning of mixtures of tetranitromethane with alcohols: /— n-hexa-
nol, //— n-octanol, ///— n-butanol, IV— n-propanol, V— isopropanol, VI— isobutanol
(according to Behrens [27]).

xide which, with 20% N 2 4 , freezes at -14°C, and with 35% N 2 4 at -30°C.
Finally, they evolved a mixture for V2 propulsion having a composition: 30% N 2 4
and 70% tetranitromethane.

The disadvantages of this mixture are its high vapour pressure resulting from
the low boiling point of nitrogen dioxide, and its erosiveness.

LIQUID EXPLOSIVES

299

According to Hannum [26] tetranitromethane mixtures with nitromethane are
desirable for practical purposes. E.g., a mixture containing 10% nitromethane
freezes at 0°C, whereas with 20% nitromethane the freezing point is — 14°C, and
with 35% nitromethane it is low as — 30°C.

Among other additives, methyl nitrate (Tschinkel [25]) may be employed, or
substances safer to handle, such as methanol and ethylene glycol monoethyl ether.

0.5 1.0 1.5

Stoichiometric ratio

Fig. 81. The rate of burning of mixtures of tetranitromethane with benzaldehyde (/)
and nitrobenzene (//) (according to Behrens [27]).

Behrens [27] examined the rate of burning in glass tubes of 5.5 mm dia. of several
mixtures of tetranitromethane with a number of organic substances : hydrocarbons,
alcohols and aldehydes. All mixtures showed a maximum of the rate of burning
at a certain stoichiometric ratio. Figures 79, 80 and 81 give typical curves of mix-
tures with hydrocarbons, alcohols and benzaldehyde or nitrobenzene, respectively.

Schwob [28] made an extensive study of burning and detonation of mixtures of
tetranitromethane and petrol. He found that burning can pass to detonation when
the tetranitromethane content is 65-95%. The burning of these mixtures should be
considered as dangerous.

However, the limits of explosibility of the mixtures under the action of detona-
tors or impact are much wider : only the mixtures with tetranitromethane content
below 40% should be considered as non-explosive.

Hazards associated with the large-scale manufacture of tetranitromethane seri-
ously hinder its use (the plant at Newark, in the U.S.A., which had produced tetra-
nitromethane on a semi-commercial scale, blew up in 1953 and was never rebuilt,
so production had to be discontinued).

HYDROGEN PEROXIDE
H 2 2

Among oxidants hydrogen peroxide is one of the richest in oxygen. Pure H 2 2
contains 47% of available oxygen. A method for the preparation of concentrated
aqueous solutions of hydrogen peroxide, containing 80-85% H 2 2 , has been worked

300 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

out since 1943 in Germany by Elektrochemische Werke at Munich. During World
War II these solutions were manufactured on an industrial scale under the name
of "T-Stoff", either as an oxidant in rocket fuel or (to a less extent) a mono-propel-
lant fuel — cathergol.

The specific gravity of an 80% solution of hydrogen peroxide is 1.34. The solu-
tion is fairly stable at room temperature and decomposes only at an elevated tem-
perature (the stability of hydrogen peroxide will be descussed later). To increase
the stability of this solution, stabilizing substances such as phosphoric acid and
its salts or 8-hydroxyquinoline were added.

The manufacture of 90% hydrogen peroxide was started after World War II in
the U.S.A. and Great Britain.

N. S. Davis and Keefe quote the following physico-chemical constants charac-
teristic of 90% hydrogen peroxide [29]:

Specific gravity at 20°C

1.39

Viscosity at 18°C

11.62 cP

Vapour pressure at 30°C

5 mm Hg

Freezing point

-n.rc

Boiling point

175°C

Heat of formation (-AHf)

of liquid

45.16 kcal/mole

of vapour (100% H 2 2 )

33.29 kcal/mole

Specific heat (between and 18.5°C)

0.58 cal/g°C

Heat of vaporization ca

330 kcal/kg

Surface tension at 18°C

75.53 dyne/cm

Conductivity at 25°C

of the chemically pure product

2 x 10- 6

of a commercial product

lOxlO- 6

Refractive index at 20°C

1.398

Non-volatile residue in a commercial

product

0.005%

(Thermal analysis of the system hydrogen peroxide-water is given on Fig. 82.)
Chemically pure hydrogen peroxide can be stored for a long time without notice-
able decomposition. Loss on storage of hydrogen peroxide may amount to 1 %
per annum.

Shanley and Greenspan [31] report the following relationship between the decom-
position of 90% hydrogen peroxide and temperature (Table 64).

Table 64

Temperature
C C

Approximate rate
of decomposition

1 % per annum

1 % per week

100

2% per 24 hr

140

Rapid decomposition

LIQUID EXPLOSIVES

301

Slight amounts of impurities may accelerate the decomposition enormously.
The effect of various substances on the decomposition of 90% hydrogen peroxide
at 100°C is tabulated below.

Table 65

Substance

Quantity

Loss of H2O2
over 24 hr

added

at 100°C

mg/1.

Without additive

AP+

Sn++

Cr3<-

0.1

Cu2+

0.1

Fe3+

1.0

Zn2+

(On the accelerating action of hydrogen on H 2 2 — vapour see p. 303.)

Hydrogen peroxide is more stable in an acidic medium than in an alkaline one,
and acids are therefore used as stabilizers. Apart from phosphoric acid, already
mentioned, boric acid, oxalic acid etc. may also be used.

20 40 60 80 100
H2O2 concentration, %

Fig. 82. Freezing temperatures of binary system H2O2-H2O [30].

Experiments have shown that the best construction material for the storage of
hydrogen peroxide is high purity (99.6%) aluminium. Aluminium containers should
be thoroughly cleaned to remove any traces of organic impurities, and washed
first with a solution of caustic soda and then with water and 10% sulphuric acid,
over a period of several hours. Finally the acid is washed out with distilled water,
after which it is desirable to re-wash the container with hydrogen peroxide. Tanks
and containers for hydrogen peroxide should be provided with safety valves, that
are ruptured by the excessive pressure produced if the peroxide decomposes.

302 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Special care should be taken to prevent hydrogen peroxide from coming into
contact with copper, chromium and lead.

Fittings, pumps, and pipelines can be made of certain types of stainless steels
which can be allowed to remain in contact with hydrogen peroxide for relatively
short periods, i.e. a few days.

A number of polymers, in particular teflon, polyvinyl chloride and, to a lesser
extent, polyethylene do not provoke the decomposition of peroxide.

If, as a result of the presence of some impurities in a vessel containing hydrogen
peroxide, decomposition is hastened, it is advisable to add an additional quantity
of stabilizer, i.e. phosphoric acid. If this proves inadequate, then, according to Shanley
and Greenspan [31], it is necessary to dilute the hydrogen peroxide with water to
67% concentration when it is no longer dangerous, otherwise, violent decomposition
may occur and the container may blow up.

When handling large quantities of hydrogen peroxide it is necessary to wear
protective goggles, overalls made of protective fabric (an apron of polyvinyl chloride),
rubber gauntlets and boots, since severe burns are caused if it comes into contact
with the skin.

As previously mentioned, concentrated hydrogen peroxide is used as a cathergol-
type monopropellant fuel.

During World War II the Germans employed 80-85% hydrogen peroxide for
launching pilotless V2 aircraft and for bringing into operation a mechanism feeding
the oxidant and the combustible from their storage tanks into the combustion cham-
bers of VI aircraft and V2 rockets. This application is based on the decomposition
reaction of hydrogen peroxide:

H 2 2 =H 2 + $0 2 + 23.45 kcal/mole (690 kcal/kg) (1)

Decomposition was initiated by the addition of a concentrated aqueous solution
of calcium and sodium permanganate to the hydrogen peroxide. The use of potas-
sium permanganate proved ineffective since KMn0 4 is insufficiently soluble in
water and the solution contains inadequate MaOf -ions for rapid initiation of
the reaction.

According to Ley [32] the British "Sprite" rocket serving the A.T.O. (assisted
take-off) of the "Comet" jet aircraft has a similar propulsion unit, i.e. it contains
a charge of 136 1. of hydrogen peroxide and about 9.5 1. of a catalyst solution,
most probably permanganate.

The use of hydrogen peroxide as a mono-propellant of the cathergol type is
based on the following thermochemical data for hydrogen peroxide of 86 and 100%
concentration (by weight) (Wood [24], Table 66).

As these characteristics show, the heat of decomposition is considerably higher
than the heat of vaporization. The vapour so produced is superheated and, at the
same time the temperature of adiabatic decomposition is high enough to make
possible a useful expansion of gases. The efficiency of such a fuel, however, is negli-
gible. Its specific impulse does not exceed 130 sec.

LIQUID EXPLOSIVES

303

A considerable amount of work [33] was dedicated to studying the mechanism
of decomposition of hydrogen peroxide in the vapour phase. It was soon recognized
that the difficulty of obtaining reproducible results is due to the action of the vessel
surface producing the heterogeneous reaction. Baldwin and Mayor [34] have shown
that the kinetics of the slow reaction between H 2 and 2 in aged boric acid-coated
vessels could only be explained by assuming that the aged surface was extremely
inert to both H0 2 and H 2 2 . Baldwin and Brattan [35] studied the reaction of gaseous
decomposition of hydrogen peroxide in an aged boric acid-coated vessel over a range
of temperatures 260-520°C. They found the decomposition and the dependence of
rate on total pressure being of the first order.

Table 66
Thermochemical data for hydrogen peroxide

Item

Heat of decomposition (kcal/kg)

Heat of vaporization (kcal/kg)

Adiabatic decomposition temperature (°C)

Gas volume on complete adiabatic decomposition (1.)

The results are in agreement with the work of Giguere and Liu [36], Forst [37]
and Hoare et al. [38].

Baldwin and Brattan also found that the addition of hydrogen to H 2 2 vapour
considerably increases the rate of decomposition of the latter.

30 40 50

NH 4 N0 3 content, %
Fig. 83. Freezing temperatures of binary system H 2 2 -NH 4 N0 3 [30].

Concentrated hydrogen peroxide was widely used during World War II as an
oxidant in a mixture with hydrazine hydrate, for the propulsion of V2 rockets.
Hydrogen peroxide mixed with hydrazine reacts spontaneously according to the
equation :

2H 2 2 + NH 2 NH 2 -H 2 0— >5H 2 + N 2 + 188 kcal

(2)

304 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

As reaction is preceded by a certain delay, this induction period was reduced by
the addition of copper salts, e.g. potassium cuprocyanide K 3 Cu(CN) 4 . This sub-
stance was supplied to the system dissolved in the hydrazine hydrate. It was found
that potassium cuprocyanide reacts with hydrazine even at room temperature to
form metallic copper which, if deposited in the pipelines, may cut off the flow of
hydrazine into the combustion chamber. To prevent this the system was modified
so that hydrazine hydrate flowed from the tank into the combustion chamber through
a cartridge containing cupric nitrate, which dissolved in hydrazine hydrate in a suffi-
cient quantity to accelerate the reaction (hydrazine and its reaction with H 2 2
will be discussed in more detail further on).

During World War II the Germans used a mixture of hydrogen peroxide with
Diesel oil in torpedo and submarine propulsion.

An idea of considerable interest advanced by Paushkin [30] was the use of a so-
lution of ammonium nitrate in concentrated hydrogen peroxide as an oxidant.
A solution consisting of 40% NH 4 N0 3 and 60% of 92% hydrogen peroxide seems
to be particularly attractive. Its freezing point is approximately — 30°C (Fig. 83).

Explosive properties of hydrogen peroxide and its mixtures

The exothermic character of the decomposition of hydrogen peroxide indicates
that hydrogen peroxide itself may have explosive properties. In fact 99.6% hydrogen
peroxide gives a lead block expansion of 75-80 cm'. In a steel tube with 34-40 mm

100% C2H5OH

\ / \

-VIA

/' '^

rA/

AA~

/ \

A \A'

A A\

/ \ /

\ /\

A / \ A

\ / \

/\ / \

/\> \/ \

/ \ /

\ /\/

/V\ / A'

x / \ ''

/ \y x >

AAA/

/\/

A/A V /

A /\

_:/ . y _ ^-V»

•^A r

"\7A

Detonation *

v'° A

— A '

• —>—

region v

V_< :

»- — V-

100% f\ /\ A,
H2O2

Fig. 84. Diagram of explosive properties of ternary mixtures H2O2-H2O-C2H5OH,
according to Shanley and Greenspan [31]; x — detonation, o — no detonation.

LIQUID EXPLOSIVES 305

dia. 94-100% hydrogen peroxide detonates completely when initiated with 50 g
of compressed pentaerythritol. With 92% hydrogen peroxide, however, detonation
is propagated only to 100 mm along the tube, away from the detonator.

Ninety per cent hydrogen peroxide does not detonate at all (Paushkin [30]).

Investigations into the explosive properties of ternary mixtures of hydrogen
peroxide and water with various organic substances, carried out by Shanley and
Greenspan [31], aroused much interest. On the basis of their results triangular
diagrams may be constructed for a number of systems: (a) hydrogen peroxide,
(b) water, (c) organic substance (such as ethanol, glycerine, acetone). In Fig. 84
a typical diagram is shown, characteristic of mixtures with ethanol. The other dia-
grams are very similar. In general, only mixtures containing a limited amount of
water possess explosive properties.

All the three organic substances mentioned above, when dissolved in 80% H 2 2 ,
give mixtures detonated by a detonating cap and a booster with a rate of about
7000 m/sec, which drops to 2300 m/sec as the concentration of hydrogen peroxide
is reduced. When weakly initiated they detonate with a rate of 750 m/sec only
(Schumb, Satterfield and Wentworth [39]).

A monograph on hydrogen peroxide was written by Schumb, Satterfield and
Wentworth [39].

HYDRAZINE
Physico-chemical and explosive properties

Anhydrous hydrazine melts at 2°C and boils at 1 13.5°C, its density is 1.0253 g/cm^
(Walden and Hilgert [40]); 1.0231 (Semishin [41]).

With water it forms hydrazine hydrate NH 2 NH 2 H 2 with a melting point
of -51.7°C and a density d% of 1.048 (Semishin).

Hydrazine is an endothermic substance. The heats of formation -AH t of anhy-
drous hydrazine and the hydrate are, according to Hughes, Gilbert et al. [42]:

N 2 H 4 liquid - 12.0 kcal/mole
N 2 H 4 -H 2 - 10.2 kcal/mole

Roth [43] found the heat of formation of liquid hydrazine at 25°C to be -AH t
= 13.8 kcal/mole.

According to Hughes, Gilbert et al. the heat of combustion of the anhydrous
substance (liquid) is -AH C = 148.6 kcal/mole.

Anhydrous hydrazine burns in air. On heating hydrazine above boiling point,
thermal decomposition of gaseous hydrazine takes place at 250-3 10°C (Elgin and
Taylor [44]; Askey [45]). Bamford [46] ascertained that hydrazine is decomposed
by an electrical spark, while Elgin and Taylor established that hydrazine vapour is
decomposed by ultra-violet irradiation (also Wenner and Beckmann [47]).

306

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

According to Bamford, the thermal decomposition and explosion of hydrazine
is expressed by equations in which free radicals are present:

(I) N 2 H 4 -► 2NH 2

NH 2 + N 2 H 4 -> NH 3 + N 2 H 3
2N 2 H 3 -^ N 2 + 2NH 3

(II) N 2 H 4 ->• 2NH 2

NH 2 + N 2 H 4 ~> NH 3 + N 2 H,
NH 2 + N 2 H 3 -> N 2 + H 2 + NH 3

(1)
(2)
(3)

0)
(2)
(3)

In addition, side reactions may occur:

- N 2 +2H 2
2NH 2 + N 2 H 4 -> 2N 2 + 4H 2

(III) N 2 H 4 -» 2NH 2

(IV) N 2 H 4 -> 2NH 2

In spite of the ease with which the decomposition of hydrazine has been estab-
lished, its explosive properties have been not established with certainty.

In relatively recent times Scott, Jones and Lewis [48] made detailed studies of
the explosive properties of hydrazine and 85% hydrazine hydrate. In the drop test,
neither hydrazine nor hydrazine hydrate were exploded by a very strong blow (5 kg
from 1 m height). Hydrazine and its hydrate are also insensitive to friction. The
authors' attempts to determine the rate of detonation of hydrazine in pipes also
failed since hydrazine is initiated by a detonator with difficulty. The reaction in the
ballistic pendulum, of charge of anhydrous hydrazine initiated by No. 8 detonator,
was 135% of that produced by TNT. Under these conditions hydrazine hydrate
does not detonate at all.

Scott, Jones and Lewis examined the ignitability of hydrazine and hydrazine
hydrate by various means.

Table 67

Hydrazine

85% hydrazine hydrate

Atmosphere

the vessel

Ignition

Induction

Ignition

Induction

is made

temperature

period

temperature

period

°C

sec

°C

sec

Pyrex glass

Air

270

3.9

292

4.2

Oxygen

204

4.9

218

5.3

Platinum

Air

226

3.0

338

3.8

Oxygen

6.0

132

19.7

Fe 2 03 (powdered

Air

0.0

—

in a glass vessel)

Nitrogen

0.0

—

Iron

Air

132

0.0

—

Nitrogen

131

0.0

—

Stainless steel

Air

160

2.0

—

Nitrogen

415

0.9

—

LIQUID EXPLOSIVES

307

They established that an electrical spark having an energy of 12.5. J when passed
through unconfined liquid hydrazine produces no signs of decomposition. Conver-
sely, hydrazine or its hydrate in a confined space undergoes explosive decomposi-
tion under the influence of a spark stronger than 2.63 J.

In an atmosphere of oxygen the ignition of hydrazine (but not its hydrate) takes
place at a lower temperature than in air. The ignition temperature is strongly in-
fluenced by the material of which the hydrazine container is made.

The authors' [48] numerical data are collected in Table 67. They are valuable
indications of the safety measures which should be observed in the storage, trans-
portation and handling of hydrazine. It can be contained in vessels of glass and
stainless steel, but under no circumstances in those made of iron. Aluminium con-
tainers are also acceptable.

The experiments of Scott, Jones and Lewis on the limits of ignitability of hydra-
zine vapours are also of great importance. The figures obtained by these authors
are tabulated below (Table 68).

The ignition of the gaseous mixture was brought about by an electrical spark.

Table 68

Content of (%)

Pressure
mm Hg

Temperature
°C

Composition of gaseous mixture

Hydrazine

Other

component

Hydrazine-air

4.67

95.33

757-758

92-101

Hydrazine-nitrogen

38.0

62.0

754

109-112

Hydrazine-helium

37.0

63.0

756-758

105-118

Hydrazine-water vapour

30.9

69.1

689-889

130-135

Hydrazine-heptane

86.8

13.2

404-327

104-133

Hydrazine is a highly toxic substance, injurious to the sight, causing temporary
blindness. The lethal dose (LD 50 ) for dogs is approximately 0.05 g/kg of body
weight. Salts of hydrazine provoke hyperglycaemia, blood clotting due to dehydration
and liver damage.

According to Raciborski [49] some moulds can assimilate hydrazine.

A monograph by Audrieth and Ogg [50] is dedicated to hydrazine.

Oxidation of hydrazine by hydrogen peroxide

This most important reaction— the oxidation of hydrazine— has not yet been
investigated fully. Work on the subject has consisted mainly of studies of the kinetics
of the process in dilute aqueous solutions. Gordon [51], studying the kinetics of
decomposition of hydrazine and hydrogen peroxide, found that the reaction rate
depends to a great extent on the pH. Its peak value is reached at pH = 10-11.

The addition of cobaltic sulphate to the reaction system raises the reaction rate
considerably.

3Q8 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Concentrated hydrogen peroxide does not react instantly with hydrazine hydrate,
only after a certain induction period. This has been the cause of a number of explosions
and accidents, produced by the accumulation of unchanged components and their
sudden reaction after the induction period has elapsed.

As previously stated, the addition of a copper salt to hydrazine reduces the
induction period practically to zero. The addition of sodium nitroprusside exerts
a similar influence.

According to McLarren [52] a mixture with methyl alcohol has frequently been
used in rockets, to react with hydrogen peroxide. E.g. in the HWK-59 jet pro-
pulsion engine and BP-20 rocket missiles 80% hydrogen peroxide is used as an
oxidant together with a combustible consisting of:

Hydrazine hydrate 30%
Methyl alcohol 57%
Water 13%

0.11% of K 3 Cu(CN) 4 is added to the fuel.

According to Fedosyev and Sinyaryev [15] a mixture of 80% hydrogen perox-
ide with hydrazine hydrate diluted with methyl alcohol in the ratio of 1 : 1 has the
following physico-chemical constants:

Heat of reaction

1020 kcal/kg or

1330kcal/l.

Specific gravity

1.30

Gas volume

940 l./g

Temperature of explosion

2330°C

Specific impulse

180 sec

Among the reactions of hydrazine used for rocket propulsion that of hydrazine
with' nitric acid is known.

It is also possible to use hydrazine hydrate alone as a monergol owing to its
high heat of decomposition. Energy and gaseous products are provided by decomposi-
tion induced by permanganates, commonly used in the solid form.

1,1-DIMETHYLHYDRAZINE (UDMH)

Among the homologues of hydrazine, asymmetric dimethylhydrazine (CH 3 ) 2 N-
•NH 2 is important. It is obtained from dimethylamine by nitrosation followed
by reduction

(CH 3 ) 2 NH 2S. (CH 3 ) 2 N-NO -^ (CH 3 ) 2 N-NH 2

or by a modification of the Raschig method of preparation of hydrazine in the
presence of dimethylamine.

It is a colourless liquid with a freezing point of ca. -56°C, a boiling point
of ca. 63°C and a density of 0.785 g/cm^.

The heat of combustion is ca. 3580 kcal/kg.

LIQUID EXPLOSIVES

309

Under the name of "Dimazine" or the abbreviation of UDMH it is used for
hypergols by mixing with nitric acid. It may also be used with liquid oxygen.

In the U.S.A. it is employed in "Nike Ajax", "Rascal" and "Vanguard" rockets
(Warren [53]).

MIXTURES WITH LIQUID OXYGEN AND OZONE

When V2 rockets were first used a mixture of liquid oxygen with 70% methyl
alcohol was employed as a fuel. This mixture, however, is not capable of self-ignition,
and it had to be ignited by means of a pyrotechnical mixture giving a hot flame.

.Mixtures containing liquid oxygen are less commonly used now for jet propul-
sion than mixtures with nitric acid. Nevertheless, for obvious reasons, they have
much prospect of success considering that liquid oxygen is a 100% oxidant.

A disadvantage of liquid oxygen is that its boiling point is very low (- 183°C),
an so is its specific gravity at this temperature (1.14). In view of the low boiling point
rockets should be filled with this liquid oxidant just before use.

As a combustible, paraffin or alcohols may be employed. Fedosyev and Sinyaryev
[15] quote the following physico-chemical constants for typical mixtures with liquid
oxygen.

Table 69

Heat of reaction

Specific
gravity

Volume

of gases

l./kg

Temper-
ature of

Fuel

kcal/kg

kcal/1.

explosion
°C

Paraffin
Ethanol (93.5%)

2200
2020

2200
2000

1.00
0.998

650

789

3280
2980

Investigations into the possibility of using liquid ozone or, strictly speaking,
mixtures of liquid oxygen with liquid ozone, have recently been carried out by the
Armour Research Foundation in Illinois. According to Platz and Hersh [54] liquid
ozone or a mixture of liquid ozone and oxygen may be obtained by introducing
oxygen, carefully purified from organic impurities, into the ozonizer, where the mix-
ture is irradiated and the ozone liquified (— 111.9°C under atmospheric pressure);
the oxygen escapes through an exit pipe. If a mixture of liquid ozone with oxygen
is to be obtained the gases discharged from the ozonizer are introduced to the
liquid oxygen after being cooled.

Liquid oxygen (100%) and a mixture of ozone with oxygen are stable enough
if the oxygen used for producing the ozone contains not more than 0.002% (calcu-
lated on C0 2 ) of organic material.

NITRIC ESTERS

Liquid and volatile nitric esters, being safer to handle and more stable chemically
than nitroglycerine or diethylene glycol dinitrate (DGDN), can be employed as
monopropellant jet fuels. Among these compounds methyl nitrate should be mention-

310

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ed first (Vol. II, p. 160). It was used for jet propulsion by the Germans during
World War II under the name of Myrol either in the pure state or as a methyl
alcohol solution (30% of methanol and 70% v of methyl nitrate). Recently, another
nitric ester, i.e. isopropyl nitrate has been suggested but for the time being there
is no further information available about its use. Sometimes ethyl nitrate is also
mentioned.

ETHYLENE OXIDE

This compound also deserves attention as a monergol. Its physical constants
are: boiling point 11 °C, freezing point -112°G, density 0.90. It decomposes exo-
thermically according to the theoretical equation:

H 2 C CH 2 -> CO + CH 4 + 32 kcal/mole (726 kcal/kg)

(1)

In point of fact the reaction is more complicated and proceeds according to the
equation:

C 2 H 4 -> *CO + yCH 4 + zC 2 H 4 + fH 2 (2)

where the coefficients x, y, z and t are less than 1. The heat of this reaction is some-
what lower than for equation (1), but in calorific value, ethylene oxide is on a par
with propellants. Table 70 gives the products of decomposition, temperature of
explosion and specific impulse according to Glassman and Scott [55] and Kruska [56].

Table 70

Pressure in chamber

Composition of products (%)

Tempera-
ture of
explosion
°C

Specific

atm

CH 4 H 2

C 2 H 4

impulse
sec

20
30
40
60

1.0
1.0
1.0
1.0

0.84
0.86
0.87
0.88

0.16
0.14
0.06
0.06

0.08
0.07
0.06
0.06

1015
1027
1033
1039

159
168
174
181

Ethylene oxide has the advantage of being safe to handle since it is not strictly
an explosive.

ATTEMPTS TO INCREASE THE ENERGY OF LIQUID MIXTURES FOR

ROCKET PROPULSION

Clearly, attempts to increase the energy liberated by liquid mixtures for rocket
propulsion are based, in the first instance, on the use of those components (com-
bustible and oxidant) which release as much heat as possible. E.g. the use of ozone
as an oxidant has been discussed above (p. 309).

LIQUID EXPLOSIVES

311

MIXTURES WITH POWDERED METALS

It has been suggested that powdered metals, e.g. aluminium should be added to
the combustible component in the form of a suspension. Stettbacher [57], for ex-
ample, suggested the following equation for the combustion of a mixture of petrol
with aluminium suspended in it in stoichiometric proportions:

C 7 H 16 + 2A1 + 12io 2 -> 7C0 2 + 8H 2 + A1 2 3
On combustion this mixture gives ca. 2545 kcal/kg or ca. 2763 kcal/1.
Stettbacher calls attention to the fact that powdered aluminium always contains

a certain amount of aluminium oxide (up to 11%), hence the heat of combustion is

lower than that theoretically calculated.

The significance of the addition of beryllium to the fuel is still rather theoretical.

E.g. paraffin with nitric acid in stoichiometric proportions gives 1440 kcal/kg, whereas

the same mixture with 7.2% and 10.0% of beryllium gives a heat effect of 2130 kcal/kg

and 2480 kcal/kg, respectively.

BORON, SILICON AND BERYLLIUM COMPOUNDS

The addition of suspensions of metals to the liquid involves difficulty in achieving
a homogeneous suspension, and as an alternative the use of organometallic and
organometalloid compounds or hydrides has been suggested. These may be combina-
tions of boron with hydrogen, boron with hydrogen and nitrogen, silicon with hydro-
gen, silicon with hydrogen and nitrogen, which all are predominantly endothermic
or only slightly exothermic.

Fedosyev and Sinyaryev [15] report the following properties of the most typical
representatives of the above compounds (Table 71).

Table 71

Compound

Melting

point

°C

Boiling

point

°C

Heat of
formation

-AH t
kcal/mole

Specific

Name

Formula

gravity

Pentaborane
Diborane amine
Trisilane
Trisilicyl amine
Beryllium ethyl

B 5 H 9

B 2 H 7 N

Si 3 H 8

(SiH 3 ) 3 N
Be(C 2 H 5 ) 2

50*

-66

-117

-106

60*
76
53
52
200*

0*
-10*
-20*
-10*

-35*

0.64*
0.70*
0.88*
0.895
0.60*

* Approximate figures.

It is to be noted that boranes react with water in such a manner that the latter
acts as oxidant, e.g.:

B 4 H 10 + 12H 2 -+ 4H 3 B0 3 + 11H 2

312

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ORGANOMETALLIC COMPOUNDS

Experiments are also being carried out on the addition of organometallic com-
pounds such as lithium ethyl, aluminium trimethyl or triethyl etc. to the fuel. These
compounds ignite on contact with the oxygen of the air or with oxidizing agents—
components or propergols— hence they can be valuable constituents of hypergols.
At the same time they liberate a large amount of heat on combustion and can thus
be used for increasing the energy evolved on the combustion of propergols.

FLUORINE AND ITS DERIVATIVES

Recently, the use of fluorine as an oxidant has been considered feasible. E.g. the
reaction of fluorine with hydrazine gives a particularly large theoretical specific
impulse (7 S ) amounting to 298 sec.

Fluorine with hydrogen gives 7 S = 352 sec, whereas oxygen with hydrogen has
a somewhat lower value (342 sec).

Nevertheless it should be borne in mind that the use of liquid fluorine has con-
siderable disadvantages. Its boiling point is -187°C. To prevent corrosion special
vessels of nickel alloys surrounded by a jacket filled with liquid nitrogen (boiling
point — 199.5°C) are required.

Operating with liquid hydrogen as a combustible also creates exceptional dif-
ficulties, due to its exceedingly low boiling point (— 253°C).

In addition to fluorine itself, fluorine compounds have also been recommended.
Chlorine trifluoride (C1F 3 ) with a boiling point of 12°C and a density of 1.77 g/cm'
is the most promising for use in rocket propulsion. Its specific gravity is 1.85,
heat of formation -AH r = 41.0 kcal/mole. It is obtainable by the action of fluorine
on chlorine in an atmosphere of nitrogen at 280°C, in a reactor of copper or
nickel.

During World War II, the Germans developed the production of chlorine tri-
fluoride as an incendiary agent.

Another fluorine compound— nitrogen trifluoride — is interesting theoretically but
difficult to manufacture. . ,

Fluorine oxide also merits attention. This is a gas liquefiable at — 144.8°C and
weakly endothermic. Its heat of formation —AH { = -9.2 kcal/mole [58].

Still another compound of great interest is perchloryl fluoride (C10 3 • F). It was
described in Vol. II. Perchloryl fluoride is distinguished by its high stability; it
causes no corrosion of commonly used materials. It reacts vigorously with oxidizable
organic compounds.

It reacts with hydrazine. In rockets such a mixture gives a specific impulse l s
of 270.

LIQUID EXPLOSIVES 313

According to Engelbrecht and Atzwanger [59], Jarry [60], and [61] the physical
properties of perchloryl fluoride are as follows:

Melting point

-146°C (-146±2°C)

Boiling point

-47.5 + 0.5°C (-46.8°C)

Vapour pressure

logio P(mm)= 18.90112- 1443.467/r

-4.09566 logio T(at - 120 to -40°C)

log P(atm)=4.46862-1010.81/r

(at -40 to + 95.15°C)

Heat of vaporization

4.6 kcal/mole

Liquid density

2.266-1.603 x 10-3 t -4.080 x 10-6T2

g/mlT

Critical temperature

95.13°C

Critical pressure

53.0 atm

Critical density

0.637 g/cm3

Heat of formation at 25°C

-J#5r=5.12±0.68 kcal/mole

Specific heat of liquid

at -40°C

0.229 cal/g°C

-10°C

0.244 cal/g°C

+ 50°C

0.290 caI/g°C

^ at 24°C

s„

1.12

MIXTURES WITH PERCHLORIC ACID

Perchloric acid (HC10 4 ) is also recommended as an oxidant for rocket fuels.
The anhydrous acid is a liquid with a specific gravity at 20°C of 1.767 and a free-
zing point of -112°C; it decomposes when heated to 9°C. It can be distilled
under reduced pressure (at 16°C under a pressure of 18 mm Hg, at 30°C under
a pressure of 50 mm Hg).

Perchloric acid forms easily hydrates:

HC10 4 H 2 0(m.p. 50°C, content of HCIO4 84.8%, sp.gr. at 20°C 1.7756)
HCI0 4 -2H 2 0(m.p. -17.8°C, content of HC10 4 73.6%)
2HC10 4 -5H 2 0(m.p. -29.8°C, content of HC10 4 69.1%)
a and hydrates: HC10 4 -3H 2 0(m.p. -37 and -43.2°C

respectively, content of HC10 4 65.0%)
2HC10 4 -7H 2 0(m.p. -41.4°C, content of HCIO4 61.5%)

The hydrates of perchloric acid (even lower ones) are judged to be unsuitable
as rocket fuels.

The heat of formation of perchloric acid —AH° is 11.1 kcal/mole.
Full decomposition proceeds with heat evolution:

HCIO4 -> iH 2 + iCl 2 + 1.750 2 + 15.7 kcal

158 kcal are evolved from 1 kg which is not sufficient to make the substance
explosive, but the addition of 3% of organic substance to perchloric acid gives an
explosive mixture. Some organic substances (amines, unsaturated compounds,
cellulose, wood, rubber) are ignited on contact with perchloric acid.

314 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Elliot and Brown [62] made extensive studies of the inflammability of mixtures of
perchloric acid with oxidizable substances. Most of the mixtures of 60% perchloric
acid when ignited in a confined space burned to detonation. The mixtures with 70%
perchloric acid and some of them with 60% perchloric acid could be ignited by
impact. Explosion was induced under action of No. 6 detonator on mixtures with
60% perchloric acid with wood meal or cotton and the rate of explosion was
found to be 3000 m/sec.

Exposure to irradiation of sunlight did not seem to affect the mixtures with
40-60% perchloric acid.

A detailed study was made by Jacquet [63] and Jacquet, Medard and Sartorius
[64] of the explosive properties of the three-component system: perchloric acid, acetic
anhydride and acetic acid. These solutions are widely used in electropolishing baths
[65]. However, Merchant [66] drew attention to the explosive properties of such mix-
tures. Indeed, a number of explosions of electropolishing mixtures occurred, and that
in Los Angeles (O'Connor plant) 1947 was particularly disastrous. Two hundred
gallons of a solution of 75 vol.% of 72% perchloric acid and 25 vol.% of acetic
anhydride exploded. Seventeen men were killed and one hundred and fifty wounded.
By calorimetric measurements [67] the heat of mixing aqueous perchloric acid
(69%) with acetic anhydride in glacial acetic acid was found to be 20.6 ± 1.8 kcal/mole.
This is the difference between the heats of hydration of acetic anhydride (34.8
kcal/mole) and of 69% perchloric acid (16.4 kcal/mole):

2.5(CH 3 CO) 2 + HC10 4 -2.5H 2 -> 5CH 3 COOH + HCIO4 + 18.5 kcal/mole

The highest explosive power would correspond to the stoichiometric mixture
composed of 66 vol.% of 72% perchloric acid and 34 vol.% of acetic anhydride.
It could decompose according to equation:

CH3COOH + HCIO4 -> 2C0 2 + 2H 2 + HC1

The calculated thermal effect in this reaction (ca. 1250kcal/kg) would fall between
those of the explosive decomposition of nitroglycerine and guncotton, and the calcu-
lated temperature would be ca. 2500°C.

Jacquet, Medard and Sartorius [63, 64] investigated the process of mixing 62.7%
perchloric acid (density 1.59) with 100% acetic anhydride. They found that no
explosion occurred when vigorous mixing was applied. This was independent of
the order of mixing but addition of the acid to the anhydride was suggested to be
the less dangerous procedure.

The stoichiometric mixture (mentioned above) was found to be very sensitive
to priming. It was detonated by a primer as weak as 0.6 g of mercury fulminate. The
rate of detonation, or more precisely of explosion, was variable (this is typical for
liquid explosives): 1300-2000 m/sec. Expansion in the lead block was found to
be 85 (picric acid = 100) which is the same value as that of dinitrobenzene (88).

LIQUID EXPLOSIVES

315

Aptitude to detonation decreased with increasing content of acetic anhydride:
mixtures containing less than 57 vol.% of perchloric acid (62.7%) could not be
detonated.

The sensitiveness to shock was also examined. 50% explosions were obtained
when 1 kg weight fall from 1.40 m on a drop of the stoichiometric liquid in a metal
capsule.

The ignition of the mixture by red hot wire or with a flame was difficult. Only those
richest in acetic anhydride could be ignited at the boiling point. With increasing

HCIO4

S?2 — Line of "complete
combustion mixtures"
,40

Acid (d: 1.70)
(Azeotrope)_ ^

Add (d:1.60k
60/—
Acid fd.-lA

50.

H2O

• % Water

(CHsCOhO

Fig. 85. Explosive properties of perchloric acid-acstic anhydride-acstic acid (and
water) mixtures, according to Jacquet [63].

perchloric acid content the ignition became more difficult. The stoichiometric mix-
ture did not burn.

The solutions used for metal polishing (see below) ignited when fine wood
shavings were added at 60°C.

Jacquet [63] summarized his results by drawing the triangular diagram (Fig. 85).
The stoichiometric mixture is marked by point (C) and the mixture which gave
the Los Angeles accident is marked (B). In the ignition zone but outside the region
of detonation are most of the compositions usually applied for electropolishing
baths. Point (.4) corresponds to the mixture in which all the water introduced by
64% perchloric acid reacted with the acetic anhydride to form acetic acid.

Handling and storage of perchloric acid must be done with particular care.

316 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

It is well known that anhydrous perchloric acid must never be allowed to come
into contact with oxidizable organic substances such as alcohols, wood, cotton,
paper, cork and most plastics otherwise ignition or explosion will result.

Anhydrous perchloric acid is liable to explosive decomposition even when
free of organic substances. On storage it gradually becomes coloured due to the
formation of decomposition products, even in the dark, and in this condition it
may explode spontaneously. Anhydrous acid which has become amber coloured (or
darker) should therefore be diluted with water immediately and discarded.

Also aqueous perchloric acid may cause fire or explosion [68].

Considerable caution must be exercised in contacting perchloric acid with
metals, e.g. the catalytic action has been reported [69] of steel particles in reducing
the explosion temperature of mixtures of perchloric acid vapour and hydrogen [69].

REACTIONS OF FREE ATOMS OR RADICALS

Another approach to the problem of increasing the energy of rocket fuels consists
of searching for possibilities to operate with combustible (e.g. hydrogen) and oxi-
dizing (e.g. oxygen) elements in the form of atoms (H, O) and not molecules (H 2 ,
2 ) at the moment of combustion, as reaction of atoms would give a much greater
heat effect than the reaction of molecules. These attempts are of no practical signifi-
cance for the present, since methods for producing free atoms sufficiently con-
centrated and storable are still unknown.

GENERAL CONSIDERATIONS

In publications including those of Latham, Bowersock and Bailey [70], and
Wood [24] the following magnitudes of the specific impulse of various mixtures,
currently used and prospective (Table 72) are quoted.

A recent publication [71] suggests mixtures of oxygen difluoride as an oxidizer.
This substance can give an I s as high as ca. 400 sec when mixed with hydrogen. The
mixture of oxygen difluoride and unsymmetrical dimethylhydrazine has a theoretical
specific impulse of ca. 330 sec.

A high performance can also be achieved with perchloryl fluoride (Vol. II) and
tetrafluorohydrazine as oxidizers.

In selecting the constituents of a fuel every effort should be made to attain the
optimum conditions likely to be created by the mixture. The optimum conditions
are attained by creating:

(1) The highest temperature of reaction

(2) The lowest molecular weight

(3) The lowest — ratio

liquid explosives
Table 72

317

Constituents

Oxidant:
combus-
tible
ratio

Specific
gravity

Specific

impulse

sec

Oxidant

Combustible

Remarks

100% nitric acid

Turpentine

4.4

221

Fuming nitric acid (FNA)

Ethanol

2.5

219

FNA

Aniline

3.0

221

FNA

Ammonia

2.2

225

FNA

JP-4

225

99% hydrogen peroxide

Ethanol

4.0

230

99% hydrogen peroxide

JP-4

6.5

233

99% hydrogen peroxide

Hydrazine

245

Fuels most

Liquid oxygen (LOX)

Ethanol

1.5

0.97

242

commonly

LOX

JP-4

2.2

1.02

248

used

LOX

Turpentine

2.2

249

LOX

Ammonia

1.3

250

Hydrogen peroxide

Nitromethane

227

N 2 4

Hydrazine

249

N 2 4

Hydrogen

11.5

0.565

279

FNA

Hydrogen

12.6

0.60

298

LOX

Hydrogen

2.9

0.23

345

70% LOX \
30% ozone ]

JP-4

2.3

253

100% ozone

JP-4

1.9

266

100% ozone

Ammonia

1.13

267

100% ozone

Hydrazine

0.63

277

Fluorine

JP-4

2.6

265

Fluorine

Ammonia

2.6

288

So-called

Fluorine

Diborane
(B 2 H 6 )

5.0

291

"Zip-fuels"
(high energy

Fluorine

Methanol

2.37

296

fuels in

Fluorine

Hydrazine

1.98

298

the U.S.A.)

Fluorine

Hydrogen

4.5

352

Fluorine

Hydrogen

9.4

0.46

371

100% ozone

Hydrogen

3.2

369

100% ozone

Hydrogen

2.65

0.23

373

Tormey [72] quotes the following three examples of mixtures which give the
optimum conditions.

(1) Mixture giving the highest temperature:

C2N2 + O3 (Cyanogen + Ozone)

Flame temperature ca. 5240°C
A = 270

318

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(2) Mixture giving the lowest molecular weight:

H 2 + F 2 (Hydrogen + Fluorine)
Molecular weight 8.9
/ s = 373

(3) Mixture giving the lowest _£ ratio:

S v
H 2 2 + petrol

— = 1.20

S.
I s =248

With the reaction of free radicals or atoms much higher magnitudes of specific
impulse may be obtained.

Table 73
Reactions of hydrogen with free radicals

Reagents

Molar ratio
H 2 /R

Specific impulse

Hydroge

Radical R

sec

H 2

2.8

410

H 2

2.2

420

H 2

5.0

492

H 2

0.5

1040

1280

Still higher magnitudes of specific impulse can be obtained theoretically by using
non-chemical reactions, e.g. ions and electrons which arrive in an electric field at
a speed close to that of light. Another method is based on the use of photon flux
with the speed of light.

Warren [53] quotes the following figures (Table 74):

Table 74

Fuel

Specific impulse (sec)
max

Chemical

Free atoms and radicals

Ions and electrons (from a nuclear reactor)

Photons (from solar radiation)

400
1200
106
1010

FINAL REMARKS

Most of the propulsion systems recorded in Tables 73 and 74 are only suggestions
for the future or in development. Some of them would require chemicals which
are produced only on a laboratory scale. Others would need chemicals whose pro-
perties and methods of production are often insufficiently known.

This is why only a very limited number of liquid compositions is in use.

They are classified into storable and cryogenic liquids [71].

LIQUID EXPLOSIVES 319

Storable liquids

They remain liquid under normal, ambient operating conditions (moderate
temperature, atmospheric pressure). Safety in storage and handling should also
be considered. A storable liquid propellant should not have an excessively high
vapour pressure at ambient temperature. The leading storable propellant uses di-
nitrogen tetroxide as oxidizer and a 50/50 mixture of hydrazine-unsymmetrical
dimethylhydrazine (UDMH) as fuel [71]. Nitrogen tetroxide and UDMH is another
storable propellant mixture in use [73].

Cryogenic liquids

They require a number of additional facilities such as a liquefying plant. The
most common cryogenic propellant now in use is liquid oxygen-RPl (rocket petro-
leum No. 1, a kerosene-cut hydrocarbon fuel). The higher energetic system liquid
hydrogen-liquid oxygen is gradually being introduced.

Dole and Margolis [73] predicted that rocket propellants in use after 1961
would include :

Storable: perchloryl fluoride and hydrazine (7 S =268).

Cryogenic: liquid fluorine and hydrazine (/ s = 316) or liquid oxygen and liquid
hydrogen (/ s = 364).

LITERATURE

1. Sprengel, Brit. Pat. 921, 2424, 2642 (1871); /. Chem. Soc. 26, 796 (1873).

2. E. Turpin, Fr. Pat. 146497 (1881); 147676 (1882); Ger. Pat. 26936 (1882).

3. K. Olszewski and S. WrOblewski, Wied. Ann. [2], 20, 243 (1883).

4. C. Linde, Ger. Pat. 88824 (1895).

'5. C. Linde, Sitzungsber. Munch. Akad. Wissenschaft. 65 (1897).

6. H. Kast and P. Gunther, Z. ges. Schiess- u. Sprengstoffw. 14, 81 (1919).

7. P. Pascal, Explosifs, poudres, gaz de combat, Hermann, Paris, 1925.

8. Hellhoff, Ger. Pat. 12122 (1880); 17822 (1881).

9. M. A. Cook, D. A. Pack and W. A. Gay, Vllth Symposium on Combustion, p. 702, Butter-
worth, London, 1958.

10. L. Medard and R. Sartorius, Mem. poudres 32, 179 (1950).

11. L. Andrussow, Chimie et Industrie 86, 542 (1961).

12. R. Levy, Chimie et industrie 57, 221 (1947).

13. R. B. Canright, Ind. Eng. Chem. 49, 1345 (1957).

14. R. Grollier-Baron and G. Wessels, Mem. poudres 36, 285 (1954).

15. W. I. Fedosyev and G. B. Sinyaryev, Vvedenye v raketnuyu tekhniku, Oborongiz, Moskva,
1956.

16. M. Barrere and A. Moutet, Vth Symposium on Combustion, p. 170, Reinhold, New York,

17. M. Kilpatrick and L. L. Baker, Jr., Vth Symposium on Combustion, p. 196, Reinhold, New
York, 1955.

18. F. McCullough, Jr. and H.P.Jenkins, Jr., Vth Symposium on Combustion, p. 181, Reinhold,
New York, 1955.

320 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

19. B. Lewis, R. N. Pease and H. S. Taylor, Combustion Processes (Vol. II of High-Speed Aero-
dynamics and Jet Propulsion), Princeton University Press, Princeton, N.Y., 1956.

20. D. Altman and S. S. Penner, Combustion of Liquid Propellants, Princeton University Press,
Princeton, N.Y., 1956.

21. M. L. J. Bernard, Vth Symposium on Combustion, p. 217, Reinhold, New York, 1955.

22. F. Bellinger, H. Friedman, W.Bauer, J.Eastes and W.Bull, Ind.Eng. Chem. 40, 1320
(1948).

23. C. W. Tait, A. E. Whittaker and H. Williams, J. Am. Rocket Soc. 83 (1951).

24. W. S. Wood, Chemistry & Industry 1959, 136.

25. J. G. Tschinkel, Ind. Eng. Chem. 48, 732 (1956).

26. J. A. Hannum, U.S. Pat. 2537526, 2538516 (1951).

27. H. Behrens, Z. Elektrochem. 55, 425 (1951).

28. R. SCHWOB, Mem. poudres 32, 153 (1950).

29. N. S. Davis, Jr. and J. H. Keefe, Jr., Ind. Eng. Chem. 48, 745 (1956).

30. Ya. M. Paushkin, Khimicheskii Sostav i Svoistva Reaktivnykh Topliv, Akad. Nauk SSSR,
Moskva, 1958.

31. E. S. Shanley and F. P. Greenspan, Ind.Eng. Chem. 39, 1536 (1947).

32. W. Ley, Rockets, Missiles and Space Travel, Viking Press, New York, 1954.

33. e.g. B. E. Baker and C. Ouellet, Can. J. Research, 23 B, 167 (1945); P.A.Giguere, Can.
J. Research,2SB, 135 (1947); R. G. McKenzie and M. Ritchie, Proc. Roy. Soc. (London}
A 185, 207 (1946); C. N. Satterfield and T. W. Stein, Ind. Eng. Chem. 49, 1173 (1957).

34. R. R. Baldwin and L. Mayor, VIM Symposium on Combustion, p. 8, Butterworth, London,
1958; Trans. Faraday Soc. 60, 80, 103 (1960).

35. R. R. Baldwin and D. Brattan, VHIth Symposium on Combustion, p. 1 10, Williams & Wilkins,
Baltimore, 1962.

36. P. A. Giguere and I. D. Liu, Can. Chem. J. 35, 283 (1957).

37. W. Forst, Can. Chem. J. 36, 1308 (1958).

38. D. E. Hoare, J. B. Prothero and A. D. Walsh, Trans. Faraday Soc. 55, 548 (1959).

39. W. C. Schumb, C. N. Satterfield and R. L. Wentworth, Hydrogen Peroxide, Reinhold,
New York, 1956.

40. P. Walden and H. Hilgert, Z. physik. Chem. 165 A, 241 (1933).

41. V. I. Semishin, Zh. obshch. khim. 8, 654 (1938).

42. V. C. Bushnell, A. M. Hughes and E. C. Gilbert, /. Am. Chem. Soc. 59, 2142 (1937);
A. M. Hughes, R. J. Carruccini and E. C. Gilbert, /. Am. Chem. Soc. 61, 2639 (1939).

43. W. A. Roth, Z. Elektrochem. 50, 111 (1944).

44. J. C. Elgin and H. S. Taylor, /. Am. Chem. Soc. 51, 2059 (1929).

45. P. J. Askey, J. Am. Chem. Soc. 52, 970 (1930).

46. C. H. Bamford, Trans. Faraday Soc. 35, 1239 (1939).

47. R. R. Wenner and A. O. Beckmann, /. Am. Chem. Soc. 54, 2787 (1932).

48. Scott, Jones and B. Lewis, U.S. Bureau of Mines Report 4460, Washington, 1949.

49. M. Raciborski, Bull. Acad. Polon. Sci., Cracovie 733 (1906); J. Chem. Soc. (Abstracts) 92,
384 (1907).

50. L. F. Audrieth and B. S. Ogg, The Chemistry of Hydrazine, J. Wiley, New York, 1951.

51. A. S. Gordon, Illrd Symposium on Combustion, p. 493, Williams & Wilkins, Baltimore, 1948.

52. McLarren, Automotive and Aviation Ind. 95, 20, 76 (1946).

53. F. A. Warren, Rocket Propellants, Reinhold, New York, 1958.

54. G. M. Platz and C. K. Hersh, Ind. Eng. Chem. 48, 742 (1956).

55. J. Glassman and J. Scott, Jet Propulsion 24, 386 (1954).

56. E. Kruska, Z. VDI 27, 65, 271 (1955).

57. A. Stettbacher, Explosivstoffe 4, 25 (1956).

58. Fluorine Chemistry, Ed. J. H. Simons, Academic Press, Vol. I, 1950; Vol. II, 1954.

LIQUID EXPLOSIVES 321

59. A. Engelbrecht and H. Atzwanger, /. Inorg. Nuclear Chem. 2, 348 (1956).

60. R. L. Jarry, J. Phys. Chem. 61, 498 (1957).

61. Chemical Processing 21, 87 (1958).

62. M. A. Elliott and F. W. Brown, U.S. Bureau of Mines Report 4196, Washington, 1948.

63. P. A. JACQUET, Metal Finishing 47, 62 (1949).

64. L. Medard, P. A. Jacquet and R. Sartorius, Rev. met. 46, 549 (1949); L. Medard and
R. SARTORIUS, Mem. poudres 32, 179 (1950).

65. P. A. JACQUET, Compt. rend. 205, 1232 (1937).

66. M. E. Merchant, Metal Progress 37, 559 (1940).

67. E. KAHANE, Compt. rend. Ill, 841 (1948).

68. J. C. Schumacher, Perchlorates, Reinhold, New York, 1960.

69. W. Dietz, Angew. Chem. 52, 616 .(1939).

70. A. Latham, D. C. Bowersock and B. M. Bailey, Chemical & Engineering News 37, No. 31,
60 (1959).

71. U. S. in Space, Chemical & Engineering News 41, Sept. 23, 98; Sept. 30, 70 (1963).

72. J. F. Tormey, Ind. Eng. Chem. 49, 1339 (1957).

73. S. H. Dole and M. A. Margolis, in The Chemistry of Propellants, p. 1, Ed. S. S. Penner
and J. Ducarme, Pergamon Press, Oxford, 1960.

CHAPTER III

BLACKPOWDER

HISTORICAL

The forerunner of all modern explosives, blackpowder, formerly often called
gunpowder, is a mixture of potassium nitrate (saltpetre), sulphur and charcoal.
The origin of blackpowder is obscure and dates back to very remote times. According
to numerous historical works, in particular that of Romocki [1] blackpowder was
invented by the Chinese many centuries B.C. The secret of its manufacture penetra-
ted from there to Central Asia and was brought to Europe by the Arabs about
the middle of the thirteenth century.

Combustible mixtures containing saltpetre, such as, for example, the famous
Greek fire with which the Greeks destroyed the Arab fleet besieging Constantinople
in 668, were already fairly widely known at that time. The secret of preparing Greek
fire was supposed to have been brought to the Byzantine capital by Kallinikos,
a Greek architect from Heliopolis of Syria. The Arab fleet was twice more defeated
with this weapon in 716 and 718. The secret of Greek fire has never been disclosed
in full but some medieval manuscripts reveal that it was a mixture containing con-
stituents of blackpowder such as saltpetre and sulphur mixed with pitch. Obviously,
this was not blackpowder but a mixture akin to it. The possibility that firearms
were used for launching the incendiary missiles with a propellent charge can be
ruled out since there is no doubt that catapults were used for throwing Greek fire.

Similarly, in a description of the siege of Niebla in Spain (1257) mention is
made of the missiles thrown by the Moslems which produced a roar and flash.
They were in all probability loaded with a mixture resembling blackpowder. At
the battle of Legnica (1241) the Tartars employed another weapon, the so-called
Chinese dragon belching fire; this was probably a type of a rocket-like launching
device for incendiary missiles.

In the book by Marcus Graecus "Liber ignium" translated from Arab sources
and published ca. 1300, there is a fairly full description of the composition of a com-
bustible mixture called "flying fire" (ignis volatilis):

1 part of resin

1 part of sulphur

2 parts of saltpetre

^ ... ■ . 13221

BLACKPOWDER 323

The mixture was kneaded with linseed oil and loaded into a piece of hollow
wood, and together these constituted an incendiary missile.

A description of the composition and principles of the manufacture of black-
powder appeared in the works of two of the greatest scientists of the Middle Ages :
Albertus Magnus (Saint Albert the Great), a Dominican Monk born in Bavarian
Swabia near 1200, and Roger Bacon of the Franciscan Order, born, according
tradition, about 1214 at Ilchester in Somerset, England.

Albertus Magnus gave a description of blackpowder in his manuscript "De
Mirabilibus Mundi".

As early as 1249 Roger Bacon described blackpowder in his manuscripts "De
Secretis" and "Opus Tertium" and gives its composition as follows:

41 parts of saltpetre
29.5 parts of charcoal
29.5 parts of sulphur

It was not until the invention of firearms that the manufacture and use of black-
powder really began to develop. This invention cannot be ascribed with certainty
to any particular person, but recent historical research has shown that there is
no foundation for the belief that Berthold Schwarz was the inventor. Arab manu-
scripts written as early as 1320 (e.g. the manuscript in the Leningrad Asiatic Museum
by Shems ed Din Mohammed) show tubes employed for shooting balls by means
of a charge of powder. It is also known that in 1326 the Republic of Venice ordered
firearms and in 1331 cannons were used by the Moors during the siege of Alicante
in Spain. Cannons were also employed on a wide scale both by the English and
the French at the battle of Crecy in 1346. In Poland the first mention of the use of
powder and firearms is to be found in the Statute of Wislica promulgated by Casimir
the Great in 1347 and afterwards, in the description of the battle of Grunwald in
1410, when the Poles employed over 60 guns. In Russia powder and firearms ap-
peared during the reign of Duke Dymitr of Don, in the late fourteenth century. The
first references appear in chronicles of 1382 and 1389.

In all probability the first "mills" for making powder in Europe were established
at Augsburg (1340), Spandau (1344) and Legnica (1348).

The composition of the powder used in the fourteenth century was:

67 parts of saltpetre
16.5 parts of charcoal
16.5 parts of sulphur

A very detailed study of the history of Greek fire and blackpowder was publish-
ed more recently by Partington [2]. The book, which is amply provided with full
quotations from the original source is the most authoritative and extensive work
on the subject.

Blackpowder was also used as a high explosive. According to Kochmyerzhevskii
[3] blackpowder was first employed in civil engineering between 1548 and 1572
for cleaning and dredging the Niemen river-bed. In 1627, in Hungary, Kacper
Weindl employed blackpowder in a mine for blasting hard coal.

324 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The first published reference to the use of blackpowder in mining was the paper
read to the Royal Society in London by Sir R. Moray in 1665 [4]. He referred to
"a way to break easily and speedily the hardest rock". According to the same paper
the method was invented by du Son in France.

J. Taylor [5] reports that in 1696 blackpowder was utilized to widen a road
in Switzerland.

From that time the use of explosives for various engineering operations such
as mining, road building, dam building, land improvement etc. became general.

According to Gorst [6] blackpowder is now employed for the following pur-
poses:

(1) as the filling for time-trains in time fuses,

(2) in the manufacture of shrapnel shells to fire the charge that expels the bullets,

(3) as a bursting charge in incendiary and star shells,

(4) in the manufacture of delay pellets and boosters,

(5) in the manufacture of powder pellets for primers,

(6) in the manufacture of primers for igniting charges of smokeless powder and
of pyrotechnic mixtures, ,

(7) in the manufacture of safety fuses in which the cores consist of blackpowder.
In addition, blackpowder is also used (although the practice is decreasing) in

sporting cartridges, in opencast coal mining and for blasting in mines where no
methane occurs.

D. A. Davies [7] suggested the use of explosive charges for rain-making. The
charges, consisting of 15 g of blackpowder plus 1.5% of silver iodide, are sent by
balloon into a cloud, where they are exploded by a time fuse. The particles of silver
iodide thus released act as nuclei on which the water vapour in a raincloud coagulates,
to fall as rain drops.

Blackpowder containing silver iodide is obtained by saturating the black-
powder with an acetone solution of potassium silver iodide and then drying it. The
required solution is prepared by dissolving 15 g of potassium iodide and 50 g of silver
iodide in 200 g of acetone.

In Great Britain and throughout the Commonwealth blackpowder was used to
prove alcohol. In this test alcohol was poured upon a small heap of blackpowder
or "proof powder" and a light was applied to burn the heap. If the mixture burnt
with explosion it was overproof, if it did not burn or burned with difficulty, the
alcohol was underproof. If the mixture burned with "slight" explosion it was proof
spirit, i.e. containing 49% alcohol by weight (Tate [8]).

COMPOSITION OF BLACKPOWDER

The blackpowder now most commonly used is composed of:

KN0 3 75%
Charcoal 15%
Sulphur 10%

BLACKPOWDER

325

Charcoal

Charcoal here implies a component with variable properties, not a specific
chemical. Hence depending on the method by which the charcoal is prepared powder
with various properties could be obtained.

As early as 1848 this attracted the attention of Violette [9], who prepared different
types of charcoal in a retort by employing different temperatures of carbonization
(Table 75).

Table 75

Effect of the temperature of carbonization on the chemical composition

of charcoal

Temperature
of

Colour

of
charcoal

Yield

Composition of charcoal (%)

carbonization

°C

1
H 1 O+N

280-300
350^00

1000

1250

brown
black
black
black

28-31

73.2

77-81

82.0

88.1

4.3

2.3
1.4

21.9

14.1
9.3

Violette found that the temperature of carbonization of wood is directly related
to the ignition temperature of the charcoal obtained (Table 76).

Table 76

The relation between the ignition temperature of charcoal and the
carbonization temperature of wood

Temperature

Ignition temperature

of carbonization

of charcoal

°C

260-280

340-360

290-350

360-370

432

ca. 400

1000-1500

600-800

The ignition temperature of charcoal obviously influences the ignition temper-
ature of the blackpowder. Blackpowder containing "cocoa-", red- or brown-coloured
charcoal is most easily ignited, while the heat of combustion of black charcoal, more
strongly carbonized, is higher than that of brown coloured material. Thus the
power of powder containing black charcoal is greater.

Strongly carbonized (black) charcoal is less hygroscopic, hence the powder
from which it is made is also less hygroscopic.

326 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The temperature and ease of ignition of charcoal is influenced not only by the
method of carbonization, but also by the type of wood used. A light, porous wood
gives more easily ignitable charcoal.

In modern times, the influence of the type of charcoal on the burning of powder
has been investigated by T. Urbahski and T?siorowski [10] and Blackwood and
Bowden [11].

Much research has been devoted to the chemical structure of charcoal and has
generally been concerned with comparing its chemical properties with those of coal
in studying the latter's structure.

As early as in 1869 to classify various coals M. Berthelot [12] treated them with
oxidizing agents such as nitric acid, potassium chlorate etc. and obtained a number
of organic acids.

Next, Dickson and Easterfield [13] oxidized charcoal with fuming nitric acid in
the presence of potassium chlorate and obtained mellitic acid in a yield of about
20% by weight. Finally, Dimroth and Kerkovius [14] made a very important ob-
servation: they converted the products of the oxidation of charcoal with nitric acid
into barium salts and in turn subjected them to decarboxylation by distillation with
barium hydroxide. In the products they detected the presence of benzene, naphtha-
lene and fluorene. The presence of naphthalene and fluorene was the only direct
evidence of the existence of condensed aromatic rings in charcoal. On the oxi-
dation of charcoal with nitric acid in the presence of potassium hydrogen sulphate
Meyer and Steiner [15] obtained pyromellitic acid. The presence of pyromellitic
acid is considered to be indirect evidence of the existence of condensed aromatic
rings in charcoal, which constitute an essential part of the chemical structure of
charcoal. Nevertheless some authors questioned this view. Philippi et al. [16, 17]
carried out experiments in which they treated charcoal with sulphuric acid at 300°C
and obtained pyromellitic acid in a yield of 1-2%. Later they increased this yield to
6-7 %, by conducting the reaction in the presence of mercury. They did not acknow-
ledge this as evidence of the existence of strongly condensed aromatic rings since
they had also prepared polycarboxylic aromatic acids by heating such aliphatic
compounds as cellulose with sulphuric acid under similar conditions. This opinion,
however, is hardly convincing since aromatic rings might have been formed on
carbonization, in the course of the treatment with sulphuric acid.

In more recent times Juettner [18] has worked extensively on the oxidation of
various types of coal, including charcoal. He examined the action of various oxidiz-
ing agents such as potassium permanganate in an alkaline medium and nitric acid.
On such oxidation of the charcoal obtained by carbonization of cellulose at 1000°C
mellitic acid in a yield of 25% by weight resulted.

Under similar conditions fluorene gives mellitic acid in a yield of about 45%,
while hexaethylbenzene yields almost exclusively carbon dioxide, without mellitic
acid. The experiments outlined above point to the existence of condensed aromatic
rings in charcoal. This structure of charcoal has been definitely proved by infra-red

BLACKPOWDER

32?

spectroscopic analysis carried out by T. Urbanski, Hofman, Ostrowski and Wita-
nowski [19]. These authors investigated the infra-red absorption spectra of the
products of thermal decomposition of cellulose and lignin at temperatures from
200-575°C. They showed that heating cellulose to a temperature above 300°C
involves breakdown of the aliphatic structure of cellulose which is replaced by.

Lignin (15mm Hg)

4000 3000 2000

(a)

WOO cm"

4000

3000 2000
(b)

1000 cm'

Fig. 86. Thermal decomposition of cellulose (a) and lignin (b) [19].

condensed aromatic rings. Thus the band 3300 car 1 of the alcoholic group (bound
by a hydrogen bond) weakens with a rise in temperature and disappears at a tem-
perature above 370°C. The band 1640 cm~i (derived from the water present in
cellulose) weakens with a rise of temperature and disappears at a temperature above
245°C.

Cellulose shows the presence of a number of acetal bands (1155, 1105 and,
1025 cm-i) which disappear in test samples heated to higher temperatures (370°C
or above). Similarly, a band characteristic of the bond C— O between carbon and,
hydroxyl group weakens with thermal decomposition and disappears at a tem-
perature above 370°C.

In test samples heated to above 300°C there appear bands characteristic of,
the aromatic system, e.g. 1600-1570 cm- 1 (aromatic ring vibrations). Above 400°C
the bands appear at 870 and 800 cm-i, characteristic of C— H vibrations in con-
densed aromatic systems.

328 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The changes in cellulose that occur on heating to various temperatures are shown
in Fig. 86. The upper curves of cellulose carbonized at a high temperature (above
500°C) resemble closely those of anthracite or even graphite.

The graph showing the decomposition of lignin is similar, differing from cellulose
in that here the change at a temperature of about 350°C is less pronounced since
the lignin itself contains aromatic rings in its molecule. Above 350°C the bands
characteristic of the aliphatic part of the lignin molecule fade, i.e. the frequencies
of alcohol and phenol groups are 3300-3200 cm- 1 , C— O of phenol 1265 cm- 1 ,
C— O of alcohol and ether 1140, 1075 and 1030 cm- 1 . Here also the general shape
of the curves for most strongly carbonized material approximates to those charac-
teristic of anthracite and graphite.

TYPES OF BLACKPOWDER

There are in fact two types of blackpowder classified according to their intended
use:

I— for filling fuses,

II— for blasting charges.

With regard to chemical composition the difference between the two types is
insignificant and lies mainly in their different rates of burning. In powders belonging
to type II the rate of burning should be as high as possible. This is achieved by
making the grains of powder of the density as low as possible.

The types of powder also differ in grain size. The powder used in small arms
■was manufactured in small grains, that for cannons being much larger. Sporting
powder, for instance, (cf. Table 79) was manufactured in France in several grades
differing in size of grain. Powder No. consisted of the largest grains, powder
No. 4 of the smallest, size being classified by the number of grains per gramme.

No. 650-950 grains per g
No. 1 2000-3000 grains per g
No. 2 4000-6000 grains per g
No. 3 8000-10,000 grains per g
No. 4 20,000-30,000 grains per g

Blasting powder consists either of grains with a density of about 1.8 or of com-
pressed cylindrical pellets with a density of 1.35-1.50 and a central perforation.
Blasting powder must burn vigorously so as to give an effect as close as possible
to detonation. It is ignited either by a safety or detonating fuse. The explosive effect
of the latter is stronger, and this may be enhanced by the "Herco blasting method"
employed by Hercules Powder Co. [20], in which a detonating fuse inserted into
the central perforation of the pellets acts as an initiating agent throughout the
charge.

The composition of the powders used for mining purposes is given in Tables
77 and 78.

BLACKPOWDER

329

Mixtures with the addition of sodium nitrate (Table 78) are somewhat stronger
giving more heat and a larger gas volume than those with potassium nitrate. They
are more difficult to ignite and burn more slowly than mixture with potassium
nitrate.

Table 77

Composition of blasting powders

Name

Composition (%)

KNOj

Strong blasting (French "poudre forte"

in the form of globules or grains)
Slow blasting (French "poudre lente" in

the form of globules or grains)
No. 1 blasting powder (Germany and

Poland, 1924)
No. 1 Bobbinite with 2.5-3.5% paraffin

(Great Britain)
No. 2 Bobbinite with 7-9% starch (Great

Britain)

40
73-77
62-65
63-66

Sulphur

8-15

1.5-2.5

Charcoal

10-15

17-19.5

18.5-20.5

(NH 4 ) 2 S0 4
and CuSQ 4

13-17

Table 78
Composition of blasting powders with sodium nitrate

Name

Composition (%)

NaNQ 3

KNO3
(instead of
NaNQ 3 )

Sulphur

Charcoal

No. 1 black blasting powder (Germany

and Poland, 1924)
American blasting powder*
Petroclastite or Haloclastite, No. 3 black

blasting powder

No. 2 black blasting powder

70-75
70-74

71-76

70-75

up to 25

up to 5
up to 5

9-15
11-13

9-11

9-15

10-16
15-17

15-19 of
coal-tar pitch
10-16 of
lignite

CC a' ?^Srt££?r, UAA - "* mafked aCC ° rding t0 the dtaensions of graphitized grains: CCC (ca. 14 mm dia.),
— «-, v., r, rr, ft t , ttFF (1-2 mm dia.).

The mixtures recommended by Raschig [21] similar to those suggested earlier
122, 23] are also worthy of note. They consist of 65-70% sodium nitrate with 30-35%
sodium benzenesulphonate or sodium cresolsulphonate, or xylenesulphonate etc.
Mixtures of this type have long been used in mining, under the name of Raschit
or White Powder. The inventor claimed that the advantage of such mixtures is their
salety m manufacture. They were prepared by evaporating an aqueous solution
ot the ingredients to dryness. Similar mixtures according to the invention of Raschig

330

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

[24] were made at Pniowiec in Upper Silesia in 1912. They consisted of sodium or
potassium nitrate (ca. 70%) and of sulphite liquor evaporated to dryness (ca. 30%).
Almost simultaneously Voigt [25] recommended using mixtures of nitrates (mainly
sodium nitrate) with salts of nitrophenolsulphonic acids (e.g. mononitrophenol-
and mononitrocresolsulphonates) or salts of mononitronaphthalenesulphonic acid.
He proposed mixing 25% of the aqueous solution with 75% of sodium nitrate.

Table 79

Composition of powders used for military purposes

Composition (%)

Form and size

Name

KN0 3

Sulphur

Charcoal

of grains

Cannon
Sporting
"Normal"*
Cannon modified

75
78
75
78

12.5

12.5
12

7-21 mm
0.1-1 mm
various
prisms (p. 360)

Delay fuse powder

I 75

10-12

13-15

grains 0.3-0.6 mm
4000-7000 per g

* Most commonly used for fire-arms e.g. as rifle powder.

The composition of blackpowder used in France for military purposes is given in
Table 79. In many cases they have retained their traditional names in spite of altered
application.

In the U.S.S.R. blackpowders for military purposes have a conventional com-
position: 75% KNO3, 15% charcoal and 10% sulphur. Their grades differ in grain
size. No. 1 cannon and rifle powder has large grains, No. 2 shrapnel of rifle powder
has smaller ones, the finest being No. 3 rifle powder. In addition sulphurless powder
composed of 80% KN0 3 and 20% charcoal is used in the U.S.S.R. as a priming
composition.

Blackpowder for small rockets (chiefly sea rescue signal rockets) is characterized

by its high content of charcoal.

It consists of:

German
rockets
Potassium nitrate 60%

Charcoal 25%

Sulphur 15%

American

rockets

59%

31%

10%

MODIFIED BLACKPOWDER

Prior to the invention of smokeless powder various attempts had been made
to improve blackpowder. In particular experiments were carried out: to obtain
sulphurless powder, ammonium powder (with ammonium nitrate instead of potas-

BLACKPOWDER 331

sium nitrate), to replace potassium nitrate with potassium chlorate and, finally, to
introduce a powder in which potassium or ammonium picrate was the combustible
(and explosive) constituent instead of charcoal and sulphur.

Sulphurless powder

Andrew Noble [26] suggested the omission of sulphur in powder mixtures. He
found a mixture of 80% KN0 3 and 20% charcoal to be somewhat stronger than
blackpowder of normal composition, i.e. containing 10% sulphur. At Noble's
suggestion sulphurless powder was introduced into England. This type of powder
consisting of 80% KN0 3 and 20% charcoal is now used in Great Britain and the
U.S.S.R. in igniters for firing pyrotechnic mixtures,

A stoichiometric mixture of potassium nitrate and charcoal contains 87.1%
potassium nitrate and 12.9% charcoal. The decomposition is expressed theoretically
by the following equation:

4KN0 3 + 5C— >2K 2 C0 3 + 2N 2 + 3C0 2

In this reaction 779 kcal/kg is evolved, the gas volume V amounts to 240 l./kg,
and the calculated temperature is 2700°C. For mixture consisting of 70% potassium
nitrate and 30% charcoal (Sulphurless Gunpowder SFG.12) the heat of reaction
is 670 ± 20 kcal/kg, according to Thomas [27].

Langhans [28] found that a mixture in which sulphur is replaced by selenium in
no way resembles blackpowder. It is slow-burning and has no practical applica-
tion.

Ammonium powder

Ammonium powder (formerly termed amide powder) was obtained in attempts
to increase the power of blackpowder. One of the causes of the relatively low power
of blackpowder lies in the fact that on explosion it produces a great quantity of
solid matter, but only a relatively small volume of gas. In working for an increase
of the volume of the gaseous products in the middle of the nineteenth century,
attempts were made to substitute ammonium nitrate for potassium nitrate. The
observation of Reiset and Millon [29] that a mixture of ammonium nitrate and
charcoal has explosive properties and explodes on being heated to 170°C was the
starting point for this work.

Gaens [30] obtained amide powder with the following composition:

KN0 3 40-45%
NH4NO3 35-38%
Charcoal 14-22%

This proved to be a more powerful explosive than blackpowder and burned with
less smoke. Shortly afterwards it was used in Germany for small calibre naval guns.

332 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Next, a similar powder was used for naval guns of larger calibre; this achieved
great success as a greater muzzle velocity was obtained with a smaller charge of
powder. Powder of this type was employed in Great Britain under the name of
Chilworth Special Powder.

Ammonium powder obtained from ammonium nitrate and charcoal was also
used successfully in guns of the Austrian navy between 1890-1896. It had however
many disadvantages, since it was difficult to ignite, gave uneven results and was
highly hygroscopic. This limited its use until it was withdrawn in all countries.

It was not until World War I that, due to the shortage of cellulose and acids
for nitration, an ammonium powder with a composition of:

NH4NO3 85%
Charcoal 15%

was used for filling cartridges for Russian and German field artillery. In Russia it
was termed SUD (Selitrougolnyi Dobavok) and in Germany Ammonpulver. The
powder was manufactured in the form of rings with the outer diameter equal to
the inner diameter of the cartridge case.

The main new feature introduced here was the use of combined charges i.e.
of nitrocellulose and ammonium powder, the latter constituting 1/3 (exceptionally
1/2) of the whole charge. Here, nitrocellulose powder not only played the role of
a propellent charge, but also acted partly as a secondary primer regulating the
burning of the ammonium powder. The greatest disadvantage of ammonium powder,
viz. the difficulty of ignition and burning, was thus overcome. It also turned out
that ammonium powder possesses the advantages of being only slightly erosive and
producing only slight muzzle flash. The hygroscopicity of ammonium powder was
overcome by employing a special packaging with hermetically sealed cartridge
cases.

Ammonium powder has the important disadvantage that ammonium nitrate
occurs in various crystallographic forms with different specific gravities,. one of the
transitions taking place at + 32°C, i.e. only just above room or summer tempera-
ture (Vol. II). When a charge of ammonium powder is heated to this temperature
a decrease in density occurs and in consequence there is an undesirable increase in
rate of burning.

During World War I 3000 tons of ammonium powder was produced monthly
in Germany.

Towards 1934 a number of patents were filed on mixtures approximating in
composition to ammonium powder and containing ammonium nitrate and gua-
nidine nitrate or nitroguanidine with the addition of ammonium bichromate. In
many cases gelatinized nitrocellulose was added as a binding agent. Compressed
charges of these mixtures had a low rate of burning and a low calorific value. They
could be applied to small jet engines, e.g. in World War II they were recommended
for torpedo propulsion and for driving certain engines and various mechanical
devices. These will be dealt with in later sections.

BLACKPOWDER 333

As a result of this work general conclusions were made as to the value and signi-
ficance of certain constituents.

It was established that the addition of kaolin to mixtures containing ammonium
nitrate raises the level of pressure at which burning is stabilized and facilitates
uniform burning of the mixtures.

The presence of guanidine nitrate proved to be preferable to that of nitroguani-
dine since the former permits the charges to be compressed to a great density.

2,4-Dinitroresorcinol proved to be an important constituent of these mixtures
as an sensitizing agent i.e. facilitating uniform burning under low pressure.

Asbestos, which increases the burning surface of powder mixtures, is also added
and a small amount of vanadium pentoxide which exerts an advantageous influence
on the uniformity of burning of a mixture.

By 1949, pressed charges began to come into use for a variety of purposes such
as operating reciprocating engines, turbostarters for aircraft, starter motors for
Diesel engines, etc. For example, a typical solid gas-generating charge for the "Wil-
liams and James" motor, designed for starting bus engines, was made from com-
pressed pellets: 10 g of guanidine nitrate plus 15% 2,4-dinitroresorcinol, with a
burning time of approximately 1 sec [5].

A similar mixture was used for starting the turbo starters of Armstrong-Siddeley
Sapphire turbojet engines. The charge weighed 5 lb and burned one-sidedly from
the central hole. The burning time was 10 sec, and the pressure developed was
750 lb/in2.

A mixture of similar composition in the form of pellets weighing 12 g each and
burning "cigar" fashion was introduced in 1949-1950 to start up 2 gallon fire ex-
tinguishers manufactured by General Fire Appliance Company. The burning time
of the charge was 1 min. A charge weighing 1.3 lb and burning for 3 min was also
•designed for a 34 gallon extinguisher [5].

A different type of ammonium powder containing ammonium bichromate was
also used to activate the "Williams and James" reciprocating engine.

This was composed of:

NH4NO3 78.5%

KNO3 9.0%

Ammonium oxalate (anhydrous) 6.9%
Ammonium bichromate 5.6%

China clay 0.7%

The 151b charge was pressed in increments, under a pressure of 5550 lb/in 2
•(370 kg/cm 2 ) into an internally-insulated steel tube so that it formed a continuous
solid column. The charge was ignited by means of an electric powder fuse com-
prising a small charge of ca. 0.32 g (5 grains) of blackpowder. The gases drove the
engine for 3 min 20 sec. The operating gas pressure rose to 280 lb/in 2 .

For larger engines a charge of 149 lb was pressed at 12,000 lb/in 2 into a contain-
er 17 in. dia. The gases drove the engine for 2 min at an average pressure of 530
ib/in 2 .

334 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The same composition was used for driving a rotary blower motor. For this
purpose an 8 lb charge was pressed into a 3.5 in. internal diameter steel tube lined
with insulating material.

Another charge used for engine-starting had the composition :

Nitroguanidine 56.0%

Guanidine nitrate 28.0%

Ammonium chromate 8.0%
Dimethyl diphenyl urea 4.0%
Beech charcoal 4.0%

8 lb of this powder, pressed under a pressure of 6000 lb/in 2 (400 kg/cm 2 ), gave
a charge that ran the "Williams and James" engine for 3 min 20 sec at a pressure rising
slowly from 100 to 190 lb/in 2 . The gas temperature leaving the charge tube was
600°C and at the engine inlet 340°C.

Another composition was :

Guanidine nitrate 94.5%

Vanadium pentoxide 0.5%
Cuprous oxide 5.0%

This charge was ignited by ca. 0.32 g (5 grains) of blackpowder igniter and a
nitrocellulose-red lead oxide-silicon mixture.

For more information see J. Taylor [5].

A quasi-powder consisting of ammonium nitrate, guanidine nitrate and nitro-
guanidine with an admixture of potassium or ammonium chromate was employed
as an agent for dispersing pesticides. This mixture burns slowly evolving a relatively
small amount of heat. When mixed with a pesticide in equal proportions, the
smoke formed on burning is rich in the pesticidal substance. Gammexane (BHC)
DDT, Parathion, azobenzene etc. may be dispersed in this way (Marke and Lilly
[31]).

Recently, mixtures containing ammonium nitrate have been recommended as
materials for rocket propulsion (p. 383).

Chlorate powder

The first attempts to substitute potassium chlorate for potassium nitrate under-
taken by Berthollet [32], who discovered potassium chlorate, were unsuccessful.
It was immediately evident that a mixture of potassium chlorate, sulphur and char-
coal was exceptionally sensitive to impact and friction, and was therefore too dan-
gerous to manufacture. The removal of sulphur from the mixture did not increase
safety and thus the powder has never found practical application.

Picrate powder

In 1861 Designolle [33] suggested using potassium nitrate mixed with metal
picrates, mainly potassium picrate, as a propellant substitute for ordinary black-

BLACKPOWDER 335

powder. It was manufactured on a rather large scale in Le Bouchet in France and
used during the Franco- Prussian War of 1870-71.

Blackpowder and small arms powder contained 9-16% and ca. 23% potassium
picrate, respectively.

In 1869 Brugere [34] and independently F. Abel [35] suggested the use of
ammonium picrate in a mixture with potassium nitrate instead of blackpowder.

By employing virtually the same procedure as in the manufacture of black-
powder grains were obtained and then slightly polished. When examined in France
in 1881 this powder gave good results for use with rifles and cannon, considerably
surpassing blackpowder in strength. Brugere powder (Poudre Brugere) consisted of:

Ammonium nitrate 43 %
Potassium nitrate 57%

The new propellant was promising but the nitrocellulose smokeless powder
invented soon afterwards superseded all mixtures containing potassium nitrate and
similar salts, that give a number of solid particles when exploded. For a time in
the United States various mixtures were still used instead of blackpowder— chiefly
for sporting purposes. E.g. Gold Dust Powder (Starke [36]) consisted of 55% am-
monium picrate, 25% potassium picrate and 20% ammonium bichromate. Soon,
however, early in the nineteenth century, the use of these mixtures was discontinued.

In World War II a variety of picrate powder consisting of ammonium picrate,
potassium or sodium nitrate and a binding agent was introduced in Great Britain
and in the U.S.A. as the propellant charge for small rockets. This was reported
more fully in a chapter devoted to mixtures for rocket propulsion (p. 365).

THEORY OF THE BURNING OF BLACKPOWDER

For a long time attempts have been made to explain why three non-explosive
substances, viz. potassium nitrate, charcoal and sulphur, when combined together
should form an explosive mixture. It was particularly incomprehensible that binary
mixtures of potassium nitrate with charcoal or with sulphur should be non-explosive
or only poorly so. The problem was the more difficult to elucidate since it involves
a reaction in the solid phase.

A number of outstanding scientists such as Descartes (1644) [37], Newton (1705),
Lomonosov (1750) endeavoured to explain the course of the reaction that occurs
when blackpowder burns. Both these and later investigators tried to analyse the
explosion products of blackpowder and to derive equations for the decomposition
process.

Gay-Lussac [38] found that the gases were composed of 52.6% C0 2 , 5% CO
and 42.4% N 2 , their volume, when the density of the powder was 0.9, being 450
times the volume of the explosive. Piobert [39] disagreed with these results, finding
a much lower value for the gas volume which he asserted was 250 times the volume
of the charge.

336 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Chevreuil [40] reported that in the bore of a gun barrel the powder decomposes
according to the equation :

2KN0 3 + S + 3C = K 2 S + N 2 + 3C0 2

If the powder burns in the open potassium sulphide is oxidized to sulphate.
When blackpowder burns slowly the products, apart from carbon, include such
components as potassium sulphide, sulphate, carbonate, cyanide, nitrate and nitrite.

A classical paper on the composition of the explosion products of blackpowder
and of the heat of reaction was published by Bunsen and Shishkov [41]. They ascer-
tained that the gases formed constitute 31% of the charge and contain approxi-
mately 50% C0 2 , 40% N 2 , 4% CO and lesser amounts (0.5-1.5%) of H 2 , 2 , H 2 S.
Solid products consist of potassium carbonate, sulphate, thiosulphate, sulphide and
nitrate with traces of potassium rhodanate, sulphur and carbon. These authors
also detected the presence of ammonium carbonate.

These investigations were repeated by Linck [42], Karolyi [43] and Fiodorov
[44]. Fiodorov's works are orginal in character: he examined the composition of
the solid products of the explosion of blackpowder in a pistol or gun barrel, and
arrived at the conclusion that the composition of these products depends on the
conditions of firing, for instance on the calibre of the gun or pistol. He also estab-
lished that the primary solid products of explosion consist of potassium sulphate
and carbonate, which then undergo reduction under the influence of the excess
carbon.

Extensive work on the products of the explosion of blackpowder in a confined
space were carried out by Andrew Noble and F. Abel [45]. They showed that there
is considerable variability in the composition of the products, depending on the
conditions under which the powder explodes.

On the basis of Bunsen and Shishkov's investigations, Berthelot [46] derived
the following equation for the decomposition of the powder :

16KN0 3 + 6S + 13C = 5K 2 S0 4 + K 2 S + 6N 2 +11C0 2

A further advance was the development of the first theory about the explosion
of blackpowder by Berthelot, who for this purpose drew largely on the experimental
work described above.

Berthelot assumed two limiting cases for the decomposition of powder :

(1) With the formation of K 2 C0 3 as a chief product of decomposition and
of K 2 S0 4 as a by-product;

(2) With the formation of K 2 S0 4 as a chief product of decomposition and
of K 2 C0 3 as a by-product.

In the first case the decomposition proceeds according to the following three
equations : *

2KN0 3 + 3C + S = K 2 S + 3C02 + N 2 (1)

2KN0 3 + 3C+S = K 2 C0 3 + C0 2 +CO + N 2 + S (2)

2KNO 3 + 3C+S = K 2 CO3 + 1.5CO2+0.5C+S + N2 (3)

BLACKPOWDER

so that i of the powder decomposes according to equation (1), a half according to
equation (2) and the rest (£) according to equation (3).

In the second case the decomposition runs according to the four equation,
namely (1), (3), (4) and (5) so that * of the powder decomposes according to equation
(1), a half according to equation (3), -I according to equation (4) and the rest f-U
according to equation (5). 24

2KN0 3 + 3C + S = K 2 S0 4 + 2CO + C+N 2 (4)

2KN0 3 + 3C + S = K 2 S0 4 + C0 2 + 2C + N 2 ,«

Debus [47] came to the conclusion that the burning of blackpowder is a two
stage process. In the first oxidation occurs according to the exothermic reaction:

10KNO 3 + 8C+3S = 2K 2 CO 3 + 3K 2 SO4+6CO 2 +5N 2 + 979kcal

The products so formed are then reduced according to the following endothermic
reactions :

K 2 S0 4 +2C=K 2 S+2C0 2 -58 kcal
C0 2 +C=2CO-38.4kcal

Potassium sulphide so formed may undergo further reactions, viz.:
K 2 S + C0 2 + H 2 = K 2 C0 3 + H 2 S
K 2 S+C0 2 +i0 2 =K 2 C0 3 +S
A part of the unburnt potassium sulphide and sulphur gives K 2 S
Much later K. A. Hoffmann [48] resumed work on the mechanism of explosion
of blackpowder On examining the behaviour of mixtures of charcoal with sulphur
he found that the reaction in blackpowder starts abcva the melting point of sulphur
at approximately 150°C, with a reaction between hydrogen present in charcoal
Zoi P u r " Hydr ° gen sul P hide thus fo ™ed reacts at temperatures between 285-

th.VMn ^"TT nitfate t0 yidd K2S °- Heat is then emitted which causes
the KN0 3 to melt. This is an essential moment in the overall reaction since molten
saltpetre reacts with molten sulphur and with carbon. The reaction proceeds the

tTaYnv * fl^^ mdting P ° int ° f the Sdt P etre ' so that a P^der with
the add-on of NaN0 3 (m.p. 313°C) ignites and burns more readily than a mixture

^ t Z 3 ,° ne (m ' P - 34 °° C) - Black P°wder containing a mixture of potassium

a h lr ™ S mtrate ignitCS Sti11 m ° re readil y since a eute ^ic mixture of KN0 3

and NaN0 3 melts at a temperature of about 220°C.

To prove the soundness of his theory Hoffmann refers to the fact that black-
powder 1S relatively difficult to ignite (as compared with nitrocellulose, for instance),
not taking fire even from a strong electric spark but only on being heated to the
temperature at which the above described reactions begin to take place

Hoffmann also performed many experiments on the significance of sulphur in
blackpowder mrxtures (this problem was previously raised by Andrew Noble r491>
to substantiate his statement that blackpowder is well fitted for use when composed
only of potassium nitrate and charcoal, hence without sulphur or with a smaller

338 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

content of sulphur, e.g. 2%. From these experiments he drew the following conclu-
sions:

(1) Sulphur facilitates an increase in the quantity of gases evolved on explosion.
In the absence of sulphur potassium nitrate with carbon gives only K 2 C0 3) but
in the presence of sulphur C0 2 is evolved and potassium forms K 2 S0 4 and K 2 S.

(2) Sulphur reduces the initial decomposition temperature. For instance a mix-
ture of two moles of KN0 3 and three gramme-atoms of carbon (in the form of
71 % charcoal) begins to decompose at 320°C and explodes at 357°C, while a mixture
of two moles of KN0 3 and one gramme-atom of sulphur begins to decompose at
3 10°C and explodes at 450°C. Finally, a mixture of two moles of KN0 3 , one gramme-
atom of sulphur and three gramme-atoms of carbon begins to decompose at 290°C
and explodes at 311°C.

(3) Sulphur intensifies the sensitiveness of mixtures to impact; a mixture of
KN0 3 and charcoal does not explode while a mixture of KN0 3 and sulphur decom-
poses when struck by a 2 kg weight falling from 45-50 cm. On the other hand,
a mixture of KN0 3 , charcoal and sulphur is exploded by a 2 kg weight from a drop
of 70-85 cm

(4) Sulphur counteracts the formation of carbon monoxide in the products of
explosion. CO and KCN occur in the products of decomposition of a mixture of
KN0 3 with charcoal, due to the reaction of charcoal with K 2 C0 3 :

K 2 C0 3 + 2C=2K+3CO

2K + 2C + N 2 = 2KCN

On the other hand, in the presence of sulphur a reaction with K 2 S0 4 takes place:

K 2 S0 4 +2C=K 2 S + 2C0 2

As there is no CO in these products no hydrogen evolves since the following
reaction does not take place :

CO + H 2 = C0 2 + H 2

When there is little sulphur in the powder, toxic CO is formed so that in black-
powder for mining purposes the content of sulphur should be not less than 10%.

Reactions between potassium nitrate and charcoal have also been investigated
by Oza and Shah [50].

Blackwood and Bowden [11] have more recently published extensive studies
on the mechanism of the initiation and burning of blackpowder and on that of the
reactions of binary mixtures, viz. KN0 3 + sulphur; sulphur + charcoal; KN0 3 +
+ charcoal.

It may be concluded from their experiments that burning occurs in a place
heated to a temperature of 130°C or higher. Heat may be applied either by the direct
effect of flame or by a hot metallic surface (e.g. glowing wire); heating may also
result from impact, adiabatic compression of the air in the spaces between the
grains of powder and finally, by mutual friction between the grains. The temper-

BLACKPOWDER 339

atureof 130°C is considerably lower than the usual ignition temperature, nevertheless
it may be sufficient to provoke an explosion if the grains are confined.

When the grains of blackpowder are subjected to a pressure of about 150
atm explosive decomposition may start at a temperature considerably lower than
usual.

The propagation of flame from grain to grain is caused by an emission of
hot molten potassium salts projected from one grain to another as the powder
burns.

The ease of ignition of blackpowder and its rate of burning are influenced by
the type of charcoal used, the decisive agents in this respect being in the opinion of
Blackwood and Bowden those constituents of charcoal which can be extracted with
organic solvents. Depending on the content of these constituents in the charcoal the
properties of the blackpowder can be varied, i.e. the smaller the content of soluble
constituents in the charcoal, the easier is the ignition and the faster the rate of burn-
ing of the powder.

Blackwood and Bowden formulate the following mechanism for the ignition and
burning reactions of blackpowder. First sulphur reacts with organic substances
present in charcoal :

S+ organic compounds ->- H2S (1)

Almost simultaneously saltpetre reacts with these compounds:

KNO3 + organic compounds -> N0 2 (2)

The following reactions may also occur:

2KNO3+S -> K 2 S0 4 +2NO (3)

KNO3 + 2NO ->. KNO2 + NO + NO2 (4)

H2S + NO2 -» H2O + S + NO (5)

Reaction (5) proceeds until all the hydrogen sulphide is used up when N0 2
appears and begins to react with the still unconsumed sulphur according to
equation (6) :

2N0 2 +2S -> 2S0 2 + N 2 (6)

The sulphur dioxide so formed may immediately react with potassium nitrate:

2KNO3+SO2 -* K2SO4+2NO2 (7)

Reactions (5) and (6) are endothermic, but reaction (7) is strongly exothermic.
Reactions (l)-(7) constitute the ignition process. According to Blackwood and
Bowden the chief reaction when the powder begins to burn is the oxidation of char-
coal by potassium nitrate.

340

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

EXPLOSIVE PROPERTIES OF BLACKPOWDER

From a comprehensive analysis of the products of explosion of blackpowder
Kast [51] derived the following equation:

74KN0 3 + 30S + 1 6C 6 H 2 (charcoal) = 56C0 2 + 14CO + 3CH 4 + 2H 2 S +
4H2+35N2+I9K2CO3 + 7K2SO4+2K2S++8K2S2O3 + 2KCNS +
(NH 4 ) 2 C0 3 + C + S + 665 kcal/kg

The heat of explosion and gas volume naturally depend on the composition of
the powder. Noble and Abel [45] in the work quoted above give the following figures
(Table 80):

Table 80
Dependence of the properties of blackpowder upon the composition

Blackpowder

Composition

Heat of

explosion

kcal/kg

Gas volume

KNO3 Sulphur

Charcoal

V cm^/g

Coarse-grained
Fine-grained (sporting)
Blasting
"Cocoa'-' powder

75
75
62
80

10
10
16

15
15

22
18

726
764
516
837

274
241
360
198

The latest studies of H. Thomas [27] quote the following figures for "Normal Gun-
powder G.12":

Potassium nitrate 75.3%

Charcoal 14.4%

Sulphur 10.3%

Heat of explosion is 735 ±15 kcal/kg at a moisture content of 0.85%. In dry
powder heat of explosion is 740 ± 15 kcal/kg.

According to Kast, the gas volume V is 280 l./kg, the specific pressure/is 2800 m
and the temperature of explosion t is 2380°C (Noble and Abel found this value to
be 2100-2200°C, Will [52]— 2770°C).

The specific gravity of blackpowder may vary within the limits of 1.50-1.80
depending on its intended use. Its apparent density is 0.900-0.980.

Blackpowder is highly sensitive to impact and friction. It explodes when struck
by a 2 kg weight falling from 70-100 cm. Its ignition temperature is 300°C. A sack
filled with blackpowder takes fire when penetrated by a rifle shot.

The rate of burning of blackpowder at atmospheric pressure is much greater,
than that of smokeless powder, but the rate and mode of burning at a pressure higher
than atmospheric depend on its compression pressure. Grains of blackpowder do
not burn by parallel layers but all over the mass of charge, if their density is lower
than 1.75. Above this density they burn by parallel layers and the burning time
then depends on the grain size (Vieille [53]). E.g. blackpowder with a density of

BLACKPOWDER

341

1.8 shows a rate of burning of about 10 cm/sec at a pressure of 1660 kg/cm2 while
nitrocellulose or nitroglycerine smokeless powder has a rate of burning of 15-30 cm
per sec.

According to Blackwood and Bowden [11] a single grain of blackpowder burns
with a rate of 0.4 cm/sec at atmospheric pressure. The flame is propagated along the
line of grains at a rate of 60 cm/sec at atmospheric pressure.

Blackpowder for demolition charges or as a mining explosive has a lower density
(about 1.67) and therefore burns very fast while blackpowder with a higher density

'•7 1.8 1.9 2.0

Density

Fig. 87. The relation between the time of burning of blackpowder and its density
(for grains 3.5-10.5 mm dia.).

(about 1.87), used as a propellant, burns slowly in parallel layers (Fig. 87). It is
very important for the blackpowder to be of adequate density since this makes it
possible to control its burning rate. The pressures required for obtaining charges
of suitable density are listed below (according to Rinkenbach and Snelling [54]).

Pressure

Densit;

kg/cm*

200

1.32

330

1.41

660

1.55

1330

1.695

2000

1.775

3330

1.84

5000

1.88

6660

1.88

Blackpowder may explode violently if it is confined in an air-tight container and
initiated with a strong initiator. Its lead block expansion scarcely amounts to 30 cm*
due to the "non-hermetic" effect of sand stemming. In an iron tube 35-41 mm dia.
it gives a rate of explosion of 400 m/sec.

342 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Andreyev [55] has made extensive studies on the transition of the burning of
blackpowder into explosion. He used for this purpose a device composed of an
iron tube, 40 mm in inner diameter and 200 mm in length, containing a 50 g charge
of blackpowder and hermetically sealed. When the charge is ignited, if the blackpowder
undergoes deflagration only, it ejects the steel base of the tube, whereas if the defla-
gration passes into explosion, the tube is blown into several pieces.

According to Andreyev blackpowder may burn at a pressure lower than atmo-
spheric. Burning, however, is the more difficult, the lower the surrounding pressure.
The limiting pressure at which blackpowder still burns is 0.1 kg/cm 2 . From 2-30 atm
the dependence of the rate of burning on the pressure may be expressed by the
equation U=72p ( >- 24 , where Uis a linear rate of burning in cm/min andp the pressure
in atmospheres.

THE MANUFACTURE OF BLACKPOWDER

RAW MATERIALS
Saltpetre

Potassium nitrate is obtainable from natural deposits in hot countries e.g. Ceylon,
Egypt, Mexico, India, Iran and some areas of the U.S.S.R. It occurs there as the
result of the microbiological oxidation of organic nitro compounds and of the re-
action of the products with the alkaline components of the soil. On being refined
by crystallization such saltpetre was (and partly still is) used for the manufacture
of blackpowder. In Great Britain for instance until relatively recently the only source
of potassium nitrate was saltpetre from India.

At one time potassium saltpetre was also obtained from manure and wood ash.
Here potassium nitrate forms as a result of microbiological processes followed by
reaction with potassium carbonate. Descriptions from various periods of history
may also be found that refer to the collection of saltpetre from the walls of cellars
situated in the vicinity of sewers. In this way, for instance, saltpetre was acquired
in Poland for the manufacture of blackpowder during the insurrections of 1830 and
1863.

From the middle of the nineteenth century potassium nitrate began to be manu-
factured from Chilian saltpetre containing 20-35% NaN0 3 . The Chilian saltpetre
was first refined to increase the content of NaN0 3 to 95% and afterwards subjected
to the exchange reaction :

NaN0 3 +KCl ^ KN0 3 +NaCl (1)

Reaction (1) is conducted in an aqueous solution at a temperature of 100°C.
It takes the desired course due to the poor solubility of NaCl at this temperature.
The sodium chloride is then precipitated and the reaction therefore shifts to the
right. The potassium nitrate so obtained was termed "converted saltpetre".

BLACKPOWDER

343

Another method of manufacturing KN0 3 was based on double decomposition
with potassium carbonate. puMUOn

At present reaction (1) is most frequently used to obtain potassium nitrate
although the sodium nitrate used is now of synthetic origin

Another method now employed consists in the treatment of potassium carbonate
or caustic potash with nitrogen dioxide. The course of the reaction is-

N 2 4 + H 2 = HN0 2 + HNO3
HN0 2 + HNO3 + 2KOH = KN0 2 + KN0 3 + 2H 2

(2)
(3)

A mixture of potassium nitrite and potassium nitrate is treated with nitric acid to
oxidize nitrite to nitrate: lo

3KN0 2 + 2HN0 3 = 3KN0 3 + 2NO + H 2 (4)

Still another method for the preparation of KN0 3 directly from KC1 and nitric
acid was used recently. Reaction proceeds at temperatures between 75-85°C accord
ing to the approximate equation: tu-ora-

4KC1+5HN0 3 =N0C1+C1 2 +2H 2 0+4KN0 3 +HC1 (5)

Potassium saltpetre prepared by any of these methods is refined i.e. it is recrvs

S" m"d r uct y a ^ Ch ° Sen ^ aCC ° rdanCe * «" «• *£

Crystallization is facilitated by the great difference, in solubility of potassium

mtra^at high and low temperatures (it is ten times more soluble at 100°C than

Potassium saltpetre for the manufacture of blackpowder must be of high purity

2LSX?SS3? for the synthetic product according to the u ss - r

standard (GOST 1939-43) is summarized in Table 81.

Table 81

TECHNICAL SPECIFICATION FOR SYNTHETIC KN0 3 USED IN THE MANUFACTURE
°F BLACKPOWDER (GOST 1939-43)

Appearance

Requirements

Min. content of KN0 3> %
Max. content of moisture, %
Max. content of chlorides, calculated as NaCl, %
Max. content of carbonates, calculated as K 2 C0 3
Max. content of water-insoluble substances, %
J^^^^lffju^tancesinsoluble in hydrochloric acid

Potassium saltpetre manufactured from Chilian saltpetre may also contain
potassium perchlorate and potassium iodate, which are very detrimental since they

344

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

increase the sensitiveness of blackpowder to friction and impact. A great many
explosions which have occurred during the manufacture of blackpowder have been
ascribed to the presence of these salts in saltpetre, although it has not been ascer-
tained experimentally that small admixtures of KC10 4 increase the sensitiveness
of blackpowder to impact and friction. Nevertheless, for the sake of safety potassium
nitrate manufactured from Chilian saltpetre must not contain perchlorates and
iodates in analytically detectable quantites.

Chilian saltpetre

Some types of blackpowder used for blasting contain sodium instead of potas-
sium saltpetre. Such a mixture is sometimes termed "explosive saltpetre" and is
considered a cheap substitute for blackpowder. The advantage of Chilian saltpetre
as an oxidizing agent lies in the fact that by weight it contains more oxygen than
potassium saltpetre. Its disadvantage consists in its relatively high hygroscopicity,
although this is caused by the presence of calcium salt since chemically pure sodium
nitrate is only slightly hygroscopic.

According to Soviet Standards GOST 828-41 sodium saltpetre for the manu-
facture of blackpowder mixtures should contain not less than 98% NaN0 3 and
less than 2% moisture.

Sulphur

For the manufacture of blackpowder the sulphur used should be of highest
purity, refined by distillation. Crude sulphur (which usually contains 2-5 % of im-
purities) is distilled from retorts heated to a temperature of 400°C. The receiver should
be maintained at a temperature above 115°C (12O-130°C), i.e. above the melting
point of sulphur (114-115°C). Under these conditions the distillate condenses to
a liquid which is then cast into sticks or blocks. This is the only form of sulphur
suitable for the manufacture of blackpowder. If the receiver temperature is lower,
the sulphur distilled from the retort condenses as flowers of sulphur which always
contain a little S0 2 , and even traces of H 2 S0 4 (the substance is easily oxidized due
to its large surface area). Sulphur in this form is therefore slightly hygroscopic and
acidic, and is unsuitable for the manufacture of blackpowder.

Charcoal

It is very important to select a suitable type of wood for the manufacture of the
charcoal used in blackpowder. It must be soft, but not resinous and should be pre-
pared from "white wood" of such trees as alder, poplar, willow, hazel etc. Before
carbonization the wood must be de-barked and cut into pieces 10-30 mm thick.
In some countries where hemp is plentiful the stems of this plant are used for making
charcoal.

BLACKPOWDER 345

The material to be carbonized is placed into sheet-iron retorts, approximately
1 m dia. and 1.5-3 m long, one end of which is closed with an air-tight Jjd and the
other fitted with an offtake for the gaseous products of distillation. These products
are usually burnt since it is not worth while recovering them. The combustion of
CO is particularly important, otherwise it may poison the atmosphere. The retorts
are heated either by exhaust gases or by superheated steam. Carbonization lasts
for 3-8 hr depending on the construction of the furnace and retorts, the temperature
and the type of material to be carbonized. After carbonization the retorts are taken
out of the furnace and allowed to cool down (with the offtake closed). The charcoal
is removed from the retort when it is cool-hot charcoal, easily ignites. The charcoal
should not be milled until the fourth day after its removal from the retort since
charcoal which is too fresh may catch fire during milling.

Three types of charcoal are obtained depending on the temperature of carbon-
ization (Table 82).

Table 82
Types of charcoal

Type

Temperature

of carbonization

°C

Content of C

"Cocoa" (red)

Brown

Black

140-175
280-350
350-450

52-54
70-75
80-85

In some blackpowder type mixtures, especially those for mining purposes, char-
coal is partially or wholly replaced by carbon black, brown coal, pitch from coal tar,
coal tar itself, naphthalene, paraffin, wood bark, cellulose or wood meal, starch,
resin etc. Thus mixtures are obtained with properties similar to blackpowder. Gene-
rally, however, they burn more slowly and are more difficult to ignite.

MILLING THE INGREDIENTS

The manufacture of blackpowder consists in thoroughly mixing the well-milled
ingredients to obtain the required particle size and density. (Fast-burning blackpowder
is fine-grained while the slower-burning type has larger grains of high density.)
Manufacture consists of the following operations:

(1) Milling of the ingredients;

(2) Mixing of the ingredients;

(3) Pressing;

(4) Corning and finishing (drying, polishing and screening).

The methods of manufacture depend upon the trend of development traditional
in a given factory or country. The variety of methods arises partly from the fact
that the manufacture of blackpowder is highly dangerous so that different countries

346

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 88. Iron ball mill.

have worked out their own methods of reducing risks. Nevertheless, there are cer-
tain agreed methods observed everywhere.

The ingredients of blackpowder may be milled by two methods :

Fig. 89. Disk mill.

(1) Each ingredient may be milled separately;

(2) Two ingredients may be milled together: charcoal with sulphur and charcoal
with saltpetre.

BLACKPOWDER

347

Saltpetre (dried at temperatures ranging from 100-1 10°C if necessary) is milled
either in ball mills, using iron balls (Fig. 88), in disk mills (Fig. 89), sr in disinte-
grators.

Charcoal is usually milled in either ball mills or disintegrators. It may be given
a preliminary treatment in edge runners.

Sulphur is milled in similar equipment, but in some countries edge runners or
rollers are used. Since sulphur is strongly electrified on milling it is advisable to add

Fig. 90. A sixteenth-century stamp mill (the stamps and one of the mortars are shown).

to it a slight amount of saltpetre to increase its electrical conductivity and reduce
the danger of an explosion of sulphur dust suspended in air. Until 1881 such dust
explosions were frequent. For the same reason all the equipment used for milling,
especially for milling charcoal and sulphur, should be well earthed.

Up to the end of the eighteenth century the ingredients of blackpowder were
milled in stone or wooden mortars in which the wooden pestles were fitted with
power drive (Fig. 90). The production capacity of such devices was, of course,
rather low. They were used too, for mixing the mass of powder, but this proved to
be dangerous, and in several countries (e.g. Great Britain) it was prohibited by law
(about 1772).

In some countries (France and Switzerland) it is believed that it is more dange-
rous to mill charcoal and sulphur separately than together in the correct propor-
tions. Saltpetre is also milled together with charcoal since the addition of such
a mixture to a mixture of charcoal and sulphur is supposed to be safer than the-
addition of saltpetre alone if the ratio of charcoal to saltpetre in the mixture added
does not exceed 1:15.

348

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Thus in France binary (two-component mixtures called "binaires") are formed:

Binary mixture containing sulphur (sulphur with charcoal);

Binary mixture containing saltpetre (saltpetre with charcoal).

Binary mixtures of charcoal with sulphur or with saltpetre, being insensitive to
friction and impact, are prepared in iron ball mills composed of drums (1.2 m dia.
and 1.5 m in length) with bronze balls. The drums (Fig. 91) are fitted inside with

^Mw////m.

Fig. 91. Iron drum for preliminary milling of the ingredients of blackpowder in the
form of two-component mixtures.

humps (7) which increase mixing and milling efficiency of the bronze balls. The
drums are fitted on their cylindrical surface with a door (2) for loading and un-
loading. To prevent escape of dust into the premises they are enclosed in a tin casing
(3), with a funnel for unloading on its lower end. The mouth of the funnel is closed
by a coarse stationary screen (4) and, situated below this, there is a fine vibrating
screen (5) with a clearance of 0.65 mm. Screen (5) is connected with the mouth of the
funnel by a leather sleeve (6).

The drum charge ranges from 80 to 180 kg:

Binary mixture containing sulphur: sulphur 50 kg

charcoal 35 kg

Binary mixture containing saltpetre: saltpetre 155 kg

charcoal 10 kg

In addition 100-150 kg of bronze balls 8-15 mm dia: are loaded into the drum,
which rotates with a rate of 18-20 r.p.m.

The binary mixtures are mixed and milled for 4-6 hr, after which the drum is
stopped, the door is opened and the contents are poured out by slow rotation. The
balls are retained by screen (4) while the material sifted through screen (5) falls into
the tin container (7) which can be hermetically sealed.

Separate drums should be used for the preparation of these binary mixtures.
Special care must be taken to ensure that the container for the sulphur mixture is
properly sealed, otherwise the sulphur which has been finely milled and heated on
milling may oxidize and even ignite in contact with air.

BLACKPOWDER

349

In the German factories at Spandau, Hanau, Ingolstadt and Gnaschwitz, milled
saltpetre, milled charcoal and a mixture of equal amounts of sulphur and saltpetre
were prepared in separate operations.

MIXING THE INGREJKENTS

The mixing of the three ingredients is one of the most dangerous operations
in the manufacture of blackpowder. If mealed powder is to be produced for pyro-
technic purposes (e.g. for signal or rescue rockets), or for initiators, mixing in a drum
is sufficient. For the manufacture of granulated or pelleted powder, the ingredients
must be mixed in an edge runner, though sometimes drums with wooden balls may
be employed.

Mixing in a drum

When mixing the three ingredients to obtain mealed powder, wooden drums
1.5-1.7 m dia. are used in a drum, provided with wooden humps. The humps enable
the drums to be rotated with a speed of 17-19 r.p.m. ; without them the speed would

Fig. 92. Wooden drum for mixing three ingredients of blackpowder.

have to be increased to 25-26 r.p.m. On completion of the mixing process the lid
is taken off the outlet, which is covered with the screen, the contents are turned
out by slowly rotating the drum, and sifted through the screen. Such a mixture
is ready for use in pyrotechnics. In some factories drum mixing is carried out as
a preliminary to kneading in edge runners.

In some factories the three ingredients are dampened with 8-10% of water and
then mixed in wooden drums with wooden balls (Fig. 92). The balls work by a
kneading action so that to some extent they replace mixing in an edge runner. The

350

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

required quantity of water is introduced into the drum by a special pipe. After being
removed from the drum, the damp mass is pressed into cakes in the same way as
the product from the edge runners.

Mixing in an edge runner

The edge runner (Figs. 93 and 94) consist of two heavy (500-600 kg) cast-iron
wheels approximately 1.5 m dia. an^*.5 m wide. The wheels rotate around horizontal

i ave

shafts connected by a clutch with a"Vertical axle (1) so as to enable clutch (2) to be
lowered or raised together with the suspended wheels (3). The vertical axle butts
against the bearing (4) situated in the centre of an iron tray (5). Near the wheels

Fig. 93. Diagram of an edge runner.

Fig. 94. General view of an edge runner.

and along the vertical axle there are bronze scrapers which push the mixture under
the wheels (the scrapers near the wheels push the material towards the centre of
the tray, those along the vertical axle push it away from the centre of the tray). The
material is thus continuously raked under the rolling wheels. The level at which the
scrapers are fixed can be regulated. The wheels may rotate about the vertical axle
at two speeds: a high speed of 10-15 r.p.m. and a low one of J-4 r.p.m. As a safety
precaution, all the nuts and other parts which might be loosened by the continuous
motion are split-pinned or fastened with copper wire to prevent them from falling
into the mass of blackpowder and thus causing an explosion. To prevent electrifica-
tion of the charge the equipment must be well earthed.

Near the vertical axle of the edge runner there is a tube for supplying the charge
with the necessary amount of water. In modern factories the water storage tank
and the cock regulating the water inflow are situated in an adjacent room from
which the apparatus is operated.

BLACKPOWDER

351

During mixing, the gap between the surface of the wheels and the tray must not
be less than 4 mm, so as to avoid the danger of seizing. This gap is regulated accord-
ing to the size of the charge and the density of mixture required. The greater the
required density the less should be the gap: the pressure of the wheels on the mixture
is then higher.

This principle of mixing by means of suspended wheels was not introduced until
the end of the nineteenth century; before then the kneading process in edge runners
was the most dangerous operation in the manufacture of blackpowder. Wooden

Fig. 95. Lay-out of a building in which edge runners operate.

trays were introduced in some factories to increase safety, but they wore out too
quickly. In many factories the use of edge runners, which caused such frequent
explosions, was completely abandoned in favour of drum mixing and the edge
runners have only been re-introduced in very few instances.

A lay-out of a factory building in which edge runners are operated is shown
in Fig. 95. The building has a "blow-out" construction, i.e. the three walls are
stout and the fourth together with the roof is light-weight with a door. One edge
runner is placed in each compartment. There is also an engine room in which the
engines and transmission gear are mounted (the edge runners usually have an over-
head drive comprising bevel gearwheels). There is a window in the wall protected
by thick panes, through which a worker operating the edge runner can watch the
run of the machine. Switches regulating the run of the edge runner and the cock
which controls the water inflow are within arms reach. Admittance to the room
where the edge runners operate is prohibited except for the loading and unloading
of material. Special felt-soled slippers, left standing by the door, should then be
put on to prevent sand from being carried in from outside. When the edge runner
is operating at high speed there is no admittance whatever into the premises.

Mixing proceeds in the following way: The ingredients, milled separately or in
binary form, are weighed out into wooden boxes or barrels fitted with grips, and
following the addition of ca. 8% of water (10% on hot and dry summer days) they
are mixed by hand with a wooden paddle. The damp material is spread evenly over
the tray of the runner. The layer of the material should not be less than 1 cm thick.

352 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The wheels are then set in motion, the rotational speed of the vertical axle being
increased gradually so that rolling is carried out with an angular-translatory motion.

Thus the wheels knead, grind and mix the material and at the same time press
it by their dead weight.

Due to friction the temperature rises to 30°C, and water evaporates so that
further addition should be made from time to time i.e. whenever the mass begins
to form dust, which indicates a drop in water content to 2-4%.

If the wheels cease to rotate about their axle, friction develops between the sur-
face of the wheels and the mixture which, apart from lowering the efficiency of the
runner, may lead to the ignition of the mixture (e.g. after the surface has dried).
To prevent this the runner is adjusted to run at slow speed (which is quite safe)
and leather is inserted between the wheels and the kneaded mixture. With the pres-
sure of the wheels on the thickened layer the friction increases and the wheels begin
to rotate. The leather may then be removed and the runner switched over to high-
speed motion.

The runners are costly to operate since they consume much energy. Care should
be taken therefore that they work for as short as possible a time. The maximum
density of material is obtained after 1-1.5 hr; further kneading reduces the density
of the mixture until a limit is reached.

Escales [56] gives the following figures characterizing the influence of duration
of kneading on the density of the product (according to data obtained from the
powder factory at Spandau).

Duration of kneading Density
lhr48min 1.63

2hr40min 1.42

5hr24min 1.36

7 hr 12 min 1.36

The French data, however, are somewhat different (Pascal [57]), namely:

Duration of kneading Density

Mining powder 30 min 1.57

Military powder F 3 2-2.5 hr 1.47

Sporting powder (78 % saltpetre) 1 .5 hr 1 .725

Sporting powder (78% saltpetre) 5 hr 1.80

If the dust sifted during the later operations is re-used for kneading, the process
may by shortened since the material is already partly kneaded. Kneading is carried
out with a moisture content of 2-4%.

PRESSING

After the kneading has been completed the mass is pressed to form press cakes
which are then dried. These cakes are hard and give a conchoidal fracture. Pressing
is done either: (a) in the same edge runners; (b) in hydraulic presses.

BLACKPOWDER

35J

Pressing in edge runners

Pressing in edge runners as carried out in France is effected by lifting the scrapers
after mixing has been completed, with the runner in slow-speed motion. The wheels
knead the mass which is not now raked by the scrapers. The material thus obtained
is in the form of a hard press cake which is then air-dried at temperatures from 20
to 30°C.

Pressing in hydraulic presses

The damp mass of powder removed from the edge runners is spread in an even
layer up to 9 cm thick over bronze or copper plates. Several layers are thus super-
imposed and compressed in a press (Fig. 96) at a pressure from 25 to 35kg/cm2 for

Fig. 96. Blackpowder press, hydraulically operated.

30-40 min. At this pressure the density of the grains obtained amounts to 1.7. By
applying a pressure of 100-1 10 kg/cm2 for 1.5-2 hr a density of 1.8 may be achieved.
The valves of the press must be handled at a safe distance, from behind a strongly
built protective wall, since pressing entails some danger e.g. a slight sliding move-
ment, under the effect of high pressure, may cause an explosion. After pressing,
the press cakes are dried, as described above.

354

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES
CORNING

Corning is accomplished by breaking up the press cakes into grains, usually in
Toll corning mills. Drum corning mills are also used but less frequently. For some
types of quick-burning powders combined corning is applied using grains from the
corning mill together with meal powder (granulation).

Roll corning mill

The dried press cakes are broken up with a mallet into smaller lumps and thrown
continuously into the corning mill (Figs. 97 and 98) consisting of several pairs of
bronze rollers arranged one above the other. After being crushed between the up-

Fig. 97. Corning mill for granulating blackpowder.

per pair of rollers the material passes over an automatic sieve, which separates
out the dust. The coarse pieces are crushed by the next pair of rollers, that operate
at slower speed, so as to form smaller grains. This process of sieving and crushing
is repeated in the following series of rollers and sieves until the material from the
last pair of rollers produces grains of the desired size. Any grains that are still too
large are recycled.

BLACKPOWDBR

355

The brass sieves under the rollers are stretched on frames so that they can be
easily taken out and, if necessary, replaced by sieves of different mesh size. The sieves
and rollers are enclosed in a plywood casing, to prevent the escape of dust. Some-
times dust absorbers are fitted (Fig. 98).

Fig. 98. Corning mill with dust remover.

Since the corning process may be dangerous, the mill is never approached while
the machinery is in operation. In modern factories loading is performed by placing
the feed pipe of the charging hopper over the corning mill in a separate cage, shielded
by a ferro-concrete barrier. Corning mills should be well earthed. The danger is
particularly great if foreign bodies e.g. metal fragments or nails, are caught between
the crushing rollers. The presence of such objects in the corning mill has caused
a number of explosions.

Drum corning mill

The drum corning mill (Fig. 99) comprises two concentric drums (7) and (2),
1.15 and 1.25 m dia. made of brass net stretched on a wooden frame with wooden
ribs inside. The inner drum (7) is made of 7.5 mm mesh net while the outer drum
(2) is of smaller mesh corresponding to the required grain size. The drums are located
in a casing (J) made of plywood. The casing is connected with the vibrating screens
(J) by a leather sleeve (4). The mesh sizes for different grain sizes (denoted by num-
ber of grains per gramme) are:

Sizes of screen meshes
(mm)
I II

1.6 1.2

1.2 0.8

Silk muslin

Number of

grains per

gramme

650-950

2000-3000

20,000-30,000

356

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The press cakes, broken into pieces with a mallet, are loaded continuously
through a hole in the drum near the axle. As the drum rotates the particles are
broken down until grains of the right size are obtained. They are then separated
by sieving.

/ /

\ \
\ \

1 1

Fig. 99. Diagram of drum granulation.

Granulating in drums. In some countries (e.g. France) the manufacture of very
fast-burning low density blackpowder for use in mining is carried out by granulating
moistened mealed powder. In the granulating process, a wooden drum 1.6 m dia.
and 0.6 m long is employed. A definite number of powder grains are taken from
the corning mill, and moistened to a 10% water content with mealed powder pro-
duced by mixing of the three ingredients. They are then poured into the drum.
As the drum rotates the mealed powder sticks to the surface of the grains to form
spherical granules. External wooden hammers are used to prevent the mass from
sticking to the inner surface of the drum. The mass heats up by friction and part
of the water evaporates, hardening the grains. The granulated powder is sieved
to obtain grains in the size range 3-6 mm. The undersized grains are recycled and
the oversized ones are ground in the corning mill and granulated again.

FINISHING

The powder from the corning mill is then finished. This consists mainly of polish-
ing, drying and grading the grains. If cylinders are required (blasting powder of
the German type) finishing is limited to pressing the grains into cylinders.

BLACKPOWDER

357

Polishing, drying and grading

After corning the grains have an irregular and rough surface which prevents
them from flowing freely and from filling a space without voids, so that their density
is rather low. To overcome this, the powder grains are polished and thus acquire
a smooth, slippery surface. The powder from a corning mill subjected to polishing
contains ca. 1.5-3% (according to German data) or 4% (according to French data)

Fig. 100. A leather drum for polishing (the latter is stretched on a wooden frame).
P indicates the wooden ribs.

of moisture. The tumbling drums are constructed similarly to that shown in Fig. 92.
They may be made of leather stretched on a wooden frame (Fig. 100). The rate of
rotation of the drum, dependent on its dimensions, is 7-16 r.p.m. (smaller drums
rotate more quickly to reach the peripheral velocity required). The drum can be
charged with 100-240 kg of powder.

At first, the work is carried out with the side door closed so that the powder
retains enough moisture to facilitate the polishing process. After a few hours 0.1-0.5%
of graphite is added to the charge and polishing is continued for several more hours.
The graphite fills the pores in the surface of the grains and coats them with a thin
layer, giving them their characteristic brightness.

During the last hours of polishing the side door in the drum is opened so that
some of the moisture escapes. The overall polishing process lasts for 4-24 hr, de-
pending on the type of powder. The finished grains are more slippery, pour more
easily and fill space better than the unpolished product.

The fine-grained powder used for filling time-trains in time fuses is sometimes
polished by coating the grains with a layer of shellac. For this an alcoholic solution
of shellac (1-2% shellac by weight) is sprayed into the drum. Powder grains coated
with this material burn more slowly than usual. The rate of burning can be regulated
by the amount of shellac used.

358

CHEMISTRY AND TECHNOLOGY OF EXi-LOSIVES

The polished powder is dried at 50°C to 0.5-1.0% moisture content (in countries
with a damp climate the permissible moisture content is higher and in countries
with a dry climate— lower). Shelf driers with a natural stream of warm air are usually
employed. The use of a pressurized air stream is permissible provided that it is

Fig. 101. A multistroke press of the Vyshnegradskii type.

warmed. The use of suction fans is inadvisable since dust from the powder may pene-
trate their mechanisms and ignite.

The polished, dried powder is finally graded by sieving out the dust. Revolving
or various types of vibrating sieves, often hand driven for safety (see Fig. 99) are
used for this purpose.

The sieve frames should be made of wood to avoid the danger of friction and
impact. Brass mesh (formerly perforated parchment) or silk muslin is used, stretch-
ed on frames so that the sieves can be easily arranged according to the grain
sizes required. Dust separated out by sieving is sent back to the edge runners.

After all these operations, i.e. corning, polishing, drying and grading, have been
completed 50 kg of the powder, ready for use, is obtained from 100 kg of mixture
supplied to the edge runner. All the waste material accumulated during processing
is returned to the edge runners. The yield is higher in the manufacture of coarse-
grained powder than in that of fine-grained powder.

BLACKPOWDER 359

Final pressing

Final pressing is applied to blasting powder of the German type which is obtained
in the form of cylindrical pellets 30 mm dia. and ca. 40 mm long with a central
hole 10 mm dia. along the axis of the charge. The hole serves for the insertion of
a detonating cup or a detonating fuse (p. 328).

Unpolished or polished powder grains containing moisture (up to 4%) are press-
ed for a short period (e.g. 30 min) in a mechanical press using two pistons: one

Flo. 102. Diagram showing the working principle of a funnel in powder blending.

applied from above, the other from below in order to achieve a highly uniform
density. Vyshnegradskii [58] designed this technique. The press is usually a multi-
stroke unit in which six to twelve pellets are compressed at the same time (Fig. 101).
The compressed pellets are finally dried at 50°C to increase their resistance to
mechanical shock. On drying, the moisture content falls below 1%. Blackpowder
in the form of cylindrical pellets is the most suitable type of explosive for blasting.
Cannon blackpowder was also once produced in the form of prisms; (this will be
discussed later, in the chapter on cannon blackpowder).

BLENDING

To obtain a uniform lot of granular powder several batches are blended in a
wooden funnel. The powder is poured into the funnel in layers after which the
bottom outlet of the latter is opened. As the contents pour out, like liquid, they
are blended (Fig. 102).

CANNON POWDER

Blackpowder for the manufacture of propellent charges has long been obsolete
although some types are still produced for other purposes. Thus the French "cube"
powder, poudre C or poudre SP, is employed as an igniter in bag cartridges.

360

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Cannon powder is manufactured by kneading in an edge runner followed by
pressing the cake to a high density. Afterwards the mass is corned, moistened to
3-3.5% water content and pressed again into cakes at a pressure of 100 kg/cm 2 .

Fig. 103. Cubing sheets of blackpowder.

The sheets are then sliced on a cutter (first into strips, then into cubes) using a rect-
angular blade (Fig. 103).

After a short polishing period to smooth over and round off the sharp and
irregular edges, the cubes are separated from dust and dried to the less than 1%
water content.

At one time prismatic powder with a very high density, burning in approximately
parallel layers, acquired an importance, but now it is only of historical interest.

It was invented in the United States by Rodman [59] and improved during the
period 1868-1882. It had the form of a hexagonal prism (Fig. 104) with a central

Fig. 104. Blackpowder in prism form.

channel and was manufactured in the usual way with the distinction that the grains
from the corning mill were finally compressed into prisms by the Vyshnegradskii
hydraulic press (p. 358). It was used for long-range gun fire. Since powder for this
purpose must burn as slowly as possible the following measures were taken to
reduce the rate of burning:

(1) Brown charcoal was used in the mixture.

(2) The sulphur content in the mixture was reduced. This gave the following
compositions :

In Germany and France

In Russia

In Belgium

KNO3 78%
Sulphur 3%
Charcoal 19%

75.5%
8.5%
16%

52%

39%

BLACKPOWDER

361

(3) Powder cakes (before corning) were pressed in a hydraulic press to a density
of 1.80 and after separating 2.5 mm grains, were pressed again into prisms with
a density of 1.86.

(4) The prisms were dried as slowly as possible (3-7 days) and at a low temper-
ature (35-40°C) to avoid the formation of internal fissures due to over-rapid
drying.

Mention should also be made of another method of reducing the rate of burning
of the powder, i.e. the addition of paraffin. In some French cube powders, paraffin
was substituted for 5% of charcoal.

SAFETY IN BLACKPOWDER FACTORIES

The manufacture of blackpowder is one of the most dangerous in the production
of explosives.

The individual operations are carried out in "danger buildings", separated by
safety distances. These distances are reduced if the buildings are mounded. For
unmounded buildings the average safety distance is 50 m. A blow-out construction

RtG. 105. System for illuminating buildings in which blackpowder is manufactured.

is usual: three walls are of stout brick while the fourth wall and the roof are of light-
weight construction, e.g. of wood or asbestos tile. Alternatively, the whole building
may be constructed of light-weight boards.

Shelters for the workers should be set up in the vicinity of each building. These
are huts covered with earth at the side of the operating building. If the buildings have
three stout walls the shelter may be located behind one of them. Doors and windows
should be large to give plenty of light, and closed from outside with a wooden

362

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

shutter. Except for windows facing north the glass should be opaque or coloured.
Doors and windows are fitted with overhanging eaves to protect the interior of the
building from rain. Inside the building, the floor may be either of wood (all crevices
being sealed with pitch) or of asphalt or concrete, close-covered with linoleum.
It is also desirable that a wooden barrier should be erected at the entrance forming
an anteroom in which those who enter the danger building put on clean felt over-
boots.

Electric cables should be carried in steel conduits on the exterior of the walls.
Incandescent lamps inside the building should be recessed into the walls and pro-

Fig. 106. Remote, control of a machine dangerous to operate (corning mill) in
a blackpowder factory [60].

tected by safety glass. Alternatively, the buildings may be lit through the windows
by reflector lamps (Fig. 105). The buildings may be heated only by hot water (not
by steam) and the radiators must be dusted frequently.

All machines must be well earthed, all their frictional parts being made of bronze,
wood etc. If steel parts must be used they should be lubricated profusely, if possible,
with solid grease (vaseline, cup grease). Particularly dangerous machines (e.g. corn-
ing mills) must be operated by remote control (Fig. 106).

In front of the building a timber floor protected by a eave should be laid. Trucks
coming and going with materials should be moved by hand. Trucks with pneumatic-
tyred wheels or barrows with bronze wheels are best for this purpose. If a narrow-
gauged track is to be used, in the vicinity of buildings holding blackpowder, the
rails should be of wood and the trucks should be fitted with bronze wheels.

Blackpowder must be packaged for transport in cloth or rubber bags and placed
in tightly closed wooden boxes.

BLACKPOWDER 363

LITERATURE

1. S. J. ROMOCKI, Geschichte der Explosivstoffe, Oppenheim, Berlin, 1895-1896.

2. J. R. Partington, A History of Greek Fire and Gunpowder, Hoffer & Sons, Cambridge, 1960.

3. W. Kochmyerzhevskii, Podrywnoye delo, St. Petersburg, 1898, according to Paprotskii,
Bolshaya Sov. Entsiklopedya, II ed., 7, 1951.

4. Sir R. MORAY, Phil. Trans. No. 5, 82 (1665).

5. J. Taylor, Solid Propellants and Exothermic Compositions, Newnes, London, 1959.

6. A. G. Gorst, Porokha i vzryvchatyye veshchestva, Oborongiz, Moskva, 1949.

7. D. A. Davies, Nature 167, 614 (1950). I*

8. F. G. G. Tate, Alcoholometry, p. 33, H.M.sH)., London, 1930.

9. Violette, Ann. chim. [3], 23, 475 (1848).

10. T. Urbanski and E. Tesiorowski, unpublished work (1931).

11. J. D. BLACKWOOD and F. P. BOWDEN, Proc. Roy. Soc. (London) A213, 285 (1952).

12. M. BERTHELOT, Ber. 2, 57 (1869).

13. G. Dickson and T. H. Easterfield, J. Chem. Soc. 74, 163 (1898).

14. O. Dimroth and B. Kerkovius, Ann. 399, 120 (1913).

15. H. Meyer and K. Steiner, Monatsh. 35, 391, 475 (1914).

16. E. Philippi and G. Rie, Ann. 428, 287 (1922).

17. E. Philippi and R. Thelen, Ann. 428, 296 (1922).

18. B. JuETTNER, J. Am. Chem. Soc. 59, 208, 1472 (1937).

19. T. Urbanski, W. Hofman, R. Ostrowski and M. Witanowski, Bull. Acad. Polon. Sci.,
sir. chim. 7, 851, 861, (1959); Chemistry & Industry 1960, 95.

20. Hercules Powder Co., according to Bacchus, The Expl. Engineer, Wilmington, April 1923.

21. F. Raschig, Ger. Pat. 257319 (1911); Angew. Chem. 25, 1194 (1912). '

22. Schwartz, Dinglers polyt. J. 226, 512 (1877).

23. P. Seidler, Ger. Pat. 78679 (1893).

24. F. Raschig, Brit. Pat. 29696 (1912).

25. A. Voigt, Ger. Pat. 260311, 260312 (1911); 267542 (1912).

26. Andrew Noble, Artillery and Explosives, London, 1906.

27. H. Thomas, according to J. Taylor [5].

28. A. LANGHANS Z. ges. Schiess- u. Sprengstoffw. 14, 55 (1919).

29. Reiset and Millon, Compt. rend. 16, 1190 (1843).

30. Gaens, Ger. Pat. 37631 (1885); Brit. Pat. 14412 (1885).

31. D. J. B. Marke and C. H. Lilly, /. Sci. Food Agric. 2, 56 (1951).

32. C. Berthollet, Ann. chim. 9, 22 (1818).

33. Designolle, according to R. Escales [56].

34. BRUGERE, Compt. rend. 69, 716 (1869).

35. According to P. F. Chalon, Les explosifs modernes, Beranger, Paris & Liege, 1911.

36. E. A. Starke, U.S. Pat. 513737, 527563 (1894).

37. Rene Descartes, Lesprincipes de la philosophie, Elsevier, Amsterdam, 1644.

38. J. Gay-Lussac, Rapports de la Comite des Poudres et Salpetres, 1823.

39. G. Piobert, Memoires sur les poudres de guirre, Paris, 1844.

40. M. Chevreuil, according to R. Escales [56], p. 406.

41. R. Bunsen and Shishkov, Pogg. Ann. 102, 325 (1857).

42. J. Linck, Ann. 109, 53 (1859).

43. M. Karolyi, Pogg. Ann. 118, 546 (1863); Jahresber. Chem. 743 (1863).

44. N. P. Fiodorov, Z.f. Chem. (1863).

45. Andrew Noble and F. Abel, Phil. Trans. Roy. Soc. 165, 49, 105 (1875); 171, 203 (1880).

46. M. Berthelot, Compt. rend. 82, 487 (1876).

47. H. Debus, Ann. Ill, 257; 213, 15 (1882).

364 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

48. K. A. Hoffmann, Sitzungsber. Preuss. Akad. Wissenschaft. 25, 509 (1929); 26, 434 (1930).

49. Andrew Noble, Fifty Years of Explosives, London, 1907.

50. T. Oza and S. Shah, /. Indian Chem. Soc. 20, 261, 286 (1943).

51. H. Kast, Spreng- u. Ziindstoffe, Vieweg & Sohn, Braunschweig, 1921.

52. W. Will, Z. ges. Schiess- u. Sprengstoffw. 1, 209 (1906).

53. P. Vieille, Mem. poudres 6, 256 (1893).

54. W. H. Rinkenbach and W. O. Snelling, in Kirk and Othmer (Ed.), Encyclopedia of Chemical
Technology, Vol. 6, Interscience, New York, 1951.

55. K. K. Andreyev, Dokl. Akad. Nauk SSSR 1945, 49; Zh.fiz. Khim. 20, 467 (1946).

56. R. Escales, Schwarzpulver und Sprengsalpeter, Veit & Co., Leipzig, 1914.

57. P. Pascal, Explosifs, poudres, gaz de combat, Hermann, Paris, 1925.

58. I. A. Vyshnegradskii, according to Bolshaya Sov.Encyklopedya 9, 541 (1951); O. Guttmann,
Die Industrie der Explosivstoffe, p. 222, Vieweg & Sohn, Braunschweig, 1895.

59. T. J. Rodman, Properties of Metals for Cannon and the Quality of Cannon Powder, C. H. Crosby,
Boston, 1861, according to J. Taylor [5].

60. Reproduced from Chemical & Engineering News.

CHAPTER IV

COMPOSITE PROPELLANTS FOR ROCKETS

GENERAL INFORMATION

Blackpowder was the oldest known and the only propellant used in rockets
up to time of World War II. It is a slow-burning powder with a high content of
charcoal (p. 330).

Blackpowder is still used in sifl^ll rockets (e.g. signal or rescue rockets for car-
rying the rope from shore to ship or vice versa). Its limited application is due to the
fact that blackpowder gives a very low specific impulse : I s = from 40 to 80 sec, which
is much lower than that of modern composite propellants and smokeless rocket
powders which give a specific impulse of 180-200 sec.

Smokqkss (double base, i.e. nitroglycerine) powder, however, has some disadvan-
tages which limit its use. It is difficult, to produce in large-size charges and produc-
tion involves costly investments. It is also dangerous to manufacture due to the
use of nitroglycerine as an ingredient. In addition the powder requires periodic
testing of the stability. It was for this reason that during World War II and
afterwards composite propellants were introduced consisting of two essential
ingredients: solid oxidant and solid fuel.

Solid composite propellants are usually rich in combustible ingredients and the
amount of oxidant is usually limited by the mechanical properties of the mixtures.
Careful choice of components is needed to obtain high loadings without jeopardiz-
ing fluidity in mixing and creating discontinuities in the binder. A common pro-
cedure consists of using oxidants in two or more sizes.

Mishuck and Carleton [1] classify rocket propellants into:

(1) Polymerizable, castable (e.g. polysulphide-ammonium perchlorate mixtures);

(2) Nonpolymerizable, castable, gel-type (cast double base powder);

(3) Vulcanizable, non-castable (rubber-ammonium nitrate).

The most popular are the polymerizable, castable systems (group 1). They offer
a great versatility in the choice of polymerizable monomers.

Rheological properties are very important in mixing and processing the composi-
tions. Thus sedimentation of suspended solid before curing ought to be negligible.
Casting must be sufficiently fluid to allow the escape of gas-bubbles. It is necessary

366

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

to know how the viscosity of a homogeneous liquid is modified by high loading
with suspended solids.

A general discusion of the factors important in choosing the composition and
methods of manufacture of the composite propellants was given by Mishuck and
Carleton [1].

The following salts are commonly used as oxidants: potassium, sodium or
ammonium nitrates and potassium or ammonium perchlorates. Lithium perchlorate
was also suggested but seems to be used in an experimental scale only.

Solid fuels are mainly plastics or elastomers which, apart from functioning as
combustibles, also serve to provide the rigidity. Occasionally a combustible and

Inhibitor

Slotted part Unstotted part

e v " *v

Cylinder star centre

Clover ieaf Rod and cylinder Slotted tube

Fig. 107. Charge shapes (cross sections).

explosive substance such as nitrocellulose in a colloid form may act as the plastic
component.

Other compositions are similar to pyrotechnic mixtures containing explosive
substances, e.g. salts of picric acid.

To increase the heat energy emitted during the reaction metallic powders may
be added, e.g. aluminium. Charcoal or soot are added sometimes to make the mix-
ture burn smoothly.

The advantage of the majority of the mixtures concerned lies in the cheapness
and stability of their ingredients, which do not decompose during storage and the
uncomplicated and relatively safe method of manufacture. Periodic chemical stability
testing for smokeless powders, which necessitates a suitable organization and entails
a high expenditure, is unnecessary.

A general requirement is that the average molecular weight of exhaust gas should
be low. This restricts the choice of solid propellant components to these containing
elements of low atomic weight. Low atomic weight elements often release a greater
heat of reaction.

Very little is known about the burning mechanism of these mixtures. There is
a "two-temperature" theory of propellant combustion [2, 3].

According to this theory the rate-controlling reactions are associated with the
gasification (pyrolysis) of the solid oxidant and the solid binder which in a composite
propellant are essentially independent of each other.

The modern grains have their outer surface protected by an inhibiting layer
and they burn from the inner surface created by differently shaped cores. Cross

COMPOSITE PROPELLANTS FOR ROCKETS

3«7

sections of some typical charge are shown in Fig. 107. The purpose of the different
shapes is to permit the rocket to operate at constant pressure by keeping the burning
surface constant. The most commonly used charge design has a star-shaped core or
some similar shape (e.g. clover leaf) [4, 5]. The thrust-burning time curves of cylindrical
and star-shape charges are given in Fig. 108.

The star-shaped charge leaves slivers which burn out at low pressure and this
leads to impulse losses. The slivers may remain unburnt and their weight is then

Time, sec Time, sec

Tube-shaped charge Star-shaped charge

Fig. 108. Thrust-burning time curves of cylinder- and star-shaped charges.

added to the payload. Rods and cylinders do not leave slivers, but supporting the
central cylinder may present some difficulty and a tube slotted at the nozzle end
seems to be a more practical design [4].

MIXTURES WITH THE SALTS OF PERCHLORIC ACID

Mixtures composed of the salts of perchloric acid and an elastomer or plastic
polymer are now the most popular of the composite propellants.

Mixtures containing potassium perchlorate are characterized by a relatively
high rate of burning, high exponent n in the equation V=kp" and a high flame
temperature. When burning they produce a dense smoke. Mixtures containing
ammonium perchlorate have a lower rate of burning, low exponent n and flame
temperature and yield less smoke. The rate of burning of these mixtures is however
higher than that of similar mixtures' with ammonium nitrate. Andersen et al. [3]
found that a composite propellant with 75% ammonium perchlorate and 25%
copolymer of polyester and styrene burns under a pressure of 6 kg/cm 2 at 15°C
with a rate of 0.64 cm/sec. i.e. 2-3 times faster than corresponding mixtures with
ammonium nitrate.

The forerunner of the rocket propellant containing perchlorates as oxidants
was the American product, Galcit, developed in the Jet Propulsion Laboratory (JPL)
of the Guggenheim Aeronautical Laboratory, at the California Institute of
Technology. It consisted of asphalt as a fuel and potassium perchlorate as an
oxidant, mixed hot. Originally the material was hot-pressed into rockets. How-
ever, at high pressures the flame penetrated between the charge and the walls

368 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

of the rocket, causing an abrupt increase of the burning surface and consequently
of the pressure, so the modification was introduced of pressing the material into
blocks, which were not in contact with the walls of the rocket.

This composition was useless outside certain temperature limits, i.e. outside the
range 4-38°C (40-100T), the mixture becoming brittle and breakable at lower
temperatures, and soft and shapeless at higher ones.

Zaehringer [6] gives the following composition for Galcit Alt 161:

Potassium perchlorate 75%

Asphalt with mineral oil or
resin 25%

Resin was added to the asphalt to raise its softening temperature.
Specific impulse I s was 186 sec at 1350-3700 lb/in 2 pressure range.

MIXTURES OF PERCHLORATES WITH ELASTOMERS.
THIOKOL PROPELLANTS

Some of these propellants are apparently very popular. They are probably the
cheapest composite propellants now in use.

T. L. Smith [7] discusses the properties required in elastomeric binders and the
mechanical properties of solid cast-in-place case-bonded propellants. He considers
that the most important mechanical property required in composite propellants
containing an elastomer and intended for case-bonded grains is a relatively large
ultimate elongation over a range of temperature and strain rate of experimental
time scale. Moreover, the propellant grain must not creep excessively during rocket
storage, and the grain must not crack or deform excessively under flight acceleration
forces.

According to T. L. Smith an ideal elastomeric binder should have low "glass"
temperature, should exhibit high elongation over a wide temperature range, should
be cross-linked preferably through stable covalent bonds, and should not crystallize
spontaneously during storage at any temperature.

The uncured binder material should be a liquid which cures with minimum heat
release and shrinkage and without evolution of gases.

The binder should not be a solvent of the oxidant and should be chemically
stable for long periods in close contact with the oxidant.

A more detailed discussion of the visco-elastic properties of rubber-like elasto-
mers for composite propellants was recently given in a paper by Landel and T. L.
Smith [8].

Among the composite propellants containing elastomers, those containing poly-
ethylene sulphide "rubber", so called Thiokol, are widely used. Since 1950 they
have been developed by the Thiokol Chemical Corporation in the U.S.A. under
the general name of "Thiokol Propellants".

CQMPOglTB PROPELLANTS FOR ROCKETS 369

These compositions consist of thiokol and perchlorate, most probably in pro-
portions :

Thiokol 20-40%
Perchlorate 80-60%

As there is no certainty that common solid thiokol produced in the form of
latex should give a sufficiently homogeneous mass, Thiokol Corporation have
worked out a method of starting with liquid thiokol which is then "cured".

As a rule, the liquid thiokol, mixed with an oxidant and an accelerator is poured
into a prepared rocket chamber and cured in situ when the charge is solidified.

Liquid thiokol

The basic ingredient of the mixtures is liquid thiokol.

According to descriptions published by the Thiokol Chemical Corporation [9]
and Gobel [10] liquid thiokol is obtained in the following way: ethylene chloro-
hydrin is condensed into dichlorodiethylformal (I) which is then treated with sodium
polysulphide to form the polymer (II):

2ClCH 2 CH 2 OH + CH z O -► ClCH 2 CH 2 OCH 2 OCH 2 CH 2 CI

NazS*

JC=from 1 to 5

HS-[C 2 H 4 OCH 2 OC 2 H4SS]„-C 2 H 4 OCH 2 OC 2 SH
II

Other dichloro compounds may be used instead of dichlorodiethylformal, e.g. :
1,2-dichloroethylene, 1,3-dichloropropylene, dichlorodiethyl ether, dichlorotriethyl
ether which, however, have never yet been put to practical application.

There are now six types of Thiokol Liquid Polymer (LP for short) in commerce:
LP-2, LP-3, LP-31, LP-32, LP-33, LP-8 (see Table 83).

Table 83

LP-2

LP-3

LP-31

and
LP-32

and
LP-33

LP-8

Physical state

Mobile

Viscous

Mobile

liquid

Colour

Amber

Specific gravity at 20°C

1.31

1.27

Viscosity at 25°C (cP)

80,000-140,000

35,000-45,000

750-1500

250-350

Average molecular weight

7500

4000

1000

500-700

n (in the formula II)

pH (water extract)

6.0-8.0

Stability (years)

over 3

Moisture content (%)

max. 0.2

370 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The difference between these polymers lies in the degree of polymerization
or cross-linking. Thus LP-2 and LP-32 are cross-linked polymers. Of these two types
the cross-linking of polymer LP-32 is weaker, therefore the modulus of elasticity of
LP-32 (after curing) is smaller. The bursting stress of cured LP-32 is twice that of

cured LP-2.

A similar difference exists between LP-3 and LP-33. Polymer LP-33 differs
from LP-3 in having a weaker cross-linking.

The viscosity of polymers LP depends to a high degree on temperature.

Thus for a sample of Thiokol LP-31:

at 26.7°C (80°F) the viscosity was ca. 120,000 cP
at 48.9°C (120°F) the viscosity was ca. 28,000 cP
at 71.1°C (160°F) the viscosity was ca. 10,000 cP

At 25°C Thiokol LP-2 had a viscosity of 35,000-45,000 cP, but at 80°C the
viscosity was only 5000 cP.

The curing of liquid thiokol is based on the reaction of the liquid polymer with
lead dioxide or peroxides. Dehydrogenation of the terminal groups of the mercaptan
chains of the polymer then takes place followed by elongation of the chains. This
reaction may be depicted as:

2RSH + Pb0 2 -> — R— S— S— R— + H 2 + PbO
III

The lead oxide so produced and the lead oxide present as an impurity of lead
dioxide may continue to react:

2 — R— S— S— R— + PbO -=► — R— SPb— S— R+ H 2

The lead mercaptide thus obtained may undergo oxidation during heat curing
or on heating with a"small amount of sulphur to form chain (III) :

— R— S— Pb— S— R + S -> — R— S— S— R— + PbS
IV III

Three to five parts by weight of Pb0 2 are usually used per hundred parts of LP.

Up to 0.5% of sulphur may also initiate the curing effect of lead dioxide. Con-
versely, the addition of fatty acids and their salts may exert an inhibitory influence
upon the curing process of polymer LP. Thus the addition of one part by weight
of stearic acid to polymer LP approximately doubles the curing time.

Peroxides (e.g. cumene hydroperoxide) are alternative curing agents. The reaction
then proceeds according to the equation:

2— R— SH + 0-> — R— S— S— R— + H 2
III

Approximately five parts by weight of hydroperoxide are used per hundred
parts of polymer LP. The reaction may be accelerated by an alkaline medium, so

COMPOSITE PROPELLANTS FOR ROCKETS

371

that a small (e.g. 0.2 part) amount of weak bases may be added e.g. benzyl-dimethyl-
amine or tri(dimethylamino)-methylphenol (DMP-30).

Also noteworthy is the addition of an agent to increase mechanical strength
(e.g. carbon black, titanium dioxide, zinc sulphide, lithopone) and plasticizers which
usually serve to disperse the Pb0 2 within the polymer.

Thiokol Chemical Corporation suggests the following compositions for B (Base
compounds) (in weight units):

Table 84

Composition of B compounds

LP-31

LP-2

LP-32

Thiokol LP
Carbon black

100

100
30

Soot or zinc sulphide or lithopone
Stearic acid

30-50
1

Sulphur

0.15

0.1

The composition of C (Curing compound) is:

Pb0 2 50%

Stearic acid 5%

Dibutyl phthalate

45%

B compound is mixed with C compound in the proportions of hundred parts
of Thiokol LP to ten to fifteen parts of C compound.

Thiokol Chemical Corporation gives the following characteristics of polymers
LP-2 and LP-32 cured in 24 hr at 26.7°C (80°F) and then pressed for 10 min at
175-190°C (287-310°F):

Table 85

LP-2

LP-32

Tensile strength lb/in 2

500

550

300% modulus lb/in 2

350

250

Crescent tear index

125

145

Shore A hardness

Low temperature flexibility (°F)

-65

High temperature resistance

(°F)

275

Ozone resistance

excellent

Sunlight resistance

excellent

Ageing resistance

excellent

Cured polymers obtained in this way show very high resistance to such solvents
as petrol, water and ethanol. On the other hand they swell and absorb ketones,
esters (ethyl acetate) and carbon tetrachloride.

372

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 109. Rocket chamber being lowered into a vapour degreasing pit to remove all
traces of grease and oil (according to Dykstra [1 1]).

1 COMPOSITE PROPELLANTS FOR ROCKETS 373

THE TECHNOLOGY OF THE MANUFACTURE OF ROCKET CHARGES
CONTA1.NIG COMPOSITE PROPELLANTS WITH THIOKOL

It is characteristic of the technology of the manufacture of rocket charges contain-
ing composite propellants with thiokol that the semi-liquid mixture is poured directly
into the rocket chamber lined from within with an insulating layer to which the
charge adheres tightly ("case-bonded" charge). This is a very cheap and rapid method
of manufacture.

Warren [5] describes the following stages of manufacture as used at the Thiokol
Chemical Corporation's Redstone Arsenal, Huntsville, Alabama:

(a) chamber preparation,

(b) oxidant preparation,
(c)' mixing,

(d) casting,

(e) curing,

(f) finishing and inspection.

Chamber preparation (Case-bonded propellants)

The rocket chamber itself is used as the mould. The inner surface of this chamber
must be very carfully cleaned so that the propellant will be well bonded to it. Any
rust or foreign material should be removed by wire brushing or sand blasting. After
that, particles loosened by this operation, together with any grease or oil which may
be present, are removed by a vapour-degreasing, trichlorethylene being the most
popular degreasing agent. With large units the solvent is poured directly into the
chamber which is then rotated and brushed at the same time (Fig. 109).

Most propellant manufactures apply an insulating lining to the chamber surfaces
to protect the metal wall. The propellant itself is an insulator during most of the
burning, so presumably the lining acts as an insulator only during the final burnout.
Another important function of the lining is to inhibit burning on the outside of the
charge where there is a poor bond to the metal case. The same binder that is used
in the propellant, but without the oxidant, makes an effective lining. It is generally
sprayed on and polymerized in place by a short curing treatment at an elevated
temperature. The thin layer provides an excellent surface to which the propellant
can be bonded.

While chamber preparation is going on the cores or "risers" are prepared to
act as moulds for the perforation of the charge. The cores can be placed in the
rocket chamber prior to casting and held in place by special supports, but usually
they are added after the propellant has been cast. Precise alignment of these cores
is of the utmost importance. A star-shaped core is given in Fig. 110.

374

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 110. Charge with a star-shaped core (Thiokol Chemical Corporation, according

to Chapman [12]).

Fig. 111. A double-motion paddle mixer for premixing Thiokol propellant (Thiokol
Chemical Corporation, Redstone Arsenal, according to Warren [5D.

COMPOSITE PROPELLANTS FOR ROCKETS

375

Preparation of the oxidant

Inorganic oxidants are used exclusively in composite propellants; their pro-
perties have already been discussed. Huggett [13] draws attention to the importance
of the particle size of the finely ground potassium perchlorate in compositions. They
should be ground with extreme care.

Fig. 112. Removal of propellant from mixer to transport vessel (Thiokol Chemical
Corporation, Redstone Arsenal, according to Warren [5]).

By using a suction mill and operating under conditions of controlled humidity,
uniform particles of 2-10 jj. can be obtained. The oxidant is ground only in the quanti-
ties needed for each batch and is never stored in a finely-ground state as the fine
particles have a tendency of "caking" and agglomerating into larger ones. The
grinding operation is not without its hazards and extreme care must be exercised,
especially with the perchlorates.

376 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Mixing

A "premix" is prepared by mixing the fuel binder (i.e. a polysulphide rubber
in the case of the Thiokol propellant) with the necessary curing agents, inert addi-
tives, and ballistic modifiers in a double-acting paddle mixer. This mixer is equipped

Fig. 113. Casting propellant around a star-shaped core, according to Dykstra [11].

with a paddle rotating on a shaft located off center in the mixing vessel. As the shaft
rotates it also revolves about the central axis of the vessel, so that it operates in
all parts of the vessel in turn. Figure 1 1 1 shows the paddle mixer in operation. (Dif-
ferent fuels may require different mixing systems.) When the premix has been thor-
oughly blended it is transferred to a larger mixer.

The final mixing takes place in a large Werner-Pfleiderer (or Baker-Perkins)
type sealed mixer of the type extensively used in smokeless powder manufacture
(Fig. 206). A capacity up to 2000 lb is used in the case of thiokol propellants. After
the premix has been added to the mixer, the finally ground oxidant is added. This
material is conveyed to the mixer in a closed container and is added so that the

COMPOSITE PROPELLANTS FOR ROCKETS

377

container and mixer remain cut off from the outside air. Mixing is continued at a con-
trolled temperature until the mixture becomes uniform. The duration of the mixing
depends on the material used. The blended material is very viscous, but still pourable

Fig. 114. The rocket chamber with the cured propellant is removed from the curing
oven, according to Dykstra [11].

and is poured into a transfer vessel for transportation to the casting room (Fig. 112).
Here, it should be de-aerated under vacuum prior to casting to ensure uniform burning
characteristics.

378 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Casting

There are several ways of filling the rocket chamber with the propellant. It can
be introduced through a long tube ("bayonet") which is lowered into the chamber
or it can be added through the bottom of the tank. The material is cast into the

Fig. 115. X-Ray inspection of the 65.5 in. dia. cast propellant, according to Dykstra [11].

chamber around the core (Fig. 113) or the core is lowered into the chamber after
the propellant has been cast in the rocket chamber. The loaded motor chamber is
then sent to the curing ovens.

Curing

The filled chamber is placed in an oven where the temperature is gradually
raised at a given rate to a definite temperature (sometimes up to 150°C). At this
temperature it is held for a determined time, the oven is then cooled to ambient
temperature and the chamber removed from the oven (Fig. 114). Curing rate and
curing time must be carefully controlled to obtain the desired physical and ballistic
characteristics. After curing is complete, the cores are removed and excess pro-
pellant cut away. The "grain" is now ready for final inspection.

Inspection

Inspection during the course of the manufacture is extremely important. By
chemical analysis the quality of most of the ingredients is maintained within specified
limits. Visual inspection is made of chamber surface and lining condition before
casting. Sieve analysis should be used-to check the particle size of the oxidant. As
with most plastic materials, control of the quality of the product can be assured
only by careful control of the variables affecting the polymerization of the material.

COMPOSITE PROPELLANTS FOR ROCKETS

379

Fig. 116. Cross-section of a Thiokol propellant for 14 ft dia. rocket 63 ft in length [14].

Temperature, time, rate of mixing and quantity of ingredients are important varia-
bles.

Once the materials have been mixed, it becomes very difficult to analyse quan ita-
tively for individual ingredients. This is especially true when curing is complete.
After curing, the propellant grain and chamber pass through a series of inspections
including weighing, radiography, X-ray (Fig. 115), fluoroscopic examination, gauge

380 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

measurements, and checking with optical comparators. Small rockets can be given
a final check by the static or flight testing of a random sample. With large rockets,
such testing becomes too expensive and must therefore be kept to a minimum.
Inspection for storage stability, a most important factor in the case of the nitric
esters, is a different matter with composite propellants. These compositions unlike
nitrocellulose do not undergo spontaneous decomposition. However, composite
systems do undergo a change in character during storage. Burning properties and
physical characteristics (e.g. tensile strength) may change continually with time.
By improving the composition and curing techniques these changes can be reduced
considerably. Figures 116 gives a cross-section of a thiokol propellant "grain" of a
14 ft dia. rocket.

MIXTURES OF PERCHLORATES WITH OTHER ELASTOMERS

Among the possible alternatives synthetic rubber may be mentioned. In particu-
lar the mixtures of potassium or ammonium perchlorates with butadiene-styrene co-
polymer are recommended.

There is, however, limited information on the method of manufacture and on
the properties of these compositions.

Among various elastomers recently recommended as combustible ingredients
of composite propellants polyurethanes are particularly important and promising
[15]. Three basic building blocks are used for polyurethane-rubber fuel-binder:
di-isocyanates, low-molecular weight triols and long-chain diols— linear organic com-
pounds terminated at each end by hydroxyl groups with molecular weights from
1000-2000.

New copolymer diols derived from ethylene oxide and tetrahydrofuran yield
linear polyurethanes of superior physical properties [16].

The chief problem in the manufacture of polyurethane propellants lies in deter-
mining the point where the liquid, uncured propellant is reduced to the uniform consis-
tency necessary to obtain reproducible ballistic performance.

MIXTURES OF PERCHLORATES WITH PLASTICS

Among the plastics which in addition to being combustible ingredients of these
mixtures serve also as binders imparting mechanical strength to the charges, the
following substances have recently become of considerable interest: methyl meth-
acrylate, polystyrene.

The use of polyethene has also been mentioned. So far the best-described
mixtures are those with methyl methacrylate. Mixtures with ammonium perchlorate
go by the name of Aeroplex N, those with potassium perchlorate are known as
Aeroplex K.

Among the few papers on the physico-chemical properties of these mixtures
that of Alterman and Katchalsky [17] is noteworthy as it contains data on a number

COMPOSITE PROPELLANTS FOR ROCKETS

381

of solid physico-chemical mixtures of methyl polymethacrylate with potassium
perchlorate.

The figures obtained by the authors are given in Table 86.

Table 86

Composition
(% by weight)

Density
g/cm3

Specific
heat

cal/g

Heat of
decom-
position
kcal/kg

Flame
temper-
ature
°K

Ignition

temperature

with 20 sec

delay

°K

Rate of burning
(cm/sec) at

Methyl

KC10 4

pressures of:

polymethacry-
late

30
atm

50
atm

100
atm

20
22.5
25
30

77.5

1.88
1.88
1.86
1.82

0.241
0.247
0.254
0.267

800
832
859
828

3750
3770
3778
3518

925
927
930
936

1.41
1.43
1.38
1.17

2.34
2.33
2.12

5.48

The very high flame temperature of the above mentioned mixtures is noteworthy.
In the majority of smokeless nitrocellulose or nitroglycerine powders it is consider-
ably lower: 2000-3000°K.

Attention has recently been paid to the possible use of alkyd type polyester
resins as combustible binders of composite propellants. The use of polyester resins
from maleic, adipic or phthalic acid has been suggested. Their main advantage is
that they can be cured in the cold by the addition of styrene or diallyl phthalate.
Curing the charges can thus be performed at a relatively low temperature.

These binders possess properties which are very important for ease of manufac-
ture and for obtaining a uniform product. Thus they have low viscosity on casting
without sedimentation of solids, have sufficient reactivity for complete low-temper-
ature cure and evolve little heat on polymerization (e.g. much lower than heat of
polymerization of methyl acrylate).

Noteworthy among the few reports in the literature on this subject is the work
of Andersen, Bills, Mishuck, Moe and Schultz [3] on the mechanism of combustion
of a mixture of 75% NH 4 C10 4 and 25% polyester with styrene. The work of Gro-
dzinski [18] who investigated the thermal decomposition of the mixtures of various
combustible substances with potassium perchlorate in a ratio of 20/80 by weight,
is also of great interest. The combustible ingredients include asphalt and polyester
resin from unsaturated (maleic) or saturated acids.

The following figures characterize the temperature at which explosion occurred
after the lapse of a certain induction period or "time lag" (see Table 87).

The ignition temperature of a sample of the mixture of KC10 4 with unsaturated
polyester resin introduced into a thermostat heated to 296°C, in relation to time, is
shown in Fig. 117 (curve I). The curve resembles that given by Roginskii for nitro-
glycerine heated at 41 °C in the presence of nitric acid as a catalyst (Vol. II, Fig. 11,
p. 48). Here the temperature of the sample just before the explosion was 298°C,

382

chemistry and technology of explosives
Table 87

Combustible ingredient

Asphalt

Saturated polyester resin

Unsaturated polyester resin

Polyethylene

Paraffin oil

Starch

Cotton linters

Graphite

Active carbon

Carbon black

Temperature i

°c ;

Induction
period

320

340

290

440

265

245

305

315

440 |

i.e. only 2°C higher than the ambient temperature. Curve 77 represents the
result of a similar experiment with a mixture of potassium perchlorate and
ethylene glycol.

20 30

Induction period, min

Fig. 117. Low-temperature explosions of mixtures of potassium perchlorate with some
combustible substances : I— with polyester resin (temperature of the thermostat 296°C),
//—with ethylene glycol (temperature of the thermostat 241 °Q, (according to Gro-

dzinski [18]).

Grodzinski ascertained that the addition of potassium or lithium chlorate to
these mixtures does not change the ignition temperature but reduces the induction
period.

The exothermic decomposition of the mixtures of potassium perchlorate with
charcoal at 300-360°C was studied by Gordon and Campbell [19].

LIIIIIJIIIPIIJ J4fl|UJ.Uim..L

COMPOSITE PROPELLANTSFOR ROCKETS

383

Some technological information on the Aerojet General Corporation's composite
propellants technology was published recently [20]. Ammonium perchlorate is
ground to the particle sizes ranging from 1 to 200 /j and then mixed to form a blend
of the various particle sizes which gives the best mechanical and ballistic properties.
Ammonium perchlorate is mixed with liquid polybutadiene-acrylonitrile fuel, liquid
plasticizer and aluminium powder (Fig. 118). The motor casing is coated internally

Fig. 118. Pre-mixing ammonium perchlorate with liquid polymsr, plasticizer or alu-
minium (Aerojet General Corp. [20]).

with an antirust compound, insulated with silica-filled nitrile rubber and finally
lined with polybutadiene-rubber polymer similar to the propellant fuel and binder
(Fig. 119). The polymer is cured in the usual way. Casting the propellant is done
through the "bayonets" from the casting pot to the case (Fig. 120).

MIXTURES WITH AMMONIUM NITRATE

Mixtures containing ammonium nitrate have recently been suggested as materials
for rocket propulsion. •

According to J. Taylor [21] potassium dichromate is an efficient catalyst of the
decomposition of ammonium nitrate. J. Taylor and Sillitto [22] found that mixtures

384

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

for rocket propulsion, with ammonium nitrate as a chief component, should contain
ammonium dichromate which facilitates the onset of the decomposition of ammonium
nitrate and subsequently supports this decomposition.

These authors suggest the use of fusible mixtures containing ammonium nitrate
(of the type described above— p. 253) from which the propellant "grains" are cast.
The grains in the form of tubes have a high density and are suitable for rockets of
small calibre.

Fig. 119. Preparation of the case (Aerojet General Corp. [20]).

The composition of two mixtures of this kind is given in Table 88.

These mixtures have a density of about 1.7. During decomposition they evolve
1150-1350 ml of gas per kg.

Other mixtures consist of ammonium nitrate activated with ammonium dichro-
mate, for instance, plus a combustible ingredient also acting as a binder.

Little is known about the composition of these mixtures.

Phillips Petroleum in the United States [23] developed a propellant composed
of ammonium nitrate as oxidant and rubber as a combustible and binding
agent. The rubber consists of synthetic rubber and such typical rubber ingredients
as carbon black (to improve the mechanical properties of rubber), an accelerator
and an inhibitor (to prevent oxidation). To endow the rubber with sufficient plasticity

COMPOSITE PROPELLANTS FOR ROCKETS

385

to yield a fairly homogeneous composition on mixing, a certain amount of plasticizer
is added.

Table

Oxidizing
mixture

Mixtures giving
complete combus-
tion

NH4N03

NaN0 3

NH4CI

Nitroguanidine

—

11.6

Dicyandiamide

—

7.2

(NH 4 )2Cr 2 7

Melting point (°C)

115-120

105

Fig. 120. Casting from the casting pot (on the top) down to the case through the
"bayonets" (Aerojet General Corp. [20]).

386 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

As ammonium nitrate is a rather slow oxidant small quantities of a catalyst,
presumably of potassium dichromate, are added to it.
The composition of the powder is :

Ammonium nitrate

83%

Combustion catalyst

2.3%

Synthetic rubber

10%

Carbon black

Curing substance

0.4%

Plasticizer

2.0%

Inhibitor

0.3%

The manufacture of this powder consists of the following operations.

Milling

Ammonium nitrate is milled until granules with a particle size ranging from
5 to 500 fi are obtained. Fineness of grinding governs the rate of burning of the
mixture.

Synthetic rubber is mixed with such ingredients as carbon black, plasticizer,
accelerator and inhibitor. Mixing is carried out on the rollers commonly used in
the rubber industry.

Final mixing

The milled ammonium nitrate and combustion catalyst are added to the fuel
so formed, which at the same time acts as a binder. This final stage of mixing is
carried out between rollers heated to 50-60°C. The temperature should not exceed
60°C. The final mixing last for \-2 hr.

Pressing

The hot, homogeneous mass is shaped in a hydraulically-operated extrusion
press of the type used for smokeless powder (Fig. 121). It is usually cruciform.
A guillotine cuts the extruded material to the required length. The outer surfaces
of the limbs of the cross are covered by a substance which does not burn readily
(e.g. strips of plasticized cellulose acetate or polystyrene 1.5-5 mm thick), and
cemented in place, to prevent uneven burning at the surface.

Curing

The final operation in propellant manufacture is the curing of the binder. The
shaped material is placed in a curing oven for 16-48 hr at 80-1 10°C. The temperature
and duration of this operation depend on the composition of the mixture, the di-
mensions of the charge and the physical properties desired.

COMPOSITE PROPELLANTS FOR ROCKETS

387

After curing, the charges are given their final dimensions. In some rockets a
high dimensional tolerance is admissible, and the final dimensions may be imparted
before curing.

Dekker and Zimmerman [24] described a cast ammonium nitrate propellant
containing polyester styrene-methyl acrylate binder.

i ■- ■::■>■■■?. ■'&!

Fig. 121. Extrusion press for extruding ammonium nitrate composite propellants,
Astrodyne, Inc. (according to Warren [5]).

A typical composition of AMT-2011 propellant was:

Ammonium nitrate 72.79%

Genpol A-20 polyester resin 9.79%

Methyl acrylate

12.22%

Styrene

2.22%

Methyl ethyl ketone

0.49%

Cobalt octoate

(1 % in styrene)

0.25%

Lecithin (10% in styrene)

0.25%

Ammonium dichromate

1.99%

The binder itself (Genpol A-20 polyester resin, styrene and methyl acrylate) was
combined with the necessary polymerization catalyst (methyl ethyl ketone peroxide)
and an accelerator (cobalt octoate or naphthenate).

The viscosity of the binder was adjusted by varying the ratio of polyester to
monomer so as to achieve a castable composition (Fig. 122). Lecithin was used to
reduce the viscosity.

The proportion of the oxidant could be as high as 82%. This gave a limit of
pourability of the mixture. This high content of solid oxidant could be achieved by

388

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Polyester resin

Styrene

Methyl acrylate

Fig. 122. Binder composition and viscosity. Smokeless and castable compositions are
indicated with an arrow (according to Dekker and Zimmerman [24]).

Coarse 100
Fine

Fig. 123. Effect of blends of fine and coarse ammonium nitrate on viscosity of a pro-

pellant containing 64% by volume solids and 0.025% by weight lecithin. Binder

vicosity was 20 cP (according to Dekker and Zimmerman [24]).

COMPOSITE PROPELLANTS FOR ROCKETS

389

selecting ammonium nitrate of a specific particle size. By using two particle size
ranges in which 70% was coarse and 30% fine (the size of fine crystals was approxi-
mately £ that of the coarse material) it was possible to obtain an oxidant with a
high bulk density. Only a small quantity (18%) of uncured binder was sufficient to
obtain a castable propellant.

Further increase of the proportion of the binder could be obtained by intro-
ducing a third distinctly different particle size.

The stoichiometric mixture was composed of 92% of ammonium nitrate and
8% of the fuel.

The specific impulse was relatively low (ca. 190 sec at 1000 lb/in 2 ).

The propellants have found an application in auxiliary or emergency power
units (JATO— jet assisted take off) for use in aircraft, with good performance in the
range of -75 to 180°F (i.e. -59 to 82°C).

Greek, Dougherty and Mundy [25] give two typical formulations for castable
and extrudable ammonium nitrate propellants (Table 89).

Table 89
Typical propellants with ammonium nitrate as oxidizer

Castable

Extrudable

Liquid polymer

10.8

Rubber polymer

12.0

Filler

—

2.5

Plasticizer

3.0

2.5

Curing agent

1.0

0.5

Anti-oxidant

0.2

0.4

Metal powder

16.0

Oxidizer

68.0

80.0

Catalyst

1.0

2.1

A general lay-out of the plant was given by the same authors (Fig. 124).

These propellants possess several advantages : they give smokeless exhausts, are
non-corrosive and non-erosive, and have a low rate of burning and low flame tem-
perature.

NEW METHOD OF MIXING INGREDIENTS OF COMPOSITE PROPELLANTS

A process for continuous pneumatic mixing of ingredients of solid composite
propellants has recently been developed by the U.S. Naval Propellant Plant, Indian
Head, Md., according to A. J. Colli [26]. The solid and liquid ingredients (oxidizers
and prepolymer respectively) are conveyed pneumatically through a porous tube.
Air flowing into the tube through the pores provides turbulent mixing. The process
is shown diagrammatically in Fig. 125.

390

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Personnel tunnel -v Service road

Warehouse

Research

exotic
materials
and high
energy test cells'

Materials and
utility tunnel

Typical
section

Casting, forming,
extruding
(cure -cool)

'>r Mandrel pull
• ' (cure-cool)

Machining

Non-destructive
Assembly ®

Fig. 124. Lay-out of the plant for ammonium nitrate composite propellants [25].

Crating and
shipping

Solids-liquid dispersion -

Air for mixing
Prepolymer

Solids with
gas carrier

Prepolymer
Air for mixing

Cross section

Manifold
Fig. 125. Continuous mixer of composite propellants' components [26].

_The mixture leaving the apparatus must be de-aerated. The air is removed from
the solids-liquid dispersion by a centrifugal separator before the propellant is vacuum
cast and cured.

The main purpose of utilizing a continuous process is to achieve a high degree
of safety with a high output: only a very small quantity of material is present in the
mixer at any given time, and the ingredients are in contact only for a fraction of a
second. While handling only 1 lb in the mixer an output of 5000 lb per hour is
achieved.

The carrier gas moves solids and droplets at random, providing intimate mixing.
The gas also prevents the material from sticking to the wall of tube. As the solid

COMPOSITE PROPELLANTS FOR ROCKETS

391

particles and associated liquid droplets move through the tube, uneven radial distri-
bution between two phases disappears.

Extensive research on the burning mechanism of a mixture containing ammonium
nitrate has been reported by Chaiken [2] and Andersen, Bills, Mishuck, Moe and
Schultz [3].

These studies were based on the "two temperature" theory of propellant burning.

According to this theory it is assumed that in a composite propellant containing
ammonium nitrate the oxidant starts to gasify at a temperature of about 600°K due
to the endothermic reaction resulting in the formation of ammonia and nitric acid.
These gaseous products then undergo exothermic redox reactions giving a flame of
about 1250°K in the vicinity of burning substance. This temperature brings about
the pyrolysis of the organic binder. The gaseous products of this reaction react in
turn with the gaseous products of pyrolysis of the oxidant to create a hot flame of
about 2400°K at a certain distance from the surface of the charge.

High temperature
flame zone
(-2400°K)

Binder

Fig. 126. Thermal layer model of combustion of solid composite propellant with am-
monium nitrate, according to Chaiken [2]; R— redox reaction flame zone (temperature
T[), u— gas velocity, <5— thickness of the thermal layer, T s — surface temperature of
oxidizer particle, ro— radius of oxidizer particle.

The diagram (Fig. 126) given by Chaiken represents a simplified picture of
this mechanism.

Andersen et ah [3] have examined a number of the mixtures of ammonium

392

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

nitrate and polystyrene with various molecular weights and determined experiment-
ally the average temperature of the surface of the oxidant (573-599°K) and that
of the binder (from 796°K for low-molecular polystyrene MW= 142,000 to 979°K
for high-molecular polystyrene MW= 500,000).

In mixtures with methyl methacrylate the corresponding figures are: 562 and
950°K.

In mixtures with copolymer polyester-styrene the temperature of the oxidant
is 582°K and that of copolymer 1020°K.

Andersen et al. obtained a similar result for mixtures of ammonium nitrate
with copolymer polyester-styrene-acrylate and with copolymer butadiene-styrene.

They also suggest that the same mechanism applies to a mixture of ammonium
perchlorate with copolymer butadiene-styrene.

The rate of burning of mixtures of ammonium nitrate with polymers at a pressure
of 1000 lb/in 2 {ca. 67 kg/cm 2 ) at 60°F (ca. 15°C) ranges within 0.21-0.36 cm/sec.

VARIOUS COMPOSITE PROPELLANTS AND THEIR CHARACTERISTICS

Napoly [27] collected the information on various composite propellants. They
are given in Table 90.

Table 90
Characteristic of some composite propellants [27]

Oxidizer
Combustible binder

NH4CIO4

Polybuta-
diene

NH4NO3

Cellulose
acetate

NH4CIO4
Polyurethane

NH4C104

NH4CIO4

Polyurethane

Polyester

14-125

30-140

236

178

12.2

17.5

0.479

0.69

1.74

1.88

NH4C104

Polyvinyl
chloride

Burning under

pressure (kg/cm 2 )
Specific impulse I s

at /> = 70 kg/cm 2
Rate of burning

(mm/sec)
under pressure

/>= 70 kg/cm 2
Exponent n in the

equation V=kp n
Density

1-140
250

11.9

0.236

1.74

1-140

171

2.2

0.50
1.55

1-140
238

0.5
1.72

1-120

225

6.5-13.5

0.4
1.64

The same author [27] gives some data on a propellant which is intermediate
between composite and double base propellants. It is composed of ammonium
perchlorate and a nitroglycerine-nitrocellulose powder (double base powder).

It gives a very high specific impulse 7 S = 250-255 (at 70 kg/cm 2 ), and has a
high rate of burning: 17.3 mm/sec (at the same pressure), exponent «=0.45 and
density 1.75.

COMPOSITE PROPELLANTS FOR ROCKETS
MIXTURES WITH AMMONIUM PICRATE

393

During World War II mixtures were developed in Great Britain with ammonium
picrate as the chief component for rocket propulsion, on the suggestion of the
author of the present book. These mixtures also contained sodium or potassium
nitrate and a combustible binder.

Similar mixtures were adopted in the United States. Zaehringer [6] quotes the
following two compositions (Table 91).

Table 91
compostion of ammonium picrate propellants

Name

Ingredients

NDRC

Type EJA

218 B

NDRC

Type EJA

480

Ammonium picrate
Sodium nitrate
Buramine resin*
Ethyl cellulose-arochlor**
Santicizer 8***

46.6

5.2

1.6

46.4
46.4

7.2

* Thermosetting urea or melamine resin.
** Arochlor-chlorinated diphenyl.
*** A plasticizer.

Riihl [28] has recently reviewed the patent literature on solid rocket propellants.

EXPLOSIVE PROPERTIES OF COMPOSITE PROPELLANTS

Composite propellants with elastomers (thiokol, polyurethane) do not detonate
readily and this is due to their non-porous texture [29].

Most of the cast propellants' charges also possess very high density and thanks
to this they are difficult to detonate (e.g. Price and Jaffe [30]). They are prone to
detonation when they are shredded and put into vessels of high degree of confine-
ment [31].

LITERATURE

1. E. Mishtjck and L. T. Carleton, bid. Eng. Chem. 52, 755 (1960).

2. R. F. Chaiken, Combustion and Flame 3, 285 (1959).

3. W. H. Andersen, K. W. Bills, E. Mishuck, G. Moe and R. D. Schultz, Combustion and
Flame 3, 301 (1959).

4. W. R. Maxwell and G. H. S. Young, /. Roy. Aeron. Soc. 65, 252 (1961).

5. F. A. Warren, Rocket Propellants, Reinhold, New York, 1958.

6. A. J. Zaehringer, Solid Propellant Rockets, American Rocket Co., Wyandotte, 1955.

394 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

7. T. L. Smith, Ind. Eng. Chem. 52, 776 (1960).

8 R T. Landel and T. L. Smith, /. Am. Rocket Soc. 31, 599 (1961).

9 Thiokol Chemical Corporation, Trenton 7, New Jersey, Liquid Polymer LP-31 and Liquid
Polymers LP-2 and LP-32.

10. G. GOBEL, Kunststoffe 49, 56 (1959).

11. P. R. Dykstra, Thiokol Astronaut 3, No. 2, 5 (1961).

12. J. C. Chapman, Thiokol Astronaut 1, No. 1, 3 (1959).

13 C Huggett, Combustion of Solid Propellants, Section M in Combustion Processes, Ed. B.
Lewis, R. N. Pease and H. S. Taylor, Princeton University Press, Princeton, N.Y., 1956.

14. Thiokol Astronaut 3, No. 1, 20 (1961).

15. H. E. MARSCH, Jr., Ind. Eng. Chem. 52, 768 (1960).

16. W. J. MURBACH and A. ADICOFF, Ind. Eng. Chem. 52, 772 (1960).

17. Z. ALTERMAN and A. Katchalsky, Bull. Res. Israel 5A, 46 (1955).

18. J. GRODZINSKI, J. Appl. Chem. 8, 523 (1958).

19. S. Gordon and C. Campbell, Vth Symposium on Combustion, p. 277, Reinhold, New York,

1955.
20 Chemical & Engineering News 42, Sept. 28, 50 (1964).

21. J. Taylor and I.C.I. Ltd., Brit. Pat. 453210 (1936); U.S. Pat. 2159234 (1939).

22. J. TAYLOR and G. P. Sillitto, Illrd Symposium on Combustion, p. 572, Williams & Wilkins,
Baltimore, 1949.

23. C. F. DOUGHERTY, Chemical & Engineering News 35, No. 40, 62 (1957).

24. A. O. Dekker and G. A. Zimmerman, Ind. Eng. Chem., Products Res. and Dev. 1, 23 (1962).

25. B. F. Greek, C. F. Dougherty and W. J. Mundy, Ind. Eng. Chem. 52, 974 (1960).

26. Chemical & Engineering News 41, Oct. 5, 48 (1964).

27. C. NAPOLY, Mem. poudres 41, 331 (1959).

28. G. RUHL, Explosivstoffe 13, 8 (1965).

29. A. B. Amster, E. C. Noonan and G. J. Bryan, /. Am. Rocket Soc. 30, 960 (1960).

30. D. PRICE and I. JAFFE, /. Am. Rocket Soc. 31, 595 (1961).

31. Bureau of Mines Annual Rep. No. 3647 (1957); Bureau of Mines Final Summary Rep. No.
3734 (1958); Bureau of Mines Annual Rep. No. 3769 (1959); Bureau of Mines Quarterly Rep.
January 1 to March 31 (1961).

CHAPTER V

MINING EXPLOSIVES

Blackpowder is a weak explosive, too slow in action to be an effective blasting
agent. Attempts have therefore been made to replace it by a more powerful explosive
(e.g. to substitute potassium chlorate for potassium nitrate). At first, however,
this produced nothing of value. It was not until the second half of the nineteenth
century when nitroglycerine had been produced that new prospects of improvement
emerged. The fact that nitroglycerine is a more powerful explosive than black-
powder was perceived at once, although how this property could be exploited fully
was not immediately clear. Both blackpowder and nitroglycerine were fired with a
fuse. A blackpowder primer was used with nitroglycerine. In both cases the explo-
sives were ignited and then, since they were burning in a confined space explosion,
but not detonation, occurred. Yet nitroglycerine is a typical explosive capable of
detonation. Thus the potential energy released by the detonation of nitroglycerine
was not utilized. The proper exploitation of nitroglycerine as an explosive became
possible only when detonators filled with mercury fulminate were introduced (1867).

Nobel's use of kieselguhr to produce dynamite [1] (Vol. II, p. 33) was a milestone
since nitroglycerine thus acquired a solid form easy to handle. Guhr dynamite,
however, failed to utilize fully the explosive power of nitroglycerine due to the
presence of a large proportion (25%) of the inert material kieselguhr. The discovery
of blasting gelatine [2] and of dynamites containing blasting gelatine as a chief
ingredient was a step forward in improving mining explosives.

Another advance in this field was the use of ammonium nitrate as a chief com-
ponent. In 1867 Ohlsson and Norrbin [3] patented a mining explosive— Ammoniak-
krut (ammonium powder, cf. p. 331) consisting of ammonium nitrate mixed with
5-10% charcoal. Ohlsson and Norrbin added to this mixture 10-30% nitroglycerine
to make detonation easier and to increase the power of the explosive. Similarly
Nobel began to add ammonium nitrate to his dynamites.

The largest application of explosives is now in coal mines. They are also widely
used in mining ores, in quarrying, for many civil engineering works such as road
building, tunnel driving, land reclaimation, canal construction, and changing the
course of rivers— even for extinguishing fires (e.g. conflagrations of oil wells). In
recent years large quantities of explosives have also been employed in seismographic
prospecting for new oil fields.

[395]

396 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

When shotfiring in mines, quarries, rock blasting, the demolition of old build-
ings etc. the first step is to bore a cylindrical shothole. The hole is loaded with one
or more blasting cartridges (usually a cartridge is 100 g in weight and 30 mm dia.).
The first cartridge is fitted with a detonating cap or an electric detonator (in coal
mines charges are fired only by electric detonators), and the shothole sealed with
stemming material, commonly moist clay.

An explosive intended for use in coal mines must be safe to handle and to operate
in the presence of material as ignitable as coal. Moreover to produce good lump
coal slow-acting explosives should be employed which displace rather than fragment
the coal. Conversely, in rock blasting greater explosive power is needed; the safety
factor is less important.

RESEARCH ON THE SAFETY OF MINING EXPLOSIVES

The increasing use of explosives in mining greatly increased coal output but
numerous gas explosions occurred in mines where blackpowder was used. The
development of nitroglycerine and more modern explosive compositions did not
eliminate this hazard. Asjthejndustry developed and the demand for coal increased
the safety of explosives used in mines become so urgent and important a problem
that in the nineteenth century many countries set up special commissions charged
with its detailed analysis.

Although firedamp explosions in mines were notified as early as the seventeenth
century in the oldest scientific establishments -the Academy of Paris and the Royal
Society of London-the scientific and technical world in general was little concerned
with them. But in 1812 an explosion at Branding Main near Gateshead-on-Tyne
in England in which ninety-two miners were killed was given great publicity. The
disaster led to the formation in 1813 of "The Sunderland Society" for preventing
accidents in coal mines. The Society succeeded in persuading Sir Humphrey Davy
to interest himself in the problem and in 1815 he devised the welf known safety "
lamp. In 1849 a further step was taken in Great Britain with the passing of an Act
for the Inspection of Coal Mines in Great Britain from which the modern era in
coal mining really began (cited by J. Taylor and Gay [4]). In 1877 the Commission
de Gnsou was formed in France. Work began on the elucidation of the conditions
under which mixtures of air and methane explode (Mallard and Le Chatelier [5]).
It turned out that these mixtures explode at 650°C after a certain induction period
which, at this temperature, is 10 sec. At a higher temperature it is shorter (e.g.
at 1000°C it is approximately 1 sec). Although the temperature of explosion of any
explosive is considerably above 650°C, the gaseous products of explosion cool
rapidly due to expansion, and furthermore a large amount of the heat of explosion
is converted into mechanical work on bursting the walls of the shothole. Thus, the
explosion of a high explosive does not necessarily produce a temperature at which
a mixture of air and methane will explode.

MINING EXPLOSIVES 397

The duration of the flame produced by the explosion is another important factor.
The flame of a high explosive is of extremely short duration and may not be sufficient
to ignite firedamp. Conversely the flame produced by burning blackpowder lasts
much longer and is therefore more dangerous. Clearly, the duration and temperature
of the flame of explosion are the factors which determine the safety of an explosive
used in mines.

On the basis of many experiments the Commission de Grisou and later, the
Commission de Substances Explosives headed by M. Berthelot and Le Chatelier
inti oduced, in 1 890, the following safety rules for the use of explosives in coal mines [6] :

1. On explosion the explosive must not leave any combustible products such
as CO, H 2 , carbon.

2. The detonation temperature computed from the heat of explosion and from
the average specific heat of the products must not exceed 1900°C for explosives
intended for penetrating rock and 1500°C for those to be used in coal mines.

On these grounds blackpowder was excluded from coal-mining works, for it
is an explosive that leaves a combustible residue and has a very high explosion
temperature (approximately 2400°C).

The theoretical calculation of the detonation temperature as a safety criterion
was adopted only in France. In all other countries (and more recently also in France)
a practical criterion has been introduced based on an experimental evaluation of
the effect of explosion of a sample of the explosive on a mixture of air and methane,
or on a suspension of coal-dust in air, under conditions similar to those existing in
mines. For this purpose testing galleries were devised, simulating the conditions of
mine galleries.

A typical one (Fig. 149, p. 441) consists of a cylindrical or elliptical tube
made of wrought iron. At one end the tube is closed with a thick plate simulating
the coal face. It is provided with a steel mortar, which imitates the shothole.
A section of the gallery containing the mortar loaded with an explosive charge is
separated by a paper diaphragm from the remainder to form an explosion chamber
which is then filled with a defined amount of methane. The explosive charge is
fired electrically with a No. 8 detonator. The course of the explosion is watched
through the observation holes along the gallery.

The first testing gallery was built in Germany, at Gelsenkirchen in 1880. It was
a tube of elliptic cross section of 2 m 2 , 35 m long. Next, about 1890 another testing
gallery was built in Great Britain, at Hebburn-upon-Tyne, by The North of England
Institute of Mining and Mechanical Engineers. It consisted of a wrought-iron tube,
101 ft long and 3 ft dia., with a paper diaphragm 22£ ft from the closed end forming
an explosion chamber. The 1^ in. dia. mortar was 42 in. long, the explosive being
fired both stemmed and unstemmed (according to J. Taylor and Gay [4]).
• The important conclusion drawn from investigations with this gallery was that
high explosives were less liable than blackpowder to ignite an inflammable mixture
of air -methane and coal-dust.

Another testing gallery was built in Great Britain at the beginning of this century

398 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

at Althofts by the Mining Association of Great Britain [7]. It was over 200 m long,
2.30 m dia. and served mainly for testing the explosibility of coal-dust and methods
of preventing it (such as stone-dust, stemming with sodium hydrogen carbonate, etc.).

A Home Office Testing Gallery Committee was appointed in 1896. The Com-
mittee's aim was to establish the best test for determining the safety of explosives
in coal mines. On its recommendation a testing station was erected at Woolwich
and the testing of explosives for inclusion in the "Permitted List" began in 1897.
The testing gallery was 27| ft long and 2\ ft dia. and was filled with an inflammable
medium: (15% coal gas +85% air). The steel mortar had a bore of \\ in. dia.

The charges were stemmed with dry clay.

The Home Office testing station was transferred to Rotherham in 1911 and a
new gallery was erected. It was 5 ft dia., with an explosion chamber 18 ft long.
The steel mortar was 120 cm long with a 55 mm bore.

In 1921, difficulties with the supply of coal-gas at Rotherham led to the tempo-
rary transfer of testing to Ardeer and eventually a new Research Station was estab-
lished at Buxton in 1922.

In Belgium the official testing gallery was erected at Frameries under direction
of Watteyne and Stassart in 1902. It was of elliptical cross section (1.85 m/1.40 m),
85 m long.

In France two testing galleries were built at Lievin in 1907. Both were cylindrical.
One of 2 m 2 cross section and 15 m long was used for testing explosives. Another
one of 2.80 m 2 section was 300 m long and was used for examining coal-dust
explosion and methods of preventing it. In Austria-Hungary two galleries came into
use in 1908.

In the U.S.A. a cylindrical gallery of 6 ft 4 in. dia. and 100 ft long was built
1909 in Pittsburgh.

Modern testing galleries are described later (p. 439 and Table 108).

In addition to the experiments in testing galleries laboratory research has also
been conducted, including a study of the flames produced by explosives when explo-
ded. These flames have been studied photographically to determine their dimensions,
intensity and duration. The following papers are noteworthy : Siersch [8], Bichel and
Mettengang [9], Wilkoszewski [10], Will [11], Taffanel and Dautriche [12].

When observing the flame projected from the mortar of a testing gallery Lemaire
[13], Payman [14] and Audibert [15] came to the conclusion that the ignition of air-
methane mixtures may also be caused by glowing particles formed on explosion and
thrown out of the shothole.

On the basis of flame studies, explosives may be divided into two principal groups:

(1) Explosives with a secondary flame which have not enough oxygen for com-
plete combustion. On the detonation of these materials combustible products are
formed such as CO, H 2 , CH 4 etc. These products become mixed * with air, producing

* Oxygen-positive explosives may also produce combustible products, although in negligible
quantity. The amount of such products is greater on detonation in the open.

MINING EXPLOSIVES

399

combustible gaseous mixtures which explode in turn giving rise to a bright secon-
dary flame (Fig. 127). Trinitrotoluene, picric acid and guncotton belong to this
group. Blackpowder which does not detonate but only explodes also gives a second-
ary flame.

**»

Fig. 127. Flame from picric acid. The primary flame (relatively small) is visible with

the secondary flame above it. The duration of both flames is shown on the scale: the

primary flame of short duration, the secondary of long duration. The material was

fired in a mortar (according to Will [11]).

(2) Explosives with a primary flame only. This group comprises all mixtures
containing more than enough oxidizing agents to provide a positive oxygen balance.
They differ from each other in flame intensity. Rock explosives (e.g. dynamites,
i.e. explosives with nitroglycerine) give a glaring flame of high-temperature and
relatively long duration (Figs. 128 and 129). Permitted Explosives* give a faint,
low- temperature flame of short duration (Fig. 130).

* English term for coal-mining explosives which have passed special tests as explosives as-
suring a sufficient safety for use in coal mines where inflammable mixtures of methane and air
occur. In the U.S.A. they are called permissible explosives.

400

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

As a result of these investigations, group (1) explosives were withdrawn from
use in coal-mines. The duration of the secondary flame is rather long so that the
gaseous mixture ignites easily.

Rock-blasting explosives of group (2) have a very high temperature of explosion.
Among these dynamites have been permitted exclusively for work in rock and in
non-fiery mines.

Fig. 128. Flame from Dynamite I (61-63% nitroglycerine). Duration 7.66 millisec. The
material was fired in a mortar (according to Bichel [16]).

Fig. 129. Flame from dynamite having a
low nitroglycerine content (Gelatine-Dyna-
mite). Duration 1.25 millisec (according
to Bichel [16]).

Fig. 130. Flame from ammonium nitrate

explosive (Ammoncarbonit). Duration 0.33

millisec (according to Bichel [16]).

The ammonium nitrate explosives of the same group may be fired without igni-
ting a methane-air mixture even when relatively large charges are used whereas
other explosives ignite the gas mixture unless used in small amounts. Thus arose the
concept of the maximum charge that can be fired without causing the ignition of
a definite methane-air mixture in a testing gallery. This is called the "charge limit".
This made clear that apart from the temperature of the flame of explosion (calcu-
lated from the heat of reaction and the mean specific heat of the products) there

MINING EXPLOSIVES 4Q|

are other agents which may provoke an explosion. One factor that increases the
danger is a very high rate of detonation and "specific pressure" /or lead block
expansion. The shock-wave produced by the detonation may involve a rapid almost
adiabatic compression of the methane-air mixture. The higher the rate of detonation
the greater is such compression and, in consequence, the probability of explosion of
the gas mixture. In addition, when the energy of the explosion of a charge is trans
formed into work only to a small extent, it may happen that the shock-wave thus
produced involves the adiabatic compression of the gas mixture referred to This
means that it is dangerous to load shotholes with a larger amount of explosive than
is necessary for the work in hand. The unused surplus of energy may cause an acci-
dent since a large quantity of heat is not transformed into work and the products
of detonation then have a higher temperature. This proves the necessity of strict
observance of the maximum safety charge (charge limit). However, the concept of
charge limit was recently subjected to revision (see p. 413).

Superficially, this seems to contradict the rule that the use of explosives which
do not detonate but only explode (or burn) like blackpowder, as mentioned above
is unsafe, but in fact this is not so. The point is that both non-detonating explosives-
with a flame of long duration flike blackpowder) and those marked by high rate
of detonation producing a violent shock-wave are unsafe. The safest explosives are
those with intermediate properties, i.e., which give a not very high rate of detonation
and lead block expansion, but in which the reaction of explosive decomposition
proceeds much more quickly than in blackpowder, which burns or explodes rela-
tively slowly.

These investigations showed that no modification of blackpowder by the addition
of "cooling salts", i.e., those that reduce the flame temperature (e.g. oxalates) will
improve its safety. On the other hand, the presence in dynamites of 35-63% of
inert (non-explosive) salts, such as sodium hydrogen carbonate, ammonium oxa-
late, or salts containing crystallization water, increases their safety not only by lower-
ing the temperature of the flame of explosion but also by reducing their explosive
power. Chlorides of alkali metals, such as potassium and sodium, are particularly
efficient in this respect.

It was also found that mixtures containing a large amount of ammonium nitrate
(ammonium nitrate explosives) ensure much greater safety than dynamites or
chlorate and perchlorate explosives.

Early in the present century attention was drawn to the danger created in mines
by coal-dust. Until 1906 it was believed in France that this hazard could be enti-
rely explained by the presence of firedamp, but even in the early nineteenth century
it was supposed that a suspension of coal-dust in the air may also explode. Thus,
as early as 1803, it was suggested in the U.S.A. that coal-dust might be involved in
mine explosions [17]. In 1844, Faraday drew attention to this (according to J. Taylor
and Gay [4]), and later Galloway [18] discussed the coal-dust danger. By the end
of the last century the possibility of explosion was proved experimentally, but it
was believed that an explosion of methane-air mixture is necessary to initiate a coal-

4Q2 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

dust explosion and that conditions were safe if there was no methane in the atmos-
phere. . .

At the turn of the nineteenth century, explosions occurred in Great Britain, in
non-fiery mines, proving the possibility of an explosion of coal-dust without methane
as an initiating agent. It was however insufficient to rouse the experts to action.

Meanwhile, in 1906, a terrible coal-dust explosion occurred at Courriere in
France, in which about 1100 men lost their lives. The mine at Courriere had no
methane at all and was considered to be safe, but as it turned out, it contained coal
which in dust form produced a very dangerous suspension. The vast extent of this
disaster aroused public indignation. It became obvious that the explosion of coal-
dust may extend over a large area under ground and that this is much more dangerous
than the explosion of methane which is usually localized within a small area.

It was also obvious that the immediate cause of explosion of both methane and
coal-dust may be the use of explosives in mines. Particular attention was paid to
the hazard involved in the use of blackpowder in mines (e.g. in Great Britain H.M.
Commissioners [19]; Royal Commission on Explosions [20]).

Attempts were made to reduce the danger in the use of blackpowder by surround-
ing the charge with a water sheath (Abel [21]), but this method was abandoned
as it was very troublesome in practice. Recently it has been revived in an improved

form (p. 489).

Another method was to dip blackpowder pellets in paraffin wax. This rendered
them waterproof, and also surrounded them with a "cooling sheath". A blasting
powder made in this form called "Bobbinite" was introduced in Great Britain.
It will be discussed later. These half-measures brought little improvement and
attention was centred on the use of ammonium nitrate explosives.

After the explosion at Courriere experiments were carried out in testing galler-
ies on blasting explosives and on suspensions of coal-dust in air. As the coal-dust
from various mines showed different inflammability, the most inflammable kind,
i.e., that with a high content of volatile matter, was commonly used for testing.

The explosion of coal-dust is usually of a thermal character. It undergoes degasi-
fication under the influence of a high temperature to form coke grains. If such
grains are found after an underground explosion it is proof that the explosion was
due to coal-dust.

It should however be pointed out that the principal source of colliery accidents
is not explosion but roof falls. Rogers [22] gives the following annual figures for
1923-59 in Great Britain (Table 92):

Table 92

1923-32

1933-42

1943-52

1953-59

Roof falls
Haulage

Explosions and fires
Shafts

500

250

450

170

250

110

166
95
30
15

MINING EXPLOSIVES

403

SAFETY EXPLOSIVES BEFORE WORLD WAR I

Certain types of safety explosives were designed on the basis of a number of
tests, although until recently there were considerable differences in the composition
of these explosives in different countries. This was due to different working con-
ditions in mines, different methods for testing the safety of explosives etc.

After World War II some standardization in composition was achieved largely
as the result of international cooperation. The annual International Conferences of
Directors of Safety in Mines Research greatly contributed to this.

The first safety explosives produced in France are tabulated below:

Table 93
Composition of earliest French safety explosives

Ingredients (%)

Name

Nitroglycerine

Nitrocellulose

Ammonium
nitrate

Potassium
nitrate

For coal work
Grisoutine Couche
Grisoutine Couche au Salpetre

12
12

0.5
0.5

87.5
82.5

5.0

For rock work
Grisoutine Roche
Grisoutine Roche au Salpetre

29.0
29.0

1.0
1.0

70.0
65.0

5.0

These explosives are characterized by an absence of "cooling substances" even
when intended for use in coal mining. In a number of countries they were considered
dangerous.

The same explosives were formerly employed in Russia, where the mining explo-
sives industry was financed by French capital. In both countries Favier Powders
(Poudres Favier) or Favier explosives (Explosifs Favier) were also employed. Their
composition is tabulated below.

Table 94
Composition of Favier explosives

Ingredients (%)

Favier Explosive

Ammonium
nitrate

Sodium
nitrate

Nitronaphthalene

Dinitronaphthalene

1A
IB
2
B

88
67
44

37.5
75

18.5
25

Chlorate explosives were also used in both countries, in mines, viz. : Miedziankit
type in Russia (p. 278) and Cheddite type in France (p. 277).

404

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In Belgium, at first explosives modelled on the French ones were used, but later
they were replaced by others containing "cooling" ingredients, such as ammonium
chloride, sulphate, or oxalate or sodium chloride. The safety explosives used initi-
ally in Belgium included mixtures with ammonium perchlorate, of the Yonckite
type. Examples of these explosives for rock and coal work are given in Table 95.

Table 95
Composition of some Belgian safety perchlorate explosives

Yonckite type,

Yonckite type

Ingredients (%)

brisant

antigrisouteuse

No. 13 (roche)

No. 10 bis (couche)

Ammonium perchlorate

Ammonium nitrate

Sodium nitrate

Barium nitrate

—

TNT

Sodium chloride

In Great Britain the first explosives which passed the test in the testing gallery
at Woolwich were approved for use in mines under the name of Permitted Explo-
sives. At Woolwich the charge of explosive in the test mortar was stemmed with
dry clay, a practice that differed from that followed in other countries.

On the basis of these tests the following substances were considered safe and
were used in mines.

(1) Bobbinite— an explosive having virtually the composition of blackpowder
but with the flame temperature of explosion reduced by the addition of salts or
a mixture of paraffin with starch.

(2) Saxonite— dynamite, made safe by the addition of about 13% ammonium
oxalate.

(3) Monobel— an ammonium nitrate explosive containing about 80% ammonium
nitrate and 10% nitroglycerine, plus 10% wood meal which adsorbed the nitrogly-
cerine.

(4) Faversham Powder— another type of ammonium nitrate explosive, without
nitroglycerine, containing, for instance, 90% ammonium nitrate and 10% TNT.

(5) Cambrite— an ammonium nitrate explosive modelled on the German "Car-
bonits" (see Table 124). It contained a small quantity of nitroglycerine, potassium
or sodium nitrate and a considerable amount of carbonaceous material (e.g. wood
meal, charcoal etc.). This material was added to prevent the complete combustion
of the carbon included in the explosive (to carbon monoxide only), to reduce the
heat of explosion and, in consequence, the temperature of the explosion.

Bobbinite and some of the other explosives permitted for use in mines caused,
however, a number of explosions of firedamp and coal-dust, and it was decided
therefore to adopt a more exacting form of test. In 1912, the Rotherham test was

MINING EXPLOSIVES

405

introduced. In this the explosive charge was fired unstemmed, as in the Continental
test. At the same time the charge limit was determined. These tests led to the exclu-
sion of Bobbinite from the Permitted List (except in some experimental mines)
and to a considerable reduction of power of all the permitted explosives, by the ad-
dition of cooling substances such as sodium chloride, ammonium oxalate etc. In
this way mining explosives having the composition presented in Table 96 were
evolved.

Table 96

Earliest British permitted explosives

Ingredients

Am-
monite

No. 1
Bellite

No. 3
Dynobel

No. 2
Cambrite

No. 3

Samso-

nite

No. 2
Viking
Powder

Tees
Powder

Nitroglycerine

14-16

22-24

50.5-52.5

7.5-9.5

9-11

Nitrocellulose

0.25-0.75

2-4

Dinitronaphthalene

4.5-6.5

Dinitrobenzene

0-0.5

DNT

0.5-2.5

TNT

14-16

Ammonium nitrate

71-75

62-65

51-54

65-69

58.5-61.5

Sodium nitrate

9-11

Potassium nitrate

26-29

Barium nitrate

3^1.5

—

Woodmeal

4-6

32-35

7-9

8-10

Starch

3.5-5.5

Potassium chloride

7-9

—

Sodium chloride

20-22

15.5-17.5

24-26

9-11

14-16

19-21

Borax

24-26

Magnesium carbonate

1.0

0.5-1.0

Moisture

0-1

0-2

3.5-6

0-1.5

0-2

Chlorate explosives, such as Colliery Steelite, which were used for a short time
in coal mines, had the following composition:

72.5-75.5% of potassium chlorate
23.5-26.5% of nitrated resin
0.5-1.5% of castor oil

They were also removed from the Permitted List.

In Austria explosives similar to those of the French Favier type were permitted
for use in mines.

In Germany, slow-action (non detonating) explosives of the blackpowder type
(blasting powder, black blasting powder— see Table 78) were used even in dusty
mines up to 1923 and typical high explosives (1910) with the composition given
in Table 97, were also employed.

The explosives which did not contain cooling salts belonged to the carbonite
or donarit groups and were intended for work in rock or in non-dusty and non-

406

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

gassy coal mines. Those containing cooling salts (Wettersprengstoffe) were used for
work in more dangerous coal mines. The composition of the Gelatine Wetter- Astralit
explosive in which dinitrochlorohydrin was added to nitroglycerine to form an
antifreezing mixture is particularly noteworthy.

Table 97
Early German safety explosives

Ingredients

Donarit

Wetter-
Astralit

Chrome
Ammonit

Ammon
Carbonit

Gelatine
Wetter-
Astralit

Nitroglycerine

. 4

3.9

Dinitrochlorohydrin

Nitrocellulose

0.1

0.5

Nitrotoluene

DNT

TNT

12.5

—

Ammonium nitrate

74.5

Potassium nitrate

Sodium nitrate

7.5

Flour

—

Woodmeal

—

Charcoal

0.5

Vaseline or paraffin

2.5

0.5

Castor oil

—

Chromium-potassium alum

Ammonium oxalate

2.5

Sodium chloride

CONDITIONS OF SHOTFIRING IN MINES

As knowledge of the danger created by the use of explosives in coal mines grew,
attempts were made to obtain new, more suitable explosives and to work out the
safest possible methods for their use, including shotfiring methods. Strict regula-
tions were laid down for charging shotholes, stemming (governing the type of stem-
ming materials, length of stemming etc.), firing charges etc. Wherever danger existed,
due to the presence of methane or coal-dust, electric detonators replaced fuses which,
if defective, might spark through the sheaths. Detailed instructions were also evolved
for firing procedure. Simultaneous firing of the charges in several shotholes (shot-
firing in rounds) was found to be safer than firing in sequence, when one shot might
raise a "cloud" of coal-dust which then might be exploded by the following shot.
On the other hand the use of electric detonators with a delay-fuse increases output
since the early shots can create free faces facilitating the work of the following shots.
It is, therefore, not surprising that short-delay detonators are now in wide use,
very often in combination with shotfiring in rounds of 2-6 shots. Research is still

MINING EXPLOSIVES 4Q 7

under way to ascertain whether when using short-delay detonators there is in danger
of ignition from the possibility that one shot might cause conditions dangerous for
another. The risk may arise due to the opening of fissures that release firedamp
and, as previously described, due to raising a cloud of coal-dust before the last
shot in the round is fired. Some authors concluded that under experimental condi-
tions ignition only occurs when the interval between successive shots exceeds 70 milli-
sec and it was therefore concluded that the usual short delay of 25 millisec should
be considered safe (Fripiat [23]).

Entirely new safeguards were also introduced to avert danger. When it was
found that coal-dust may explode it was suggested that the cartridges would be
stemmed by filling them with water (MacNab [24]) or with a porous mass mixed
with water. The water, dispersed by the detonation, forms a cloud that may reduce
the probability of explosion. Stemming with incombustible substances in powder
form was also tried. On detonation, such stemming was intended to form an air
suspension that would prevent explosion of the coal-dust. Cybulski [25], however
proved the inefficiency of this method by showing that stemming with dust does not
produce a cloud capable of providing protection against explosion. In addition
a powder stemming is ejected more quickly from the shothole than one made of
clay.

Watteyne and Lemaire [26] recommended that outside the shothole, or strictly
speaking outside the stemming, a paper bag should be placed, containing an incom-
bustible material, such as ground sand, salts containing water of crystallization (e g
sodium sulphate), salts volatile at a high temperature (e.g. sodium chloride), or
salts decomposing at a high temperature (e.g. ammonium sulphate). This was later
developed by Lemaire into the use of a sheath of inert substance around the cart-
ridges (see p. 429). *

Numerous tests indicated the factors that influence the results of shotfiring and
showed, in particular, that variations in results are caused by changes in the compo-
sition of the methane-air mixture (e.g. when gas produced by coal carbonization,
which is liable to variations in composition, or natural gas which varies in composi-
tion depending on its origin, was used for testing instead of pure methane). This
emphasized the need to use a test gas of constant composition.

It was also found that the coal-dust in the presence of which explosives were
tested varied in ignitability; to achieve comparable results therefore coal-dust
Irom the same seam of the same mine must be used. If the same explosive is always
used, the ignitability of coal-dust in a given mine may be tested to determine to what
extent work in the mine is dangerous. It is obvious that changes of atmospheric
humidity also affect the results.

Numerous tests were also carried out to examine other, less tangible factors
which affect the accuracy and reproducibility of the results obtained in a testing
gallery. The following factors were found to be of importance:

(1) The density of loading of the shothole (greater density facilitates ignition of
methane or dust);

408

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(2) The diameter of the explosive charge (larger diameter favours easier igni-
tion) ;

(3) The shape and surface area of the testing gallery (elliptical shape and smaller
surface favour easier ignition);

(4) The wrapping of the explosive charge, e.g. paraffined paper, parchment
paper, tinfoil etc. (this factor was tested over a long period and some of the results
are discussed further on p. 424).

MINING EXPLOSIVES USED DURING WORLD WAR I

In World War I the belligerents reserved nitric acid and its salts for military
purposes. The mining industry could obtain only small amounts of ammonium nitrate
and other salts of nitric acid, and therefore had to employ other oxidizing agents.

For blasting rock and for work in non-gassy and non-dusty coal mines, chlorate
explosives were widely used, e.g. Miedziankit in Russia and Germany, Cheddites
in France. In Belgium and Germany perchlorate rock explosives were also in general
use: in Belgium, ammonium perchlorate (Yonckites), in Germany, potassium per-
chlorate (the composition of these explosives differs slightly from those included in
Table 121, which were introduced immediately after World War I). During the war
an explosive "Blastin" was used in Great Britain for rock work. It consisted of:

Ammonium perchlorate

60%

Sodium nitrate

22%

TNT

11%

Paraffin

For coal work, especially in gassy and dusty mines a safety perchlorate explosive
with the composition tabulated below (Table 98) was manufactured in Germany.

Table 98
German safety perchlorate explosives

Wetter

Neuleonit

Ingredients

Permonit

Persalit

Perchlorit 4

Potassium perchlorate

32.5

Ammonium nitrate

34.5

Sodium nitrate

DNT

—

TNT

—

Nitroglycerine

Flour

Woodmeal

Glue solution in glycerine (1 : 7)

—

Charcoal

—

Sodium chloride

MINING EXPLOSIVES 409

These mixtures, however, brought no new advantages and did not fully meet
requirements with regard to safety in gassy and dusty mines, sensitiveness to impact
and friction and mining efficiency. In addition, perchlorate explosives, like chlorate
ones, have the defect that under certain conditions they do not detonate, but burn
out in the shothole, which may lead to a catastrophic fire or explosion. Perchlorate
explosives were then withdrawn from coal work after World War I and restricted
exclusively to rock work. Japan seems to be the only country where some perchlorate
explosives are accepted as "permitted explosives" (see p. 474).

RESEARCH AFTER WORLD WAR I

After 1918 research started before the war was continued intensively, by erecting
further testing stations and extending the sphere of their activity. Table 108 shows
the characteristics of the most important testing galleries in use between and after
the wars as official testing stations. Here the new explosives are studied to determine
their suitability for use of mines; safe methods of shotfiring are sought; and ways
of preventing explosions of methane and coal-dust investigated. The classification
of mines from the view point of coal-dust safety is also carried out at these testing
stations. Testing galleries are also in operation at explosive factories. These are used
to control the manufacture of explosives, and to test new safety explosives. Dia-
grams and descriptions of some galleries will be found in a later section (Table
108, p. 440).

The intensive, many-sided work of these testing stations contributed greatly
towards overcoming the hazards of using explosives in mines. They led to the pro-
duction of explosives possessing a high degree of safety and to the formulation of
safety regulations laid down on the basis of long experience so that coal mining is
little more hazardous than work in other branches of industry. The problem of the
safe use of explosives in coal mines was developed into a new branch of applied
science, which exceeds the scope of this book. Here only a few fundamentals will
be given.

The flame of explosion. Experiments by Hiscock [27] and T. Urbafiski [28]
showed that the flame of safety explosives in firedamp and coal-dust is very small
and its intensity insignificant. These explosives, differed little in the dimensions
and intensity of flame produced in these tests.

Audibert [29], Payman [30] and Beyling [31] studied the flame of explosives
detonating in a steel mortar with or without stemming, and showed that a stemming
12 cm or so long is sufficient to quench a flame perceptible in a photograph. On
shots fired without stemming the authors made a number of important observations:

(1) In a shothole of definite depth with the detonator placed at the end of the
charge facing the opening, the flame decreases with increasing charge.

(2) With a constant explosive charge the flame increases with the depth of the
shothole i.e. as the space between the charge and the opening increases.

410

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(3) With a constant free space between the charge and the opening, the dimen-
sions of the flame are independent of the size of the charge.

(4) The flame is increased by moving the detonator from the open end of the
shothole to the bottom of the shothole.

No definite relationship has been found between the dimensions of the flame
and the ability of the explosive to ignite the methane-air mixture. According to
Beyling, for instance, if the detonator is situated at the open end of the shothole,

■B:

Fig. 131. Detonation flame of Metanit B, according to Cybulski [25]. Diameter 32 mm,

charge 300 g, suspended in air. Time interval between the photographs 0.051 millisec.

Total duration of the flame 0.306 millisec.

as is usual in shotfiring, no charges within the range 50-100 g ignites the gas provided
that the free space between the charge and the opening lies within the limits 0-200 cm.
It has also been found that a decrease in the free space decreases safety,
creating a greater probability of explosion in spite of the decrease of flame (case 1).
Thus, Payman found that a smaller charge of ammonium nitrate explosive (No.
2 Viking Powder, see Table 96) is more likely to explode the gas mixture than a larger
one. This would be compatible with observation 1 ; on the other hand, on moving
the charge nearer to the mouth of the shothole the flame is removed almost com-

MINING EXPLOSIVES

411

pletely, although, as Payman reports, this increases the chance of explosion, contrary
to observation 1.

A detonator placed in the lower section of the charge is more likely to explode
a methane-air mixture. This is attributable to the increase of flame, in accordance
with observation 4.

Such contradictory test results show that the safety of explosive cannot be as-
sessed solely in terms of the dimension of the explosion flame.

It is the opinion of many authors that high speed photographs of the flame
should be supplemented by information about its position at a given moment in
relation to the shock wave and the hot products of explosion.

Nevertheless examination of the flame on high-speed film makes an important
contribution to the understanding of the properties of explosives and is often employ-
ed as an auxiliary test for the determination of the safety of explosives intended for
use in mines.

The two sets of photographs illustrated were taken by Cybulski. Both refer to
safety explosives used in Poland, viz., ammonium nitrate type Metanit B (Fig.
131) and nitroglycerine (dynamite) type, Barbaryt AGI (Fig. 132).

Fig. 132. Detonation flame of Barbaryt AGI, according to Cybulski [25]. Diameter

32 mm, charge 100 g, suspended in air. Time interval between the photographs 0.051

millisec. Total duration of the flame 0.153 millisec.

The effect of solid particles. In his studies of the detonation products of explosives
stemmed in different ways, Audibert [29, 32] paid attention to the fact that in char-
ges insufficiently stemmed a certain amount of the explosive remains in the form
of small particles which may undergo further explosive decomposition according
to conditions, i.e., temperature and ambient pressure. If these particles are ejected
into a space filled with a methane-air mixture they may lead to the explosion of
this mixture. The possibility of the existence of particles of undecomposed explo-
sive in ejected detonation products has been disputed by some authors (Segay [33]),
but many others have proved that it can occur. T. Urbahski [34] found that a thin
layer on the periphery of a cylindrical charge of ammonium nitrate explosive is

412 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

scattered by the explosion of the charge and does not explode. Beyling [31] showed
that undetonated particles of ammonium nitrate explosive may cause ignition of
the methane-air mixture in a testing gallery.

There is also a certain probability that fragments of the walls of a shothole
ejected by detonation may join these unexploded particles. Further, it has been
suggested that the remnants of a metallic detonator case cause ignition of me-
thane-air mixtures. Payman's [14] photographic studies proved that the particles of a
copper detonator case cannot provoke ignition. Being always surrounded by detona-
tion products they cannot penetrate the shock wave created by the explosion products
and come into contact with the methane-air mixture. Conversely, the particles of
an aluminium detonator case may penetrate into the methane-air mixture through
the shock wave. Hence the hazard in the use of aluminium cases, which has been con-
firmed frequently in practice. In addition, the very high heat of combustion of alum-
inium, giving high temperatures, must be taken into consideration as a factor liable
to increase the probability of a methane explosion.

Shock wave. An early assumption, originally unsupported by direct experience,
that the detonation of an explosive may involve the explosion of a methane-air
mixture even when the detonation is not accompanied by a large, hot flame, has
been confirmed experimentally. It was found that a mixture of 6.5% methane with
air, for instance, may ignite under rapid compression by mechanical means up to
54 atm.

Perrott [35] ascertained photographically that on the detonation of an explosive
in a methane-air mixture, the ignition of the latter starts at a spot not yet reached
by the flame of detonation of the explosive. Thus, in all probability the ignition
is caused by a shock wave travelling ahead of the flame.

Payman et al. [36, 37] showed the substantial validity of this observation. By
a schlieren photographic method Payman found that the detonation wave has the
same direction as the gaseous products when the detonator is situated at the bottom
of the shothole. With this arrangement, the ejection of gases from the shothole is
more intense, the velocity of the shock wave is higher and the probability that the
methane-air mixture will ignite is also greater. If, on the other hand, the detonator
is placed in the normal way, i.e. near to the stemming, the direction of the detona-
tion wave is opposite to that of the gaseous products. The latter then form a sort of
"gaseous stemming" around the detonator which holds back the shock wave and
thus reduces the probability of explosion of the gas mixture.

Investigations into the shock wave arising on the detonation of an explosive
as part of the problem of safety in the use of explosives attracted attention to the
rate of detonation of mining explosives as a factor directly influencing the velocity
of the shock wave. The earlier observation that a mining explosive, to be safe with
methane, should be relatively weak, was thus confirmed. The reason why explosive
charges with a large diameter are more dangerous than those with a small diameter
also became clear. This is due to the higher rate of detonation of charges with a larger
diameter.

MINING EXPLOSIVES 413

GENERAL CONSIDERATION ON SAFETY OF EXPLOSIVES

From many tests carried out in testing galleries and from mining practice it has
been shown that an explosive which is safe against methane, should have a rate of
detonation of approximately 2000 m/sec at the given density and should give a lead
block expansion not exceeding 200 cm 3 . In 1924 a special commission in Germany
established a lead block figure of 235 cm 3 at 15°C as the maximum value admissible
for the explosives used for coal mining. This is now considered too high.

Dubnov [38] postulated the causes of the explosibility of methane by the detona-
tion of an explosive. On being ejected from the shothole the products of detonation
of an explosive charge are mixed with the air and the methane present in the atmo-
sphere. The probability that the mixture will explode depends on the following
factors :

(1) Time t, which elapses before the combustible ingredients achieve a concentra-
tion capable of being exploded.

(2) Induction period z, which is necessary to bring about the explosion of this
gaseous mixture at a given temperature.

Since after time t has been surpassed, the composition of the gaseous mixture
tends towards a decreasing concentration of combustible ingredients while the
temperature of the mixture falls, the safety condition is fulfilled when z>t.

Until recently the concept of a safe "charge limit" was regarded as one of the
fundamentals in the use of explosives in coal mines. However, Cybulski's [39] in-
vestigations have shown that the thickness of the clay layer used as stemming is
critical for the safety of the explosive, which is relatively little affected by the type of
explosive. Also the charge weight seems to be unimportant according to these ex-
periments. Therefore doubt arises as to whether the concept of charge limit, as a
measure of safety of the explosives, is of real value. It appears that proper stemming
is of greater significance.

Figure 133 shows the relationship between the safety of an explosive (expressed
by a charge limit for the non-ignition of the methane-air mixture) and its power (ex-
pressed in per cent of the power of blasting gelatine), according to British investi-
gations [4, 40]. Since the usual initiation near the mouth of the mortar does not give
a sufficiently clear picture (the majority of safety explosives then fail to ignite metha-
ne), detonation was initiated at the back of the mortar. As the two curves show,
there is a greater probability of the ignition of methane in a drift with a smaller
diameter. It is obvious that the relationship between charge limit and power is com-
parable only for explosives of the same type (e.g. ammonium nitrate, gelatinous,
semi-gelatinous etc.). The power of an explosive must not therefore be considered in
the abstract, apart from its composition, and the above-mentioned view, that limi-
tation of the rate of detonation is recommended, may be right only within definite
limits governing the composition of the explosive and under definite experimental
conditions. J. Taylor [41] reports the following experimental results :

Polar Ajax, a gelatinous explosive was fired as usual in the mortar of a testing

414

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

^20

Gallery 14"dia. 9" long
Gallery 10"dia. 9" long

25 50

Power, %BG

Fig. 133. Relationship between charge limit and power of explosives (J. Taylor and Gay

[4]; Titman and Wilde [40]).

3500

3000'

cr 2500 -

2000-

1500

1000

10 20

Charge limit, oz

Fig. 134. Relationship between charge limit and rate of detonation of explosives
suspended in methane-air mixture, according to J. Taylor [41]; /—explosive contain-
ing no cooling salt; //—explosive containing 10% sodium chloride; ///—explosive
containing 50% sodium chloride; IV— explosive based on sodium nitrate.

MINING EXPLOSIVES 415

gallery at two different rates of detonation : 2400 m/sec by initiation with an ordinary
detonator and 4800 m/sec by priming with a tetryl pellet. Initiation was from the
bottom of the bore. The charge limits were respectively 14-16 oz and 10-12 oz
a difference which is almost negligible. Finely ground TNT, with a rate of detonation
of 4100 m/sec, and coarse grained TNT, with a rate of detonation of 1680 m/sec
behave similarly: their charge limit is the same: 2-4 oz.

However, J. Taylor found that the converse may also hold good: on the detona-
tion of an explosive freely suspended in a chamber filled with a mixture of 9°/
methane in air (i.e. not fired in a mortar), the relationship between the charge limit
and the rate of detonation (Fig. 134) is clearly marked in explosives of similar com-
position. Murata [42] came to the same conclusion. Likewise, if detonation is initia-

Fig. 135. Free suspension test gallery, according to J. Taylor and Gay [4].

ted near the mouth of the test mortar, i.e. conditions similar to those of firing an
explosive suspended in a methane-air mixture, the rate of detonation has a pro-
nounced effect on the charge limit, i.e. the higher the rate, the smaller the charge.
The procedure for testing with the explosive charges suspended in the methane-air
mixture, as carried out at Ardeer, Great Britain, and described by J. Taylor and Gay
[4] is shown in the photograph (Fig. 135). The diagram (Fig. 136) shows the test
lay-out.

This kind of situation usually does not occur in coal mining practice, but in
fact there have been instances when an explosive charge has been fired across gaps
filled with methane inside the shothole. The conditions characteristic of the firing
of explosive charges suspended in a methane-air mixture are then very similar to
those met in practice.

Experiments have shown that the ejection of detonation products perpendicular
to the axis of the charge (angle detonation) is more dangerous than ejection
from an ordinary mortar, i.e., in the direction of the mortar's axis (Payman and

416

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Wheeler [43]; J. Taylor and Hancock [44]). A special mortar, the "angleshot mortar"
is used for determining the effect of this phenomenon. It is shown diagrammatically
in Fig. 137. This is a diagram of equipment used at the German testing station at
Dortmund-Derne (J. Taylor and Gay [4]). A steel ricochet plate from which the de-
tonation products are reflected is situated alongside the mortar. The angle of reflection
of these products depends on the angle of the mortar with respect to the ricochet plate.

Explosive

Paper
diaphragm

Gas chamber containing
9% methane-air

Fig. 136. Diagram of free suspension test gallery, according to J. Taylor and Gay [4].

Profile
piece

Baseplate

Adjustable
angle plate

''<

-Adjustable
angle plate

Fig. 137. Diagram of angle-shot mortar test at Dortmund-Derne, according to

J. Taylor and Gay [4].

The latter may be 90° or less, as is shown in Fig. 137. The reflection of the detonation
products from a barrier such as the ricochet plate involves a violent rise in temper-
ature at this spot, due to the adiabatic compression of gases. A general view of angle
shot mortar at Derne is given on Fig. 138.

MINING EXPLOSIVES

417

In Poland another system is also known that works on the same principle as the
angle-shot mortar. This is a mortar with a horizontal slit. Figure 139 shows diagram-
matically the arrangement in a rectangular section gallery at Mikol6w.

Fig. 138. General view of angle-shot mortar at Dortmund-Derne [45].

According to Cybulski's studies [46, 47] the severity of this test is increased by
the prolonged effect of high temperature on the cloud of coal-dust: in the first stage
volatile matter is released from the coal-dust, and only afterwards does the igni-
tion of the explosive system occur. Ignition of coal-dust may be achieved more
easily by the simultaneous detonation of several shots. Firing in the angle-shot
mortar gives results which approach the natural condition of working in coal mines
to a much greater degree than firing in the classical mortar. According to Cybulski [39]
the classical test gives less reliable results than the angle-shot method.

Deflagration of mining explosives. Long experience in the use of explosives in
mining shows that some high explosives undergo deflagration instead of detonation
in a shothole. Delayed shots then occur, commonly called "blown-out shots" (ge-
kochte Schiisse).

Such shots are characterized by the fact that immediately after firing the deto-
nator there is no report which is the normal acoustic effect. Instead, after a certain

418

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

delay, the gas pressure in the shothole increases due to combustion of the explosive*
and the stemming is ejected with a report similar to that of a normal shot. The
course of deflagration may differ with the conditions. Deflagration, however, always
creates a great danger of ignition of methane since the deflagrating explosive gives
a flame of long duration and of very large dimensions. Moreover, the explosive does not
carry out the normal work of blasting, but only achieves the insignificant result of
ejecting the stemming. Since, therefore, the heat of the explosive decomposition is
not converted into work, the temperature of the deflagration flame is very high and
hence more dangerous.

2.50 m-

-1.59 m-

10m »

Fig. 139. Diagram of a mortar with slit in a gallery at Mikolow (Poland).

It has long been known that inflammable explosives, such as dynamites, and
especially chlorate and perchlorate explosives are particularly prone to deflagration.
This is one of the reasons why chlorate and perchlorate explosives have either been
withdrawn from use in mines in a majority of countries or restricted to use in
rock, where there are no inflammable gases.

The amonium nitrate explosives are much harder to deflagrate and are there-
fore generally recognized as the safest, although they may deflagrate under certain
conditions.

The first systematic studies which elucidated the conditions under which explo-
sives and especially ammonium nitrate ones undergo deflagration, were carried out
by Cybulski [48]. This work indicated that deflagration of explosives, including
ammonium nitrate ones, may occur in mining practice, if they happen to be mixed
with coal-dust in the vicinity of the detonator. The sensitiveness of an explosive to
detonation is then reduced, but its ignitability is increased, hence the explosive takes
fire instead of being detonated. Ammonium nitrate explosives are readily detonated,
but are ignited with difficulty. Clearly, this is their great advantage, but nevertheless,
under the conditions described above (admixture with coal-dust) they may be sub-

MINING EXPLOSIVES

419

ject to deflagration. Mixing with coal-dust may result from incorrect loading of the
shothole and from negligence over freeing the shothole from coal-dust. Careful
loading of the shothole and the use of ammonium nitrate explosives is a sufficient
guarantee for the prevention of delayed shots.

Cybulski established that penetration of coal-dust into the explosive near the
detonator is most dangerous, creating conditions particularly favourable to defla-
gration.

Explosives containing nitroglycerine and also chlorate explosives are more lia-
ble to deflagration than those containing ammonium nitrate without nitrogly-
cerine.

When a partial misfire occurs in the shothole, there is great danger that the charge
will deflagrate. The hot detonation products act on the undetonated remnant of the

Parallel steel plates

Fig. 140. Diagram of break gallery, according to J. Taylor and Gay [4].
(By permission of Controller H.M. Stationery Office.)

charge and may initiate it. It may happen that detonation of a part of the charge opens
the shothole by ejecting the stemming or by shattering some coal. The deflagration
of the remaining part of the charge may then be prevented.

A partial misfire may have various causes. For instance, the charge may be
contaminated by coal-dust, as described above. The incomplete detonation of hard-
ened cartridges of ammonium nitrate explosive occurs widely. It is obvious that
the incomplete detonation of damp explosives is also possible.

Cybulski's [48] extensive investigations in coal and rock have also solved the prob-
lem of the double detonation. It is now known that the deflagration of an explosive in
the shothole may pass into detonation after a certain time. This occurs with particular
ease in nitroglycerine explosives. The results of Cybulski's studies [48] are now taken
into consideration in Poland in defining which explosives may be used in mining.

420 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In Great Britain a test was devised for simulating the effect of blown-out shots
in a shothole. Small charges of explosive were fired, from a suspended position or
from a gun, in a methane-air mixture between two steel plates 2-4 in. apart in a "break
gallery" (Fig. 140) developed by Shepherd and Grimshaw [49].

Explosibility of coal-dust. It is of great importance to know the conditions gov-
erning the explosion of coal-dust in order to develop methods of preventing it.
Extensive, fundamental studies have been carried out in leading coal producing
countries [50]. In Poland they have been carried out by Cybulski since 1925, mainly
on a great scale in an underground testing gallery 200 m long and also in smaller galler-
ies 44 and 100 m long.

Cybulski [51] found that there is a straight line relationship between the explo-
sibility of coal-dust and the specific surface area of the dust. This may be expressed
by the formula :

2'=0.0010119F+0.7836578, where

Z— degree of explosibility of the dust,

F— the specific surface area of dust.

Cybulski has shown that very fine coal-dust contains a higher proportion of vola-
tile matter than the average sample. This increases the danger created by fine coal-dust.

Cybulski [52] has also shown that very fine coal-dust may explode even when it
contains less than 12% combustible volatile matter. The presence of a significant
amount of water in coal-dust may lead to its explosion even if free from volatile
matter. This indicates the particularly hazardous nature of fine coal-dust. Cybulski
has also shown that under particularly unfavourable conditions the coal-dust must
contain as much as 95% incombustible matter (e.g. by admixture with non-inflam-
mable stone dust) if explosion is to be avoided.

More recently attention has been paid to the problem of the explosive properties
of sulphur dust suspended in air. This is of great importance to work with explosives
in sulphur mines.

Sulphur dust is more dangerous than coal-dust, because of the low ignition
temperature of sulphur suspensions in air. According to Dubnov [53] 100 g charges of
the U.S.S.R. explosives Ammonit No. 1 and 8 ignited sulphur dust. The same explo-
sives did not ignite a methane-air mixture when the quantities were 400 and 500-
650 g respectively. In sulphur mines explosives of very low detonation temperature
should be used.

FUNDAMENTAL COMPONENTS OF MINING EXPLOSIVES

The ample theoretical and practical material derived from investigations of
mining explosives in testing stations and from mining practice led to the conclusion
that the composition of mining explosives should be subject to certain rules aimed at
the highest possible safety in the use of the explosives together with the conservation
Of maximum working efficiency. These rules are stated below.

MINING EXPLOSIVES 421

OXYGEN CARRIERS

Nitrates. The chief oxygen carrier, employed in all modern explosives, is ammo-
nium nitrate (it is discussed more fully in Vol. II, p. 450).

In some countries (Great Britain) ammonium nitrate is used with an admixture of
triphenylmethane dye, Acid Magenta which inhibits the transformation of one
crystalline form into the other at 32°C (Vol. II, p. 454).

Substances that reduce hygroscopicity are also frequently added. This is discuss-
ed more fully in Vol. II (p. 453). The substances used are: carboxymethylcellulose
and calcium stearate (J. Taylor and Silljtto [54]). Silicone resins are used in
France.

In mixtures in which sodium chloride occurs together with ammonium nitrate,
there is a possibility of molecular exchange and of the formation of the more stable
system: sodium nitrate-ammonium chloride:

NH4NO3 + NaCl — > NaN0 3 + NH4CI

Kreyenbuhl and Sartorius [55] found that this exchange proceeds slowly if the
moisture, content is less than 1 %. If it is 1 % or more so, the exchange is complete
after 3-6 months and is accompanied by decreased explosive power. The addition of
a moisture absorbent (e.g. urea resin foam) inhibits this exchange (French explosive
Noburex, p. 552).

Alternatively, the use of mixtures with sodium nitrate and ammonium chloride
have been proposed in place of those containing ammonium nitrate and sodium
chloride (see below). Other nitrates -sodium, potassium, barium— are of minor
importance. They are generally used in insignificant amounts in admixture with
ammonium nitrate. The use of potassium nitrate has been prohibited in Germany since
1925, when it was thought that the presence of potassium nitrate in some ammo-
nium nitrate explosives may cause the ignition of methane-air mixtures. It was-
believed that the decomposition of potassium nitrate during detonation proceeds.
too slowly and may cause a long duration of flame. Although this was not proved
experimentally, the prohibition has remained in force and potassium nitrate has been
replaced by barium nitrate.

Sodium nitrate is often employed as an oxygen carrier in dynamites.

For some time mixtures of safety explosives containing sodium nitrate with
ammonium chloride were recommended instead of ammonium nitrate with sodium
chloride since the system NaN0 3 +NH 4 Cl is more stable than NH 4 N0 3 +NaCl
(see above). According to the investigations of Hackspill, Rollet and Lauffenburgier
[56] the system sodium nitrate-ammonium chloride is stable at room temperature.
These observations led to studies of explosive mixtures with sodium nitrate
and ammonium chloride. At first, the results were encouraging. Thus Ahrens [57J
established that the safety of these explosives seems to be greater than that of explo-
sives with ammonium nitrate and sodium chloride. It was believed that this is due
to formation of free sodium ions at high temperature. This can be represented

422 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

diagrammatically :

e e

NH4CI + NaN0 3 — > Na + CI + [NH4NO3]

detonation
The power of explosives containing sodium nitrate and ammonium chloride was
found to be the same as that for equivalent compositions containing ammonium
nitrate and sodium chloride [58]. The compositions with ammonium chloride-
-sodium nitrate are now very popular in France, Germany, Belgium and Great Brit-
ain.

However with poor ventilation they seem to give more persistent smoke than
compositions based on ammonium nitrate and sodium chloride. Analysis of the
smoke showed it to be due to finely divided ammonium chloride suspended in
moist air, and although not harmful it may be objectionable to the miners. In
addition, calcium sulphate was included to suppress the smoke (J. Taylor and Gay
[4]). Calcium nitrate found only a limited use as a substance for lowering the temper-
ature of detonation flame.

Murata [42] showed that the addition of a few per cent of potassium and sodium
nitrates somewhat increases the safety of an explosive with respect to methane but
in larger quantities their effect is almost negligible.

Perchlorates

Perchlorates are not used now as oxygen carriers in safety explosives. Perchlorate
explosives were used temporarily during World War I, and later a small admixture
of potassium perchlorate was introduced into ammonium nitrate safety explosives
such as Bradyt F (p. 475), used for a certain time in Poland. Potassium perchlorate
was added on the assumption that during explosive decomposition it may promote
the formation of an additional amount of potassium chloride, a substance inhibit-
ing the explosion of methane-air mixtures. Bradyt F did in fact prove to be
highly safe on examination in a testing gallery. Moreover, an admixture of perchlo-
rates increases the sensitiveness to detonation in ammonium nitrate explosives.
At present, perchlorates are used as ingredients of ammonium nitrate explosives only
in Belgium, France and Japan.

Chlorates

Chlorates have been little used. Only potassium chlorate has been employed as
a non-hygroscopic ingredient in the simplest rock explosives, including Miedziankit
(p. 278). Experience of many years proves that such explosives often burn out
in the shothole, creating a great risk of fire. As was pointed out earlier (Vol. II)
chlorates should never be mixed with ammonium salts (such as ammonium nitrate),
because of the formation of highly unstable ammonium chlorate. At present, chlorate
explosives are banned in underground mines in the majority of countries and their
use in coal mines is prohibited everywhere.

MINING EXPLOSIVES 423

ACTIVE INGREDIENTS AND COMBUSTIBLES

All ammonium nitrate explosives containing non-explosive salts (e.g. sodium
or potassium chloride) should include at least 4% nitroglycerine to increase their
sensitiveness to detonation. Small amounts of nitroglycerine (up to 6-8%) do not
require gelatinization with nitrocellulose; it is enough to introduce a certain quantity
(up to 2%) of woodmeal into the mixture to absorb nitroglycerine and to prevent
its exudation. In addition, the woodmeal stabilizes nitroglycerine explosives by
absorbing impurities and the product of decomposition of the ingredients. Explo-
sives containing woodmeal therefore possess ephanced chemical stability.

If the content of nitroglycerine exceeds 10%, the explosives are called dynamites.
It is then necessary to bind the nitroglycerine with highly viscous "dynamite collodion
cotton" (Vol. II). Dynamites thus acquire a plastic consistency which is a very
convenient form of mining explosive, facilitating the introduction of a detonator
with a fuse. In countries where the explosive may be exposed to frost, a mixture of
nitroglycerine with nitroglycol or nitrodiglycol should be used, since the explosive
then retains its plasticity even at low temperatures. In the U.S.S.R. another type of
plastic explosive with nitroglycerine was recently introduced containing a 3-5% so-
lution of methyl polymethacrylate in nitroglycerine.

Woodmeal is of particular importance among combustible ingredients since
apart from its stabilizing effect, discussed above, it enables the mixture to be retained
in a loose state with a low density. Charcoal is another ingredient sometimes added.

The addition of aluminium to increase the heat of the explosion and, in conse-
quence, the power of the explosive, as practised for some time, is now believed to be too
dangerous. Ammonium nitrate explosives with aluminium are permitted only for
rock work in opencast mining (e.g. quarries) or underground, where there is no
methane.

Ammonium nitrate explosives usually contain a certain amount of nitro com-
pounds from mononitro (e.g. nitronaphthalene) to trinitro compounds (e.g. trinitro-
toluene).

Some ammonium nitrate explosives used in the U.S.S.R. contain a small amount
of paraffin (0.5-1 %) to reduce the hygroscopicity of the ammonium nitrate.

Recently Hino and Yokogawa [59] found that by adding less than 1 % surface
active agents, an improvement of transmission of detonation of ammonium nitrate
explosives can be achieved.

OXYGEN BALANCE

The maintenance of an adequate oxygen balance in a mixture is a problem of
high importance. Former German explosives of the Carbonit type *, with incomplete
combustion, have been banned. Modern mining explosives are required to have an
oxygen balance equal to or higher than zero. Investigations in the U.S.S.R. have

* Carbonits had a negative oxygen balance on the wrong assumption that incomplete com-
bustion gives greater safety for use in mines, by lowering the heat and temperature of explosion.

424

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

shown that a highly positive balance is disadvantageous since it favours the formation
of nitric oxides in the products of an explosion. For earth moving, farming, and
opencast mining the use of explosives with a negative oxygen balance (e.g. picric
acid, TNT etc.) is permitted.

According to Soviet findings a concentration of carbon monoxide of less than
0.00025% is admissible in the atmosphere of a coal mine, after ventilation of the
shotfiring position. Before ventilation, -immediately after shotfiring, the content
of CO should not exceed 0.02%.

A positive oxygen balance markedly reduces the quantity of carbon monoxide
formed. Assonov and Rossi [60] report the following results of experiments on the
effect of the oxygen balance of explosives on the composition of the gaseous
products of explosion (Table 99).

Other agents also examined in the U.S.S.R. exert an influence on the combustion
of the products of an explosion. According to Assonov and Rossi [60] increase in the
moisture content of an explosive involves the increase in the content of gaseous pro-
ducts.

Table 99

Composition of gaseous products of explosion in explosives with
a different oxygen balance

Explosive

Oxygen balance

°/

Content of toxic gaseous products
(1. per kg of explosive)

carbon monoxide

nitric oxides

2 T Ammonit
No. 1 Dinaftalit
T Ammonit
Dynamon

+9.0
+0.6
-1.0
-9.8

15.0
7.0
7.5

95.6

25.1
7.3
6.2
6.5

Other factors include the density of loading (with a larger density a smaller
quantity of nitric oxides is obtained), the diameter of the cartridge (a larger diameter
favours the reduction of the amounts of carbon monoxide and nitric oxides); e.g.
after the explosion of 62% dynamite* 8.5 l./kg of carbon monoxide and 4.2 l./kg of
nitric oxides were found in the gaseous products, with cartridges 30 mm dia. With
cartridges 20 mm dia. the figures were 9.0 l./kg and 12.5 l./kg respectively.

Cybulski [61] and Assonov and Rossi also found that if made of paper saturated
with paraffin, the cartridge case affects the composition of the gaseous products,
considerably increasing the quantity of carbon monoxide.

Assonov and Rossi report that with a proportion of 2 g of paper per 100 g of
explosive, the content of paraffin should be less than 2.5 g. With this paraffin content
the amount of toxic gases formed is 26.4 l./kg of carbon monoxide and 37.7 l./kg of
nitric oxides.

* 62% dynamite means a "straight" dynamite containing 62% nitroglycerine or a dynamite
of the same power as one with 62% nitroglycerine, according to U.S.A. nomenclature.

MINING EXPLOSIVES

425

If the content of paraffin is 4.0 g, 29.5 l./kg of carbon monoxide and 59.7 l./kg of
nitric oxides are formed.

According to Tananel [62] paraffin paper also reduces the safety of an explosive
towards methane. He established this for explosives with a high oxygen balance (Gri-
sounaphthalite and Grisoudynamite, with oxygen balance +13 and +18% respec-
tively). As shown by Murata [42] this phenomenon occurs only in explosives with
a highly positive oxygen balance. If an explosive has only a slightly positive oxygen
balance, the presence of paraffin paper tends to enhance its safety (Fig. 141).

6 8

Wood meal, %

+18.31

-0.51

+12.01 +5.71

Oxygen balance, %

Fig. 141 . Effects of paraffined paper cartridges upon the safety of explosives of various
oxygen balances: /-cartridge of paper only; //-paraffined paper cartridge (ac-
cording to Murata [42]).

Good stemming, occupying at least $ of the shothole, prevents the formation of
large amounts of toxic gases.

Experiments carried out in the U.S.S.R. have shown that the content of carbon
monoxide and nitric oxides in the products of an explosion is influenced by the me-
dium in which the explosion takes place. The quantity of carbon monoxide and ni-
tric oxides is higher with shotfiring in coal and lower in copper and iron ore seams.
The figures given by Assonov and Rossi are summarized in Table 100.

Table 100

Quantity of toxic oxides forming in the products of explosions occurring in

various seams

Shotfiring

Explosive

in coal

in iron ore

carbon monoxide
l./kg

nitric oxides
l-/kg

carbon monoxide
l./kg

nitric oxides
l./kg

2 Ammonit
62% Dynamit

40.7
54.3

29.8
4.8

4.7
3.3

7.9
8.0

426

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Calculation of the oxygen balance is carried out by taking into consideration
the quantity of oxygen supplied by the oxygen carrier and that required by the com-
bustible ingredients of the explosive mixture. The figures for the most frequently
used ingredients are tabulated below.

Table 101

Quantity of oxygen given or required by ingredients of
mining explosives

Quantity of oxygen (kg)

Oxygen carrier

given by 1 kg

of the substance

Inorganic salts

Ammonium nitrate

+ 0.200

Ammonium perchlorate

0.340

Barium chlorate [Ba(C10 3 )2-H 2 0]

0.298

Barium nitrate

0.306

Calcium nitrate [Ca(N0 3 ) 2 -4H 2 0]

0.339

Potassium chlorate

0.396

Potassium nitrate

0.396

Potassium perchlorate

0.462

Sodium chlorate

0.450

Sodium nitrate

0.470

Organic substances

Nitroglycerine

0.035

Tetranitromethane

0.490

Quantity of oxygen (kg)

Substances requiring oxygen

required for total combus-

tion of 1 kg of the substance

Aluminium

-0.890

Cellulose (C 6 Hi O 5 )

1.185

Charcoal

2.667

Cyclonite (RDX)

0.216

Diglycoldinitrate (DGDN)

0.408

Dinitronaphthalene

1.394

Dinitrotoluene (DNT)

1.144

Nitrocellulose (11.95% N)

0.387

Nitroglycol

0.000

Nitroguanidine

0.308

Nitronaphthalene

1.988

Penthrite

0.100

Sulphur

1.000

Trinitrotoluene (TNT)

0.740

Trinitroxylene

0.896

Woodmeal (C 6 Hi O 5 )

1.185

MINING EXPLOSIVES

INERT INGREDIENTS INCREASING SAFETY

427

Long experience has shown that the most efficient non-explosive ingredients for
increasing the safety of explosives towards methane and coal-dust are sodium and
potassium chlorides. The latter is more efficient due, no doubt, to the action of the
potassium ion which inhibits explosive reactions in the gaseous phase. This will be
considered further when discussing the problem of securing flashless discharge from
guns (p. 544). An explosion of a methane-air mixture is a phenomenon of the same
type as the explosion of a mixture of muzzle gases with air. It is also beyond ques-
tion that as in the latter case, the inhibition of an explosion of a methane-air mixture
is to a less degree a thermal phenomenon which results from the temperature fall
of the igniting flame. To a much higher degree it is based on the spraying in
gas of a substance containing potassium ions. A number of authors have shown that
this is not only a thermal phenomenon. For instance Dubnov [38] proved this by
experiments in which he added sand instead of sodium chloride to the explosives.
He matched the quantity of sand so that the heat capacity was the same, and found
that the presence of sand cannot prevent explosion of a methane-air mixture.

Dubnov [53] also reports the results of experiments in a testing gallery. The
charge consisted of 90% of 80/20 ammonit (80% ammonium nitrate and 20% TNT)
and 10% of various salts. A steel mortar was loaded with a 200 g charge, without
stemming, and fired in an atmosphere containing 8-10% methane. The percentage
number of ignitions of methane was then determined (Table 102).

Table 102

Salt

Specific heat

Percentage ignition
of methane

0.200

KH 2 P0 4

0.208

KC1

0.162

K2SO4

0.178

KNO3

0.221

NaCl

0.206

Na 2 C0 3

0.273

CaF 2

0.215

CaC0 3

0.189

100

NH4CI

0.363

100

Pb(OOCCH 3 ) 2

0.134

100

These figures show that a high specific heat is not correlated with inhibition of
explosion of methane. E.g. the high specific heat of ammonium chloride does not
help this salt to inhibit methane explosions.

As mentioned previously (p. 421) ammonium chloride and sodium nitrate in
stoichiometric proportion has also been suggested, as an effective source of sodium
ions which possess marked ability to inhibit methane explosions.

428

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Dubnov examined the effect of various salts on the ignition temperature and
on the explosion lag of methane-air mixtures (Table 103).

Table 103
Effect of the addition of various salts on explosive properties

Minimum

ignition

temperature

°C

Explc

ssion lag (sec) at different temperatures

Salt

750°C

760°C

770°C

780°C

790°C

800°C

810°C

820°C

No salt

added

710

1.8

1.0

0.8

—

CaC0 3

710

1.4

1.1

0.9

0.7 •

—

NaCl

730

3.2

1.9

1.4

1.0

0.8

0.6

—

LiCl

790

4.2

2.2

1.95

KC1

800

—

5.6

3.9

1.85

CaCl 2

810

4.0

3.0

These experiments confirm that the efficiency of potassium chloride is higher
than that of sodium chloride in inhibiting explosion. (On inhibition of other gaseous
explosive mixtures by potassium chloride see papers by Prettre [63]).

Highly interesting results were obtained by Murata [42], showing very high effi-
ciency of potassium chloride and somewhat lower efficiency of sodium chloride.
Investigations on the effect of potassium and sodium nitrate are also noteworthy.
They were found to have a beneficial influence only when a few per cent were used ;
in greater quantity they have practically no effect.

According to Dubnov [53] ammonium chloride seems to be particularly suitable
as a cooling agent in explosives for sulphur mines.

Dubnov [53] also describes experiments which prove that the activity of an inert
salt increases with its fineness (Table 104). The figures in the Table refer to sodium
chloride. An explosive charge containing 35 or 40% sodium chloride was detonated
in a methane-air mixture.

Table 104
Activity of inert salt as a function of its fineness

Average size of the

particles of sodium chloride

Content of
sodium chloride

Percentage
ignitions

Power of explo-
sive according to
lead block test

1.38

100

221

1.16

100

219

0.70

210

0.38

216

below 0.14

192

1.38

100

197

0.70

187

0.38

181

MINING EXPLOSIVES

429

Recently "ultra-safe" explosives were introduced in Great Britain. They contain
cooling salts ground to a finer size than it is customary (p. 467).

Dubnov [53] reports figures depicting the relationship of charge limit, power,
heat of detonation and temperature and sodium or potassium chloride content for
an explosive containing ammonium nitrate and TNT in the proportion 80 • 20
(Table 105).

Table 105
Charge limit and power of explosives as a function of sodium or potassium chloride

CONTENT

Content of inert
salt

Charge limit
g

Lead block
expansion

cm 3

Heat of deton-
ation
kcal/kg

Calculated

temperature

°C

NaCl

KC1

NaCl

KC1

NaCl

KC1

NaCl

KC1

, 10

15
20
25
30

100
200
350
450
550

150
250
450
550
650

340
296

272
245
214

342
298
279
254
226

817
745
673
600

525

845
774
703
632
560

2100
2040
1980
1920
1860

2120
2070
2023
1975
1926

It is interesting to note that the compositions containing potassium chloride
gave higher calculated values of the heat and temperature of explosion. In spite
of this, the efficiency of potassium chloride as an agent increasing the safety
of the use of the explosive in presence of fire-damp is higher. It confirms the
former observation that potassium ions are efficient inhibitors of gas explosions.

Murata [42] reported similar results for explosives containing ammonium nitrate,
8% of nitroglycerine, 0.3% of nitrocellulose and 8% of woodmeal. Sodium or po-
tassium chloride in proportion 1-15% and 1-10% respectively replaced the corre-
sponding part of ammonium nitrate.

In 1914 Lemaire [64] proposed the use of a sheath made of incombustible and
inexplosive substances in explosive charges. As an inert material he recommended:
milled sand, salts containing water of crystallization (e.g. sodium sulphate), salts
volatilizing at high temperatures (e.g. sodium chloride, sodium fluoride, cryolite),
salts decomposing at high temperatures (e.g. ammonium chloride). Sheaths were
made by placing the substance between the dual walls of a paper tube.

The preliminary experiments indicated the importance of this discovery. It was
found, for instance, than even 800 g of blasting gelatine fitted with a sheath of
sodium chloride or sodium fluoride weighing as much as the explosive itself failed
to ignite methane or coal-dust. On the other hand 100 g of blasting gelatine, without
an inert sheath, was enough to cause explosion of methane or coal-dust.

After investigations carried out in Belgium at the Institut National des Mines
de Paturages in 1930, an inert sheath with the following specification was intro-
duced:

430 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

thickness of mantle: 3 mm

composition: 25% of binder (usually calcium sulphate or clay)

75 % of salts : sodium or calcium fluoride or sodium chloride with an
admixture of at least 35% of sodium fluoride.

The total weight of material forming the sheath was 65 g per 100 g of explosive.

Later, the thickness of the mantle was increased to 5 mm in safety explosives
(S.G.P.— p. 447) and to 6.5 mm in rock explosives.

In 1947, sodium hydrogen carbonate was introduced as the most efficient salt.

About 1933 work on the use of sheaths of this kind was also started in other
countries. In the U.S.S.R. sheaths of 9 mm thickness composed of powdered
salts in paper tubes were adopted.

In the Home Office Gallery at Buxton (Great Britain) the effect of the following
groups of substances used for filling the sheaths was tested :

(1) Substances preventing explosion of methane-air mixture, e.g. potassium
chloride ;

(2) Cooling substances with high thermal conductivity, e.g. iron chips ;

(3) Substances evolving incombustible gases, assumed to form a gaseous sheath
around the explosion, e.g. sodium hydrogen carbonate;

(4) Reducing substances which may combine with atmospheric oxygen, thus
removing the ingredient necessary for the explosion of methane, e.g. sodium hydro-
sulphite.

Experiments in a testing gallery showed that addition of the majority of these
substances increases the charge limit to such an extent that it cannot fit into the
mortar. For this reason, a very strict test was carried out on the basis of the detona-
tion of an explosive directly suspended in a methane-air mixture. Under such con-
ditions the charge limit was diminished nearly four times as compared with that
determined for a mortar.

Table 106

Substance in a sheath

Charge limit
g

No sheath

Kieselguhr
Lead sulphate
Sodium thiosulphate
Sodium hydrogen sulphite
Sodium formate

113
170
170
170
226

Manganese dioxide
Ferrous oxalate

226
226

Hydrated sodium carbonate
Sodium hydrosulphite
Sodium hydrogen carbonate
Iron chips

226
283
283
340

a and b Belgian mixtures
c Belgian mixture

170
226

MINING EXPLOSIVES 431

Table 106 presents the results of some experiments on the Tees Powder explosive
(p. 405) provided with a sheath 4 mm thick.

On the basis of these experiments a sheath was introduced in England filled
with sodium hydrogen sulphite, a cheap substance and fairly efficient when in the
form of a mantle about 2.5 mm thick, protected from the outside by non-impregna-
ted watertight paper. The explosives which may be used with a sheath of this kind
are designated by the symbol "S" (sheathed).

The use of sheaths very slightly weakens the strength of an explosive but does
not reduce its sensitiveness to detonation. On the other hand it considerably in-
creases its safety for use in mines with methane and sensitive coal-dust.

Fig. 142. Rings of sodium chloride or sodium hydrogen carbonate for making rigid
sheaths (Courtesy Dr. L. Deffet, Sterrebeck, Belgium).

The production of sheaths has led, however, to some new and unexpected diffi-
culties. Thus, the explosive core was sometimes not sufficiently well centered, or the
substance filling the sheath moved to one end and formed a barrier to the trans-
mission of detonation from one cartridge to another.

It is also evident that the increased safety produced by using sheathed explosives
is lost if the sheath is removed deliberately. The latter may happen, for instance, if
the hole into which the charge is being loaded is too narrow.

To avoid this danger various solutions were suggested and introduced, for
instance, the use of rigid sheaths. In Belgium these were made by the compression

432 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

of rings of sodium chloride or sodium hydrogen carbonate (Fig. 142) or by the
extrusion of mixtures of sodium chloride with wet clay which was then dried.

This arrangement has the additional advantage that it facilitates detonation.
In the U.S.S.R. Seleznev suggested [60] a rigid sheath composed of china clay,
sodium chloride and calcium sulphate.

Another method of increasing the mechanical strengths of sheaths was suggested
by Fleming [65]: it consisted of making them from "bicarbonate felt"— 90% sodium
hydrogen carbonate and 10% paper pulp.

In Great Britain an alternative solution was found. The inert material was directly
and uniformly incorporated into the explosive itself, instead of forming a separate
sheath (J. Taylor [66]). These explosives were marked "Eq. S.", i.e. equivalent in
safety to sheathed explosives of the same class.

In addition to the ordinary inert sheath, an active sheath in Germany and the
U.S.S.R. consisting of "cooling" salts admixed with nitroglycerine has been de-
veloped, for the following reasons:

(1) The substance of a sheath is better sprayed after the detonation owing to
the presence of nitroglycerine taking part in the detonation and since it is then
better mixed with the detonation products of the explosive;

(2) Being an explosive, the sheath does not weaken the charge and does not
inhibit the transmission of detonation from one cartridge to another.

In Germany sheaths consisting of 12 or 15% nitroglycerine, 33 or 35% sodium
chloride and 50 or 55% sodium hydrogen carbonate are used. Each cartridge com-
prises 70 g of explosive and 50 g of sheath.

In the U.S.S.R. a sheath composed of nitroglycerine, sodium chloride, potas-
sium chloride, sodium nitrate and woodmeal is employed.

In Belgium "Bicarbite" special charges have been introduced, consisting of 85%
sodium hydrogen carbonate and 15% gelatinized nitrocellulose. Cartridges of this,
composition are placed in the shothole alternately with explosive cartridges. By
altering the proportion of explosive cartridges to Bicarbite ones the degree of safety
towards methane and coal-dust can be modified.

Recently Boucard and Deffet [67] examined the action of organic compounds
containing halogens, such as : carbon tetrachloride, hexachloroethane, y-hexachloro-
cyclohexane, chlordane (C 10 H 6 C1 S ), iodoform and also iodine.

It was found that the organic compounds in question could inhibit the explosion
of methane-air mixture particularly if they were first subjected to a thermal decom-
position. It seems advisable to add these compounds to explosives so that they
undergo some decomposition in the course of the detonation. They do not seem
to form substances increasing the safety by themselves.

INERT NEUTRALIZING AGENTS

Since ammonium nitrate always contains a small amount of free nitric acid a
certain amount of substances to neutralize the acid should also be added to ammon-

MINING EXPLOSIVES

433

ium nitrate explosives containing nitroglycerine. Since soluble bases (e.g. sodium
hydrogen carbonate) exert a negative influence on ammonium nitrate and other
ingredients of explosive mixtures such as TNT, insoluble bases, e.g. calcium or
magnesium carbonates, should be employed. Ferric oxide may also be used, e g
in the form of pyrites ash.

In explosives which contain a small amount of nitroglycerine (4-8%) the addi-
tion of woodmeal (p. 423) provides adequate stability.

To reduce the density, a larger amount (6-12%) of light woodmeal, peat meal
or vegetable fibre (e.g. cellulose) is sometimes added. This brings about a con-
siderable reduction in density (to 0.70) and increases the sensitiveness of the explo-
sive to detonation.

Powdered plastic foam may also be used for this purpose.
In the U.S.S.R. a cartridge of a basic substance was added to charges of
explosives to neutralize the acid and harmful products which might result from
incomplete detonation (e.g. nitrogen dioxide). This was a cartridge comprising
calcium hydroxide in an amount equal to 25% weight of the whole stemming. This
reduced the quantity of acid gaseous products (C0 2 and N0 2 ) by 10-20%.
The addition of water to the stemming has a similar effect.
An important ingredient for improving the composition of the products of
explosion used in Great Britain is calcium sulphate.

The effect of the addition of barium sulphate has also been tested in Great Britain.
This substance has the exceptional property of sensitizing a nitroglycerine gel im-
mersed in water, so that the explosive is capable of propagating an explosion with a
detonation rate of 6000 m/sec, under water at depths as great as 2750 ft. According
to J. Taylor and Gay [4] this phenomenon may be due to the reflection of the shock
wave at the interface between the explosive and the sensitizing powder, an effect
which occurs only when the latter is compressed above a certain minimum density.
This has been exploited in the design of explosives for deep seismic explosions

TESTS FOR MINING EXPLOSIVES

Any explosive which is intended for use in mines has to satisfy certain require-
ments and to ensure overall safety. All mining explosives, both for coal and rock
must be subjected to the tests before being licensed for use. The tests vary from
one country to another. Those described below are based mainly on methods applied
in Poland.

TRANSMISSION OF DETONATION

To examine this property of explosives, it is necessary to determine the maximum
distance between two cartridges (weighing 100 g each, 30 mm dia.) at which detona-
tion will propagate from the cartridge fitted with a No. 8 detonator to the second
cartridge which has no detonator. This distance should not be less than 3 cm, and

4 34 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

for special permitted explosives 4 cm; for ammonium nitrate explosives it is usually
4-8 cm and for dynamites it is more. Sometimes the transmission of detonation
is expressed in terms of units, based on the calibre of the cartridge (e.g. 32 mm).

SENSITIVENESS TO DETONATION

To determine the sensitiveness to detonation 3 cartridges of explosive each
weighing 75 g and 25 mm dia., or 4 cartridges each weighing 100 g and 30 mm
dia. may be used. The cartridges are set out in a row so that they touch each other.
One of the end cartridges is then detonated by the weakest possible detonator.
The explosives tested should be detonated completely under these conditions by
a weak detonator*. The weakest detonator producing the detonation of the above
mentioned cartridges is characteristic to the sensitiveness to detonation of a given

explosive.

Testing transmission of detonation and sensitiveness to detonation is intended
to establish whether or not a given explosive ensures complete detonation in a
shothole. An incomplete detonation is always undesirable since it involves a con-
siderable waste of time and effort and may cause an accident during subsequent
shots. Both tests are always employed as inspection tests during production. Freshly
manufactured explosive should not be used in these tests, since much better results
are obtained after a few days' storage. Thus, according to T. Urbanski's experiments
{68], the permitted explosive Bradyt F (p. 475) initially gives a transmission of
detonation value of 14-15 cm. The next day, the same cartridges give values not ex-
ceeding 7 cm. This occurs because the freshly-mixed explosive contains a consider-
able quantity of air which greatly facilitates the transmission of detonation. After
a few days or even a few hours, when the excess of air has escaped, the transmission
of detonation deteriorates. These changes in the explosive, however, are not so
great as to affect the rate of detonation and expansion in the lead block.

In dynamites and in blasting gelatine, containing a large amount of nitroglycer-
ine gelatinized with collodion cotton, the changes in properties caused by the
presence of air persist for a very long period of time (often about a year) since the
air dispersed in the viscous colloid of blasting gelatine remains there much longer
than in explosives with a powdery consistency. E.g. blasting gelatine, containing
91 % nitroglycerine and 9% collodion cotton gives a lead block expansion of 580 cm^
immediately after mixing and 545 cm3 two days later. After longer storage (more
than a month) this figure may fall to 500 cm^ or to even less.

The rate of detonation undergoes similar variations. E.g. dynamite consisting
of 62.5% nitroglycerine, 3.5% collodion cotton, 28% potassium perchlorate and
6% woodmeal detonates with a rate of 7000m/sec immediately after mixing. In

* To determine the sensitiveness to detonation of mining explosive in some countries (including
Poland) special standard detonators have recently been introduced, containing 0.05, 0.1, 0.15, 0.20 g
etc. of silver azide. The stronger contains 0.60 g of silver azide or an equivalent quantity of
lead azide.

MINING EXPLOSIVES

435

six months the rate of detonation may decrease to 2750 m/sec. If the explosive is
kept in warm premises the rate may decrease still more— up to 1800 m/sec. If the
cartridges harden due to the setting of ammonium nitrate or potassium chlorate,
transmission of detonation becomes worse and detonations may be impaired.

An explosive intented for tests on the transmission and rate of detonation must
therefore be seasoned for at least 24 hr.

With dynamites the tests should be repeated after a few days and then after a
few weeks.

The detonators employed for these tests should be reliable and their initiation
strength must be checked.

It should be pointed out that the transmission of detonation depends not only
on the composition of the explosive and the previously mentioned factors but also
on the diameter of the cartridges, their confinement and their density.

The transmission increases with the diameter up to a certain diameter at which
it becomes constant.

An investigation of the probability of transmission of detonation as a function
of the diameter of the cartridge was carried out by Hino [69]. This is depicted by
the diagram in Fig. 143.

Diameter, mm

Fig. 143. Probability of detonation (in %) as a function of cartridge diameter, ac-
cording to Hino and Yokogawa [59]. R t — diameter which does not give any transmis-
sion of detonation; R c - "critical" diameter (50% probability of transmission of
detonation); R m — minimum diameter which gives 100% probability of transmission

of detonation.

The transmission of detonation in a confined space (in a tube or shothole) is
higher than in the open air. E.g. an ammonium nitrate explosive with 4% nitro-
glycerine which ordinarily has a transmission value of 10-15 cm, in a mortar gives
the value 19-23 cm, according to T. Urbanski [70]. This author studied change of
transmission as a function of density. The results for ammonium nitrate explosives
are given in the diagram in Fig. 144.

The decrease in transmission values with density beyond a certain value (density
higher than 1.1) seems to be of particular significance. T. Urbanski showed as early
as 1926 [70] that the shock wave produced by detonation of an explosive can move
along a shothole, or a mortar, at a rate higher than the rate of detonation (e.g.

436

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

for the rate of detonation of 3220 m/sec in a tube 35/42 mm, the rate of the shock
wave was 3830 m/sec). The shock wave may "dead" press the cartridges to a degree
which renders their transmission of detonation difficult.

According to Hino and Yokogava [59] an addition of surface active agents
(0.5-1 %) to mixtures of ammonium nitrate with liquid coal-tar improves the trans-

s 14

cfff
o

%10

Bradyt C

Bradyt F

Ammonit 4

°s.

0.8 0.9

Fig.

1.0

1.1

1.5

1.6

1.7

1.8

1.2 1.3 1.4

Density

144. Transmission of detonation as a function of density, according to T. Urban-
ski [70].

mission of detonation and increases the rate of detonation. These authors assume
that the surface active agents increase the rate of reaction between the ammonium
nitrate solid phase and the coal-tar liquid phase.

160

a
1

to
2

150

140

130

120

110V

100

< I I i

. ... : I I |

1 2

Storage time

32
Weeks

Fig. 145. Change in the power of explosives on storage [71]; /—ammonia gelatin,
type B with 35% NG; //—ammonium nitrate explosive, type II with 6% NG and 76%

NH4NO3.

MINING EXPLOSIVES

437

Wetterholm [71] gave a number of diagrams illustrating the change of explosive
properties of mining explosives during storage.

Figure 145 suggests a very small reduction |n the power of dynamite and ammo-
nium nitrate explosives on storage, measured in a ballistic mortar. Figure 146 indi
cates the change in the transmission of detonation during storage. The decline in trans

1 2

Storage time

32
Weeks

Fig. 146. Change in the transmission of detonation of explosives on storage [711- /-
ammoma gelatin, 1 in. (25 mm) dia., type B with 35% NG; //-ammonium nitrate
explosive, 1 ,n. (25 mm) dia., type II with 6% NG and 76% NH4NO3; ///-semi-
plastic explosive, ll in. (30 mm) dia., type S with 54% NG.

mission of a gelatine explosive (35% nitroglycerine) can be halted and restored by
re-kneading the explosive. The change in transmission of detonation with tem-
perature is depicted on another diagram (Fig. 147). The change of the rate of
detonation of these explosives on storage is given in Fig. 148. Wetterholm also
gives the value of the rate of detonation at various temperatures (Table 107).

Table 107

Temperature

Ammongelatine

(35% NG)

m/sec. No. 8

detonator

Ammonium

nitrate

explosive

(6% NG) m/sec,

No. 8 detonator

-25

+ 35

2300
2450
2700

3350
3350
3600

438

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

-30

-20

-10 . +10 +20

Temperature, "C

+30

■30 -20 -10 +10 +20 +30 +40 +50 +60 +70 +80 +90 +100
Temperature, °F

Fig. 147. Change in the transmission of detonation of explosives with temperature
[71]; /—semi-plastic explosive, type S with 54% NG; //—ammonium gelatine, type B
with 35% NG; ///—ammonium nitrate explosive, type II with 6% NG and 76%

NH4NO3.

YUUU

i i

i
/

i i

§5 6000

16 5000

—

=| 4000

€ 3000

III

° 2000

"§ 1000

i i

1 i

1 2

8 16

Storage time

Weeks

Fig. 148. Change in the rate of detonation of explosives on storage [71]; /—ammonia
gelatin, iron tube 1 in. (25 mm) dia., type B with 35% NG; //—ammonium nitrate
explosive, iron tube 1 in. (25 mm) dia., type II with 6% NG and 76% NH4NO3;

///-

• ammonium gelatine, cardboard tube 1 ^ in. (40 mm) dia., type B with 35% NG.

POWER OF EXPLOSIVES

The power of explosives is determined in a conventional manner. The most
usual test is the determination of the lead block expansion and the rate of detona-
tion. In some countries the crushing of lead and copper cylinders is determined.

MINING EXPLOSIVES 439

In Anglo-Saxon countries a test in a ballistic mortar is also used. Tests for explosive
power should be repeated occasionally for inspection purposes in production.

SAFETY TESTS WITH METHANE AND COAL-DUST

Explosives for use in coal mines containing methane or ignitable coal-dust
must pass a test in a testing gallery. Different procedure is used in different coun-
tries. General data on galleries and methods of testing are summarized in Table 108.

Experimental gallery in Poland

The experimental gallery for testing explosives for use in the mining industry
(Fig. 149) is a wrought-iron tube, circular in cross-section, 44.2 m long and 2.0 m in
inner diameter. One end of the tube is open, while the other is closed with a concrete
block. There is a cavity inside the concrete block in which a steel mortar is mounted
(Fig. 150). The cavity is connected through the safety valve, vent shaft and air damper
with a powerful electric ventilator for airing the gallery after the shot. The safety
valve and air damper close the vent shaft and protect the ventilator from accidental
damage when firing takes place.

Throughout the length of the gallery there is a floor of sheet iron with an empty
space below filled with concrete except for the first three meters from the closed end,
where the heating pipes are situated.

Inside the gallery, at a distance of 3.4 m from the closed end a ring is fixed to
which a paper partition is fastened so as to form a chamber of 10 ml In the chamber
a mixture of methane and air is prepared. In winter this is heated to a temper-
ature of above 0°C by steam in the heating pipes beneath the floor.

In the wall of the gallery facing the observation post there are observation win-
dows at intervals of 1 and 2 m along its whole length, through which the course
of firing is watched.

Between the first windows at the firing end there are apertures for the firing
leads and arrangements for measuring the temperature inside the gallery.

After the cables have been laid, the paper partition is fastened in place, and
the cables are connected through the short-circuiting unit with the observation
post.

The concrete block closing the firing end is situated in a compartment in which
the methane dispensers and the ventilators are located.

At a distance of about 10 m from the gallery, half way along its length, an ob-
servation post is located, from which the charges are fired and the results observed.

In Poland the following standards established by Cybulski have been in force
since 1950. Tests are made with cartridges 32 mm dia., placed in a mortar of 50 mm
dia., unstemmed. When testing permitted explosives* in the presence of firedamp,

* i.e. safe towards firedamp and coal-dust.

440

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

n
<

<*H

. ° <3

i is ff |

r»

n m

en ro

C<1

E ^

&0 .tS

h, 60 60

60 60

•5

* g
§ o

o >,

.3
E
E

.S * M
P 60 .2

« E u
" E S
o «> ""

o .s .g

60 i i

Poo

C (A (A

« o o

a G
E 'E
E E

O V

-*-> -4-*

l/l VI

E
E

60 5?

■i!

2 §

o o

ft ^

G in ri

go G C

a c

rt o

1 E
E S
S

•S 53

42 t;

in in oo

in in

»n

in vi co

»n in

iri

»n

o 2

£ E

O <u

•a

ja ji

60 E

o o ts ■*

O O

cfl «*

c o

v> no cs m

r-i

fS NO

t=:

hj (J

7 !

O O

& E
•3 8

O oo

1 +

a S

•n

■*

ON ON

a\ 1 1

ON 1

•o

*—*
1

0O 00

O J-

ex ex o.

u u

«D

II 1

3 „

G G

•5 •S

"3
hi

■*-; vi
G 60

o a3

E E

«3
E

+-J VI

G 60

t3
E

>, hi

i-H

ipacit

f gas

ambe

r-;

\o o

I-l

•»\

t 15

u^ in

"3 _

E E

-a

vo — . o © o

>n m

o o

r— 1

s >>

o h

••5 O
VI ^

2 BO

s <~

-: So 'g, <n c4

!
»n

T-H

f E

O ■* fS <N

Tf >n

<S (N — i — i

*— i

en <N

•a

60 t3 ■=
=« > S

a, O >

e
o

3
ffl

ttl C3

'g
o

CO CO

§2

o
3 ja o

■a

•<

a<3 1

^ S3

O 3

° a
>, o

2 Ji
Si M

'3 *

O 6

a ■-

MINING EXPLOSIVES

441

-lUSL'l

a
"o

1-1

442

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

10 shots, each of 500 g, are fired with the detonator placed at the front of the charge
(i.e. initiated directly). The content of methane in the air is 8-9.5%. With special
permitted explosives the requirements are much more stringent. The test involves
1 kg of explosive, fired by means of a detoriator placed at the rear of the charge
(i.e. inversely initiated). The number of shots should not be less than 20.

Fig. 150. The closed end of the experimental gallery simulating a coal-face, at

Mikolow (Poland).

When testing the safety of normal and of permitted explosives (i.e. permitted
Metanits and Karbonits) in the presence of coal-dust 10 shots, each of 500 g, flirectly
initiated, are fired. A cloud of coal-dust is formed by firing small charges (5-7 g)
of permitted explosive in bags with coal-dust, suspended in the explosion chamber
of the gallery. This creates the most favourable conditions for the ignition of coal-
dust, and is therefore a most rigorous test for the explosive. The shots in the bags
are fired simultaneously, just 0.5 sec before the test shot.

MINING EXPLOSIVES

443

The classical method of firing explosives in a testing gallery with the charge
inside the mortar does not present a true picture of the safety of explosives towards
coal-dust since the explosives commonly used fail to ignite coal-dust under such
conditions. The angle shot mortar (Fig. 138) is therefore sometimes used as an
unofficial auxiliary test.

Special permitted explosives are tested under much more drastic conditions.
As previously stated 1 kg of inversely-initiated explosive is fired creating a cloud
of coal-dust. The number of test shots should not be less than 20. Safety in the.
presence of coal-dust is also tested by a special method worked out by Cybulski,
which consists in firing two charges, each weighing 1 kg, simultaneously from two
opposite mortars. The distance between the mortars is 1 m. A cloud of coal-dust
is obtained as described above. The charges are inversely initiated. The number
of shots is 20. The test is considered exceptionally stringent. Such tests are repeated
from time to time for inspection purposes.

In addition the flashes produced when the explosive charges are fired in the air,
are recorded photographically, as an auxiliary test.

Experimental gallery in U.S.S.R.

In the U.S.S.R., tests to authorize permitted explosives for use in dangerous
mines are carried out according to the standard specification— GOST 7140-54.

Each lot of the explosive manufactured is tested by drawing samples from 2%
of the boxes.

Gas chamber.
Aperture
Mortar^
t

Stirrer Diaphragm

Mortar spraying
coal-dust

.Windows

Fig. 151. Diagram of the testing gallery in the U.S.S.R., according to GOST 7140-54.

A testing gallery is shown diagrammatically in Fig. 151. The wall which closes
the gallery is fitted with an aperture 300-400 mm dia., to which the mortar is brought
on a truck.

The testing of explosives for use in the presence of methane-air mixture is carried
out by firing a charge of 600 g. For those used in the presence of coal-dust a charge
of 700 g is used.

Three tests are carried out on each lot of explosives intended for use in methane-air
mixture and two tests on each lot for use in the presence of coal-dust; in one of
these no coal-dust is present in the mortar, in the other 100 g of coal-dust are placed
in the bore, most of it between the charge and the mortar's mouth.

444

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Should the methane-air mixture or the coal-dust be ignited, repeat tests are
carried out in duplicate on samples drawn from twice the number of boxes. If these
tests again give unsatisfactory results, even only for one substance (gas or dust),
the lot is rejected.

At least once a year every production plant sends samples (of not less than 60 kg
each) of all permitted explosives manufactured there, to the Makeyevskii Scientific
Institute of Safety in Coal Mines Research (MAKNll) for a check inspection.

Samples of permitted explosives for use in the presence of methane-air mixture
or of coal-dust are tested by the MAKNII according to paragraph 5 of GOST
7140-54 and repeated in accordance with paragraph 6 of GOST 7140-54.

MAKNII carries out tests on experimental new substances or pilot production
lots (to be authorized for handling and use) by firing 10 shots from each batch in
methane-air, and 10 from each lot in coal-dust (5 with and 5 without coal-dust
(100 g) in mortar).

The stemming used when testing explosives in the presence of methane is
prepared as follows :

Plastic clay is dried to a moisture content of less than 5% and stemming material
is made from it by mixing four parts by weight of the clay with one part by weight
of water. The mixture is stirred carefully until a uniform paste is obtained.

50-55 g of clay so prepared containing 20-25% water is pressed into a cake
10 mm wide and 55 mm dia. to make the stemming (Fig. 152).

Explosive charge

Bore-hole

Stemming.

Electric
detonator

Fig. 152. Charged mortar with clay stemming, according to GOST 7140-54.

Method of testing in the presence of methane. An explosion chamber lO-llm^
in capacity is separated by a gas-proof diaphragm within the experimental gallery.

The cartridges of explosive being tested are placed in the mortar end to end so
that the charge reaches the bottom of the hole, with the last cartridge facing the
mouth of the mortar not less than 5 cm away from it (Fig. 152).

A No. 8 instant fulminate-tetryl electric detonator is fixed to the cartridge next
to the mouth of the mortar and the hole hermetically sealed with stemming of
plastic clay.

The loaded mortar is moved through the window into the explosion chamber
of the experimental gallery.

MINING EXPLOSIVES 445

Gas (natural or synthetic methane) containing not less than 90% methane and
not more than 8-10% hydrogen is introduced into the chamber.

Before firing the shot the methane-air mixture is stirred by a fan or a mixer and
the methane content determined by a Sager pipette or gas analyser.

The temperature of the methane-air mixture in the chamber should lie between
-10 and +30°C.

Method of testing in the presence of coal-dust. Testing in the presence of coal-
dust is carried out in an experimental gallery without a diaphragm. The lay-out of
the explosive cartridges and the electric detonator in the mortar is the same as that
described above.

The coal-dust used for the tests is obtained by milling either coal from the Make-
yevskii seam of the Tchaikino-South mine, which contains 29-35 % volatile matter,
less than 9% ash and less than 2% moisture, or another coal of the same composition.

According to GOST 4403-48 the coal should be ground so that the residue on
a No. 15 sieve is less than 10% and that more than 50% of the dust passes through
a No. 16 sieve.

Coal-dust is injected into the experimental gallery by means of a spray gun
500-700 mm long with a 150-220 mm bore, loaded with 50±5g of the explosive
under test. The charge is fitted with a No. 8 fulminate-tetryl electric detonator in a
paper or copper tube. 6 + 0. 1 kg of coal-dust is spread over the charge.

The spray gun is placed :

at a distance of 11.5 m from the bottom of a gallery 1500 mm dia.;

at a distance of 10.0 m from the bottom of a gallery 1600 mm dia.;

at a distance of 8.0 m from the bottom of a gallery 1800 mm dia. ;

and at an angle of 20° to the horizontal axis of the gallery.

The charge of the explosive being tested, weighing 700 g, is located in the mortar.

A No. 8 fulminate-tetryl electric detonator in a paper or copper tube is lo-
cated in the first cartridge. Coal-dust spraying is commenced 5-10 sec before the
charge of explosive being tested is initiated.

If a 700 g charge fails to ignite coal-dust under the conditions described above,
the explosive is considered to be safe in the presence of coal-dust.

If any residue of the explosive is found in the bore of the mortar or in the explo-
sion chamber, the test is considered a failure irrespective of whether or not the coal-
dust is ignited.

APPLICATION OF STATISTICS TO GALLERY
TESTING OF EXPLOSIVES

During World War II Dixon and Mood [72] developed a special experimental
method for testing the sensitiveness of explosives to impact. The method gave a
statistical estimation of the mean value. It became known as the Bruceton up-and-
down method. The primary advantage of the method is that it increases the accuracy
with which the mean value can be economically determined. The method requires
fewer tests than other methods.

446

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

It was used for dust-explosion gallery tests on "permissible" explosives at
Buxton at the S.M.R.E. by Gibson, Grimshaw and Woodhead [73]. It is now ac-
cepted in the U.S.A. as a standard procedure for determining the charge limit of
a safety explosive. Figure 153 gives a general idea of the method. Usually the
charge (W 50 ) giving a 50% probability of explosion of firedamp or coal-dust is
determined by this method [75].

580

Is, 540

500

460
420

X XX

/\ AA

n JT, n n

x / X/

X p n n

j — i — i — i — L_i i i i 1 i i i

\/\/\/

X-

n n

5 10

Trial number

Fig. 153. Determination of incendiveness of a permitted (permissible in the U.S.A.)
explosive in a firedamp atmosphere, by the up-and-down method, according to Grant

and van Dolah [74].

STABILITY OF MINING EXPLOSIVES

The chemical stability of explosives depends on their ingredients. Those which
do not contain nitroglycerine may be stored for long periods. Ammonium nitrate
explosives must be protected against moisture and, if this is done, they may be
stored for periods up to 6 months. In practice, all explosives containing nitroglycerine
should be utilized fairly quickly (within 1-3 months, depending upon the regulations
in force in a given country). It is very important to remember that ammonium
nitrate always contains a certain amount of free nitric acid which, if not neutralized
by the addition of suitable ingredients, has a deleterious effect on nitroglycerine,
leading to its decomposition. Apart from chemical stability the constancy of physical
properties is also of importance. E.g. cartridges of ammonium nitrate explosives
often become hard and difficult to detonate and impossible to use due to the impracti-
cability of introducing detonators into them (Vol. II, p. 459). Nitroglycerine explo-
sives which do not contain nitroglycol or aromatic nitro compounds in a propor-
tion sufficient to lower the freezing point may freeze at low temperatures and need
to be thawed. This may involve the separation of nitroglycerine, thus creating an
additional hazard (Vol. II, p. 72).

MINING EXPLOSIVES USED IN VARIOUS COUNTRIES

In the following lists the mining explosives used in the inter-war period and
in current use in different countries are cited.
The lists are presented in alphabetical order :

MINING EXPLOSIVES

447

1, Belgium. 2, Czechoslovakia. 3, France. 4, Germany. 5, Great Britain. 6 Hungary
7, Japan. 8, Poland. 9, U.S.A. 10, U.S.S.R. '

BELGIUM
In Belgium earlier mining explosives had the composition given in Table 109.

Table 109
Belgian mining explosives

Ingredients

RII

Centralite

Bbis

Sabulite

Ammonium nitrate
Sodium nitrate
Ammonium perchlorate
Potassium perchlorate
TNT
DNT

Nitroglycerine
Nitrocellulose
Cellulose
Naphthalene
Carbon black
Sodium chloride
Sodium oxalate

Vbis
Flamivore

Antigrisou
Yonckite

—

0.05

—

4.85

0.1

The explosives in Belgium are now classified into 4 types, according to l'Associa-
tion de Fabricants Beiges d'Explosifs et le Centre de Recherches Scientifiques et
Techniques pour ITndustrie des Produits Explosifs [76] (Tables 110-112).

Type I -dynamites and rock explosives. The chief ingredients of the latter are
ammonium nitrate, nitroglycerine and TNT. Their rate of detonation at a density
of 1.1 is about 3800m/sec, their transmission distance 6 cm.

Type II-explosives which, when a 1400 g charge is fired in a steel mortar in an
experimental gallery, fail to ignite a 9% methane-air mixture or coal-dust. They
are used with an inert sheath in rock workings. Their rate of detonation is about
3300 m/sec (charge 30 mm dia.) at a density of 1.1, their transmission distance 6 cm.

Type III— the S.G.P. (Securite-Grisou-Poussieres) explosives. These are used
exclusively with an inert sheath and must not ignite methane or coal-dust in the
following two tests:

(1) A shot of 1000 g, inversely initiated, fired in a mortar without stemming;

(2) A shot of 1400 g in an angle shot mortar fired at an angle of 90°.

Their rate of detonation at a density of 1.1 and 30 mm dia. is 3000 m/sec, their
transmission distance 8 cm.

Type IV- explosives extra safe in use. Explosives in this group undergo the

448

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

most stringent tests. In an experimental gallery a 2200-2400 g charge (depending
on dimensions of the gallery) must not ignite methane or coal-dust.

Mixtures of this type are represented, among others, by Bicarbite mixtures,
which are not explosives strictly speaking, but are used in cartridges together with
cartridges of genuine explosives, to increase their safety (p. 431).

Common blackpowder or improved "poudre H" is also used in Belgium in
opencast mines. The latter contains a certain amount of ammonium nitrate and
is used only in mines when large charges of explosives are used. It is initiated
exclusively by a detonating fuse.

The composition of modern Belgian explosives is given below.

Type I ammonium nitrate explosives may be represented by Ruptol consisting of:

Ammonium nitrate 73%
Nitroglycerine 10%

DNT 5.5%

Cellulose 5.5%

Woodmeal 4.0%

Aluminium 2.0%

The cartridges are 22, 25 and 30 mm dia.

Table 110
Type II Belgian explosives

Ingredients

Explosives

C. A. Fractorite

B Ruptol

003 Sabulite

Nitroglycerine

Ammonium nitrate

74.5

DNT

TNT

14.2

Sodium chloride

Calcium silicide

Kieselguhr

1.5

Woodmeal

6.5

1.8

Cellulose

1.0

Metal soap

0.5

—

Density

1.1

1.04

1.1

Diameter (mm) (of explosive core and

with sheath)

26/36

23/36

23/37

Date of approval

1958

1957

CZECHOSLOVAKIA

The explosives used in Czechoslovakia over the inter-war period were similar
to those used in other Central European countries.

As rock explosives Dynamites and Donarits with the same composition as
German Donarit were employed (the difference between Czechoslovak and German

*■" -.-.vHWWflfcW

MINING EXPLOSIVES

449

a.
XI
u

3
j

u
n

EL
>>

z
>

o a
.5 8
z §

a
a
o
o
U

I I I § I

2 I

I I I

^| \o

l I I

2 I

l l I

O ^0

O vo

O \©

2 l

! I I

-S '

O \o

o
o

s &

s «>

° s

Z Z

5 E

„ k c
« — .2

.2 S g

S .3 o
o - « -

•o

E J

•2 o

s o -a z -g -o

< Cu H

ti i- g rt g

H ;8 £

■a
c

o
"a

8 6 •-
S .2 *
Q Q

450

chemistry and technology of explosives
Table 112
Type IV Belgian explosives

Explosives

Ingredients

Arionite

Ch. 41 Char-
brite

VIII Nitro-
cooppalite

Bicarbite

Nitroglycerine
Nitrocellulose
Sodium nitrate
Ammonium chloride
Sodium hydrogen
carbonate
Kieselguhr
Cellulose
Metal soap

9.9

0.1

55.6

34.4

1.5
1

54.5
34

0.5

10.11
0.09

55.5
34.3

14.9
0.1

Density

1.13

1.1

1.0

1.44

Date of approval

1957

(pp. 267 and 457) Donarits was that the former contained 6% nitroglycerine and
12-16% nitroglycol).

As examples of safety explosives, Metanits may be cited (Table 113).

Table 113
Czechoslovakian permitted explosives

Ingredients

Metanit A

Metanit N

Nitroglycsrine

24-26

Nitrocellulose

—

0.6-1.0

Aromatic nitro compounds

Ammonium nitrate

71.5

23-27

Sodium chloride

20.5

42.8^49

Woodmeal

Calcium nitrate, 50% solution

2.5-3.5

Palatinol A

0.1-0.5

Density

0.98

1.65

Power (lead block test, cm 3 )

240

168

Transmission (cm)

Rate of detonation (m/sec)

3850

5000

Rock and coal explosives with the composition tabulated below are now used
(Table 114).

In addition, Gelatine-Donarit 1 (p. 455) is used as a rock explosive. TNT is
also employed in opencast coal working and blackpowder is used where low power
and a slow explosion giving large lumps are required.

mining explosives
Table 114

451

Rock explosives

Coal explosives

Ingredients

Plastic*

Gelatine

Dynamon I

Astralit I

Metanit N

Bikarbit *

Nitroglycerine and nitroglycol

12.5

Nitrocellulose

0.8

—

0.8

Nitro compounds

14.1

Ammonium nitrate

61.9

82.9

Calcium nitrate, 50% aqueous

solution

—

Sodium chloride

—

45.9

Sodium hydrogen carbonate

—

<*>

40.5

Woodmeal

2.75

2.5

Dyes

0.3

0.25

0.5

Diethylphthalate 1

0.3

—

Density

1.52

0.96-0.97

1.04

1.4

1.3

Oxygen balance (%)

+ 1.1

+ 3 to +5.5

+ 5.4

+ 0.4

Power (lead block test, cm 3 )

370-390

320-355

270-280

150-170

20-25

Transmission (cm)

15-20

2-6

2-3

Rate of detonation (m/sec)

6200-6500

3900-4200

3800-4800

3200-3800

2350

Heat of detonation (kcal/kg)

902

800

902

107

• Bikarbit is often used as an active sheath. The cartidges are composed of 52% Metanit and 48% Bikarbit.

A recent addition to the list of Czechoslovakian rock explosives [77] includes
Geldonarit composed of:

Nitroglycol 22.0%

Collodion cotton 0.8%

Ammonium nitrate 61.9%

Cyclonite 15.0%

Dye 0.3%

Its properties are:

Density 1.52

Oxygen balance +6.8%

Lead block expansion 370-390 cm 3

Transmission of detonation 15-20 cm
Rate of detonation 6350 m/sec

Heat of detonation 960 kcal/kg

FRANCE

The more modern mining explosives employed in France (Table 115) do not
differ in the main from those used earlier (see p. 403).

There have been modifications in explosives for coal work, in which 5% am-
monium nitrate was replaced by potassium nitrate. These were called the saltpetre
explosives (salpetre).

452

chemistry and technology of explosives
Table 115

Explosifs couche

Explosifs roche

(for coal work)

(for rock work only)

Ingredients

Grisou-

naphthalite

dynamite

naphthalite

dynamite

couche

roche

Ammonium nitrate

87.5

91.5

Nitroglycerine

Nitrocellulose

0.5

TNT

Dinitronaphthalene

8.5

Charge limit

500

1000

Apart from explosives of this kind, others similar to those commonly used in
other countries, i.e. permitted explosives containing sodium chloride, have more
recently begun to be used in France.

There has also been a growing interest in ammonium chloride with sodium
nitrate as ingredients of mining explosives in France.

Table 116
Group I. French explosifs couche ameliores

Dynamites

NType

Grisou-

Explosifs

Ingredients

dynamite

chloruree

n°15

Noburex

grisou

chlorures

n°16

Nn°64

Nn°65

Nitroglycerine

12.3

DNT

0.7

Dinitronaphthalene

1.5

Penthrite

—

Ammonium nitrate

35.5

Sodium chloride

58.5

Woodmeal

—

3.5

Urea-resin foam

—

Peat

Properties

Density

1.3

0.8

1.0

1.3

1.25

Power* (lead block test,

picric acid =100)

Transmission (cm)

10-50

6-12

5-10

4-8

3-7

Rate of detonation (m/sec)

2300

2000

2100

3000

3300

Date of approval

1949

1952

1954

1949

1950

* CUP -Coefficient d'utiUtatlon pratique.

MINING EXPLOSIVES

453

Finally, attempts are being made to use such ingredients as penthrite and cyclonite.
The most modern mining explosives employed in France are classified into
four groups with regard to their safety in use:

I. Improved explosives for coal work (explosifs couche ameliores). They may
be used with delay detonators provided that the delay between the first and the
last shot is less than 5 sec. In particularly dangerous places short-delay detonators
should be used. The minimum initial charge is 1500 g in coal^and 2000 g in rock.

II. Explosives for coal work (explosifs couche). They may be used only with
instantaneous detonators. The minimum initial charge is 500 g in coal and 1000 g
in rock.

III. Explosives for rock work (explosifs roche).

IV. Explosives permitted only in places where safety is not an important factor
(explosifs a l'usage restreint).

With regard to composition they are divided into:

Dynamites, in which nitroglycerine is the chief ingredient;

Type N explosives, in which ammonium nitrate is the chief ingredient.

The most widely used of these explosives is Grisou-dynamite chloruree n°15.

Table 117
Group II. French explosifs couche

Ingredients

Dynamites

Grisou-dynamite I Grisou-dynamite
chloruree n°l I chloruree n°a

Nitroglycerine
DNT

Ammonium nitrate
Ammonium nitrate of low

density
Sodium nitrate
Sodium chloride
Ammonium chloride
Woodmeal
Urea-resin foam

Properties
Density

Power (picric acid =100)
Transmission (cm)
Rate of detonation (m/sec)

Date of approval

20.5
55.5

21.5
2.5

1.25
77

10-40
3300

1933

20.5

55.5

21.5

2.5

0.90
77

6-20
2500

Minuret

1954

20
17

0.85
74
5-15
2300

NType

1952

Nn°7

7
76

15
2

1.0
87
2-6
3700

1933

The most extensively used of these explosives are grisou-dynamite chloruree n°l
and N n°7.

454

chemistry and technology of explosives

Table 118
Group III. French explosifs roche

nts

Dynamite

N type

Ingredie

Grisou-dynamite roche

a la cellulose ou

cellamite

N n°lb

grisou-naphthalite*

roche

Nitroglycerine
Dinitronaphthalene
Ammonium nitrate
Woodmeal

67.5

* 2.5

8.5
91.5

Properties
Density

Power (picric acid=
Transmission (cm)

100)

1.2
111
10-30

1.0
103
2-6

Grisou-dynamite roche is in relatively wide use for blasting rocks.

Table 119
Group IV. French explosifs a l'usage restreint

Dynamites

N types

Dyna-

Ingredients

mite

gomme

mite
gomme
B Am

Nitro-
baronite

Nn°31

Nn°30

Nn°21

N n°0

N n°C

Nitroglycerine

22.75

Nitrocellulose

DNT

TNT

9.8

10.6

19.7

21.3

Dinitronaphthalene

12.6

Penthrite

2.5

4.9

Ammonium nitrate

78.5

80.2

75.4

78.7 •

87.4

Woodmeal

—

1.25

—

Aluminium

Bran

9.2

Properties

Density

1.60

1.55

1.35

1.0

1.1

1.0

Power

(picric acid = 100)

155

145

125

138

132

123

120

111

Transmission (cm)

1-15

5-15

8-12

6-10

8-12

6-10

3-€

Rate of detonation

(m/sec)

2000*
7800

2000*
7500

4200

4000

5100

4200

* Depending on the strength of the detonators.

Dynamite gomme B Am is much used. Explosives n° 31, 30, 21 and are very
little used in coal mines.

MINING EXPLOSIVES

455

GERMANY

German mining explosives are commonly divided into rock and coal explosives,
the word "wetter" being added to their name if safe towards the firedamp ("Schlagen-
des Wetter" or "Schlagwetter").

The following explosives are for rock :

(1) Blackpowder type;

(2) Dynamites;

(3) Ammonium nitrate type; ^^f

(4) Perchloratetype; £j$0B»

(5) Chlorate type.

During the first few years after W$rld War I these explosives were permitted
for coal work if there was no methane or dangerous coal-dust in the mine. Soon,
however, their use was restricted exclusively to rock work or in safe and opencast
mines. Explosives of the blackpowder type (Table 120) were used in Germany even
in mines with dangerous coal-dust, which led to a great catastrophe at the Heinitz
mine in 1923. From then on the use of these explosives was restricted to rock work
only. Some of the most dangerous explosives such as Ammonite 5 (with aluminium)
were not permitted for use in rock work in coal mines but only in quarries. The
most important types of German rock explosives are summarized in Tables 120
and 121.

Table 120
German explosives of the blackpowder type

Composition, %

Name

Potas-

Brown

sium

Sodium

Char-

char-

Pitch

Sulphur

Cel-

Carbon

nitrate

coal

lulose

black

Sprengpulver 1

73-77

10-15

8-15

Sprengpulver 5

70-73

5-7

17-19

5 7

Sprengsalpeter 1

0-25

40-75

10-16

9-15

Sprengsalpeter 2

0-5

65-75

10-16

9-15

Sprengsalpeter 3

0-5

66-76

0-10

15-19

9-11

Sprengsalpeter 4

0-40
and
0-3
FeS0 4

25-70

8-12

10-15

Sprengsalpeter 5

0-40

30-75 J

5-7

17-19

5-7

During World War II Gelatine-Donarit 1 was extensively used, as was Dyna-
mite 1 and Ammonite 1. The composition of Gelatine-Donarit 1 was:

Nitroglycol

Nitrocellulose

TNT

DNT

Ammonium nitrate

Sodium nitrate

22%

0.8%

6%
55%
10%

456 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Woodmeal 1 %

Pyrites ash 0.2%

The oxygen balance of this explosive is +3.7%, its density 1.53. Its explosive

properties are :

Heat of explosion 1030 kcal/kg

Rate of detonation 6150 m/sec

Lead block expansion 380 can 3
Transmission 10 cm

The coal explosives, safe towards methane and coal-dust, are classified into 3
groups:

(1) Ammonium nitrate;

(2) Semi-gelatinous;

(3) Gelatinous.

Semi-gelatinous and gelatinous explosives are dynamites, adopted for coal work
in the presence of methane and coal-dust due to their relatively high contents of
"cooling" salts, mostly potassium or sodium chlorides (sometimes in the form of
an aqueous solution of calcium nitrate).

Owing to the reduced sensitiveness to detonation of the ammonium nitrate
coal explosives in the presence of the chlorides of alkaline metals, at least 4% nitro-
glycerine is added to these explosives. During World War I, when nitroglycerine
was in short supply, it was sometimes replaced on the suggestion of Kast, by an
admixture (5-10%) of potassium perchlorate as a sensitizing agent. The composi-
tion and properties of German coal explosives are listed in Table 122.

As can be seen, the gelatinous explosives of the Nobelite type, safe in the presence
of methane, contain a small amount of calcium nitrate solution. Calcium nitrate
was added to Nobelites to reduce the temperature of the flame of explosion. After
World War I, small quantities of calcium nitrate in the form of a concentrated
aqueous solution were added to the milled nitroglycerine powder (from the post-war
surplus) used as a rock explosive. This was done to counteract dustiness; e.g. Nitro-
glycerinpulver 1 explosive had the following composition :

94-96% of milled nitroglycerine powder
4-6% of 50% calcium nitrate solution

Next Schwanke [78] published a patent in which he suggests the impregnation
of woodmeal with a solution of calcium nitrate. The mixture is dried at temperatures
from 90 to 130°C and compressed into tubes (resembling those of black blasting
powder).

According to Chemisch Technische Reichsanstalt [79] a well-dried (64 hr at a
temperature of 120°C) mixture having the composition:

Calcium nitrate

57%

Woodmeal

42%

Pyrites ash

MINING EXPLOSIVES

457

>
S
O

pa
w

- 8

3 s
H g

D
g

b
s

w
O

o
o

8 '3

5 sa

8 o

o 3

.-a-o

(2 8.

s «

•2 2
•o is

3 lH

\£ rn i-H ^h

•«d- t-h i— i
3 vi r!,

T ^7

"ft <n ih in

d *i ^ o

JH ON ■* VO

*o •/-> tr> «

I I I I

05 o ■* ■*
cs m "o i-~

«N CM

^h cs m in

«> © o o

S I S S

a S a s

re re re a

RRRR

QOQQ

>-l —1 © -H

o « m

H O VI CI

2' 1 ■

00 00

I I

O o

I J, 4 I

2 1

. o

t— — ■

cs _

.3 S © o

c» re T 1 T

o s

r- r- -3-

00 rj-oo

h O 0O
v© t- vo

3.S <i '
o «
Oh

o
o

5 .3 5 m
■& -3 2 op
8 8-2 <i

Oh O

00
00

1 ©
3S

458

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

2
a.

>
'«

"S.

Char-
ge
limit

SO 00

Trans-
mis-
sion
cm

Lead
block
expan-
sion
cm 3

o
rN
tN

in in

O 00
tN rt

Rate
of
deto-
nation
m/sec

o o
in m
[— so
in in

so
o

SO Os
so so

Heat of
explo-
sion
kcal/kg

SO
in

O O
-tf r-
so in

Oxygen
balance

r-l tN

-<fr so"

+ +

OS
en

a
E
o

Calcium
nitrate,

50%
solution

1 1 1

rsi « 1 1 1

Sodium
or po-
tassium
chloride

00 3
-"

sium
chloride

19
sodium
chloride

3
O

<L>

O m
u

m + >n

so' o m -2 d

CO -J" g{ 2

■

"° is

8 §

tN m

+
tN

coal-
dust
0.5

"! "! S m f-.

— i O O C3 C3 O

(agar-
agar
0.5

Nitro-
cellu-
lose

1 1 1

'a
'%

"3
o

so oo r^
odd

Nitro-
glycol

1 i 1

■*. 'I °°.
in o\ i-^
IN tN tN

o
en

Nitro-
glyce-
rine

SO Tj-

■*

III 1

H
Z

1 «N

tN 1 1 1 1

tN 1

III 1 <N

Ammo-
nium
nitrate

<N tN

d
oo

in in

tN so' d
en tN en

OS
tN

u
E

Ammonium nitrate
Wetter Detonit A

(1936)
Wetter Detonit B

1
1

t-t

2
pq

t-
u

Gelatinous
Wetter Nobelit A
Wetter Nobelit B
Wetter Wasagit A

MINING EXPLOSIVES

459

gives an expansion in the lead block of 165-175 cm' when a charge of 29-30 g is
used; a charge of 46-47 g of black blasting powder gives a lead block expansion
of 145-160 cm3.

The disadvantage of mixtures containing calcium nitrate lies in their high hygro-
scopicity, though the hygroscopicity of calcium nitrate is lower than that of ammoni-
um nitrate. Hence more modern explosives containing calcium nitrate also contain
ammonium nitrate. Such mixtures include Calcinite 1, an explosive safe towards
methane, with the following composition :

35.5% of ammonium nitrate
38% of calcium nitrate

7.2% of TNT

4.8% of DNT

6% of nitroglycerine

8% of woodmeal

0.5% of pyrites ash or dark dye or pigment

The more recent list of Germany rock explosives is given in Table 123 [80].

Table 123

Nitroglycerine

Name

and nitroglycol

Am-

Sodium

Aromatic

Wood-

with collodion

monium

nitrate

nitro

meal

Dye

cotton

nitrate

compounds

Dynamit 1

(GFR)

Gelamon 1

32-35

4(M5

12-15

4-6

3-6

0-2

(GDR)

Gelamon 2

27-30

47-52

10-13

7-10

2-5

0-1

(GDR)

Amnion Gellit 2

25-29

57-62

8-12

1-2

0-2

(GFR)

Ammon Gellit 3

18-23

61-67

10-15

1-2

0-1

(GFR)

Gelatine-Donarit

20-22

55-57

10-15

9-12

0.5-2

0-0.5

(GDR)

Donarit 1

4-6

78-81

12-16

2-4

0.3-0.5

(GFR and GDR)

Donarit 2

4-6

80-82

3-5

8-10

0-0.5

(GFR and GDR)

Dekamon and Ammonex (ammonium nitrate with 5-6% Diesel-oil) are also
in use in the German Democratic Republic and the German Federal Republic,
respectively.

At present, permitted explosives in Germany are divided into 3 classes with
regard to their safety in the presence of methane and coal-dust [80, 81].

Class I safety explosives (unsheathed) should fulfil the old safety requirements,

460

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

namely that 600 g fired unstemmed in a 55 mm dia., 600 mm deep shothole in a steel
mortar must not cause ignition of 9% methane-air mixture in five consecutive shots.

Class II safety explosives must fulfil new conditions when tested in an angle-
shot mortar: 350 g of explosive fired in this mortar at a distance of 65 cm from the
wall and at an angle of 40° must not ignite a 9% methane-air mixture. The charge
is increased by increments of 25 g. The charge limit should not give an ignition in
any of five consecutive shots.

Class III safety explosives must be safe in a 9% methane-air mixture when
fired with the maximum number of cartridges that can be placed in a row in the 2 m
long groove of the angle-shot mortar. The experiment starts with 1800 g charge.
It is increased by 200 g increments. The charge limit is determined; this should not
give any inflammation in five consecutive shots.

From the point of view of composition, German safety explosives are classified
as formerly into three groups:

(1) Ammonium nitrate mixtures with a mimmum of 4% nitroglycerine and
cooling salts,

(2) Semi-gelatinous, with ca. 12% nitroglycol or nitroglycerine and nitroglycol
and nitrocellulose.

(3) Gelatinous, with 25-30% nitroglycol or nitroglycerine and nitroglycol and
nitrocellulose.

They do not differ essentially from those given in Table 122.
Interesting innovations are Wetter-Astralit and Wetter-Carbonit B (West Ger-
many). Their composition is given in Table 124.

Table 124

Wetter-Astralit

Ingredients

(used as an active
sheath)

Wetter-Carbonit B

Nitroglycerine and nitroglycol

8.7
(with nitrocellulose)

Ammonium chloride

31.0

Potassium nitrate with tetryl

59.0

"Guar-meal"

0.95

Metal-soap

0.05

Aluminium hydroxide

0.3

Sodium chloride

Sodium hydrogen carbonate

Properties

Oxygen balance (%)

Density

1.45

Lead block expansion (cm 3 )

No data available

Rate of detonation (m/sec)

1750

Transmision (cm)

Heat of detonation (kcal/kg)

106

MINING EXPLOSIVES 461

In East Germany explosives trials should be completed by examination in the
ballistic mortar (as in Anglo-Saxon countries).

GREAT BRITAIN

In Table 125 a list is given of explosives safe in the presence of methane and
coal-dust (Permitted Explosives) used in Great Britain after World War I.

According to Cybulski [82] Polar Viking explosive with a density of 1.01 has
the following rates of detonation depending on the test conditions :

unstemmed, charge 22 mm dia. 1930 m/sec
unstemmed, charge 44 mm dia. 2505 m/sec
in a steel tube 31.7/38 mm 3580 m/sec

The mining explosives in Great Britain are known by conventional names based
on their composition. They may be divided into the following groups according
to their chemical composition :

Dynamites, gelatine dynamites or gelatines— explosives with high contents of
nitroglycerine. In the second group collodion cotton is used to form a gel with
nitroglycerine.

Gelignites— plastic explosives with a lower content of nitroglycerine than dy-
namites and formerly including potassium nitrate as an oxidizing agent; this has
gradually been superseded by sodium nitrate. The presence of sodium nitrate in
explosives is indicated by the letters "N.S.", e.g. N.S. Gelignite.

In all explosives containing ammonium nitrate the names mentioned above
are preceded by the prefix "Ammon", e.g. Ammon Gelatine Dynamite, Ammon
Gelignite etc.

"Sheathed explosives" with an inert sheath, usually containing sodium hydrogen
carbonate, are also employed. After World War II "Eq. S" explosives were intro-
duced. These have already been mentioned (p. 429).

Under The Coal Mines (Explosives) Act 1951, Statutory Instruments No. 1675:
1951 states that all mining explosives used in Great Britain are divided into two
groups with regard to their safety in use :

Non-permitted, for ordinary use where special safety precautions are not re-
quired (Table 126).

Permitted, which have passed special statutory tests as safety explosives, for
use in coal mines, where methane-air mixtures or inflammable coal-dust are likely
to be present.

In 1957 the National Coal Board [83] centralized the purchase of permitted
explosives in Great Britain and introduced standardization of composition, car-
tridge sizes etc.

The testing gallery at the Ardeer factory of Imperial Chemical Industries Ltd.,
which is identical with the Official Home Office Testing Gallery, is shown in Fig. 154
and the testing mortar in Fig. 155.

462

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

a
<

D
Q

<
O

Z
<
w
z
<<
a

H
W

ft
O
a
o

W
a.
w

So
O
J
a.
X
w

z
z

2
pa

^ 2P

ta n

•n

o ^

t- '

1 1 1

1 oo 1

^H ' '

Oh >

u .« m

Pola
Saxon

No.

1 1 J^

1
in

■*'
tN

1 1 1
>n ~-

1
•n

tN en

HEM

r*^

Pola

amsoi

No.

1 1 1

1
in

1 , ,
m

1 1 J.

»n o

<Z)

«n

»n

t~-

J2 o

ft, >>

1 1

i i 7
©"

Q NO
m ^j-

CN tN

1 1 1
•* O
tN

■"S tn

1-H

V£>

^ .

^ 1
in

m »-h

r~i tN

£ o

-i '

i : i

1 1 J.
in o

tS >H

o » «

^ ©

S u -a

VO i-t

<N tN

h u £

> r - O

1 1
in oo

i i i

1 1 1

ON O

LLl Oh

2 l-

•n

C3 &

"oi "5

*- IN

2 *

O O

o '

1 3 '

i i i

1 1 J.

in o

Q sl,

t-*

co G i_.

o >> v

»n

in
tN

,!• J3-P

, «S .

tN tN

Antii
Penr
Pow

1 1

i i I

1
in

I 1 1
in

1 1 J.

»n o

I-~'

r-'

»n

'5 <s

o .

— ' tN

1 1

IT)

1 ! 1

I I I

4 ' i

(O o

■s °

"rt

■O J3

C 2

E £

"^ ^

8 S

'eb 'el

J- 3

"C ft

1 E

3 C

'3 c
2 5

'3
E

3 ccj o

U ^3 -4-»

E § 3

g 3

l<8

3 H h | g S S o §

5 z Z 1 a 5 ,S 5 fl

■2 E .2

i! i

IfPfPPPip^fBPIHP'BPBlfi?

MINING EXPLOSIVES

.... -UP! »<■■ .i- T .~ 1

463

Fig. 154. Experimental gallery at I.C.I.'s Ardeer factory, according to J. Taylor and
Gay [4] (a replica of Home Office Testing Gallery).

The free suspension test and break gallery, described previously, are shown
in Figs. 135, 136 and Fig. 140, respectively.

The Buxton Test, often called the Home Office Test, was introduced in 1932
and is at present the British Official Gallery Test.

When testing in the presence of firedamp two series are fired :

(1) Five shots of 8 oz each, unstemmed;

(2) Five shots of 28 oz each, stemmed with a 1 in. clay plug;

Fig. 155. Gallery mortar at I.C.I.'s Ardeer factory, according to J. Taylor and Gay [4].

464 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

For testing behaviour in coal-dust:

(3) Five shots of 28 oz are fired into a coal-dust suspension each stemmed with
a 1 in. clay plug.

The permitted maximum shot-firing charge in dangerous mines was made the
same for all explosives, i.e. 28 oz in any one shothole, but an exception is made
in the case of certain low density explosives, where the maximum permitted charge
is 18 oz.

The explosive has to satisfy certain requirements concerning sensitiveness to
propagation of detonation.

A new test was introduced in 1953 for new types of permitted explosives: Eq. S.
and sheathed explosives (see p. 429). For testing in a gas mixture it specifies:

(1) Five shots, each of 20 oz inversely initiated without stemming. (In sheathed
explosives the charge includes the sheath.)

(2) Five shots directly initiated and stemmed with a 1 in. clay plug. For Eq. S.
each shot is 36 oz and for sheathed explosives each shot is 28 oz (18 oz for low-
density explosives.)

(3) The test applied for sheathed explosives only consists of five shots, each
of 8 oz directly initiated, and fired unstemmed. (The weight does not include the
sheath.)

For testing in a coal-dust suspension it specifies :

(4) Five shots, each of 20 oz inversely initiated are fired without stemming.
(In sheathed explosives the charge includes the sheath.)

No ignition must occur with any shot. Since 1962 a new test was adopted. It

involves mortar shots into gas, allowing no more than 13 ignitions out of 26 shots.

Modern Permitted Explosives used in Great Britain are divided into five groups.

I. TNT-Ammonium nitrate powders. These do not contain nitroglycerine.
A typical representative of this group is Douglas Powder. Sheathed Douglas Powder
was replaced by Unirend, where the equivalent weight of sodium hydrogen
carbonate in the sheath was replaced by sodium chloride which was incorporated
in the explosive. The safety of freely suspended cartridges is much less than that
normally obtained with similar explosives containing nitroglycerine.

According to J. Taylor and Gay [4] this is probably due to a relatively high rate
of detonation of mixtures manufactured in an edge-runner mill. The influence of
edge-runner mixing on rate of detonation has already been discussed.

Explosives of this group are used for general coal work. They are medium to
low water-resistant.

II. Nitroglycerine gelatine explosives. The nitroglycerine content (25% and over)
and density (ca. 1.5) are high. A typical example is Polar Ajax. Sheathed Polar
Ajax was replaced by Unigex and Unigel. Explosives of this group possess the
highest power per unit volume. They are used for hard rock and blasting in hard
coal or anthracite. They are highly water resistant.

m. Nitroglycerine semi-gelatinous explosives. The nitroglycerine content is
ca. 15%, and density ca. 1.1. They are used for dealing with fairly hard coal. They

OMM^i^jM^^

MINING EXPLOSIVES

465

466

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

show medium water resistance. A typical example is Dynobel No. 2. An Eq. S.
explosive equivalent to sheathed Dynobel is Unibel.

IV. Nitroglycerine powders. They contain ca. 10% nitroglycerine which is not
gelatinized. The density is ca. 1.0. A typical example is Polar Viking. Its sheathed
Eq. S. Explosive is Unifrax.

They are used for work with soft coal. They possess medium-low water resistance.

Table 127

Composition and properties of sheathed permitted explosives and corresponding

Eq. S. explosives (according to J. Taylor and Gay [4])

Explosives

Ingredients

Sheathed

Unigex

Polar Ajax

Unigel

(initial)

(modified)

Low-freeze nitroglycerine

27.5-25.5

28.3-26.3

17.7-15.7

18.0-16.0

Nitrocellulose

1.1-0.1

1.4-0.4

1.1-0.1

Nitrotoluene

2.8-0.8

2.2-0.2

TNT

Ammonium nitrate

42.0-39.0

32.5-29.5

75.4-42.4

Sodium nitrate

46.5^19.5

Sodium chloride

25.5-23.5

30.3-28.3

27.7-25.7

Ammonium chloride

30.0-28.0

China clay

3.5-1.5

5.3-3.3

5.0-3.0

4.7-2.7

Woodmeal

3.0-1.0

3.7-1.7

Wheat flour

1.5-0.5

Oat husk meal

4.5-2.5

2.4-0.4

Diammonium phosphate

1.1-0.1

Acid Magenta

0.05-0.001

0.1-0.001

0.5-O.005

A. S. No. 2

0.01-0.001

Alcohol

0.5-0.0

Calcium sulphate

4.7-2.7

Sodium carboxymethyl cellulose

1.5-0.5

1.1-0.1

Calcium stearate

1.1-0.1

Barytes

5.3-3.3

Volatile matter

2.0-0.0

Gallery tests

Charge limit

Suspended in 9% methane-air

(oz)

10-12

Fired from 47 in. bore mortar

into gallery 9% methane-air

(oz)

20-28

Density of unsheathed cartridge

1.12

1.7

1.3

1.15

Power (as % of that of blasting

gelatine)

unsheathed

ca. 40

sheathed

...,;A.A,.^.a^.

MINING EXPLOSIVES

467

V. Nitroglycerine low-density powder. This is a modification of Group IV
Low density (ca. 0.7) is obtained by incorporating vegetable fibre or peat. A low
density form of ammonium nitrate and sodium chloride may also be used. A 1 Rou
kol is an example of an explosive of this group. Sheathed Rounkol has been replaced
by Unikol. They are used for soft coal, for maximum lump production. Their water
resistance is the same as that of Group IV.

Table 128

Composition and properties of sheathed permitted explosives and corresponding
Eq. S. explosives (according to J. Taylor and Gay [4])

Expl

osives

Ingredients

Sheathed

Polar

Dynobel No. 2

Unibel

Sheathed
Douglas
Powder

Unirend

Low-freeze nitroglycerine

16.0-14.0

12.4-10.4

Nitrocellulose

1.1-0.1

Nitrotoluene

2.2-0.2

1.3-0.3

TNT

Ammonium nitrate
Sodium chloride
China clay

64.0-61.0
16.2-14.2
1.25-0.25

50.0-47.0
36.9-33.9

16.0-14.0
70.5-67.5
17.0-15.0

12.7-10.7
52.1^19.1
39.2-36.2

Woodmeal

6.0-4.0

4.8-2.8

Acid Magenta
A. S. No. 2

0.05-0.01
0.01-0.001

0.1-0.001
0.01-0.001

Volatile matter

2.0-0.0

0.5-0.0

Gallery tests

Suspended in 9% methane-air

Charge limit

(oz)
Fired from 47 in. bore mortar

8-12

24-30

into gallery containing 9%
methane-air (oz)

Density of unsheathed cartridge
Power (as % of that of blasting
gelatine):

1.15

1.2

unsheathed
sheathed

56
43

Recently Imperial Chemical Industries Ltd. developed a class of explosives
known unofficially as "ultra-safe". The general principle of these compositions is
that they contain a higher proportion of cooling salts, ground to a finer size, than
is customary with other permitted explosives. They have proved considerably safer
than any other permitted explosive.

However, increased safety has been accompanied by considerably reduction in
power. In hard material the results were poor, but firing by delay detonators gave

468

chemistry and technology of explosives
Table 129

Composition and properties of sheathed permitted explosives and corresponding
Eq. S. explosives (according to J. Taylor and Gay [4])

Ingredients

Explosives

Sheathed
A 1 Rounkol

Unikol

I Sheathed
I Polar Viking

Unifrax

Low-freeze nitroglycerine
Ammonium nitrate
Sodium nitrate
Sodium chloride
Woodmeal
Plant fibre

Diammonium phosphate
Resin ^»

Acid Magenta
A. S. No. 2

11.0-9.0
57.0-54.0
11.0-9.0
13.0-11.0

12.5-10.5
0.5-0.0
0.5-0.0
0.1-0.05

11.0-9.0
37.1-35.1
14.5-12.5
31.9-29.9

11.0-9.0
0.5-0.0
0.5-0.0
0.1-0.05

Gallery tests

Suspended in 9% methane-air

(oz)
Fired from 47 in. bore mortar

into gallery containing 9%

methane-air (oz)

Charge limit

11.5-9.5
72.2-69.2

11.0-9.0
9.8-7.8

0.05-0.01

Charge limit

13.2-11.2
51.4-48.4

32.0-29.0

8.2-6.2

1.1-0.1

0.05-0.005

0.01-0.001

Charge limit

Density of unsheathed cartridge
Power (as % of that of blasting
gelatine) :

unsheathed

sheathed

0.7

38'

0.8

0.98

0.8

satisfactory results. Their production and use still seems to be at the experimental
stage (Wildgoose [84]). The new classification [98] includes: P-l ordinary permit-
teds, P-2 sheathed permitteds, P-3 Eq. S., and P-4 the new class.

HUNGARY

The composition of coal mine explosives manufactured and used in Hungary
is given in Tables 130 and 131.

Dynamite explosives of novel type ("Nidin") with nitroglycerine and nitro-
glycol mixtures seem to be particularly popular.

Ammonium nitrate explosives, both rock and permitted types, do not differ
from those generally used in Central Europe.

JAPAN

As in European countries the development of explosives in Japan has had a
chequered history. By the end of the nineteenth century dynamites and gelignites
were used in mines. In 1899 a dreadful explosion occurred at Toyokuni resulting

mining explosives

Table 130
Hungarian dynamite type explosives

469

Ammon-

Nidin

Ingredients

dinamit

Nitroglycerine

30+1

Nitroglycerine and

nitroglycol

33 ±0.5

40 ±0.5

50 ±0.5

60 ±0.5

80+0.5

Collodion cotton

1.25±0.1

1.5 ±0.2

2 ±0.2

3.1 ±0.2

4 ±0.2

7 ±0.3

Nitrotoluenes

3.75±0.1

0±0.3

—

TNT

5 ±0.5

—

Woodmeal

1±0.1

2 ±0.3

4.5 + 0.3

3.7 ±0.2

5.5 + 0.2

2 ±0.2

Ammonium nitrate

58.8 ±1.5

57.6 ±1

51.8±1

42.5 ±1

31.8±1

—

Sodium nitrate

10.3 ±0.5

Magnesium oxide

0.6 ±0.02

0.3 ±0.01

—

Aluminium hydroxide

0.6 ±0.02

0.1+0.01

0.2 ±0.02

Ferric oxide

0.3 ±0.01

0.3 ±0.02

0.3 ±0.01

Sodium hydrogen

carbonate

0.1 ±0.01

0.4 + 0.01

0.2 ±0.02

0.2 ±0.01

Glycerol

1.2 ±0.2

—

Dibutylphthalate

1±0.2

Density

1.4 ±0.2

1.42 ±0.2

1.5 ±0.2

1.48 ±0.2

1.54 ±0.2

1.56 ±0.2

Oxygen balance (%)

+0.38

+ 1.68

+ 2.6

+4.8

+ 1.61

+0.3

Lead block expansion

(cm3)

415 ±20

390 ±20

370 ±20

420 ±20

420+20

480+20

Transmission of

detonation (cm)

7±2

10±2

Rate of detonation

(m/sec)

5600 ±200

5500 ±200

5800 ±200

5900 ±200

6100+200

6600 ±200

Heat of detonation

(kcal/kg)

1160

1140

1180

1240

1345

1270

in the loss of 210 lives. Further great catastrophes ensued in 1903, 1906 and 1907,.
the last one (also at Toyokuni) causing the death of 365 people. It was caused
by the explosion of coal-dust initiated by the explosion of a methane-air
mixture.

This led to the manufacture of Anzen-Bakuyaku ammonium nitrate safety
explosives on the lines of European compositions, but later modified to suit the damp
climate of Japan (Yamamoto [85]). The production of the following new explosives,
began in 1913:

Ume (plum bossom) dynamite derived from British Saxonite; Matsu (pine tree)
dynamite, a kind of blasting gelatine; Ran (orchid flower), derived from Carbonite
and Kaede (maple leaf), a modification of Belgian Grisoutite. The last two were
soon withdrawn from use due to their undesirable products of explosion. The
composition of the above explosives is tabulated below (Table 132) (according to>
Yamamoto [85]).

470

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Table 131
Hungarian ammonium nitrate explosives

Ingredients

Paxit 3

Nitrocertuszit

Ammonium nitrate

82±1.5

72±1.5

Di- and trinitrotoluene (1 : 9)

5.4+0.5

4.9 ±0.5

Nitroglycerine

5 ±0.5

4 ±0.5

Collodion cotton

0.1 ±0.002

1.0 ±0.002

Woodmeal

3 ±0.2

0.5 ±0.1

Charcoal

1.7 ±0.2

Wheat meal

2.5 ±0.2

1.5 ±0.2

Sodium chloride

17+1

Aluminium hydroxide

0.1 ±0.02

Lubricating oil

0.3 ±0.02

0.7 ±0.01

Ferric oxide

0.1 ±0.002

1.0 ±0.002

Density

1.04

1.08

Oxygen balance (%)

+4.55

+ 7.02

Lead block expansion (cm 3 )

385 ±20

235 ±20

Transmission of detonation (cm)

Rate of detonation (m/sec)

3950 ±200

3600 ±200

Heat of detonation (kcal/kg)

955

603

Table 132
Japanese mining explosives

Explosives

Ingredients

Anzen
Bakuyaku

Ume

Ran

Kaede

No. 1

No. 2

dynamite

Nitroglycerine

Nitrocellulose

—

DNT

—

Ammonium nitrate

—

Potassium nitrate

—

Sodium nitrate

—

Barium nitrate

—

Woodmeal

—

Kieselguhr

—

Ammonium oxalate

—

Sodium sulphate

Recently rock explosives mainly for use in opencast mining, composed of am-
monium nitrate with fuel oil (as invented in the U.S.A. see p. 482) were introduced
in Japan.

In 1915 the Regulation for Preventing Coal Mine Explosions was issued. In
1917 an official experimental gallery was erected at Nogata.

MINING EXPLOSIVES

471

After World War I it was found that the mining explosives used till then in
Japan did not pass the more stringent tests in the experimental gallery (e.g. those
with a 400 g charge). Over the period 1922-24 new, safer explosives appeared which
were classed into three groups.

(1) Ume dynamite which was improved by adding a considerable quantity of
borax thus making it similar to the British Samsonite.

(2) Shoan dynamite (Shoan is the abbreviation of Shosan-ammonia, i.e. am-
monium nitrate). Shoan dynamite is semi-gelatinous and resembles British Dynobel.

(3) Shoan Bakuyaku, a kind of Ammonite (Table 133).

A new modern experimental gallery (modelled on the one at Buxton in Great
Britain) was constructed in 1927 and the explosives which passed the gallery test
with the charge of 400 g were designated Permitted Explosives (Kentei Bakuyaku).

During World War II the composition of some explosives was altered due to
lack of raw materials such as borax. Ume-dynamite was therefore replaced by
Shiraume-dynamite into which ammonium nitrate and sodium chloride were in-
troduced.

After World War II new testing galleries were erected at the factories, facili-
tating improvements in the manufacture of explosives. The following new types
were also introduced. Taketoyo Factory (Japan Oil and Fat Co., Ltd.) introduced
new kinds of Toku (special) permitted explosives which contain dried pulverized
seaweeds as a cooling agent. Seaweeds contain halogen (CI, Br, I) and alkali (Na, K)
ions (ca. 1 0% by weight) in a finely divided form.

Table 133
Newer Japanese mining explosives

Explosives

Ingredients

Ume

Shoan

dynamite
A

dynamite

dynamite
B

Bakuyaku
A

Bakuyaku
B

Nitroglycerine
Nitrocellulose
DNT

Ammonium nitrate
Potassium nitrate

49-51
1.5-2.0

8-12

19-21
0.5-0.7

44^18

7-11
0.1-0.3

62-65

4.0-4.5

6-8

70-72

7-9
74-76

Woodmeal

1-3

3-6

4-7

2.5-3.5

1.6-2.5

Starch

Sodium chloride

Borax

34-36

1.0-2.5
26-28

1.5-2.5
19-22

19-21

14-16

Charge limit (g)
Ballistic pendulum

600

800

700

600

500

swing (mm)

56 1

Another improvement consisted of preparing low density explosives (L.D.
explosives) which contain ammonium nitrate in the form of bulky, porous crystals

472

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(Vol. II, p. 460). Some of the L.D. explosives can be classified as Eq.S. explosives.
S-Shoan dynamite contains powdered talc besides ordinary cooling salts and pos-
sesses a high degree of safety.

The following tests for safety in the presence of methane and coal-dust have
now been introduced in Japan [86].

Gas test. The charge loaded into the mortar is 400 g. A No. 6 electric detonator
is detonated near the mouth of the mortar. The test is repeated ten times without
stemming. No ignition should occur.

To investigate new explosives, various charges over 400 g are used. Some of
the permitted explosives do not cause any ignition with a charge of up to 800 g.

These explosives are classified as:

"no ignition at 500 g"
"no ignition at 600 g" etc.

Coal-dust test. 1500 g coal-dust is uniformly scattered on the four shelves fixed
to the walls of the explosion chamber of the gallery, and stirred by the air current
produced by a fan. One minute after stopping the fan, the charge is fired. Relative
humidity in the explosion chamber is kept below 80%. The test is repeated five
times under the same conditions as those in the methane test. No ignition should
occur.

Ballistic pendulum test. 100 g charges should give a swing of more than 40 mm
to the five tons pendulum.

Table 134

Japanese permitted gelatine and semi-gelatine dynamites

Explosives

Ingredients

Shiraume

Shoan

Toku Shoan

L. D. Shoan

dynamite

Shin dynamite

dynamite

Nitroglycerine (and nitroglycol)

33.0

8.0

Nitrocellulose

1.3

0.2

0.3

0.2

Ammonium nitrate

32.8

63.0

69.7

66.6

Woodmeal or starch

2.4

8.0

4.0

8.2

Oil

0.5

—

Seaweed

—

10.0

Sodium chloride

30.0

20.8

8.0

17.0

Properties

Density

1.57

0.95

0.90

0.70

Charge limit (g)

—

700

Swing of ballistic pendulum (mm)

Rate of detonation (m/sec)

6030

3000

2200

Lead block expansion (cm 3 )

260

250

240

Relative power (% of blasting

gelatine

63.1

56.0

57.0

Gap test : cartridge diameters

(32 mm)

MINING EXPLOSIVES

473

More stringent tests, not yet approved officially, are applied to new explosives
showing a high margin of safety.

The tests listed above resemble those employed in Europe, and consist of:

(1) detonating a charge freely suspended in methane-air atmosphere,

(2) detonating a charge mounted in various ways: plain steel plate, an angle
bar, steel rail, an angle-shot mortar, or an angle-shot mortar with a ricochet plate,

(3) detonating a charge suspended in a kraft-paper tube, with or without a slit,

(4) firing from a mortar into the gallery by inverse priming, with an expansion
chamber.

The composition and properties of the most typical Japanese explosives are
summarized in Tables 134 and 135 (according to Yamamoto [85] and Yokogawa [86]).

Table 135
Japanese permitted ammonium nitrate explosives

Explosive

Ingredients

Ko Shoan

Toku Sho-

L. D. Sho-
an Bakuyaku

Shin
D Shoan

Bakuyaku

an Bakuyaku

E 2

Bakuyaku

Ammonium nitrate

64.5

72.8

76.2

Low density ammonium nitrate

—

73.9

Nitroglycerine

5.96

4.7

5.0

Nitrocellulose

0.04

0.1

DNT

—

1.92

2.0

TNT

2.0

—

3.0

Nitronaphthalene

—

1.28

Dinitronaphthalene

6.0

Woodmeal

7.0

Starch

5.5

6.2

2.0

2.8

Seaweed

4.0

—

Sodium chloride

18.0

5.0

10.0

5.0

Potassium chloride

8.0

7.0

Properties

Density

0.96

0.95

0.70

1.00

Swing of ballistic pendulum

Rate of detonation (m/sec)

3010

4000

Relative power (1 % of blasting

gelatine

63.8

—

67.4

Gap test: cartridge diameters

(32 mm)

For blasting and quarrying rock blasting blackpowder or ammonal type explo-
sives are used. A typical feature of Japanese explosives is the admission of a few
ammonium perchlorate explosives, named Carlits (inventor Carlson). E.g. Midori

474

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(green) carlits are permitted for coal mining while Kuro (black) and Murasaki
(purple) carlits are rock explosives unsuitable and suitable, respectively, for under-
ground use. Toku Kaba (special brown) carlit— -also a rock explosive— is marked
by producing non-toxic fumes and is recommended for underground work where
ventilation is inadequate.

The composition of some carlits is tabulated below.

Table 136
Japanese perchlorate explosives

"Carlit" explosives

Ingredients

Kuro

Murasaki

Toku Kaba

Midori

III GO V GO

Ammonium perchlorate

72-77

81-86

46-51

5-9

4-9

Ammonium nitrate

58-63

59-64

Sodium nitrate

—

33-31

3-7

2-7

Dinitronaphthalene

8-13

1-5

TNT

6-11

6-10

Oil

1-5

1-3

—

Woodmeal

4-9

3-7

2-6

1-5

1-6

Ferrosilicon

14-18

8-12

—

Sodium chloride

11-16

10-15

Properties

Density

1.05

1.11

1.15

1.05

1.06

Swing of ballistic

pendulum (mm)

68 67

Rate of detonation

(m/sec)

4400

4000

3000-4300*

3000^1300*

Lead block expansion

(cm 3 )

485

465

320

285

Relative power (% of)

blasting gelatine

Gap test :

cartridge diameters

(32 mm)

4.5

* Different figures according to various sources.

Explosives with a safety sheath are also used. The safety sheath must correspond
to Japanese Industrial Standard M 7609 (1952). Two kinds are used: (1) a sheath
wrapper into which the cartridge is rolled, and (2) a sheath tube made in the form
of a pipe into which a cartridge is inserted.

Safety sheaths undergo gas and coal-dust tests in the gallery in which the sheath
covers 450 g of a standard explosive composed of:

Ammonium nitrate 83.7%
Nitroglycerine 8 %

Collodion cotton 0.3%
Woodmeal 8.0%

MINING EXPLOSIVES 475

The power of sheathed explosive is tested using the sheath to cover a standard
explosive composed of:

Ammonium nitrate 64.8%
Nitroglycerine 8.0%

Collodion cotton 0.2%
Woodmeal 7.0%

Sodium chloride 20.0%

The swing of the ballistic pendulum should be 55 mm or more and the gap
distance 3.5 times the diameter or more.

POLAND

After World War I German explosives were used to some extent in Poland.
Detonating powder, black blasting powder, dynamites, ammonites and chloratits
were used as rock explosives. Chloratit 3 with a composition corresponding to that
of Miedziankit of Laszczynski (p. 278) ranked high among rock explosives. Ex-
plosives safe in the presence of methane and coal-dust, such as ammonium nitrate,
Lignozyts, Bradyts, semi-gelatinous and gelatinous Bradyts were also employed.

Polish safety explosives for coal mining included Bradyt F. Its composition
was developed by T. Urbanski [87] :

77.5 % of ammonium nitrate 4% of nitroglycerine

4% of potassium perchlorate 9% of sodium chloride

4% of TNT 1.5% of woodmeal

A feature of Bradyt F was the introduction of potassium perchlorate which
increased the safety of the explosive by virtue of the potassium chloride formed
on explosive decomposition. This proved effective in the experimental gallery.
However, after being in use for many years, explosives of this type were withdrawn
because of their comparatively high power (lead block expansion ca. 280 cm3)
which was considered much too high for coal working, when more stringent regula-
tions were introduced.

There are now four groups of mining explosives in use:

A. Rock explosives, which are divided into 3 subgroups according to their
chemical composition :

(1) Ammonium nitrate explosives (ammonites);

(2) Nitroglycerine explosives (dynamites);

(3) Blackpowder.

They may be used only for rock, where there is no coal and no risk of gas. They
may also be used, exceptionally, in mines where gas can be present, provided that
the content of methane is less than 0.1% at faces situated 50 m away in slightly
dusty areas, and 150 m away in very dusty ones.

B. Coal explosives. They may be used only in non gassy coal seams.
Their use is limited to strictly determined areas where dangerous coal-dust occurs.
They may also be used instead of explosives of group A.

476

chemistry and technology of explosives

Table 137
A. Polish rock explosives (Amonits)

Explosives

Ingredients

Amonit

skalny

Ammonium nitrate

84.5

83.5

80.9

Nitroglycerine

DNT

—

TNT

11.8

7.5

5.5

Woodmeal

—

1.5

—

Ferric oxide

—

0.2

—

0.1

—

Aluminium

7.5

Lead block expansion (cm 3 )

330

—

400

Rate of detonation (m/sec)

3300-3700

Transmission of detonation (cm)

Table 138
A. Polish rock explosives (Dynamits)

Explosives

Dyna-

Dynamit

Ingredients

Dyna-

mit

Dynamit

skalny

Dynamit

mit

skalny

3GH

skalny

skalny 1

2GI

waterproof

5G1

5G2

Nitroglycerine

45.75

17.1

17.2

Nitroglycol

—

15.25

4.9

15.4

—

4.8

Nitrocellulose

4.5

0.8

1.3

DNT

—

TNT

—

Ammonium

nitrate

—

68.5

—

Sodium nitrate

—

14.8

15.6

Potassium nitrate

—

19.3

—

Woodmeal

5.7

Ferric oxide

—

0.1

Glycerine

—

0.4

—

Glycol

0.3

Density

—

1.45

1.37

1.45

Lead block

expansion (cm 3 )

—

390

330

440

320-350

Rate of detona-

tion (m/sec)

—

5000

6100

2400

2500

Transmission of

detonation

(cm)

MINING EXPLOSIVES

477

C. Permitted explosives, which are divided into 2 subgroups according to their
chemical composition:

(1) Ammonium nitrate explosives;

(2) Nitroglycerine explosives (permitted dynamites).

Both may be used in coal mining where the content of methane in the air is
less than 1%. In more dangerous areas only ammonium nitrate explosives are
permitted. Both may also be used wherever A and B explosives are permitted.

D. Permitted special explosives. They may be used in coal mines where the
content of methane in the air is less than 1.5% and in mines where particularly
sensitive coal-dust occurs.

The explosives used in Poland should not be prone to deflagration (this is dis-
cussed on p. 417).

Those permitted for use in gassy and dusty mines should meet the requirements
outlined on p. 439.

Table 139
B. Polish coal explosives (Karbonits)

Explosives

Ingredients

Karbonit

we.glowy

w^glowy

we.glowy

w^glowy

D29

D49

Ammonium nitrate

68.5

Nitroglycerine

Nitroglycol

—

DNT

2.5

1.5

TNT

6.5

5.5

Woodmeal

3.5

Sodium chloride

10.5

Density

1.03

—

Lead block expansion

(cm3)

255

—

270-285

270

Rate of detonation

(m/sec)

2600

1900-2200

Transmission of

detonation (cm)

4-5

Since these explosives may be used safely only in mines where little or no gas
occurs, Cybulski has recently formulated explosives which may be used in mines
where much gas is present and where ordinary permitted explosives are not generally
approved. Two such explosives are used: permitted special Metanit A and per-
mitted special Metanit B. They are ammonium nitrate explosives with a very high
content of cooling salt (e.g. up to 50% sodium chloride). Permitted special Metanits
are very safe in the presence of firedamp and coal-dust. They stand up to the tests
described above and do not ignite firedamp even with a 1.5-2 kg charge suspended

478

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

z
<

H
W

•*

1-1

-1

-J
o

SI*

2 §

^ « s
2 !

113

t? -2 CN

2 !°

S «) A

s I

-o
s

»n >o

^- m n

to <n

m O ^h t^- rn

I _

I 1

I I

I
00

BO S I

1 I

11:1

I I

I ! I

m m o ~h ro m in

<2 ■* I —

I I

III

I I

I ! i I

1-1 o

3 ™

o 8 v

BO S I

C
a!

o
3

o
■2

•^

rcf

</i

mining explosives

Table 141
C. Polish permitted explosives ( dynamite-type Barbaryts)

479

Explosives

Ingredients

Barbaryt

powietrzny

Barbaryt
powietrzny
| AG1

Barbaryt

powietrzny

AG2

Nitroglycerine
Nitroglycol

1 2 i

! 17.2
1 4.8

14
8

Nitrocellulose

DNT

TNT

0.5
3

0.5

! 2 -5

0.8

Ammonium nitrate
Sodium nitrate

32.5
4

! 30.5
4

21
6

Sodium chloride

35.5

Talc

Glycerine
Glycol

Saturated aqueous solution of
ammonium nitrate

1
0.5

1.5
1

- 1
1

1.2

These ingredients are
used to "soften" the
mixtures and to
facilitate mixing and
cartridging

Density

Lead block expansion (cm 3 )
Rate of detonation (m/sec)
Transmission of detonation

1.45
190
2300

1.45 :

190
2200

1.45
160
2200

(cm)

4-10

3-8

Table 142
P. Polish permitted special explosives (special Metanits)

Explosives

Ingredients

Metanit po-
wietrzny
specjalny B

Metanit po-
wietrzny
specjalny C

Ammonium nitrate
Nitroglycerine
TNT

Woodmeal
Sodium chloride

42.5

3.5

3
45

47.5

3.5

3
40

Density

Lead block expansion (cm 3 )

Rate of detonation (m/sec)
Transmission of detonation (cm)
Charge limit (g)

1.0

130-170

(average 140)

1600

4-6

1000

in an explosive mixture of firedamp. Permitted special Metanits correspond to the
British Eq. S. explosives.

480

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In Poland, as in all parts of the world, blasting gelatine consisting of 92-94%
nitroglycerine and 8-6% dynamite collodion cotton is used for work in hard rock.

U.S.A.

The rock explosives used in the U.S.A. are similar in composition to those used
in Europe. In addition to blackpowder (with Chilian saltpetre) dynamites with the
composition tabulated below are used extensively. The oldest are the "straight
dynamites" designated by percentage expressing the content of nitroglycerine. The
latter is not gelatinized but only adsorbed by woodmeal and the like.

Table 143

u.s.a. straight dynamites

(% expresses the relative power of the explosive)

j 20%

30%

40%

50%

60%

Nitroglycerine

20.2

29.0

39.0

49.0

56.8

Sodium nitrate

59.3

53.3

45.5

34.4

22.6

Sulphur

2.9

2.0

Carbonaceous combustible

material (e.g. woodmeal, starch)

15.4

13.7

13.8

14.6

18.2

Antacids (calcium carbonate,

zinc oxide etc.)

1.3

1.0

0.8

1.1

1.2

Moisture

0.9

1.0

0.9

1.2

A mixture with nitroglycol is now generally used instead of pure nitroglycerine.
Gelatine dynamites are similar in composition. They contain nitroglycerine
(and nitroglycol) and are gelatinized with nitrocellulose.

Table 144
u.s.a. gelatine dynamites (% expresses the relative power of the explosive)

20%

30%

40%

50%

60%

80%

100%*

Nitroglycerine (+ nitroglycol)

20.2

25.4

32.0

40.1

49.6

65.4

91.0

Nitrocellulose

0.4

0.5

0.7

0.8

1.2

2.6

7.9

Sodium nitrate

60.3

56.4

51.8

45.6

38.9

19.5

Sulphur

8.2

6.1

2.2

1.3

Carbonaceous combustible

material

8.5

9.4

11.2

10.0

8.3

10.1

Antacids

1.5

1.2

1.1

1.7

0.8

Moisture

0.9

1.0

0.9

1.0

0.9

0.7

0.1

* 100% gelatine dynamite is usually called "blasting gelatine".

Since the high price of nitroglycerine makes the use of straight or gelatine dy-
namites rather costly in relation to their power, it may be partly substituted by

ii i ■ uiyipywpif*.,™. ■*"" -

MINING EXPLOSIVES

481

ammonium nitrate, giving "ammonium dynamites" and "ammonium gelatines"
(Tables 145 and 146. The percentage expresses the relative power of the explosive.)

Table 145

U.S.A

AMMONIUM DYNAMITES

20%

30%

40%

50%

60%

Nitroglycerine

12.0

12.6

16.5

16.7

22.5

Ammonium nitrate

11.8

25.1

31.4

43.1

50.3

Sodium nitrate

57.3

46.2

37.5

25.1

15.2

Sulphur

6.7

5.4

3.6

3.4

1.6

Carbonaceous combustibles

10.2

8.8

9.2

10.0

Antacids

1.2

1.1

0.8

1.1

Moisture

0.8

0.7

0.9

0.7

Table 146
u.s.a. ammonium gelatines

30%

40%

50%

60%

80%

Nitroglycerine

22.9

26.2

29.9

35.5

38.3

Nitrocellulose

0.3

0.4

0.7

0.9

Ammonium nitrate

4.2

8.0

13.0

20.1

34.7

Sodium nitrate

54.9

49.6

43.0

33.5

19.1

Sulphur

7.2

5.6

3.4

Carbonaceous combustibles

8.3

8.0

7.9

4.3

Antacids

" 0.7

0.8

0.7

0.8

0.9

Moisture

1.5

1.4

1.6

1.7

1.8

It is characteristic of many of these explosives that most of them contain sulphur,
unlike European explosives.

As in Europe, ammonium nitrate explosives consist of mixtures of ammonium
nitrate with various ingredients :

(1) explosive, such as nitroglycerine or nitrostarch,

(2) combustible, but non-explosive.

In the first group there are mixtures with the following composition:

Ammonium nitrate 40-80%
Nitroglycerine or nitrostarch 3-10%
(less frequently aromatic nitro compounds are employed)

made up with woodmeal and antacid ingredients.
The second group is represented by a mixture of:

Ammonium nitrate 92.5%
DNT 4.0%

Paraffin wax 3.5%

482

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Its rate of detonation varies with the diameter of the cartridge from 3500 to 5000
m/sec. The gap test gives a value of 12-45 cm.

In the U.S.A. in 1956 a new class of rock explosives was tried for the use in
opencast mines, based on a patent taken out by Lee and Akre [88]. They are low
cost explosives composed of ammonium nitrate and an inexpensive fuel such as
carbon black. Rapid development took place in 1957-58 and their use was extended
to certain underground operations. Fuel oil was next used as a combustible ingre-
dient (AN-FO) explosives.

AN-FO explosives are usually made in the mine ("do-it-yourself" explosives)
thus saving the cost of explosives transport.

Usually prilled ammonium nitrate is used as it facilitates detonation of the
explosive and the porosity of the granules help to retain the liquid ingredient.

1000

5 6
Mineral oil. %

+20 +1S

+12

-12

-16

8+4 0-4

Oxygen balance, %

Fig. 156. Explosive properties of ammonium nitrate-mineral oil mixtures as a function
of mineral oil content [71].

The ratio of ammonium nitrate to fuel oil giving a zero oxygen balance is ca.
95/5. The highest rate of detonation (ca. 3300 m/sec using a 100 mm dia. cartridge
in a steel tube) is reached when the fuel oil content is 5-6% [89]. The rate of detonation
can be raised to 6000 m/sec by detonating a PETN fuse all along the cartridge.
Very often initiation is strengthened by adding a cartridge of a readily detonating
explosive, e.g. 60% dynamite. The most easily initiated mixture contains 2% of
fuel oil. With 10% of fuel oil the ease of detonation is considerably impaired.

Wetterholm [71] gives a summary of explosive properties of ammonium nitrate-
mineral oil mixtures as a function of the oil content (Fig. 156). Another diagram

MINING EXPLOSIVES

483

by the same author (Fig. 157) gives a minimum charge diameter for stable deto-
nation of ammonium nitrate-Diesel oil mixtures with 5% water or without water,
against the composition of the mixtures (Diesel oil content).

It is advisable to use ammonium nitrate with as little as possible of the inert
material usually added to ammonium nitrate to prevent its caking. According to
the same source, the quantity of inert material should not exceed 0.4% of NH 4 N0 3
(fertilizer grade ammonium nitrate usually contains 2-5% added matter).

The diameter of the cartridges used for opencast mining is usually 100 mm,
which evidently facilitates detonation.

w.
4J0-

3.5

3.0-

a
■5 2.0

1.0
0.5
J

3 4 5

Diesel-oil, %

Fig. 157. Minimum charge diameter for stable detonation in iron tubes (wall thickness

3 mm). Initiation by No. 8 detonator and 27 g compressed TNT ; 5% water : f — detonation,

• — miss; 0% water: H detonation, o — miss [71].

As mentioned before, preparation of AN-FO explosives is usually carried out
where they are to be used. The methods are described later in the chapter on the
manufacture of mining explosives.

AN-FO explosives are approximately twice as cheap as the usual ammonium
nitrate rock explosives and three times cheaper than dynamites.

They have become very popular not only in the U.S.A. but also in Canada,
U.S.S.R., Japan and Sweden.

More information is to be found in papers presented at an International Sym-
posium in Rolla, Missouri [90].

A modified type of AN-FO explosive has been suggested in the U.S.A., made
by mixing ammonium nitrate with nitromethane (ca. 5%)which is itself explosive.
They have a higher rate of detonation than AN-FO (e.g. 3900 m/sec). However,
they have not been widely used because of their higher price.

484 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The other types of cheap ammonium nitrate explosives to made on the spot by
mixing ammonium nitrate with cheap solid fuel such as carbon black, peat, brown
coal etc. were also suggested in the U.S.A. They are much less popular than AN-FO.

Safety explosives

Safety exsplosive are known in the U.S.A. under the name of Permissible Explo-
sives or simply permissibles. As in European countries their chief ingredient is
ammonium nitrate. Permissibles may or may not contain nitroglycerine.

According to Taylor and Rinkenbach [91] permissible Monobel contains:

Ammonium nitrate 80%

Nitroglycerine 10%

Combustible and other material 10%

No detailed information is available about the compositions of permissibles,
which are kept secret by the producers.

An ammonium nitrate explosive in widespread use is Nitramon (Kirst, Wood-
bury and McCoy [92]) containing ammonium nitrate with DNT. It is not clas-
sified as "permissible".

U.S.S.R.

Mining explosives are classified in the U.S.S.R. into the following groups:
(I) Approved for opencast works;
(II) Approved for opencast and underground works in non-gassy and non-dusty
mines;

(III) Approved for rock work in gassy and dusty mines;

(IV) Approved for rock and coal work in gassy and dusty mines ;
(V) Approved for sulphur mines.

Groups (I) and (II) comprise non-permitted explosives, groups (III)-(V)— permit-
ted explosives.

The permitted explosives are tested in experimental galleries in the presence
of methane and coal-dust (explosives of groups (III) and (IV)), or in the presence
of sulphur dust (explosives of group (V)). The explosives intended for use in oil
fields are tested for safety towards petroleum vapours.

The non-permitted explosives and those permitted for works in sulphur and
oil fields have cartridges 31-32 mm dia.

The explosives of group (IV) have cartridges 36-37 mm dia. The compressed
Ammonits have a cartridge diameter not less than 36 mm.

The composition and properties of some of these explosives are given in Tables
147-153.

Explosives for underground work are not usually allowed to contain more than
0.5% moisture. The requirements for explosives for opencast work are not so strin-
gent and a moisture content of up to 1.5% is approved.

MINING EXPLOSIVES

485

Table 147

u.s.s.r. non-permitted explosives
Rock explosives for opencast work (Group I)

Ingredients

Explosives

Ammonit No. 9

Ammonit No. 10

Ammonium nitrate
TNT

Woodmeal or powdered peat, or powdered
cottonseed cake or asphaltite

87
5

85
8

Density

Lead block expansion (cm 3 )

Transmission of detonation (32 mm dia.) cm

Rate of detonation (m/sec)

Heat of detonation (kcal/kg)

0.8-0.9

300-330

2-3

3000-3500

857

0.85-0.95

300-330

2-4

3200-3600

905

TNT in various forms (flakes, pressed or cast charges) is also approved for
opencast mining.

Table 148

U.S.S.R. rock explosives for opencast work

Ingredients

Ammonium nitrate
TNT
DNT
Powdered bark

Explosives

Ammonit
No. 6

Ammonit
No. 7

Dinaftalit
No. 1

79
21

Density

Lead block expansion
(cm 3 )

Transmission of detona-
tion when dry (cm)
when moist (cm)

Rate of detonation (m/sec)

Heat of detonation
(kcal/kg)

Form

Powder

Corned

Pressed

1.0-11

5-10

3-5

3«XM2O0

1028

1.0-1.15

360-380

4-8

3600-3800

1028

1.25-1.35

5-10

4600-5500

1028

1.0-1.1

350-370

4-6

2-3

3600-3900

962

1.0-1.1

320-360

3-6

2-5

3500-4500

950-976

Dinitronaftalit No. 1 Zh V and No. 1 V and Ammonits No. 6 Zh V and No. 7
Zh V comprising ammonium nitrate partly waterproofed are also used. All these
explosives are required to transmit detonation after storage under water at a depth
of 1 m for 1 hr. The transmission figures when moist are given above.

Rock Ammonits No. 1, No. 1 Zh V (water-proof) and No. 2 are also used.

486

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

They are more powerful than the Ammonits described above and are manufactured
in the form of compressed cartridges 36 and 45 mm dia. (250 and 400 g of weight,
respectively) or powdered material 60-210 mm dia. The compressed cartridges
are supplied together with the initiators fitted with detonator pockets.
The properties of rock Ammonits are given in Table 149.

Table 149
U.S.S.R. rock Ammonits

Ammonits

Properties

Rock No. 1

Rock No. 1 ZhV

Rock No. 2

Powder

Pressed

Powder

Pressed

Density

0.95-1.1

1.45-1.5

0.95-1.1

1.45-1.51

1.5-1.6

Moisture (max)

0.2

Oxygen balance (%)

-2.2

-0.7

+0.13

Lead block expansion

(cm 3 )

450-480

420-440

Transmission of detona-

tion when dry (cm)

6-10

4-8

7-12

5-9

7-10

after being kept

in water (1 m depth)

for 1 hr (cm)

4-6

3-8

5-8

5-9

7-10

Rate of detonation

(m/sec)

4000-5000

6000-6500

4000-5000

6000-6500

6500-7000

Heat of detonation

(kcal/kg)

1270

1290

1166

Another group of very powerful non-permitted explosives are the Ammonals :
VA-2, VA-4, VA-8. All are water-proof. They are characterized by the presence of
aluminium powder. Their properties are similar to those of rock Ammonits.

The non-permitted explosives include low-freezing 62% dynamite composed of:

Nitroglycerine 3 7 %

Nitrodiglycol (DGDN) 25%
Nitrocellulose 3.5%

Sodium or potassium nitrate 32%
Woodmeal 2.5%

It is manufactured in cartridges with diameters of 31-32 mm (200-250 g) and
44-45 mm (500-550 g).
Its properties are:

Density 1.40-1.45

Lead block expansion 380-420 cm 3
Transmission of detonation when dry 8-10 cm
and after being kept

in water (the same)

Rate of detonation 6000-7000 m/sec

Heat of detonation 1200 kcal/kg

MINING EXPLOSIVES

487

Low density Ammonit No. 14 is a relatively new explosive — it was first tested
in 1954. It contain;, low density ammonium nitrate and 7% nitroglycerine. Its pro-
perties are :

Density 0.75-0.80

Oxygen balance —0.2%

Lead block expansion 320-350 cm 3

Transmission of detonation 4-8 cm

Rate of detonation 1800-2500 m/sec

Heat of detonation 900 kcal/kg

Cheap ammonium nitrate-fuel oil explosives (p. 482) are also in use in the
U.S.S.R. under the name of Igdanit.

Table 150

u.s.s.r. permitted explosives

Permitted Ammonits

Ingredients

Ammonium nitrate
TNT

Woodmeal
Powdered bark
Sodium chloride
Potassium chloride

Density

Lead block expansion (cm 3 )
Transmission of detonation (cm)
Rate of detonation (m/sec)
Heat of detonation (kcal/kg)

Explosives

Ammonit No. 8 Ammonit AP-1 | Ammonit AP-2

1.00-1.15
240-280

3-5
2500-3000

690

65
14

1.5
19.5

1.00-1.15
260-290

4-7

3000-3500

800

68.5
15
1.5

1.00-1.15
285-310

4-8
3200-3700

855

A water-proof modification of a permitted Ammonit is water-proof Ammonit PZh
V-20. It is more easily detonated than Ammonit No. 8. The other water-proof Am-
monits are: AP-4 ZhV and AP-5 ZhV. They are used in damp conditions.

The permitted explosives Pobedits (Pobeda= Victory) are widely used: rock and
coal Pobedit PU-2; water-proof Pobedit VP-1; Pobedit No. 6; water-proof
Pobedit VP-2.

All are of low nitroglycerine content (5-9%).

Their properties are given in Table 151.

Pobedit P-8 is an ammonium nitrate sheathed explosive with 7% nitroglycerine.
The sheath is composed of potassium chloride powder.

The properties of this explosive are:

Density

Lead block expansion

Rate of detonation

1.0-1.15
285-310 cm 3
3500-4000 m/sec

488

chemistry and technology of explosives

Table 151
pobedits

Properties

Pobedit

PU-2

VP-1

No. 6

VP-2

Density

1.00-1.10

1.15-1.30

1.00-1.10

1.15-1.30

Oxygen balance (%)

-0.53

+ 0.51

-0.14

+ 0.3

Lead block expansion (cm 3 )

250-280

265-290

285-310

320-340

Transmission of detonation

when dry (cm)

5-10

6-20

5-15

6-20

after storage in water (cm)

5-15

5-14

Rate of detonation (m/sec)

2800-3500

3200-3700

3500-4000

3800-4300

Heat of detonation (kcal/kg)

800

813

870

910

The explosive part of the cartridge has 28 + 0.5 mm dia., and a weight of 132 + 7 g.
With the sheath its diameter is 36 ± 1 mm and weight 225 ± 15 g.

The transmission of the detonation figure for the complete cartridge is 4-7 cm.

Pobedit P-8 is tested not only in a steel mortar in a gallery, but also suspended
free in methane atmosphere and in the presence of coal-dust.

Ammonit No. 15 is a low density permitted explosive, containing 8% nitrogly-
cerine. Its properties are:

Density

Oxygen balance
Lead block expansion
Transmission of detonation
Rate of detonation
Heat of detonation

0.7-0.8
-0.03%
240-260 cm 3

3-6 cm
1700-2300 m/sec
800 kcal/kg

It is supplied in two dimensions:

31-32 mm dia., weight 150 g
and 36-37 mm dia., weight 200 g

Table 152
u.s.s.r. sulphur ammonits

Properties

Sulphur
Ammonit No. 1

Sulphur
Ammonit No. 2

Density

Oxygen balance (%)

Lead block expansion (cm 3 )

Transmission of detonation (cm)

Rate of detonation (m/sec)

Heat of detonation (kcal/kg)

0.95-1.05

-0.8

200-220

5-8

2500-3000

720

0.95-1.05

-0.5

150-170

4-7

2000-2500

600

MINING EXPLOSIVES 489

Permitted explosives for various purposes

In sulphur mines. Sulphur Ammonits No. 1 and No. 2 are used. They contain
5% nitroglycerine. They are recommended when the sulphur content of an ore is
lower and higher than 20%, respectively.

Their properties are given in Table 152.

In oil fields. Petroleum Ammonits are ammonium nitrate explosives with a low
(4-9%) nitroglycerine content. Their properties are given in Table 153.

Table 153
u.s.s.r. petroleum ammonits

Properties

Neftyanoy
Ammonit No. 1

Neftyanoy
Ammonit No. 2

Water-proof

neftyanoy

Ammonit No. 3

Density

Oxygen balance (%)

Lead block expansion (cm 3)

Transmission of detonation when dry

(cm)
after storage in water at a depth of 40 cm

for 1 hr (cm)
Rate of detonation (m/sec)
Heat of detonation (kcal/kg)

0.95-1.05
-0.2
220-230

5-8

2000-2500
634

0.95-1.05
+0.1
230-250

5-10

2500-3200
690

1.1-1.3

-0.3

220-240

3-7

2-5

2500-3200

700

Petroleum ammonits No. 1 and No. 2 are recommended where there is a danger
of petroleum vapours or of petroleum plus methane.

COMBINED BLASTING AND WATER INFUSION
FOR COAL BREAKING

As a further stage in the effort to reduce the risk involved in the use of explo-
sives in mines, an old idea was recently reintroduced. It was decided that safety in
use may be increased by surrounding the charge with a layer of water (e.g. in a pa-
per container, circular in cross-section). As long ago as 1876 this method suggested
by MacNab [24] was recognized as effective for diminishing the danger of gas ignitions.

Water infusion applied at high pressure in holes drilled in the coal face was
used to mine coal in some Westphalian collieries in 1914, and in some mines in
South Wales in 1942.

Besides being safe in themselves, these methods considerably reduce coal-dust
concentrations.

Demelenne [93] experimented in Belgium with small charges of explosives,
using the gas pressure that developed as the motive force to obtain the desired
infusion. A considerable reduction of coal-dust (by ca. 40%) was claimed.

490

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The new technique of pulsed infusion shot-firing was then introduced in Great
Britain. It required the development of highly water-proof explosives. Polar Ajax
was originally suggested, but it was only capable of withstanding a water pressure
of ca. 20 lb/in 2 for a few hours, whereas ability to withstand up to 800 lb/in 2 water
pressure was required. Earlier work had shown that barium sulphate increases the
sensitiveness of nitroglycerine gelatine explosives under high water pressure.
This property had been applied in geophysical investigation and existing explosives
for deep' seismic prospecting therefore provided a basis for the new development.

Charge and
detonator-

Water seal

Leading wires
Infusion tube

Fig. 158. Blasting coal, employing combined blasting-water infusion technique, ac-
cording to J. Taylor and Gay [4].

Eventually a new explosive Hydrobel was developed by Haslam, Davidson
and Hancock [94] (Table 154). Submarine-type electrical detonators for short holes
(Fig. 158) and Cordtex detonating fuse for long (up to 150 ft) holes are used to

Table 154
Hydrobel

Nitroglycerine and nitroglycol

40.9-37.9

Nitrocellulose

3.0-1.0

Ammonium nitrate

21.0-19.0

Sodium chloride

28.1-26.1

Barytes

10.7-8.7

Chalk

1.0-0.1

Diammonium phosphate

0.8-0.1

Woodmeal

1.6-0.6

Acid Magenta

0.05-0.001

A. S. No. 2 (optional)

0.01-0.001

Volatile matter

2.0-0.0

Density

1.7

Power (in % of blasting gelatine)

Rate of detonation of unconfined ex-

plosive in 1 Vi in. diameter cartridges

(m/sec)

6000

MINING EXPLOSIVES 491

detonate the charge. A Cordtex detonating fuse consists of a high-explosive core
surrounded by textile layers made water-proof by an outer plastic. To increase
safety, the fuse is coated with a balata composition, in which one sixth of the
weight consists of finely ground cryolite.

The new technique of mining coal is based on the introduction of water at a pres-
sure ranging from 100 to 400 lb per sq. in. The water is forced by the infusion pump
into the shothole which is loaded with Hydrobel explosive and sealed with stemming.

This technique of coal-getting by combined blasting and water infusion elimi-
nates the need for undercutting. It is known that undercutting is largely responsible
fcr the formation of coal-dust. The hazard is considerably reduced by wetting the
coal-dust before it is suspended in the air. In addition the toxicity of the fumes
is reduced as some of the toxic constituents, such as N 2 4 , are soluble in water.

LIQUID OXYGEN EXPLOSIVES (OXYLIQUITS)

Oxyliquits are explosives consisting of combustible materials impregnated with
liquid oxygen. They were invented by Linde [95] and were originally prepared by
impregnating such substances as sawdust, carbon black etc. with liquid air. These
mixtures, however, were hard to detonate since, as appeared later, they were incapa-
ble of adsorbing a sufficient amount of the liquid and so did not contain enough
oxygen. Substances with highly adsorptive properties then began to be added, for
instance activated coal, or incombustible adsorbents such as kieselguhr.

Fig. 159. Section of the oxyliquit charge.

The disadvantage of employing liquid air is that it usually contains less than
33% oxygen which is not always enough for complete combustion. Moreover the
composition of liquid air undergoes continuous alteration on storage. Liquid air
has therefore been superseded by 98% liquid oxygen.

A charge of oxyliquit is shown in longitudinal section in Fig. 159. Adsorbent
combustible mixture (/) is enclosed in a cotton or paper bag (2) 30-50 mm dia.
(the use of smaller diameters is not recommended since the charge looses rapidly
oxygen by evaporation). The bag with the combustible mixture is placed within a do-
uble-walled envelope (5) made of corrugated and ordinary cardboard which stiffens
the cartridge and constitutes a thermal insulator. The outer diameter of the cartridge
plus envelope is 35-60 mm, its length— 300 mm. In ignition cartridges the inner

492

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

bag includes a detonator with electric fuse (4). Safety fuses cannot be fired in coal
mines for fear of premature explosion which may result from the sparking the
safety fuse or ignition of the molten pitch insulating the braid of the safety fuse.

Before firing the prepared cartridges are dipped in a Dewar vessel containing
liquid oxygen. The cartridge absorbs the oxygen and gradually sink in it, after
which the explosive charge is ready for use (impregnation lasts for about 20 min).
The prepared cartridges are placed into the shothole (one of them should have
a detonator with an electric fuse) and further procedure is then routine. The shot
must not be fired later than 10-15 min (depending on the composition of the cart-
ridge) after the cartridges have been taken out of oxygen.

The efficiency of oxyliquits depends on:

(1) the composition of the absorbent (combustible);

(2) the quantity of oxygen adsorbed;

(3) the time from the moment of impregnation with oxygen to the moment
of shotfiring;

(4) the losses due to the evaporation of oxygen.

The composition of the combustible mass. The power of oxyliquits depends on
the composition of the absorbsnt. D oxyliquits are very strong explosives with
a performance resembling that of dynamite; A oxyliquits have a performance equal
to that of ammonium nitrate explosives; P oxyliquits have a relatively slow action
similar to blackpowder.

Kast and Haid [96] report figures characteristic of the absorbent on the explo-
sive properties of some oxyliquits (Table 155).

Table 155
d oxyliquits

Properties

Density of prepared explosive
Heat of explosion (kcal/kg)
Gas volume (V , l./kg)
Calculated temperature of explosion

(°Q
Lead block expansion (cm 3 )
Rate of detonation in the open, m/sec)

Absorbent

Karben*

1.04

2180

615

5750

535

4760

Carbon
black

0.72

1995

535

6500

530

4680

Cork
dust

Woodmeal

Peat

0.63

1660

700

4195

510

3300

0.82

1535

700

4095

450

3610

0.53

1670

700

4385
485

3275

* Alias "Kupren" - absorbent obtained by the polymerization of acetylene on a copper catalyst at temperatures
200-280°C.

The explosives enumerated in Table 155 are very powerful type D oxyliquits.
Their power may be reduced by the addition of inert cooling substances, e.g. sodium
chloride, kieselguhr etc. An explosive is then obtained with a performance similar to

MINING EXPLOSIVES

493

that of coal explosives, e.g.

Petrol

Carbon black
Kieselguhr

12%
63%
25%

This oxyliquit has a rate of detonation of 3430 m/sec.

If the power of oxyliquits is to be increased to obtain rock explosives, this may
be achieved by the addition of aluminium.

A disadvantage of mixtures of liquid oxygen with combustible substances lies
in their high sensitiveness to impact and friction. This property depends largely
on the composition of the combustible material used. Oxyliquits containing sawdust
show the lowest sensitiveness, while those containing hydrocarbons (e.g. naphtha-
lene) pitch or petrol are more sensitive.

Explosives with liquid oxygen were used in Poland in some Upper Silesian
mines during the inter-war period. To increase their safety, cartridges of Badowski's
invention composed of 35% woodmeal and 65% sodium chloride were used. A cylin-
drical piece of ice was also introduced into the bottom of the shothole. A similar
extinguisher was inserted at the mouth of the shothole. This did in fact give greater
safety against coal-dust.

The amount of oxygen absorbed. The effect of the oxygen content absorbed
by the combustible mass in the liquid phase on the explosive power is represented
in Table 156.

Table 156

Oxygen content in liquid phase

Lead block expansion
cm 3

no explosion

147

384

The amount of oxygen adsorbed by various combustible substances and, for
the sake of comparison, that required for the combustion of the substance is shown
in Table 157.

Table 157

Amount of oxygen

Combustible

adsorbed

required for com-

substance

by 1 g of the

bustion of 1 g to

substance, g

C0 2 +H 2 0, g

Sawdust

2.4

1.37

Cotton

3.0

1.18

Carbon black

2.3

2.7

Charcoal

2.67

Kieselguhr

3.0

494 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Losses due to the evaporation of oxygen. The time which elapses from the
moment of removing the cartridge from the Dewar vessel up to shotfiring affects
the explosive considerably, since it weakens with the loss of oxygen. In addition
a considerable amount of toxic carbon monoxide may be evolved on explosion.
The greatest possible efforts should be made, therefore, to minimize the evapora-
tion of oxygen. This begins immediately after removal of the cartridge from the
liquid. The rate of evaporation depends on the type of adsorbent used and the
insulation. The larger the specific surface of the adsorbent, the less evaporation
occurs; e.g. charcoal and carbon black, with a very large surface, give an explosive
which loses oxygen, and thus explosive power, ,very slowly. On the other hand,
sawdust loses oxygen much more quickly, since the oxygen is rather weakly adsorbed.
The rate of evaporation of oxygen depends also on the diameter of the cartridges
and the stemming.

In the U.S.S.R. an absorbent of milled cane is used since it has very good ad-
sorptive properties. The relationship between the rate of evaporation of oxygen from
the cane cartridges, their diameter, and type of stemming is shown in Table 158.

Table 158

Diameter of charge and mode
of stemming

Time of evaporation to
give combustion

to C0 2 to CO

32 mm in the open
180 mm in the open
180 mm in the shothole

10 min

42 min

4hr

18 min

70 min

7hr

Charges of oxyliquit, 33 mm dia., containing carbon black (density about 0.3)
give complete combustion to C0 2 within 5 min after removal from liquid oxygen
and have a power 15% higher than that of standard dynamite; in 25 min their
strength is 35% lower than that of dynamite.

A mixture of kieselguhr and petrol, in the ratio of 60 to 40, has 10% less power
than dynamite, with a rate of detonation of approximately 3000 m/sec; after 45 min
its power falls to 45% that of dynamite.

Loss of oxygen by evaporation does not particularly affect the rate of detonation
of an explosive. For charcoal oxyliquit, for instance (Kast and Haid [96]), the fol-
lowing figures have been found:

After

Rate of detonation

3 min

4930 m/sec

6 min

4670-4750 m/sec

10 min

4780 m/sec

The fall in the rate of detonation is insignificant.

Despite the drawback arising from the need for haste, the loss of power caused
by the evaporation of oxygen has a great advantage. In the event of a misfire, the

MINING EXPLOSIVES 495

unfired charge loses its explosive power in a few hours, so that a shothole may be
drilled again in the same place.

The other advantage of oxyliquits lies in their safety during transport: the ex-
plosives are manufactured on the spot, just before the use. One of their most serious
drawbacks, as mentioned before, lies in the fact that an explosive which has lost
too much oxygen, may give rise to a considerable amount of toxic carbon monoxide
on detonation.

SOME OTHER PEACEFUL APPLICATIONS OF EXPLOSIVES

Among various applications of explosives to engineering work such as tunnel
and road building, water front, harbour and river regulation etc., explosive working
of metals was recently added to the list of peaceful uses of explosives [97].

Another application is in geological investigation by seismographic methods,
much used in the search for underground sources of liquids, such as oil.

These applications, which form the subject matter of specialist books, will not
be discussed in the present work.

LITERATURE

1. A. Nobel, Brit. Pat. 1345 (1867); Swedish Pat. 102 (1867); U.S. Pat. 78317 (1868).

2. A. Nobel, Brit. Pat. 4179 (1875); U.S. Pat. 175735 (1876).

3. C. J. Ohlsson and J. H. Norrbin, Swedish Pat. of 31 May 1867.

4. J. Taylor and P. F. Gay, British Coal Mining Explosives, Newnes, London, 1958.

5. Mallard and H. Le Chatelier, Ann. mines [8], 4, 274 (1883); 14, 197 (1888); 16, 15 (1890) .
Mem. poudres 2, 355 ( 1884-1889).

6. Reglementation des explosifs a employer dans les mines..., Paris, 1890; Mem. poudres 5,39 (1892).

7. According to A. Breyre and E. Kedesdy, Z ges. Schiess- u. Sprengstoffw. 4, 1. 22 (1909).

8. SlERSCH, Osterr. Z. Berg- u. Huttenw. 44, 4 (1896).

9. C. E. Bichel and Mettengang, Z. Berg-, Hiitten- u. Salinenw. 50, 669 (1902).

10. Wilkoszewski, Z. ges. Schiess- u. Sprengstoffw. 2, 141 (1907).

11. W. Will, Z ges. Schiess- u. Sprengstoffw. 4, 323, 343 (1909).

12. J. Taffanel and H. Dautriche, VHIth Intern. Congress of Chemistry, New York; H. Dau-
triche, Mem. poudres 15, 164 (1909-1910); J. Taffanel, Compt. rend. 151, 873 (1910).

13. E. Lemaire, Ann. mines 23, 649 (1922).

14. W. Payman, Trans. Inst. Min. Eng., London 75, 191 (1928).

15. E. Audibert, Ann. mines 15, 213 (1929).

16. C. E. Bichel, New Methods of Testing Explosives, Griffin & Co., London, 1905.

17. G. S. Rice, Trans. Amer. Inst. Min. Met. Eng. 71, 130 (1925).

18. R. L. Galloway, A History of Coal Mining in Great Britain, p. 261, MacMillan & Co., London,
1882.

19. H.M. Commissioners, Accidents in Mines, Final Report, H.M.S.O., London, 1886.

20. Royal Commission on Explosions from Coal Dust in Mines, H.M.S.O., London, 1894.

21. F. A. Abel, Accidents in Mines, p. 61, Institution of Civil Engineers, London, 1888.

22. T. A. Rogers, Colliery Guardian 202, 26 (1961).

23. J. Fripiat, Vllth International Conference of Safety in Mines Research, Paper No. 33, Safety

496 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

in Mines R.B., Sheffield, 1952; VHIth International Conference of Safety in Mines Research,
Paper No. 44, Safety in Mines R. B., Sheffield, 1954.

24. W. MacNab, according to Daniel, Dictionnaire des explosifs, Paris, 1902.

25. W. Cybulski, The International Mining Congress, Dortmund, 1935.

26. V. Watteyne and E. Lemaire, Ann. mines Belgique 16, 937 (1911); 18, 781 (1913).

27. HlSCOCK, Colliery Guardian 128, 819 (1924).

28. T. Urbanski, Roczniki Chem. 6, 838 (1926); Z. ges. Schiess- u. Sprengstoffw. 11, 270 (1927).

29. E. Audibert, Ann. mines 6, 63 (1924).

30. W. Payman and H. C. Grimshaw, Safety in Mines Research Board, London 69 (1931).

31. Beyling (Gelsenkirchen), Safety in Mines Research Board, London 74 (1932).

32. E. Audibert, Comite Central, Note Technique 11, 23 (1926).

33. SEGAY, Ann. mines Belgique 30, 1347 (1929).

34. T. Urbanski, Roczniki Chem. 13, 130 (1933).

35. G. S. J. PERROTT, Ind. Eng. Chem. 19, 1293 (1927).

36. W. Payman, D. W. Woodhead and H. Titman, Proc. Roy. Soc. (London) A 148, 604 (1935).

37. W. Payman and D. W. Woodhead, Proc. Roy. Soc. (London) 163, 575 (1937).

38. L. V. Dubnov, Ugol (1949).

39. W. Cybulski, Archiwum Gornictwa No. 4 (1963); XI Conference Internationale des Directeurs
de Stations d'Essais, Aix-le-Bains, 1963.

40. H. Titman and D. G. Wilde, Safety in Mines Research Establ, Sheffield, 1953, Research
Report No. 73.

41. J. TAYLOR, Trans. Inst. Min. Eng., London 109, 358 (1949-1950).

42. T. Murata, J. Ind. Expl. Soc. Japan 15, 1 (1954).'

43. W. Payman and R. W. Wheeler, Iron Coal Trans. Rev. 136, 223 (1938).

44. J. Taylor and J. C. HANCOCK, Trans. Inst. Min. Eng., London 106, 678 (1946-1947).

45. F. Lebrun and L. WATERLOT, Explosifs 15, 85 (1962).

46. W. CYBULSKI, Komunikat GIG No. 159 (1954); No. 197 (1957).

47. W. Cybulski, Komunikat GIG Nos. 245, 255 (1960).

48. W. Cybulski, The Investigations into the Phenomenon of Deflagration of Polish Blasting Ammonium
Nitrate Explosives, Sosnowiec, 1937. (In Polish)

49. W. C F. Shepherd and H. C. Grimshaw, Safety in Mines Research Establ., Sheffield, 1954,
Research Report No. 8.

50. P. Beyersdorfer, Staub-Explosionen, Steinkopf, Dresden & Leipzig, 1925.

51. W. CYBULSKI, Komunikat GIG No. 57 (1949).

52. W. Cybulski, Komunikat GIG No. 1C9 (1952); No. 198 (1957); No. 227 (1959).

53. L. V. Dubnov, Predokhranitelnye vzryvchatyye veshchestva v Gornoy Promyshlennosti, p. 68,
Ugletekhizdat, Moskva-Leningrad, 1953.

54. J. Taylor and G. P. SlLLITTO, Trans. Inst. Min. Eng., London 109, 991 (1949-1950).

55. A. Kreyenbuhl and R. Sartorius Chimie et industrie, No. special 245 (1954).

56. L. Hackspill, A. P. Rollet and Lauffenburgier, Compt, rend. 198, 1231 (1934).

57. H. Ahrens, VHIth Intern. Conference of Directors of Safety in Mines Research, 1954, paper
No. 30.

58. Safety in Mines Research Establ., London, 1952, H.M.S.O. Research Report No. 65.

59. K. Hino and M. Yokogawa, Intern. Symposium on Mining Research University of Missouri,
Rolla, Missouri, 1961.

60. According to V. A. Assonov and B. D. Rossi, Spravochnik po burovzryvnym rabotam,
Ugletekhizdat, Moskva-Leningrad, 1949.

61. W. Cybulski, private information.

62. J. Taffanel, Z. ges. Schiess- u. Sprengstoffw. 5, 305 (1910).

63. M. Prettre, Mem. poudres 25, 531 (1932-1933); 26, 239 (1934-1935).

64. E. Lemaire, Ann. mines Belgique 19, 587 (1914).

MINING EXPLOSIVES 497

65. J. S. B. Fleming, Brit. Pat. 416586 (1934).

66. J. Taylor, Colliery Guardian 179, 329 (1949); Trans. Inst. Min. Eng., London 109, 2 (1949-l95<y,

67. J. Boucard and L. Deffet, Explosifs 17, 33 (1964). '

68. T. Urbanski, unpublished work (1925).

69. K. HlNO, Theory and Practice of Blasting, Nippon Kayaku Co. Ltd., 1959.

70. T. Urbanski, O/Schlesischer Berg- u. Huttenman. Vereins Z. 65, 217 (1926).

71. A. Wetterholm, Explosives for Rock Blasting, Atlas Copco AB, Stockholm and Sandwiken*
Jernwerks AB, Sandviken, 1959. '"wucens

72. W. J. Dixon and A. M. Mood, J. Amer. Statistical Assoc. 43, 109 (1948); W. J. Dixon and
F. J. Massey, Jr., Introduction to Statistical Analysis, McGraw-Hill, New York 1951

73. J. W. Gibson, H. C. Grimshaw and D. W. Woodhead, Safety 'in Mines Research Establ.
1952, Research Report No. 47; see also: N. E. Hanna, G. H. Damon and R. W. Van Dolah
Meeting of the Directors of Safety in Mines Research, CERCNAR, Verneuil, 1958; U.S Bureau
of Mines, Report of Investigation 5463 (1959).

74. R. L. Grant and R. W. van Dolah, University of Missouri School of Mines and Metal/ Bull
Technical Series, 97. • •»

75. e.g. R. L. Grant, L. M. Mason, G. H. Damon and R. W. Van Dolah, U.S. Bureau of
Mines, Report of Investigation 5486 (1959).

76. L'Association de Fabricants Beiges d'Explosifs et le Centre de Recherches Scientifiques et
Techniques pour lTndustrie des Produits Explosifs, Explosifs No. 4 (special) (1958).

77. A. A. Podgornova, in Novye vzryvchatyye veshchestva (Ed. L. V. Dubnov), 44/1 p 64
Gosgortekhizdat, Moskva, 1960. ' '

78. C. Schwanke, Ger. Pat. 497212 (1930).

79. Jahresber. Chem.-Techn. Reichsanstalt (1929).

80. T. Urbanski, Chemie und Technologie der Explosivstoffe, Bd. Ill, Verlag f. Grundstoffindustrie
Leipzig, 1964.

81. H. Ahrens, Vllth International Conference of Safety in Mines Research, Paper No. 27, Sheffield,

82. W. Cybulski, Badania detonacji materialow wybuchowych metodq kamery z wirujqcym zwier-
ciadlem, Katowice, 1948.

83. National Coal Board Specification No. P 112/1957, Permitted Explosives.

84. A. B. Wildgoose, Colliery Guardian ISO., 493 (1961).

85. S. Yamamoto, Xth Intern. Conference of Directors of Safety in Mines Research, Pittsburgh,
U.S.A., 1959; Additional Copies and Supplements, University of Tokyo, 1959.

86. M. Yokogawa, Nippon Kayaku Co. Ltd., private information (1959)'.

87. T. Urbanski, unpublished work (1925).

88. H. B. Lee and R. L. Akre, U.S. Pat. 2703528 (1955).

89. S. R. Brinkley and W. E. Gordon, Explosive properties of the ammonium nitrate-fuel oil
system, Hanna Coal Co., Cadiz Ohio, U.S.A.

90. International Symposium of Mining Research, University of Missouri, Rolla, Missouri, February
21-24, 1961.

91. C. A. Taylor and W. H. Rinkenbach, U.S. Bureau of Mines Bull. No. 219 (1923).

92. W. Kirst, C. A. Woodbury and McCoy, U.S. Pat. 1992217-7 (1935).

93. E. Demelenne, Ann. mines Belgique 52, 56 (1953).

94. R. Haslam, S. H. Davidson and J. C. Hancock, Trans. Inst. Min. Eng., London 114 87
(1954-1955).

95. C. Linde, Sitzungsber. Munch. Akad. Wissenschaft. 1899, 65.

96. H. Kast and A. Haid, Z. angew. Chem. 37, 973 (1924).

97. J. S. Rinehart and J. Pearson, Explosive Working of Metals, Pergamon Press, Oxford, 1963.

98. R. Westwater, Colliery Guardian, February 25, 1966.

CHAPTER VI

THE MANUFACTURE OF MINING EXPLOSIVES

THE MANUFACTURE OF AMMONIUM NITRATE EXPLOSIVES

Raw materials

THE properties and purity standards of ammonium nitrate — a chief ingredient
of these mixtures-have been discussed earlier (p. 463, Vol. II), and so has the
purity of potassium nitrate (p. 343). The other ingredients should meet the require-
ments of high purity demanded in the commercial products.

The quality of an explosive depends to a great extent o[ the uniformity of the
mixture, i.e. on the milling of the ingredients and on their mixing. The careful pre-
paration of ingredients is therefore an essential requirement. Mills of the disinte-
grator type are best suited for this purpose (Excelsior, Perplex etc.). Not infrequently
the ingredients are dried before milling. The drying of non-explosive ingredients,
such as all salts other than ammonium nitrate, is fairly straightforward. It is usually
carried out in ordinary shelf driers heated from below by warm air convection.
Since the salts become lumpy on drying, they are then screened, if necessary, being
crushed between the rollers before screening. To prevent hygroscopic salts (sodium
nitrate, sodium chloride) from absorbing moisture, a temperature of 25-30°C is
maintained in the premises.

More recently driers based on the fluid-bed principle were introduced. Figures 160
and 161 give diagrammatic presentations of sodium nitrate and woodmeal fluid-bed

driers.

Ammonium nitrate may also be dried in the same driers. Since, however, very
large quantities of this substance are processed in factories, the use of a high output
drier is necessary. The very low sensitiveness of ammonium nitrate to friction and
impact permits the use of steel machinery.

Various drum type driers are very popular. A typical example with a stationary
drum is shown in Fig. 162. In this arrangement the substance passes through the
mill to the elevating conveyer which carries it to the end of the drier. A screw con-
veyer inside the drum moves the substance through the drier. Hot air (60-80°C)
is sucked from above by a fan. The dried substance is poured into the screening

[498]

THE MANUFACTURE OF MINING EXPLOSIVES

5 4 3 2

499

\~\"'.-y>,'>?/i-

Qv-vav.-v ' .w/u-

7ZZ7Z227ZZZZZ^pZZZZZZZZZZZE:

izzzzzzzzzzzQ

Fig. 160. Sodium nitrate fluid-bed drier (Courtesy H. Orth G.m.b.H., Ludwigshafen-
Oggersheim, G.F.R.) ; / -filter tube, 2- mill, 3— lift, 4- feeder, 5- outlet cyclone, 5- fil-
ter tube, 7— end-mill, 8 — discharge funnel, 9 — ventilator, 10 — calorifuge, 11 — fluid-bed

drier.

drum. Lumps formed during drying are recycled to the mill and drier. Rotary driers
(Fig. 163) and shelf driers of the Schielde type are also in use (Vol. II).

More modern fluid bed driers are now much in use. A diagrammatic presenta-
tion of such a drier is given in Fig. 164 and the general view in Fig. 165.

Mixing of ingredients

There are various methods of mixing. They depend on the traditions of a given
factory or country and on the ingredients included in the mixture. Explosives con-
taining nitroglycerine are mixed differently from those without nitroglycerine.
Nitroglycerine explosives are usually mixed in two stages: first all the ingredients
with the exception of nitroglycerine are mixed together and then the nitroglycerine
is added in a different mixer.

500

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Mixing of ingredients without nitroglycerine. These may be mixed either cold
or hot. In the cold, the ingredients previously dried and milled either in iron drums
with wooden balls or in kneaders of the Werner-Pfleiderer type (Figs. 206, 207,
208) are mixed for about 1 hr. In France an edge runner is commonly used of a
construction similar to that for the manufacture of blackpowder (Fig. 94). In the
edge runner the explosive is not only mixed and milled but also crushed. It may
thus reach a higher density which is particularly advantageous when intended for

- ca. 7.60m

Fig. 161. Woodmeal fluid-bed drier, 200 kg/hr (Courtesy H. Orth G.m.b.H., Ludwigs-
hafen-Oggersheim, G.F.R.) ; / - pre-drier, 2 - final drier, 3 — blowing engine, 4, 9, 14 -
electric motors, 5— air filter, (5-air heater for the pre-drier, 7— air heater for the final
drier, 8-feeding and discharging valves, 10 and 77-cyclones, 72-air stream
breakers 13— air screw wheels, 15— air coolers.

military purposes (making Amatol 80/20 for shell filling). Mixing in an edge runner
makes it possible to obtain a higher rate of detonation which is of importance for
rock explosives and ammonium nitrate mixtures intended for military purposes.
Schneiderite, for instance, consisting of a mixture of ammonium nitrate and
dinitronaphthalene, is mixed in an edge runner in batches of 6Q kg. The mixing

THE MANUFACTURE OF MINING EXPLOSIVES

SOI

Fig. 162. Schematic view of a drier with a stationary drum.

lasts for about 40 ruin. Since it is necessary to protect ammonium nitrate against
the absorption of atmosphere moisture a temperature of approximately 30°C must
be maintained and the material must be dried to about 0.4% moisture content.
Two edge runners can manufacture 1500 kg of Schneiderite in 24 hr. If the explosive
is to be used for filling shells, and thus requires a considerable density, the charge
from the edge runner is sifted through a 5-mm mesh sieve; coarse grains are pressed

Ventilator

v^//^/-Xs>//xs/,.<.^,:<v//A^.;.$s,:x\>:-rA^::>K^

Fig. 163. Schematic view of a rotary drier.

502

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

a
o

B
E

Ml
BO

•a

PL,

THE MANUFACTURE OF MINING EXPLOSIVES

503

in a hydraulic press (60 kg/cm2) while fine grains, together with dust, are recycled
to the edge runner. The layer of explosive to be pressed must be such as to achieve
a thickness of 6.5 mm by the end of the operation.

Fig. 165. General view of a fluid-bed drier for ammonium nitrate (Courtesy H. Orth
G.m.b.H., Ludwigshafen-Oggersheim, G.F.R.)

The pressed cakes of explosive are broken up in a corning mill, the construction
of which is shown in Fig. 166. After passing through breaking rollers (/) and (2) the
material passes through a leather sleeve to sieves (3) (13 mm mesh) and (4) (6 mm

504

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

mesh). Material with grains of the correct size constitutes 50% of the whole. It is
collected in receiver (5). The dust (from container (5)) and coarse grains (from
vessel (7)) are recycled to the edge runner.

Schneiderite so obtained has a specific gravity of 1.5-1.6, number of grains per
kilogramme 1500-2000 and a rate of detonation approximately 5400 m/sec.

Fig. 166. Schematic view of a corning mill for ammonium nitrate explosive.

Fig. 167. Disk mixer.

Hot mixing may be conducted either in kneaders of the Werner-Pfleiderer type
with a heating jacket or in disk mixers (Fig. 167). These mixers consist of a round
plate of copper, brass or bronze, approximately 2 mm dia., with beaded edges,
which is heated with hot water to a temperature of 80-90°C. The material is mixed
with bronze blades and kneaded with a bronze roller.

THE MANUFACTURE OF MINING EXPLOSIVES

505

The operation in kneaders or disk mixers begins with the mixing of all the in-
gredients, except those which sensitize the mixture to impact and friction, e.g.
potassium or ammonium perchlorates. Molten TNT or another nitro compound
(DNT) is then poured in, the whole is carefully stirred and cooled while stirring.
Only when the temperature has fallen to 40-50°C, are the perchlorates added. When
the material is cooled down to a temperature of 25-30°C, it is stirred for a certain
time, then unloaded and transported in sheet metal or wooden boxes to the store-
room.

It usually takes a long time Q-4 hr) to cool a mixture, especially when operating
the kneaders, so that some factories have adopted a more economic method of
mixing by preparing separately, hot, a mixture of ammonium nitrate (two parts)
with TNT (one part), called "triamon".

A certain amount of this mixture is stored at temperatures of 25-30°C, to be
used cold as the need arises for the preparation of the finished product by one of
the methods described above.

Medard and Le Roux [1] examined the influence of various methods of mixing
on the properties of ammonium nitrate explosives ("Explosifs du type N"). They
found that mixing in heavy (5 ton) edge-runners gives explosives of higher sensi-
tiveness and rate of detonation than the same explosives mixed in kneaders of the
Werner-Pfleiderer type.

The difference can be seen from the Table 159 which refers to Explosive N
No. 31 (see Table 119).

Table 159

Mixed

in kneaders

in edge-runners

Density

0.95

1.00

Sensitiveness to initiation in

grammes of fulminate

at density 1.00

0.30

0.25

1.30

2.00

0.50

1.50

no detonation

0.80

Transmission of detonation (cm)

at density 1.00

6.5

1.30

0.5

Rate of detonation (m/sec)

at density 1.00

3350

3900

1.30

3690

4800

Mixing either hot, or hot and cold alternately, has the advantage over mixing
only in the cold that the crystals of ammonium nitrate are coated with a layer of
nitro compound due to contact with molten TNT or DNT, which somewhat re-
duces their hygroscopicity. Mixing in the hot requires careful temperature control
as the explosive leaves the kneader or mixer after being cooled. This temperature

506 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

should under no circumstances be higher than that of the change of crystal form of am-
monium nitrate, i.e. + 32.TC for pure substance. If the explosive has a temperature
higher than + 32°C after leaving the kneader, the change in crystal form may produce
caking in later stages of manufacture, i.e. in the stored boxes or (at worst) in the
paper cartridges.

Caking during storage is not of great consequence except that additional screening
is then needed. On the other hand, caking in the finished charges is much more
serious, since charges of hardened explosive are not suitable for use as their sensi-
tiveness to detonation is impaired and it is difficult to introduce detonators into
them.

Mixing of ingredients with nitroglycerine. Ammonium nitrate explosives con-
taining 4-6% nitroglycerine are mixed in two stages. First, all the ingredients other
than nitroglycerine are mixed by one of the methods described above and, secondly,
nitroglycerine is added at a temperature of 30-32°C and carefully stirred. Mixing can
also be carried out at 40°C and ended at 30°C. The temperature change is required
for the reason given above. Mixing with nitro glycerine is carried out either in
Werner-Pfleiderer (Fig. 206) or Drais type (Fig. 173) kneaders.

Cartridging

Free-flowing ammonium nitrate explosive is loaded into cartridge cases usually
32 mm dia., made of paraffin paper. The cartridge cases are handmade or manufac-
tured on Hesser machines (in which they are numbered and paraffined at once)
or on Niepmann machines (Fig. 179). The cartridges generally produced weigh
100 and 50 g. For loading, machines are usually employed in which the explosive
is poured into a funnel fitted with a rotating screw that pushes the explosive to-
wards the mouth of the funnel onto which the paper cartridge case is placed. There
another rotating screw feeds the explosive into the cartridge case. After filling the
open end of the cartridge case it is sealed (Fig. 168).

Paraffining and packaging

Formerly cartridge cases were made of non-paraffined paper and were paraffined
after filling by dipping into melted paraffin, lignite wax (montan wax) or a mixture
of both. A mixture of paraffin with rosin in the ratio of 70/30-90/10 was preferable
to paraffin alone. The amount of coating should not exceed 2.5 g per 100 g of explo-
sive. It was found, however, that paraffin may penetrate too deeply inside the cart-
ridge and desensitize the explosive. Moreover, hot paraffin may induce a change of
crystal form in the ammonium nitrate and cause the cartridges to harden. This
method was therefore abandoned and the cartridge cases are now paraffined before
being filled. The prepared cartridges are then placed into cardboard boxes, wrapped
with paper and paraffined, together with the whole box. The latter usually contain
2-2.5 kg of explosive.

THE MANUFACTURE OF MINING EXPLOSIVES

507

To help identify the explosives according to their safety, and to avert serious
errors, cartridges are packed in various colours which in Poland are:

rock explosives— red,

coal explosives— blue,

permitted explosives — cream-coloured,

special permitted explosives — cream-coloured

with the two black stripes and black bordering.

Fig. 168. Cartridging machine for free-pouring mining explosives.

In the U.S.S.R. :

explosives approved for opencast mines have white paper cartridges,
explosives approved for non-gassy and

non-dusty mines —red,

rock explosives approved for gassy and

dusty mines —blue,

coal and rock explosives approved for

gassy and dusty mines —yellow,

explosives approved for sulphur mines —green.

508 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

AMMONIUM NITRATE-FUEL OIL MIXTURES
(AN-FO)

Various methods have been used for mixing ammonium nitrate with fuel oil
on site at opencast mines. Some were very primitive and consisted of simply adding
fuel oil to a polyethylene bag of ammonium nitrate and allowing the mixture to
stand long enough for the ammonium nitrate to be soaked with oil. Then the mixture
was poured into vertical shotholes.

Another method consisted of charging the shothole with a definite quantity of
ammonium nitrate and pouring a measured quantity of fuel into it through a funnel.

Although these methods do not require any special equipment, they cannot
guarantee uniform and reliable results.

Some users organized a plant where the AN-FO mixture was prepared, packed
into multiwall paper or polyethylene bags, and stored until moved by company
trucks to the mine or quarry.

The mixing and packaging equipment and procedures were as follows [2]. Dry
ammonium nitrate prills are dumped into a sheet-metal hopper, approximately
4x4x4 ft in size with a sloping bottom, which feeds a 5 in. pipe through which
an auger, driven by an electric motor through reducing pulleys, move the prills
past a second, smaller bin into which oil is sprayed. The flow of oil is controlled by
a hand-operated valve. A short distance beyond the point at which the oil is added,
the auger discharges into a bucket elevator which raises the mixture to the top of a
hopper approximately 5 X 5 X 5 ft in size. The top of this hopper is approximately
10 ft above the floor. At the bottom of the hopper, four filling tubes, closed by
means of simple slide valves, control the flow of the mixture into packages. Usually
the mixture is repackaged into the multiwall paper bags in which the ammonium
nitrate had been received or into polyethylene tubes.

Sewing, heat sealing and tying equipment is available for closing the packages
of explosive.

Fire followed by a disastrous detonation occurred in a plant near Norton, Va.
According to van Dolah and Malesky [2] the accident demonstrated the importance
of the many safety recommendations issued earlier by the Bureau of Mines [3].
In brief, all typical safety regulations for explosives factories and stores should be
observed, i.e.: blasting agents should not be stored with ammonium nitrate or
explosive mixtures, no smoking or open flames should be permitted in any of the
buildings, the floors should be cleaned frequently to prevent any accumulation of
ammonium nitrate, fuel oil or explosive mixtures, no more than one day's produc-
tion of fuel-mixed ammonium nitrate should be permitted in or near the mixing and
packaging plant, all electrical equipment should conform to safety regulations and
all switches, controls, motors etc. should be outside the buildings.

Some opencast mines in the U.S.A. and Canada worked out special equipment
for production of AN-FO. Some of these designs are given below, according to the
literature [4].

THE MANUFACTURE OF MINING EXPLOSIVES

509

Figure 169 gives an idea of the apparatus which seemed to be in use at Bingham
Mine, Utah. It permits charging horizontal, skew and vertical shotholes with an
ammonium nitrate-fuel oil mixture. Ammonium nitrate (45.4 kg) is placed in the
vessel. Fuel oil (3.78 1.) is introduced from a measuring tank. The ammonium
nitrate and fuel oil are moved by compressed air (2.6 and 1.9 atm respectively).
Mixing occurs in the lower part of the vessel and immediately afterwards the
mixture is fed into the shothole.

Fig. 169. Diagrammatic presentation of mixing equipment used in Bingham Mine,

Utah, U.S.A. [4].

Another design, used in Canada, is shown in Fig. 170. It consists of a funnel
provided with a number of nozzles supplying the liquid fuel. Ammonium nitrate
passes through a sieve (i) and an exit (2) on a divergent cone (3) into the mixing
chamber where the liquid fuel is injected by four nozzles— one of them (4) is shown.
Uniform flow of the liquid is automatically maintained.

Another method also used in Canada is presented diagrammatically in Fig. 171.
Ammonium nitrate passes through a mill and a sieve (/) and is transported to a
mixing house through a stainless steel tube by means of compressed air. Liquid fuel
enters the same mixing house from a tank. Mixing occurs in a rotating hopper.
The mixture is loaded onto a special lorry (Fig. 172), which is equipped with a
compressor, a screw conveyer, and a dosage valve. Shotholes are loaded straight
from the lorry. The explosive from the lorry is delivered to the shothole by means
of a compressed air.

510

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 170. Diagrammatic presentation of mixing equipment used in Canada [4].

Fig. 171. Diagrammatic presentation of a mixing plant in Canada [4]; 1— ammonium
nitrate compartment, 2— fuel oil tank, 3— mixing house.

Fig. 172. Lorry for ammonium nitrate-fuel oil explosive [4]; 1 — screw conveyer,
2— lifting device, 3— dosage valve.

THE MANUFACTURE OF MINING EXPLOSIVES 511

THE MANUFACTURE OF DYNAMITES
Raw materials

Methods for the manufacture of raw materials of requisite purity for nitro-
glycerine, nitroglycol and nitrodiethylene glycol production and for dynamite
collodion cotton have already been described (Vol. II, pp. 87, 145, 152 and 409).

The oxidizing agents, such as sodium or potassium nitrates, were described in
the section on blackpowder (p. 342) and ammonium nitrate in Vol. II, p. 450.

The other ingredients should meet the requirements for commercial products of
high purity. Salts should be dried and milled.

Collodion cotton is a dynamite ingredient of very great importance. Before use
each lot of collodion cotton should be checked. For this purpose a small sample
of blasting gelatine made of nitroglycerine, or of nitroglycerine with nitroglycol
used in manufacturing should be prepared. The suitability of nitrocellulose for
manufacturing purposes is judged by the properties of the gel, notably by its chemical
stability and consistency particularly at an elevated temperature (40-50°C). The
gel should not give any exudation of nitroglycerine at this temperature.

Collodion cotton is supplied wet to dynamite factories with a water content up
to 30-35%. Some factories use it directly in this form, although the presence of
water is detrimental to the uniformity of the gel produced. In the majority of fac-
tories nitrocellulose is dried before use.

As will be discussed later (pp. 573, 642) the drying of nitrocellulose is dangerous
since when dry it is ignited with unusual readiness by sparks, friction or impact. The
drying operation therefore requires great care and is usually carried out in a separate
building surrounded by safety walls. The amount of nitrocellulose to be dried at
a time must not be too large (less than 100 kg). In the premises a temperature
of 45-50°C should be maintained by radiators heated with hot water. The
temperature is controlled from outside. The moist nitrocellulose is spread in a thin
layer (4-5 mm) over the shelves of wire cloth stretched on carefully-earthed, metallic
frames. Wooden frames with muslin stretched over them may also be used, provided
that a lower temperature of drying is maintained (e.g. 40°C) and too fast an air
circulation (e.g. mechanical ventilation) is avoided, otherwise high static charges
may build up on the nitrocellulose. Drying lasts for approximately 24 hr and is
ended at a 1.2-2% moisture content since complete drying may be dangerous.
After the drying process is completed, the heating is turned off, the drying room
opened and the nitrocellulose is unloaded after cooling.

Mixing of ingredients

The dynamite ingredients are mixed with a solution of collodion cotton in nitro-
glycerine prepared separately or in a mixture of nitroglycerine with nitroglycol.
The drying and milling of the ingredients are preparatory operations.

512

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Dissolution of collodion cotton. The collodion cotton used for the manufacture
of dynamite gives solutions of very high viscosity. Since it dissolves very slowly, to
avoid delays in production, dissolution is usually divided into two stages, viz. initial
dissolution and mixing.

Formerly, the initial dissolution was carried out in the following way. Nitro-
glycerine was weighed in ebonite or bronze cans and was poured into a wooden
tub, lined with ebonite or lead sheet, having a capacity of 25 kg of nitroglycerine.
Next collodion cotton was poured in and the whole was stirred by hands protected
with rubber gloves so as to obtain a mass as uniform as possible. The contents of
the tub were allowed to stand for a few hours (the last batches over-night). During
this time the nitrocellulose dissolved in the nitroglycerine and the mass became
transparent, but not yet uniform. A uniform solution was obtained only by using
stirrers or kneaders.

Fig. 173. Diagrammatic illustration of a Drais mixing and kneading machine with two

tanks. The drive and switchgear is in an adjacent room for safety reasons (Courtesy

Draiswerke Maschinenfabrik G.m.b.H., Mannheim-Waldhof, G.F.R.).

The initial dissolution is now usually carried out in copper tubs with a water
jacket which maintains a temperature of 45-50°C in the tub. The tubs contain
100 kg of a mixture of nitroglycerine and nitrocellulose. Their contents are stirred
with wooden paddles and then allowed to stand for 20-30 min. This is sufficient
time for complete dissolution of nitrocellulose at this temperature.

Mixing. For mixing the solution of collodion cotton with the other dynamite
ingredients vertical kneaders of the Drais type are commonly used (Figs. 173 and
174). The kneader of this type consists of stirrers and a vertical cylindrical tank
provided with thermal insulation. The stirrers are fitted with a mechanism enabling
them to be lifted and lowered. The toothed wheels which drive the mechanism are

Fig. 174. General view of a Drais mixing and kneading machine (Courtesy Draiswerke

Maschinenfabrik G.m.b.H., Mannheim-Waldhof, G.F.R.); a— with stirrers dipped in

the tank, b— with stirrers lifted, tank removed.

514

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

protected from dust by means of an air-tight casing. Drais type kneaders with two
or three replaceable tanks are also known (Fig. 175).

The tanks have a capacity of 300-450 1. and hold approximately 200-300 kg
of dynamite. They are suitable for the initial dissolution of collodion cotton (in a
separate room) and for efficient mixing of the collodion cotton with nitroglycerine
by hand (in rubber gloves) or with a hand-operated paddle. After the initial dissolu-

Fig. 175. Drais mixing and kneading machine with three tanks (Courtesy Draiswerke
Maschinenfabrik G.m.b.H., Mannheim-Waldhof, G.F.R.).

tion stage the tank is moved to the stirring equipment, where the other ingredients
are added and the final mixing is carried out. The remaining dynamite ingredients
are then added and stirred with the same stirrers. Since one set of stirrers is provided
for several tanks, continuity of work may be maintained and the prolonged initial
dissolution does not cause any delays in the manufacture.

THE MANUFACTURE OF MINING EXPLOSIVES

515

For the manufacture of dynamites, kneaders of the Werner- Pfleiderer type may
also be employed. They are widely used in the manufacture of smokeless powder
(Fig. 206). Drais kneaders however, have the advantage over the Werner-Pfleiderer
type in that they are safer to handle, and hence more suitable for mixing materials
sensitive to friction and impact, such as dynamites.

Safety in the operation of vertical kneaders is achieved by ensuring that :

(1) The points where there is friction between the stirrers and their bearings
are not in contact with the explosive mass.

(2) The distance between the paddles and the interior of the vat in which the
stirring is carried out is relatively large.

Cast iron

Fig. 176. A kneader used in France for the manufacture of blasting gelatine and

dynamites [5].

In Werner-Pfleiderer kneaders this distance is very small so that if a hard object
(metallic part, screw, nut etc.) penetrates into a kneader, an explosion may occur
due to seizing. Werner-Pfleiderer kneaders have a capacity of 50-200 kg of dynamite.

In some countries other types of plant are used for mixing dynamites. In France,
for instance, a kneader constructed as shown in Fig. 176 is used for the manufacture
of blasting gelatine and dynamites rich in nitroglycerine. Here a copper vat (/),
lined with lead and located in a heating jacket of sheet steel, rotates about a vertical
axis. A bronze stirrer, (2), can rotate about another vertical axis. The charge in the
kneader is 166 kg of nitroglycerine and 11-12 kg of collodion cotton. At the begin-
ning a temperature of 15-20°C is maintained and it is then raised to 45-50°C.

After 15 min during which the collodion cotton is dissolved, the remaining

516

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ingredients (e.g. 24 kg of potassium nitrate) are added. After stirring, the con tents
of the vat are cooled to a temperature of 20-25°C.

For mixing dynamites, in particular those containing less nitroglycerine, wooden
edge runners of a special design are employed in France (Fig. 177). The tub (7)
has the base of ebonite plates (2). The wheels of the runner are also covered with
ebonite bands (3).

^ ^ fLsi^imimm^^^fs^^jsiSie^^^^^^jX^p

Wood (1)

Fig. 177. Wooden edge runner for the manufacture of dynamite [5].

Cartridging

Dynamite from edge runners is transported in trucks to a separate room for
cartridging. Figure 178 represents a diagram of one of the oldest and simplest designs
for a hand operated machine for cartridging. Dynamite of plastic consistency is
loaded by hand into the tapered body (/), through funnel (2). The rotation of the
crank handle driving shaft (4) with worm (5) pushes out the explosive mass through
the nozzle (5). There are ribs (6) inside the tapered body which prevent the mass from

THE MANUFACTURE OF MINING EXPLOSIVES

517

turning. The mass in funnel (2) should be pressed with a peg to facilitate its exit
through nozzle (5). This nozzle has a brass orifice that holds a paper cartridge case.
The dynamite passes through the orifice, pushing the bottom of the paper cartridge
case and filling it while in motion. When the paper cartridge case has moved forward
a distance corresponding to the required length of the cartridge, the dynamite is
cut off and the cartridge closes. A new paper cartridge case is then attached to the
orifice and the operation is repeated.

The explosive may also be extruded in the form of a long rod which is then cut
into pieces of the required length on a table covered with a copper or brass sheet.
The pieces are then put into paper cartridge cases which should be made of paraffined
paper.

Fig. 178. Machine for the cartridging of dynamite [6].

In machines of the design described above a certain danger is created by the
pressure caused by the decreasing area of taper. The increase of the pressure leads
to greater friction and this, in turn, may cause the ignition of the dynamite. In a
machine of better design, therefore, the body in which the worm rotates is cylindrical.
It may be equipped with several brass orifices, so that several cases may be cartridged
at a time. Since this attachment is safer, it may be fitted with a power drive. Loading
the funnel with dynamite and fitting the orifices are the only manual operations.

Highly efficient machines are also available for mechanized cartridging. In the
U.S.A., for instance, there are machines in use, operated by two people and capable
of processing 6800 kg of dynamite in 8 hr. Mechanization, however, is not always
advisable since it may decrease safety.

In Germany a machine of the Niepmann type has been adapted for cartridging
dynamite [7]. A schematic view of its operation is shown in Fig. 179. The explosive
supplied through funnel (i) is pushed out through worm (2) into conduits (4) and
(5), from which it is ejected by pistons (3). The empty paper cases are fed by con-
veyer (9) into slots on the rotating table (7). When the cases reach position (8), a
charge of plastic material, extruded from the orifice (6), is slipped into them. Further
rotation severs the rods of dynamite while the cases, already filled, are passed to

518

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

slanting receptacles (//) and (72). As there are few movable parts this method of
cartridging is fairly safe. The machine processes 2400 cartridges of 100 g each,
per hour.

A general view of a modern dynamite plant in the U.S.A. is given in
Fig. 180.

ferm

Fig. 179. Schematic view of the operation of a Niepmann type machine for dynamite

cartridging [7].

Thawing of dynamites

Dynamites that contain nitroglycerine but no nitroglycol can easily freeze at
low temperatures. Temperatures below 6°C should be regarded as undesirable and
any temperature below 0°C will freeze the dynamite cartridges.

Although crystalline nitroglycerine is less sensitive to shock and can detonate
at a higher rate than the liquid substance, the hard frozen cartridges are troublesome
•to handle, as it is difficult to insert a detonator into them. When such cartridges
are warmed crystalline nitroglycerine separates from the gel as it melts and readily
flows out of the cartridges. This can create a danger as nitroglycerine spilled on to
the floor is subject to shock and friction.

THE MANUFACTURE OF MINING EXPLOSIVES

519.

520

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Several accidents have occurred in the course of thawing frozen dynamite cart-
ridges. In France, 1250 kg of dynamite caught fire and detonation occurred in a coal
mine at Merlebach, in 1925. Commission des Substances Explosives [8] investigated
the accident. The wooden cases with frozen dynamite were placed over steam heat-
ed calorifuges. In spite of the fact that the temperature in the compartment for
defreezing the dynamite did not exceed 21°C, according to the recording thermo-
meter, the dynamite caught fire after 40 hours' warming. Most likely this was due
to superheating.

Thermoregulator

Insulation
sheet

.Calorifuges

s Wooden floor
Fig. 181. Schematic design of a building for thawing (decongelation) of dynamites [8].

The Commission recommended several safety precautions, such as :

(1) keeping the temperature of calorifuges below 100°C preferably by heating
with hot water to ca. 50°C,

(2) keeping the temperature inside the defreezing compartment below 30°C,

(3) using compartments for thawing dynamite so designed that the cases contain-
ing explosive are placed away from the calorifuges (Fig. 181).

Although modern dynamites contain mixtures of nitroglycol-nitroglycerine they
should be considered to be only comparatively unaffected by a low temperature.
Particularly heavy frost may freeze these explosives and the recommendations of
the Commission des Substances Explosives should be observed.

THE MANUFACTURE OF CHLORATE AND PERCHLORATE

EXPLOSIVES

To avert the danger arising from the mechanization of mixing appliances, chlorate
mixtures are usually prepared either by primitive manual methods or in machines
made chiefly of wood.

Chlorate explosives of the Cheddite type (p. 277, Table 61) are manufactured
in an enamelled vat with a double bottom, heated by steam. First the organic in-

THE MANUFACTURE OF MINING EXPLOSIVES 521

gredients are mixed and melted at 80°C (e.g. nitro compound with castor oil, paraffin
with vaseline etc.). Finely-crushed chlorate is then added while stirring continuously
with a wooden paddle. After a uniform mixture has been obtained, the hot mass
is discharged onto a table covered with sheet brass and rolled with a wooden roller.
As the mass solidifies, it becomes brittle and crumbles. It is then rubbed through a
sieve. Finally the sieved mass is graded by sifting out the dust which is recycled.

If the explosive contains a liquid organic substance, it is prepared, as stated
above, by filling the paper cases with chlorate and dipping them into the liquid.

Miedziankit is manufactured in this way, i.e. the milled and dried potassium
chlorate is filled into paper cases by the machines for cartridging (Fig. 168). The
cases are made either of non-glued, absorbent paper or of ordinary, densely-perfo-
rated paper. The cartridges filled with chlorate (90 g of weight and 30 mm dia.) are
taken to another room in which they are saturated with kerosene. Here troughs
mechanically fed with kerosene from a movable tank are placed on tables. The
cartridges of chlorate are put into the troughs which rotate every few minutes so
that each cartridge is uniformly saturated. After half an hour, when the cartridges are
sufficiently saturated with liquid, they are taken out of the troughs and placed in
containers made of parchment paper.

Careful, fine milling of the potassium chlorate is a very important factor in the
manufacture of chlorate explosives. Opinions are divided on the question of the
safety of the milling of chlorate. In Germany chlorate is milled like ammonium
nitrate in common mills of the disintegrator type. A pre-requisite for such a proce-
dure, however, is a high purity of chlorate (free from traces of combustible material,
especially organic substances); milling impure chlorate is exceedingly dangerous.
In France wooden edge runners are used for milling chlorate similar to those for
dynamite (Fig. 177).

When manufacturing chlorate explosives, it is of extreme importance to observe
the safety code which demands the highest possible purity, the removal of all chlorate
dust settling on furniture, clothes etc. The workers' foot-wear should be wooden
soled, since nails in soles not infrequently cause accidents. The floor in the premises
should be covered with linoleum or magnesia cement.

The manufacture of perchlorate explosives is similar, but due to the lower sensi-
tiveness to impact of perchlorate explosives, some operations as e.g. mixing, may
be conducted in Drais kneaders (Figs. 173-175) or even in kneaders of Werner-
Pfleiderer type (Figs. 206-208).

CARDOX, HYDROX AND AIRDOX CARTRIDGES

The hazards involved in the use of explosives in coal mining have necessitated
constant efforts to improve the safety of blasting in coal mines, using other methods,
i.e. the Cardox, Hydrox and Airdox methods of blasting. The Cardox and Hydrox
cartridges function in a way similar to explosives.

522

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Cardox. These are steel cartridges (Fig. 182) containing liquified or solidified
gas and a pyrotechnic composition. On burning, the pyrotechnic mass emits a great
amount of heat and converts the liquid or solid into a gas at a high pressure sufficient
to blow out the walls of the shothole. Cardox was invented in the U.S.A. in 1920

Filling valve-

Electrical ,
connections

-Firing head

Chemical energizer

Release disks

detaining pawls ^

-Tube

- Carbon dioxide
charge

^Discharge head

Fig. 182. Cardox blasting cartridge, according to J. Taylor and Gay [9].

by Helmholz, Farrel and Crawford and received Bureau of Mines U.S.A. ap-
proval in 1928 [10].

In Cardox charges liquid carbon dioxide is the ingredient providing the gas
pressure. The pyrotechnic charge consists of chlorate mixtures that evolve a great
amount of heat, such as either sodium chlorate, powdered charcoal and aluminium

THE MANUFACTURE OF MINING EXPLOSIVES 523

dust or 10% naphthalene and 90% potassium chlorate. The mass is ignited by an
electric detonator. The reaction proceeds according to the equation :
C 10 H8 + 8KClO3^10CO 2 + 4H 2 O + 8KCl + 471kcal

In Great Britain a safer mixture was evolved (Payman [11]) with the following

composition :

Potassium perchlorate 85 %

Asbestos

1.5%

Nitrotoluene

4.5%

Kerosene

8.5%

Castor oil

0.5%

Potassium perchlorate 84%

Phenol-formaldehyde resins 16%

Due to the heat of reaction carbon dioxide evaporates very quickly. A high
pressure (700-2000 atm) is produced which blows out a steel disk closing the chamber.
Theoretically the pressure may increase to 4000-5000 atm. The steel cartridge is
used again since it is not destroyed. Due to high cost the use of Cardox has nearly
disappeared.

Hydrox. Hydrox charges (Fig. 183) are loaded with a mixture that usually consists
of 56% sodium nitrite and 44% ammonium chloride. They carry a charge of cap
mixture ignited electrically, the composition of which is similar to that of pyro-
technic mixtures. The heat of combustion of the latter charge causes the mixture
of sodium nitrite and ammonium chloride to react according to the equation:

NaN0 2 + NH 4 C1 -> NaCl + 2H 2 + N 2

Another possible reaction mechanism is :

NaN0 2 + NH4CI -> NH 3 + HG1

NaN0 2 + HC1 -> NaCl + HONO

HONO + NH 3 -* N 2 + 2H 2

The pressure of nitrogen and steam produced shears the bursting disk and re-
leases the gases into the shothole.

The reaction may stop at an intermediate phase. In fact, the presence of ammonia
can usually be detected in the reaction products.

Theoretically reaction takes place at a temperature of 1300°C. In reality the tem-
perature in the cartridge only slightly exceeds 800°C due to its high thermal capacity.

Since the temperature may become too high in certain cases and may cause the
ignition of methane in mines (especially when the cartridge is defectively assembled,
without the bursting disk, that encloses the charge) a mixture of ammonium nitrate
with organic combustible substances and salt hydrates is sometimes employed.
Such a mixture is incapable of self-sustained reaction at atmospheric pressure but
will decompose in the cartridge, when the pressure rise due to the combustion of an
initiating charge composed of guanidine nitrate and an alkali metal persulphate. The
initiating charge is fired by a pyrotechnic mixture, initiated, in turn, by means of an
electric current.

524

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

About 1953 a modification of the Hydrox device, called Chemechol, was intro-
duced in the U.S.A. In this design the electric detonator of the initiator is protected
from ignition by stray currents. By this method a pressure of 18,000-22,000 lb per
sq. in. (1200-1500 kg/cm 2 ) is attained. However it has been withdrown from the
market mainly due to high costs.

Discharge head

'Bursting disk

-Main charge

Steel tube

Primary charge

Hydrox
igniter No. 5

Electrodes

'0 'ring

Firing head

Fig. 183. Hydrox blasting cartridge, according to J. Taylor and Gay [9].

THE MANUFACTURE OF MINING EXPLOSIVES

525

A mixture of sodium nitrite with ammonium chloride may decompose with time
to form unstable ammonium nitrite. The presence of moisture and of acids favours
the decomposition of a mixture of sodium nitrite with ammonium chloride. On the
other hand, alkaline reaction and absence of moisture stabilize the system. Exper-
iments have shown that for all practical purposes the mixture is best stabilized by the
addition of 2% sodium carbonate. Ammonium carbonate or magnesium oxide may
also be used.

Thermochemical data on a Hydrox mixture stabilized with the addition of 2%
sodium carbonate are tabulated below (according to J. Taylor [12]).

Table 160

State of
water

Conditions

Heat of
reaction
at 25°C
kcal/kg

Temperature

of reaction

°C

Gas

volume
l./kg

Vapour
Liquid

constant pressure
constant pressure

420
589

1120
1120

537

Vapour
Liquid

constant pressure
constant pressure

435
594

1310
1310

537

Products of complete reaction of Hydrox mixture are:

NaCl 46.7%

N 2 22.4%

H 2 28.8%

Na 2 C0 3 2.1%

The total amount of gas produced comprises 51.2% (by weight) of all products.

A Hydrox mixture can be initiated either by heating it at one point by a heat-
producing fuse, by an electrically-heated wire or a blackpowder igniter. A few drops
of acid produce a self-sustained reaction, after an induction period. The induction
period depends on the nature of the acid solution, its concentration and on the amount
of alkaline stabilizer in the Hydrox powder.

Concentrated hydrochloric acid induces vigorous reactions which are partly
"Hydrox reactions", and partly reactions which give rise to the evolution of NO
(according to J. Taylor [12]).

3NaN0 2 + 3HCI -> 3NaCl + HNO3 + 2NO + H 2
3NaN0 2 + 2HC1 + NH 4 C1 -> 3NaCl + NH4NO3 + 2NO + H 2

(1)
(2)

Neither hydrochloric nor nitric acid initiates a self-sustained reaction in powder
stabilized with sodium carbonate or magnesium oxide.

This is probably caused by resulting reactions which destroy the acids too rapidly.

J. Taylor has described a method of igniting Hydrox powder stabilized with 2%
sodium carbonate: it consists of adding 0.2 cm' of 50% solution of chromic acid.

526 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Both Cardox and Hydrox devices should be subjected to tests before permission
is granted for their use in coal mines, because the essential components of these
devices are mixtures similar in nature to explosives. Faulty assembly, such as omis-
sion of a bursting disk may bring hot gases in contact with a methane-air mixture
or coal-dust.

The greatest advantage of Cardox and Hydrox methods lies in their safety towards
methane and coal-dust. They also contribute to the getting of coal in large pieces.
In comparison with explosives, however, shotfiring with Cardox and Hydrox devices
is more expensive.

Before World War II, in the U.S.S.R, another kind of Hydrox was suggested
by Komar [13]. It consisted of modifying Cardox cartridges by replacing carbon
dioxide with water. Water has a critical temperature (374°C) and total heat of
evaporation much above the critical temperature (31°C) and heat of evaporation of
carbon dioxide. Therefore water requires a much higher heat to be supplied by the
burning composition. This created difficulty and therefore Hydrox with water did
not seem practicable.

Podbelskii [14] suggested using salts which readily undergo dissociation into
gaseous products instead of carbon dioxide.

Airdox. The use of Airdox charges is based upon an entirely different principle.
A steel cartridge, closed at one end and connected with an air-compressor at the
other, is introduced into the shothole. The air is fed in up to 700 atm and when
it reaches the required pressure the bursting disk closing the charge is ejected so
that the air is discharged and the shothole shattered.

The Airdox method was introduced in mines in the U.S.A. in the early 1930s.

Towards 1938 a modification of this method was introduced under the name
of the Armstrong Airbreaker [9]. In this system the air is compressed up to 800
kg/cm 2 and introduced by opening a blow-out valve; this ruptures the bursting
disk and blows up the shothole.

The Airdox device is very safe towards methane and coal-dust. Safety is es-
pecially increased by the improvement of the bursting disk which is fitted with a
plastic plug closing the charge.

The Airdox device is widely used in the U.S.A. The Armstrong Airbreaker has
been employed in very gassy mines in Great Britain, France the U.S.A. and Poland.

LITERATURE

1. L. Medard and A. le Roux, Mem. poudres 34, 195 (1952).

2. According to R. W. van Dolah and J. S. Malesky, U.S. Bureau of Mines Report of In-
vestigation 6015 (1962).

3. U.S. Bureau of Mines Information Circulars 7988 (1960) and 8179 (1963).

4. L. I. Baron, G. P. Demidyuk and N. F. Adryanov, in Vzryvnoye deb, No. 45/2 (Novoye
v teorii i praktike vzryvnykh rabot), ed. L.N. Marchenko, Gosgortekhizdat, Moskva, I960.:

THE MANUFACTURE OF MINING EXPLOSIVES 527

5. L. Vennin, E. Burlot and H. Lecorche, Les Poudres et Explosifs, Beranger Paris-Liege
1932.

6. O. Guttmann, Die Industrie der Explosivstoffe, Vieweg & Sohn, Braunschweig, 1895.

7. BIOS Final Report 833, Investigation of German Commercial Explosives Industry.

8. Commission des Substances Explosives, Ann. mines 9 (1927).

9. J. Taylor and P. F. Gay, British Coal Mining Explosives, Newnes, London, 1958.

10. U.S. Bureau of Mines Report of Investigation 2947 (1929).

11. W. Payman, Iron Coal Trans. Rev. 121, 448 (1930).

12. J. Taylor, Solid Propellants and Exothermic Compositions, Newnes, London, 1959.

13. According to V. A. Assonov, Svoistva i tekhnologya vzryvchatykh materyalov, ONTI, Moskva-
Leningrad, 1938.

14. G. N. Podbelskii, Ugol No. 3 (1948).

CHAPTER VII

SMOKELESS POWDER

HISTORICAL

A few years after the discovery that the treating of cellulose with nitric acid con-
verts it into a combustible substance, the idea arose of using nitrocellulose as a
propellant instead of blackpowder. Schonbein's experiments [1], repeated by Pelouze
[2], showed the high energy of nitrocellulose. It was found that a charge of nitro-
cellulose endows a projectile with a penetrating effect similar to that of a triple
charge of blackpowder.

The primary difficulty in exploiting this property lay in finding a method for
manufacturing nitrocellulose. It was not until large-scale manufacture of nitro-
cellulose was achieved by Lenk [3] that propellant charges could be used for Austrian
artillery. Shortly afterwards, however, this method proved to be unsatisfactory, due
to' the variable results obtained and to the excessive pressure developed when firing
nitrocellulose charges, which in many instances damaged or even blew up the cannon.

Earlier observations that nitrocellulose burns very quickly in a confined space—
much more quickly than blackpowder— were confirmed. Since it is difficult to
reduce the burning rate by physical methods only, e.g. by compressing the nitrocel-
lulose to increase its density, attempts were made to slow down the rate of burning
by the addition of "phlegmatizing" substances, such as glues, waxes, fats etc.

Partly successful results were obtained by Schultze [4] who prepared his powder
by the following method. Wood cut into 1-2 mm grains was purified by boiling in
sodium hydroxide solution and bleaching with calcium hypochlorite. It was then
nitrated with a mixture of nitric and sulphuric acids. The nitration product was
stabilized by boiling in a sodium carbonate solution, then dried and impregnated
with a solution of either potassium or barium nitrate. After drying the grains were
polished in a drum with paraffin wax to form a powder of the following composition:

50% nitrocellulose and nitrated hemicelluloses
13% non-nitrated wood pulp
33% potassium and barium nitrates
4% paraffin

This powder, however, was still too fast-burning for use in military rifles, but

SMOKELESS POWDER 529

was found suitable for use in shot guns, and was a forerunner of propellants of
the "Schultze type" used in some countries (chiefly Great Britain) as sporting
powders. B

A few years later it was discovered that nitrocellulose dissolves in organic sol
vents, such as acetone, ethyl acetate and in mixtures of alcohol with ether leaving
on evaporation of the solvent a highly dense, transparent film, which burns more
slowly than nitrocellulose itself (Hartig [5]).

Some investigators tried to make use of this property. Volkmann [6] improved
Schultze's powder by dipping nitrated grains of wood into a mixture of ether and
alcohol and then either mixing them with blackpowder to prevent caking and
coating them with a layer of this explosive or compressing the sticky grains into
larger cubes. In spite of the encouraging results obtained in using this powder (the
size of charges required was half of that of blackpowder) the Austrian authorities
stopped manufacture on the formal grounds that the plant concerned infringed
their blackpowder monopoly.

A number of patents were then registered for various methods of using solvents
to prepare granular powder from nitrocellulose (Spill [7], Reid [8], Wolf and Forster
[9]). None of those methods, however, found practical application, except for a
short time in the work of Duttenhofer at Rottweil [10]. Duttenhofer nitrated
slightly carbonized cellulose, stabilized it and saturated the nitrocellulose so ob-
tained with ethyl acetate until a gelatinized mass was formed. After being dried
the horn-like mass was broken up in a corning mill and the grains so obtained were
graded. Clearly, Duttenhofer employed virtually the same production method as
that used to manufacture blackpowder.

Duttenhofer's powder was used for a certain time in Germany under the name
of RCP (Rottweil Cellulose Pulver). Its greatest disadvantage was the irregularity
of the shape of the grains which prevented it from burning as uniformly as the
smokeless powder [11] invented by Vieille at about the same time.

Vieille. developed his powder as the result of systematic investigations. In 1879
he began a study of the burning of explosives in a manometric bomb which he
invented together with Sarrau.

In the course of studying the burning of blackpowder Vieille found that it can
burn in parallel layers provided that its specific gravity is approximately 1.80 or
more (p. 340). He extended his experiments (1882-1884) to nitrocellulose, and
tested its behaviour at various densities. Since it turned out that high specific
gravity nitrocellulose cannot be achieved simply by pressing, Vieille made use of the
recognized method of increasing its specific gravity by treatment with various
solvents. He formed the dough-like mass into flakes and thin sheets which on
drying showed a fairly high specific gravity (about 1.65). By experiments in the
manometric bomb, Vieille demostrated that the flakes of the new powder burn in
parallel layers and that this property makes their time of burning dependent upon
their smallest dimension i.e. upon their thickness. Hence by altering this the total
time of burning of the flakes may be controlled, and the "coefficient of the vivacity"

53 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(coefficient de vivacite) of the powder

dp
dt

where p is pressure produced by burning the powder, t is time of burning, may be
determined in the manometric bomb. Thus powders of an adequate vivacity adjusted
to a given calibre of arms may be easily standardized.

Shortly afterwards (1885) Vieille's powder was introduced in France under
the name of B powder (Poudre B). Vieille utilized two types of nitrocellulose for
its manufacture: collodion cotton CP 2 (Vol. II, p. 373), soluble in a mixture of ether
and alcohol and forming the powder dough; cotton CPj , insoluble in a mixture of
ether and alcohol, incorporated into the powder mass in the form of unchanged fibres.

In Russia, Mendeleyev [12] worked out a method for the manufacture of smoke-
less powder from pyrocellulose, i.e. relatively high-nitrated (12.5% N) nitrocellulose
soluble in a mixture of ether and alcohol. In 1892 the manufacture of this powder
was started for naval guns. Nitrocellulose powder of this type was soon adopted
for military purposes in the U.S.A., where nitrocellulose powder became known
as "single base powder".

A second type of smokeless powder, ballistite, was invented by Alfred Nobel
[13] in 1888. He took advantage of the ability of nitroglycerine to dissolve nitro-
cellulose and thus replaced a volatile, non-explosive solvent (ether and alcohol in
former powders) by a non-volatile explosive solvent— nitroglycerine. The ratio of
nitrocellulose to nitroglycerine was 45:55. This is a relatively small amount of
nitroglycerine which dissolves nitrocellulose with difficulty. Abel and Dewar [14]
however succeeded in adapting acetone for the manufacture of nitroglycerine pow-
der. This is a solvent of both the active ingredients: nitrocellulose and nitroglyc-
erine. The product— British Cordite— has not been used outside the British Common-
wealth. The powders made of nitrocellulose, nitroglycerine and a mixture of ether
and alcohol as a solvent achieved only temporary success. The use of any solvent was
troublesome and proved a drawback in manufacture, so that nitroglycerine powder
without a volatile solvent, derived from ballistite, aroused much greater interest.

Work on the improvement of nitroglycerine powder without a volatile solvent
was aimed at the reduction of its content of nitroglycerine. By the selection of a
suitable nitrocellulose and by the addition of non-volatile solvents ("gelatinizing
agents"), of the so-called "centralite" type ("carbamite" according to English
nomenclature) as in Claessen's [15] patents, a new type of nitroglycerine powder,
the so-called RP-12 or RPC-12, was manufactured from 1912 onwards (p. 652).
This powder was used extensively during World War I since it could be produced
much more quickly than nitrocellulose powder. The manufacture of this powder
contributed largely to the long resistance of the Central Powers in World War I.
The enormous consumption of smokeless powder during this war led to diffi-
culties in producing a sufficient quantity of nitroglycerine. In Russia and Germany
attempts were made to replace part of the nitroglycerine by aromatic nitro compounds

SMOKELESS POWDER 531

such as DNT or "liquid TNT" (an oily mixture of DNT and TNT with isomers
of TNT). This powder had several advantages. In comparison with nitroglycerine
nitro compounds give a powder with a lower temperature of explosion that produces
less erosion and flash. Powders containing nitro compounds with nitroglycerine
were later adopted in the U.S.S.R.

Attempts to replace nitroglycerine partly or wholly by nitroglycol had little
success due to the high vapour pressure of the latter which facilitates volatization
and, consequently, reduces its ballistic stability (ballistic properties change as
nitroglycol volatilizes). Later, diethylene glycol dinitrate was tried (nitrodiglycol,
DGDN) and was shown to have great advantages over nitroglycerine. With nitro-
glycerine, good gelatinization of the nitrocellulose may be obtained if the ratio of
nitroglycerine to nitrocellulose is not less than 60:40, whereas with nitrodiglycol
this ratio may be much lower, viz. 20-45 nitrodiglycol to 80-55 nitrocellulose, since
nitrodiglycol is a better solvent of nitrocellulose than nitroglycerine. This facilitates
manufacture and, at the same time produces a more uniformly gelatinized mass.
Various alterations may also be introduced in the composition of powder such as
an increase in the content of nitrocellulose or on addition of insoluble ingredients,
serving, for instance, to suppress flash.

Solventless powder without nitroglycerine (G powder, p. 660) has a lower heat
of explosion, and consequently causes less wear on the bore.

Gallwitz [16] reports the following data on the influence of the heat of explosion
upon the bore wear. With a nitroglycerine powder containing no solvent and giving
a heat of explosion of 950 kcal, the barrel stands up to 1700 rounds while with a
similar powder giving a heat of explosion of 820 kcal, it withstands 3500 rounds.
The reduction of the calorific value of the powder by 130 kcal therefore doubles
the useful life of the barrel.

Further reduction of the calorific value of nitroglycerine powder proved to be
impossible. But by using nitrodiglycol instead of nitroglycerine, a powder was
obtained with a heat of explosion of 690 kcal, which prolonged the life of the barrel
considerably, i.e. to 15,000-17,000 rounds.

Nitroglycerine or nitrodiglycol powder was known in the U.S.A. as "double
base powder".

A further development led to the invention of flashless powder. Tests carried
out in various countries, included the addition of aromatic nitro compounds to
nitrocellulose powders and of potassium salts to nitroglycerine powders. Nitro-
diglycol powder with an addition of 2% K 2 S0 4 produced a small flash. During
World War II in Germany and Great Britain it was the custom to add a consider-
able amount of nitroguanidine to nitrodiglycol powders. In Germany this was called
"gudol" powder (German Gudol Pulver).

Other attempts to improve nitrodiglycol powders were based on the introduction
of substances such as penthrite (German Nipolit Pulver) and cyclonite. In both
cases a powder with a high calorific value was obtained. The manufacture of these
powders never went beyond the pilot plant scale.

532 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

PROPERTIES OF SMOKELESS POWDER

PHYSICAL PROPERTIES

The specific gravity of semicolloidal nitrocellulose powder with a mixture of
ether and alcohol as a solvent usually ranges from 1.54-1.63, although the specific
gravity of nitrocellulose itself is 1.66. This indicates a certain porosity of the powder.
The pores are filled either with air or with traces of residual solvent. According
to Brunswig [17] 100 g of nitrocellulose rifle flake powder contains 4-8 cm 3 of the air.

Nitroglycerine powder— completely or almost completely wholly colloidal— is
less porous, therefore more difficult to ignite than nitrocellulose powder and requires
a stronger priming.

The gravimetric density* of a powder depends upon the dimensions and shape
of the grains. A suitably high gravimetric density may be obtained by polishing
the grains and coating them with graphite. Its value determines the extent to which
a cartridge case or powder chamber can be filled with powder. Generally, to make
the most of the capacity of a cartridge case or powder chamber the gravimetric
density should be as high as possible. E.g. by raising the value for smokeless powder
from 0.770 to 0.830, the charge for a rifle was increased from 2.65 to 3.2 g and this
led to an increase in muzzle velocity and range.

Occasionally it may happen that the gravimetric density of a powder is too
high, in which case a considerably part of the chamber space or cartridge case
remains empty. Then ignition may prove unreliable and burning not quite uniform.
According to Brunswig [17], for instance, powder with a gravimetric density of
0.850 gave 0.07% misfires and 1.2% hangfires per 22,000 rounds, while a powder
with a gravimetric density of 0.820 showed no such disadvantage.

Nitrocellulose powders (semicolloidal) are moderately hygroscopic; in an at-
mosphere saturated with water vapour they attain 2.0-2.5% moisture content.
Nitroglycerine powders (colloidal) are virtually non-hygroscopic. (The hygroscop-
icity of nitrocellulose was discussed in fuller detail in Vol. II, p. 283.) This is ac-
counted for by the fact that the nitrocellulose in the latter powders has a colloidal
form of low hygroscopicity. By using such gelatinizing agents as DNT instead of
nitroglycerine, the hygroscopicity of nitrocellulose powders was reduced. In the
U.S.A. these powders were known under the name of NH-Powder (non-hygro-
scopic powder).

EXPLOSIVE PROPERTIES
Products of decomposition

The products of decomposition of smokeless powder resemble those formed
by the decomposition of its ingredients, i.e. nitrocellulose or nitrocellulose with
either nitroglycerine or dinitrodiglycol.

* Gravimetric density = the weight of the charge in a unit volume (e.g. kg in litre).

SMOKELESS POWDER

533

In nitrocellulose powder the chief products of explosive decomposition are
inflammable gases: CO, H 2 and C0 2 , H 2 0, N 2 . In nitroglycerine powder the
average composition of the gases is similar, but owing to the more advantageous
oxygen balance the amounts of complete combustion products (C0 2 and H 2 0)
are higher. In the products of the decomposition of smokeless powder methane
is also found in small amounts and sometimes also hydrogen cyanide or carbon.

The composition of the decomposition products of powder varies, depending
on many factors, the most important of which is the pressure in the powder chamber
and in the bore of the gun. The pressure, in turn, depends mainly upon the density
of loading. Thus, for instance, the decomposition products obtained from the same
powder vary according to the density of loading (Table 161): the amount of C0 2
and CH 4 increasing, and that of CO and H 2 decreasing as density increases.

Table 161

Amounts of decomposition products of powder in relation
to density of loading

1 Pressure
1 kg/cm 2

co 2

CH 4

H 2

N 2

H 2

0.1
0.3

730
3200

9.6
16.4

44.8
38.4

0.7
5.5

20.7
13.2

10.3
13.3

13.9

13.2

This may be deduced from the following equation, which takes place with
reduction in volume :

CO+3H 2 & CH 4 +H 2 0+57.8 kcal

The reaction is exothermic, hence as the gases cool the reaction equilibrium is
shifted to the right.

The decomposition products of a given powder differ at various distances from
the muzzle since temperature and pressure decrease considerably with the move-
ment of the projectile along the bore. Brunswig [17] reported the following figures
for the German M/88 rifle (Table 162).

Table 162

Travel of the base

of the projectile

Temperature

of gases

°C

Pressure

kg/cm 2

200

1426

1385

300

1202

834

400

1060

577

500

965

434

600

877

339

693 (muzzle)

818

280

534

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In addition to the above reaction the following dissociation reactions also take

place:

2C0 2 *±2CO + 2 -135.2 kcal

2H 2 0*±2H 2 +0 2 - 115.6 kcal

Here rise in temperature favours the displacement of the equilibrium to the
right, whereas increase of pressure displaces the equilibrium to the left.

Poppenberg and Stephan [18] showed that pressure affects the system more
strongly than temperature, so that the content of C0 2 grows and that of CO falls
as the projectile moves towards the muzzle. The ratio C0 2 :CO at different posi-
tions along the bore varies as shown in Table 163.

Table 163

Travel of the projectile
cm

co 2

0.298

0.324

0.362

0.393

Among the reactions which proceed inside the bore, the following is noteworthy:

CO + H 2 0?±C0 2 + H 2

The reaction is exothermic and the equilibrium is displaced to the left as the
temperature rises (Table 164).

Table 164

Temperature
°C

[CO][H 2 0]
[C0 2 ] [H 2 ]

1600

4.24

1405

3.48

1086

2.04

886

1.19

686

0.52

The nitrogen present in the form of nitrate groups in the ingredients of the
powder is almost completely transformed into molecular nitrogen N 2 in the final
products of the reaction. A small amount of nitrogen, however, may remain in a
form of nitrogen oxides, especially if the powder is burnt at a low pressure.
Sometimes muzzle gases contain ammonia which is easy to detect by smell. Am-
monia is formed by the reaction

N 2 + 3H 2 *±2NH 3 + 22.0 kcal

which proceeds while there is still considerable pressure as the gases cool. This
reaction is promoted by the presence of iron particles in the propellant gases.

SMOKELESS POWDER

535

Diagrams (Fig. 184) show the effect of the density of loading of nitroglycerine
and nitrocellulose powder upon the amount of C0 2 , CO, H 2 and CH 4 evolved
(according to Andrew Noble [19]) in the explosion products.

Nitroglycerine powder— cordite— gives a better proportion of more completely
oxidized products.

0.10 0.20 0.30

0A0 0.50 0.10
Density of loading

0.20 0.30 0.10 0.50

Fig. 184. Proportion of C0 2 , CO, H 2 and CH 4 in products of explosion of cordite and
nitrocellulose powder as a function of density, according to Brunswig [17].

Gases may escape from the barrel through the breech end after the breech block
has been opened. They may be toxic to the gun detachment if the breech end is
situated in a confined space (naval gun turret, concrete pillbox, tank etc.). A flow of
air to remove them is therefore essential. Similar hazards arise due to backflash.

Knight and Walton [20] examined the products produced by burning powder
in a confined space, simulating a naval gun turret. Ten seconds after the ignition of
32 kg of powder, the estimated composition of the gases in the test container (about
25 m 3 in capacity) was:

N0 2

co 2

H 2
CH 4

N 2

17%

28%

37%

After 20 sec the powder had burnt completely and fresh air was introduced
into the container. The hot combustion products exploded again on being mixed
with air. The estimated composition of the products after the second explosion

536

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

was:

no 2

co 2

o 2

12%

N 2

67%

Due to the presence of N0 2 and CO the gaseous products were very toxic to
experimental animals.

Heat of explosion, volume of gases and temperature of the explosion products

The heat of explosion depends chiefly on the composition of the powder, e.g. in
nitrocellulose powder on the content of nitrogen in the nitrocellulose and in nitro-
glycerine powder on the content of the nitroglycerine.

The effect of the composition of the powder, in particular of the content of
nitroglycerine upon the volume of gases, heat and temperature of explosion is il-
lustrated by the figures in Table 165 (according to Brunswig) and by the curves in
Fig. 185 (on the basis of another series of experiments by the same author).

10 20 30 40 50 60
NitroQlL/cerine content in the powder, %

Fig. 185. Gas volume and heat of explosion of nitroglycerine powder as a function
of the proportion of nitroglycerine in the powder [17].

Table 165

Composition of powder

Nitroglycerine
Nitrocellulose

Other non-explosive ingredients
(centralite, vaseline etc.)

30
65

40
50

47
53

58
37

100

Volume of gases, Vq (l./kg)
Heat of explosion (kcal/kg)
Temperature, / (°Q

913
1030
2470

910
935

900
1005

810
1090
2850

875
1250
2825

934

924

2230

Generally speaking, nitroglycerine powders give a higher heat of explosion, so
that the temperature of their products is higher than that in nitrocellulose powders.

SMOKELESS POWDER

537

This means that nitroglycerine powders are more erosive (greater bore wear) (p. 548)
and more flashy (p. 544).

Non-explosive substance (e.g. vaseline) are added to nitroglycerine powders
to reduce the heat of explosion and the temperature of the flash. The addition
of "cool" explosives such as nitroguanidine has the same effect.

The phenomenon of burning smokeless powder has been the subject of many
investigations. Usually they are described in text-books on ballistics [21]. Here
only a few essentials will be dealt with.

According to the Soviet authors Belayev and Zeldovich [22] the burning of
colloidal or semicolloidal powder consists of the following stages:

(1) decomposition of the solid and formation of gases,

(2) reaction between the gases leading to a considerable increase of the tem-
perature. The temperature of the solid surface remains relatively low.

This theory is diagrammatically represented by Fig. 186, according to Zeldovich
[21]. The thickness is denoted by X.

Pomter ~—U-+ Gases T

Fig. 186. Temperature distribution in a powder grain of thickness X, according to

Zeldovich [21]; T - temperature inside the grain, T s - temperature on the surface,

T c — temperature of the combustion flame.

The chemical reactions of gas formation occur in zone (1) of X p thickness;
reaction between the gases occurs in zone (2). The temperature inside the powder
is T , in zones (1) and (2) is T s and T c respectively.

Inside the powder grain a superheated layer exists of thickness X b . Here de-
composition starts in the part denoted X s .

The thickness of X s is very small and forms only about 5% of the total thickness
X h . Heat is transmitted by conductivity, convection and radiation. The rate of
burning of smokeless powders is determined by the rate of transmission of the
energy from the products of combustion to the powder itself.

It is generally accepted that colloidal and semicolloidal powder burns in parallel
layers. In fact this should be considered as an approximation as the burning surface
is uneven and covered with pits products by the "hotter" spots.

538

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Rideal and A. J. B. Robertson [23] suggest that initiation of the thermal explosion
of nitrocellulose is preceded by liquefaction. Very likely the same would apply to
nitrocellulose powder.

Huffington [24] studied the combustion of solventless cordite containing 56%
nitrocellulose (12.2% N), 29% nitroglycerine, 4.5% carbamite and 10.5% dinitro-
toluene plus 2.45% cryolite and 0.35% volatile matter.

He concluded that the burning of cordite is essentially an intermittent pheno-
menon in which periods of quiescence are followed by abnormally rapid burning.
Under certain circumstances, there is a tendency towards the end of a rapid burning
stage for the more stable components of the cordite (such as dinitrotoluene) to
remain undecomposed or only partially decomposed relative to the other components.

These experiments appear to throw a new light on the structure of solventless
cordite. It seems that nitroglycerine is present as minute globules and that the sec-
ond type of intermittent burning is due to the non-thermal explosion of these globules.

A simple exponential equation proposed by Vieille [1 1] is very often used to denote
the rate of burning (v) of colloidal and semicolloidal powders (i.e. powders burning
approximately in parallel layers);

v=kp",

where k is a constant, p pressure, n an exponent.

For semicolloidal powder Vieille found «=0.67. Zabudskii [21] later established
n=0.93.

Since smokeless powder burns by parallel layers the shape of the powder grains
can have a decisive influence on the mode of burning: neutral (with an approximately
constant burning surface), progressive (increasing surface), degressive (decreasing
surface) (Fig. 187).

CnO>

5-S

3 3

QQ to

Prog^

50 100

100 50 100
% of the burned arain

50 100

100

Fig. 187. Modes of burning of perforated grains and cords. The dotted lines represent
approximately the positions of the inward and outward burning, according to [25].

The neutral and progressive types are most valuable.

Klein, Menster, Elbe and Lewis [26] recorded the flame temperature of burning
smokeless powder using thermocouples and found in the vicinity of the burning
surface layer about 0.6 mm in thickness in which there was a sudden jump in tem-

SMOKELESS POWDER

539

perature to 1200°C at pressures 25-50 kg/cm2. They calculated from the figures
obtained that the temperature of the surface of the powder immediately before
burning was 250°C.

The temperature of the explosion products of smokeless powder (flame tempera-
ture) is dependent on the density of loading, increasing as the density increases.
This is illustrated by the diagram in Fig. 188. A particularly notable rise of tempera-

3200
3000

„- 2800

I 2600

flj 2400

2200
2000

~i — i — i — i — i — r

Ballistjte

-i — i — i — i i i i i

0.1 0.2 0.3 0.4 0.5
Density of loading

Fig. 188. Temperature of the explosion flame of powders as a function of density,
according to Brunswig [17].

ture with density of loading is observed in nitrocellulose powders or those nitro-
glycerine powders which are poor in nitroglycerine and hence in oxygen.

A knowledge of the heat of explosion of smokeless powder is of considerable
practical importance since the muzzle velocity of the projectile depends largely on
this factor. For this reason the heat of explosion of manufactured powder must
be checked from time to time. If the heat of explosion of a given lot of powder
deviates from the specified value, then the necessary size of charge should be calcu-
lated to match the standard muzzle velocity. For this the empirical formula

CQ = K

is used, where C denotes the charge of powder, Q — the heat of explosion of 1 kg
of powder, K — a constant quantity for a given projectile and a given muzzle veloc-
ity, irrespective of the type of powder.

Thus, for instance a charge of 4.3 kg of powder with a heat of explosion equal
to 820 kcal/kg gives the same muzzle velocity as a charge of 6.0 kg of powder with
a heat of explosion equal to 590 kcal/kg.

J. Taylor [27] drew up the following table of the heat of explosive decomposition,
the gas volume and rate of burning of British nitroglycerine (double base) and
American nitrocellulose (single base) powders (Table 166).

540

chemistry and technology of explosives
Table 166

Nitrocellulose powders
(U.S.A.)

Heat of explosive

decomposition

(water liquid)

kcal/kg

Gas volume

(water vapour)

l./kg

Linear coefficient
for rate of burn-
ing from one sur-
face in./sec/in 2

Pyro Cannon Powder
NH Cannon Powder
FNH Cannon Powder

875
765
740

955

978

1005

0.40
0.38
0.355

Nitroglycerine powders
(Great Britain)

Cordite MD
Cordite WM
Cordite SC
Cordite NQ
Cordite N

1025

1013

970

880

765

940

934

957

1001

1058

0.595

0.575

0.50

0.375

0.315

In the German Army nitroglycerine powders (double base powders) possessed
the following calorific values: 1250, 1150, 950 and 820 kcal/kg.

The diglycoldinitrate powders were characterized by calorific values: 930, 740,
690 kcal/kg (according to Gallwitz [16]).

Sensitiveness to impact and friction

Smokeless powders have low sensitiveness to impact and friction. They do not
ignite when hit by rifle bullets and are thus fairly safe to handle in war-time.
Nevertheless there have been accidents caused by the sudden ignition of nitro-
cellulose powder brought about by the violent friction between the sharp edge of
a heavy bin and powder scattered on the floor.

Nitrocellulose powders is particularly sensitive to friction when hot. Accidents
have been caused by the ignition of hot powder on removal from the drier, in all
probability due to friction. Here electrification of the powder on drying also plays
an important part (see below, pp. 542, 616) giving rise to the possibility of an electric
discharge while it is being taken out. This had led to the regulation that the powder
must not be taken out of the drier before it has been cooled.

Nitrocellulose powder is more sensitive to friction than nitroglycerine powder,
although the latter is more sensitive to impact (due to the presence of nitroglycerine).

Sensitiveness to detonation

According to Kast [28] nitrocellulose powder does not detonate even when
very strongly initiated (e.g. 50 g of picric acid or 100 g of tetryl), but may explode
with a rate from 1000 to 1800 m/sec.

Conversely, T. Urbanski and Galas [29] found that smokeless powder detonates
with a rate from 3800 to 7000 m/sec when initiated by 20 g of picric acid, the rate

SMOKELESS POWDER

541

of detonation depending either on the density of loading or, in wide strip powders,
on whether the detonation wave travels perpendicularly or in parallel with the sur-
face of the strips. The figures compiled in Table 167 are characteristic of the pheno-
menon observed (the powder was placed in iron pipes 26/33 mm dia., and 20 g of
picric acid was used for initiation).

Table 167

Rate of

Type of powder

Way of loading

Density

detonation
m/sec

Nitrocellulose powder

Rifle flake non-polished powder

loosely poured

0.79

3800

Rifle flake polished powder

loosely poured

0.93

5300

Cannon strip powder French type BC,

circles laid perpen-

0.55 mm thick

dicularly to the tube

axis

1.45

7010

Nitroglycerine powder

Ballistite, 1.3 mm thick

circles laid perpen-
dicularly to the tube

axis

1.53

7445-7615

Ballistite, 3 mm thick

circles laid perpen-
dicularly to the tube

axis

1.52

7720-7125

If powder strips or tubes are laid along the axis of the bore the powder is much
more difficult to detonate. A stronger initiator is required and not infrequently
explosion occurs with a rate of about 1500 m/sec, instead of a detonation. The
figures listed in Table 168 refer to the same conditions of initiation (20 g of picric
acid) in a pipe 26/33 mm.

Medard [30] also described experiments which were carried out in France by
Vieille as early as 1906, Dautriche (1913), Burlot (1920-26) and by himself in 1938.
Dautriche found that nitrocellulose powder BM 17 D2 in bands of thickness ca.
44 mm detonated at the rate of 6560-7200 m/sec when initiated by 50 g of picric
acid.

Burlot found similar figures. He also examined the effect of the impact of a fal-
ling weight or of a rifle bullet. Only deflagration occurred — there was no detona-
tion. However in his later experiments Burlot [31] has found that nitrocellulose pow-
der (poudre BM9— in strips) can detonate under the shock produced by a rifle
bullet "D" (caliber 8 mm, 7.5 g) having a velocity above 1200 m/sec (e.g. 1266
m/sec).

Ignitabiiity

The ignition temperature of nitroglycerine smokeless powder is approximately
180°C and that of nitrocellulose powder about 200°C. Ignition with a direct flame,

542

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

especially of nitroglycerine powder, is rather difficult. In practice an indirect
priming of blackpowder is used, the latter being ignited with a priming cap.

It has long been known that smokeless powder is readily electrified and that
an electric discharge may cause its ignition. When dried it is particularly readily
electrified by the frictions produced by the current of warm, dry air. The powder
is also subject to very strong electrification when polished due to friction of the
grains against each other and against the wooden balls.

Table 168

Rate of

Type of powder

Way of loading

Density

detonation
m/sec

Nitrocellulose powder

BC strip powder 0.55 mm thick

strips laid along

the bore

0.95

1495-1150
(explosion)

Nitroglycerine powder

Ballistite 1.3 mm thick

strips laid along

the bore

1.23

no detonation

Ballistite 3 mm thick

strips laid along

the bore

1.24

2250-1965

According to Nash's [32] studies, nitrocellulose EC powder is electrified under
the effect of an air current at a temperature of 65°C. The drier the powder, the
more it is electrified. The dependence of voltage on the moisture content of the powder

V
300

200

100

12 3

Moisture content, %

Fig. 189. Electric (static) charge of nitrocellulose powder as a function of moisture
content, according to Nash [32].

1
-o\o

9^-—.

■m 1 .O

is shown in the diagram (Fig. 189). As can be seen, powder with a moisture of over
1% cannot easily be electrified. Dry EC powder can be electrified up to 300 V.
Colloidal powder is electrified even more strongly. E.g. pyrocellulose powder gelati-
nized with a mixture of alcohol and ether, when heated to 50 C C, can be charged up
to 3200 V under the influence of a warm air current. After being cooled to 15°C

SMOKELESS POWDER 543

and subjected to a cold air stream the voltage falls in 5 min to 2200 V, and after
20 min to 900 V, after which the latter voltage remains constant. Powder electrified
to 2300 V maintains a charge of this voltage for a long time if not subjected to a
cold air stream.

Nash proved experimentally that a discharge of accumulated electricity may
cause EC powder to ignite if the voltage reaches 20,000 V. On this basis he came to
the conclusion that EC powder cannot take fire in the drier under the influence of
electric discharges since for its ignition a charge of higher potential is needed than
that to which it can be charged on drying. Nevertheless the many accidents which
have occurred by the ignition of hot powder when taken out of the drier appear
have been caused by the discharges of static electricity (sensitization of the powder
to friction due to heating may also be possible here, as mentioned above
p. 540).

A number of explosions of smokeless powder during polishing are also attribut-
able to electrification and subsequent discharge of electricity. In practice the careful
earthing of the drums used for polishing is a sufficient safeguard. A danger can
also be produced if the solvent (particularly ether) is charged with static electricity
and then discharged with sparking. This is dealt with later on p. 589.

Langevin and Biquard [33] showed that the evaportion of a liquid (alcohol,
ether, benzene) does not electrify the residual solvent. Hence evaporation of the sol-
vent in driers cannot lead to electrification of the powder. In the opinion of these
authors nitrocellulose powder may be ignited by the discharge of the condenser at
a voltage of 3000 V, if the condenser charge is greater than 0.3 juF.

According to the work of Brown, Kusler and Gibson [34] dry nitrocellulose can
be ignited by a discharge of 0.062 J energy, nitrocellulose smokeless powder may
be ignited by a discharge of 4.7 J energy whereas graphitized powder is much more
difficult to ignite, requiring an energy of over 12.5 J. English authors (Morris [35])
believe these figures should be considerably lower : 0.3-0.6 J.

The French Comite Scientifique des Poudres [36] collected statistical data indica-
ting that 75% of the accidents involving ignition occurred with completely dried
powder; of these 41% were caused by the electrification of the powder by friction,
and 34% by other, undefined reasons (e.g. insufficient chemical stability of powder).
The remaining 25% of all accidents involving ignition were caused by the fact that
the powder contained a considerable amount of solvent. Of these cases, 19.5%
arose from electrification of the powder itself by friction, and 5.5% by electrification
of the solvent. Generally speaking, electrification proved to be the source of 60%
of the accidents in which the powder ignited.

MECHANICAL PROPERTIES

Mechanical properties of double base powders mainly for large rockets will be
discussed later (pp. 675 and 678).

544 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

FLASH AND METHODS FOR SUPPRESSING IT

The discharge of a gun is almost always accompanied by a glaring flash which
at night discloses its position. Experiments, particularly using cine-films, give the
following picture of how flashes arise.

The "primary flame" in the form of a dark-red taper is created by inflammable
gases escaping from the bore. These gases mix with the air to form an inflammable
mixture. If the temperature of the mixture is sufficiently high, it ignites at the end
furthest from the muzzle (Fig. 190).

Fig. 190. Muzzle flame: 1— primary flame, 2— secondary flame, 3— initiation of the

secondary flame.

The burning of the mixture may end with an explosion (detonation) if its composi-
tion is suitable. The explosion of the gas mixture is accompanied by a bright flame.
This "secondary flame" is elliptical and visible for a great distance. The dimensions
of the flame and the intensity of the flash depend to a great extent on the calibre
of the gun ; e.g. a shot from a 30 cm naval gun gives a secondary flame up to 50 m
long, visible for a distance of 50 km.

The following circumstances favour the formation of a secondary flame :

(1) The presence of gases that form inflammable mixtures in the gaseous products
formed by the explosion of the charge (in the muzzle gases).

(2) High temperature of the muzzle gases.

On the other hand the presence among the products of explosion of substances
that prevent the explosion of the gas mixtures reduces the possibility of the second-
ary flame.

The most important inflammable components are hydrogen, carbon monoxide
and methane, which form explosive mixtures with air. Gas composition has been
discussed earlier in this chapter (p. 532).

According to Roszkowski [37] the explosive limits for these gases with air are:

Hydrogen

9.2-68.5%

Carbon monoxide

13.0-77.6%

Methane

5.5-13.2%

Methane has the narrowest limits so that a shot in which a large amount of
this gas is generated is unlikely to produce a secondary flame. A. high pressure in
the bore favours the formation of a large proportion of methane (p. 533), hence a high
charge density reduces the probability of the secondary flame.

Another factor in reducing the likelihood of a secondary flame is an increase in
the concentration of non-flammable gases (C0 2 and N 2 ) in the products formed

SMOKELESS POWDER

545

by explosive decomposition of the powder. Since the nitrogen content in these pro-
ducts is limited by the nitrogen content of the powder, the only factor that can be
varied to any extent is the content of carbon dioxide. High pressure and low tempera-
ture promote reactions for the formation of C0 2 (p. 534). Coward and Hartwell
[38] claim that the explosive limits of a methane-air mixture are narrowed by adding
C0 2 or N 2 (Fig. 191). Methane-air mixtures containing about 50% C0 2 are non-
explosive.

Atmospheric oxygen, %

20 19 18 17 16 15 14 13

i 1 1 1 1 r

5 10 15 20 25 30 35 40
C0 2 (/) or N 2 (//) added, %

Fig. 191. Change of limits of explosibility of methane-air mixtures under the influence
of added C0 2 (curve /) and N 2 (curve //) according to Coward and Hartwell [38].

The presence of water vapour in a methane-air mixture may exert the same effect
as the presence of C0 2 , but small amounts of water vapour increase the possibility
that the gas mixture will explode. This accounts for the observation that the secondary
flame develops much more easily when the firing takes place in a moist atmosphere.

Another component favouring explosion in gas mixtures is nitrogen dioxide
which may occur in small amounts as a result of incomplete decomposition of the
powder.

The temperature of the propellent gases depends on the heat of explosion and
on the gas composition. The greater the heat of explosion, and hence the tempera-
ture of gases, the more readily the secondary flame arises. Powders of a low calorific
value may therefore give, no secondary flame if the temperature of the gas mixture
is lower than the temperature of ignition.

The ignition temperatures of the most important gaseous components of the
explosion products (H 2 , CO and CH 4 ) in methane-air mixtures he within the follow-
ing limits :

Hydrogen 390-620°C

Carbon monoxide 610-725°C

Methane 730-790°C

546

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

This shows clearly the advantage of shifting the equilibrium towards the forma-
tion of a larger proportion of methane.

From a large number of practical data, it has been deduced that a powder will
be fiashless if on decomposition the heat emitted is below a certain limiting value.
The heat effect of the explosion depends on the type of ordnance, i.e. on the calibre
and length of the barrel, ranging from 800 kcal/kg for big guns to 1000 kcal/kg
for smaller ones. "Flashlessness" can therefore be achieved either by adding active
(explosive) components that lower the heat of explosion, (e.g. nitroguanidine,
DNT) or inert (non-explosive) ones such as vaseline, centralite etc., to the propellant.

Substances inhibiting the development of the secondary flame are those which
inhibit burning reactions. The strongest of the substances known to possess this
property is the potassium ion. Its ability to prevent the development of a second-
ary flame was demonstrated by Dautriche [39] as early as 1908. Since then potassium
nitrate has been employed in the manufacture of flash-reducing charges added
to the charges of ordinary propellant.

Fauveau and Le Paire [40] proved that the addition of substances such as chlor-
ides of alkali metals and alkaline earths to the charge reduces their tendency to
develop a secondary flame; here the result is achieved by inhibiting the reaction
of explosion, not by reducing the temperature of the gases. This is proved by the
fact that salts of alkaline earths (e.g. CaCl 2 ) and of metals which have a high heat
of evaporation and a considerable heat of decomposition at a high temperature,
do not prevent the formation of secondary flame. Conversely, a typical inhibitor
of gas explosions is potassium chloride which has neither a high heat of evaporation
nor dissociates at temperatures reached within the barrel.

Table 169
Effect of KC1 on the ignition temperature of mixtures

OF CO WITH AIR

Content of KC1 in mg

Ignition temperature (°C) of the mixture

per litre in the gas

with air containing

mixture

24.8% CO

44.1% CO

67.3% CO

656

657

680

0.4

—

750

800

0.5

730

820

0.7

—

810

900

1.0

790

850

1020

1.3

810

2.0

890

950

2.5

—

1000

3.0

970

—

3.5

1010

Prettre [41] found that potassium chloride sprayed in a mixture of carbon mon-
oxide and air considerably raises the ignition temperature of these mixtures (Table

SMOKELESS POWDER 547

169), but it does not affect the ignition temperature of mixtures of hydrogen with
air.

It has been shown that potassium chloride lengthens the induction period (the
interval between the time the heating of gas mixture begins and the moment of
explosion). Pease [42] found that in a vessel with walls coated internally with potas-
sium chloride, the reaction of hydrogen with oxygen undergoes considerable inhibi-
tion, the induction period being lengthened one thousand times. The action of
potassium ions is now explained by their capacity to initiate the combination of
free H and O atoms and OH radicals into inert molecules, thus breaking chain
reactions.

Free atoms of hydrogen and oxygen and OH radicals arise as a result of the
following chain reactions :

H 2 + 2 ->20H

OH + H2 ->H 2 + H

H + 2 -^OH +0

+ H 2 -*OH +H

CO + OH-»C0 2 +H

Examination of the flame spectrum of a mixture of hydrogen and oxygen does
indeed point to the fact that hydrogen atoms and hydroxyl radicals take part in
this reaction. The presence of OH radicals as intermediate products of the reaction
between hydrogen and oxygen was demonstrated by an examination of the absorp-
tion spectrum of a mixture of these gases heated to a high temperature (over 1000°C)
and by an examination of the absorption spectrum of water vapour at a temperature
of 1500°C and higher (Bonhoeffer and Reichardt [43]; Avramenko and Kondratyev
[44, 45] ; Dwyer and Oldenberg [46]). Examination of the course of gaseous explosive
reactions by kinetic absorption spectroscopy, invented by Norrish and Porter [47]
is particularly noteworthy. By this method Norrish et al. [47-49] established the pres-
ence of the OH radical as an intermediate product in the reaction of hydrogen
with oxygen, in the burning of hydrocarbons in an atmosphere of oxygen and
in the other reactions in which water is one of the final products. Potassium ions
promote the following chain-breaking reactions :

H + H -^H 2
H + OH->H 2

O + O ->o 2
CO + O -^co 2

Investigations showed that the salts of other alkali metals are not so efficient
in suppressing secondary flame as potassium salts. Fairly numerous experiments
were carried out to clear up whether or not known antiknock substances, such
as tetraethyl lead or nickel carbonyl prevent the development of a secondary flame.
They proved to have no effect on its development. In practice, two methods for re-
moving gun-flash may be employed, i.e. either a special flashless powder is produced,
containing nitroguanidine or DNT and a small admixture of potassium sulphate,

548 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

or special flash-reducing charges containing potassium salts are added to the charges
of smokeless powder. Potassium salts are often mixed with a weakly explosive
substance (e.g. DNT) to facilitate the dispersion of potassium salt in the mixture
of gases. The majority of the methods for removing flash, however, lead to an
increase in smoke. The best results, in this respect, are obtained by the addition
of nitroguanidine to the propellant. This either does not increase the smoke at all,
or increases it only to an insignificant extent (smoke formation will be discussed later).

The combustible ingredients in the products formed when a propellant charge
explodes give rise to backflash. This may arise when the breech-block of the gun
is opened due to the air-current which is then created in the barrel. This air-current
may be intensified by wind blowing in the direction opposite to the direction of
fire. Then ignition of the hot gases mixed with air often results from the smoulder-
ing residue in the barrel (smouldering remainder of the cartridge bag, glowing
soot). Sometimes the gases do not stop burning after a shot and continue to burn
after the opening of the breech-block. Backflash creates a great danger for the gun
crew since it may cause the ignition of the charges of powder prepared for subse-
quent shots or introduced into the barrel.

As with secondary flame, the addition of potassium salts to the powder prevents'
the development of backflash. In modern guns of heavy calibre the development
of backflash is prevented by blowing either air, or a stream of water through the
barrel, immediately after each shot.

SMOKE FORMATION

Nitrocellulose and nitroglycerine powders should properly be called "slightly
smoky"; the name "smokeless" is inexact. The smoke from nitrocellulose and
nitroglycerine powders is composed chiefly of water vapour. Shots from small arms or
cannons of small calibre are slightly smoky or almost smokeless. Conversely guns of
heavy calibre often give a considerable amount of smoke. The presence of metal
torn off from inside the barrel and from the driving band in the products of com-
bustion of the propellant is a partial cause of smoke.

It has been observed that the majority of remedies for preventing the development
of flash lead to an increase of smoke (e.g. potassium salts give white smoke, aromatic
nitro compounds give black-grey smoke due to the presence of unburnt carbon).
Nitroguanidine is the only additive that does not appreciably increase smokiness.
The burning of blackpowder in the primer produces an insignificant amount of smoke.

EROSIVENESS OF SMOKELESS POWDER

Every powder produces erosion or bore wear to some extent. After a large number
of shots the wear may be considerable, especially in large guns, and this reduces
their accuracy. The erosiveness of a given powder depends, first of all, on its flame

SMOKELESS POWDER

549

temperature. Nitroglycerine powders cause particularly severe erosion due to their
high heat of explosion; they have a higher flame temperature than nitrocellulose
powders. Attempts to reduce the heat of explosion of nitroglycerine powder by the
addition of explosively inert substances (vaseline, centralite) or active substances
of a lower heat of explosion (DNT) were primarily intended to decrease erosion.
Vieille [11] conducted extensive experiments on erosion in a manometric bomb,
closed except for a small orifice, about 1.3 mm dia. in a metal plug (Fig. 192). Hot
gases escaping through the orifice, eroded it to an extent determined by weighing
the plug before and after each experiment: the loss of weight was taken as a measure

Fig. 192. Erosion bomb of Vieille [11].

of the erosiveness of the powder. Vieille showed that pressure lower than 1000 kg/
cm 2 had little effect on erosion. Erosion increased with pressure over the range
from 1000 to 2000 kg/cm 2 after which a further increase of pressure from 2000 to
4000 kg/cm 2 had no effect.

The erosion of metals varied depending largely on the melting point. For various
metals erosion expressed in mm' of metal removed from the orifice per 1 g of powder
was:

Platinum 59 mm'

Platinum-iridium 74 mm 3

Brass 326 mm 3

Zinc 1018 mm 3

Vieille considered that erosion is caused by fusion of the metal and subsequent
expulsion of the molten substance. The results of Vieille's experiments are summar-
ized in Table 170. The weak erosive action of nitroguanidine, attributable to its low
temperature of explosion, is noteworthy.

Apart from physical agents such as temperature and the mechanical action of
gases, chemical agents also cause the erosion. Monni [50] noticed that the erosive-
ness of smokeless powder decreases with the addition of charcoal, probably because
the additional quantity of carbon thus introduced serves to carbonize the steel
which may undergo decarbonization under the influence of C0 2 at a high tem-
perature, according to the reaction:

C0 2 + C^2CO

550 chemistry and .technology of explosives

Table 170
Erosiveness of various powders according to Vieille [11]

Type of powder

Charge

Pressure

kg/cm 2

Calculated

temperature

°C

Erosion

mm 3 /g

Nitrocellulose powders

Pistol T

3.55

2500

2675

7.4

Rifle BF

3.55

2200

2675

6.4

Nitroglycerine powders

Ballistite 50% NG

3.55

2360-2540

3385

24.3

Cordite 38% NG

3.55

2500

18.1

Sporting (78% KN0 3 )

8.88

1960

3530

4.5

Rifle (75% KNOj)

10.00

2165

2910

2.2

Various explosives

Blasting gelatine (94% nitroglycerine)

3.35

2460

3545

31.4

Nitroguanidine

3.90

2020

970

2.3

The decarbonization of steel increases the porosity of the metal. The adsorption
of gases in the pores may intensify the erosive action; since the gases enter the pores
at high pressures and temperatures and then expand, they "blow up" the pores.
The following factors reduce erosion :

(1) Low maximum pressure ;

(2) Low temperature of explosion ;

(3) Uniform ignition of powder, uniform burning and complete combustion ;

(4) The largest possible content of H 2 and the smallest possible content of C0 2
and CO in the explosion products.

To minimize erosion it is desirable to use powders which give a heat of explosion
of about 700 kcal/kg and a temperature of about 2100°C. The effect of the heat of
explosion on the lifetime of the barrel has been dealt with earlier (p. 531).

STABILITY OF SMOKELESS POWDER

Soon after the manufacture of nitrocellulose smokeless powder began it was
established that the powder obtained by the partial dissolution of nitrocellulose
in a mixture of alcohol and ether (partly colloidal powder) has a chemical stability
inferior to that of the nitrocellulose from which it derived. Thus Vieille [11] reports
that on heating to a temperature of 110°C CPi guncotton undergoes denitration
with the evolution of 0.04 cm' NO/hr/gramme whereas the powder obtained from
these substances without a stabilizer undergoes denitration at more than twice the
rate, namely 0.10-0.15 cm 3 NO/hr/gramme of substance.

Originally, the cause of this phenomenon was unknown. In all the countries
producing smokeless powder methods for improving the stability of the powder
were sought. Similar research was also initiated by Vieille but in spite of systematic

SMOKELESS POWDER 551

experiments he arrived at the wrong answer. Vieille examined powders which had
been stored for a long time and in which it was assumed that decomposition had
started. He noticed that these powders contained less residual solvent than fresh
powders, i.e. freshly manufactured powders contained approximately 1 % of re-
sidual solvent (mainly less volatile ethyl alcohol) whereas the old powders contained
considerably less than 1 %. He therefore concluded than the stability of powder
is lowered by the loss of residual solvent and its stability may be improved by adding
a less volatile substance, chemically similar to the solvent. Indeed an improvement in
stability was achieved (1896) by adding 2% amyl alcohol (the proportion being referred
to the dry weight of the nitrocellulose). Ten years later, however (1906) it was noticed
that these powders showed signs of decomposition, so the amount of alcohol added
was increased to 8%. These powders were marked with the letters AM and a number
indicating the content of amyl alcohol, thus AM 2, AM 8.

In 1905, on the Japanese battleship "Mikasa" some English nitroglycerine
powder containing a certain amount of mercuric chloride, exploded. This substance
interferes with the heat test (Vol. II, p. 23) and makes it impossible to detect the
begining of decomposition in the powder. (Mercuric chloride inhibits the darkening
of the potassium iodide-starch test paper in the presence of N 2 4 .) Next, in 1907,
a disaster occurred which aroused public protest all over the world, especially in
France. This was the explosion of the ammunition store on the French battleship
"Jena". The investigation into the cause of this explosion was still in progress when,
in 1911, another similar catastrophe occurred when some powder manufactured
in 1906 containing 8% amyl alcohol blew up on the battleship "Liberie". It became
plain that the decomposition of smokeless powder during storage was caused by
some process other than the volatilization of residual solvent (Buisson [51]).

The accidents described above showed that nitrocellulose cannot be considered
wholly safe even when carefully purified, if it is partly or wholly colloidal in form.

Subsequent investigations have shown that the stability of powder at tempera-
tures higher than room temperature depends upon many factors. Thus, an examina-
tion of the effect of the web thickness of flakes or tubes upon the stability of powder
at temperatures of 75-80°C showed that the larger the web thickness the higher
the stability. This relationship is shown in Brunswig's [17] diagram (Fig. 193). Curve
I shows losses in weight (in %) of the fine tubular powder (for a 3.7 cm cannon),
curve 77 refers to a somewhat thicker powder (for a 7.5 cm gun), curve 777 refers to
a thicker powder (for a 15 cm gun), curve IV refers to a very thick powder (for
a 20 cm gun), curve V to the thickest powder of all (for a 30 cm gun). A loss of
25% weight occurred in the finest powder after 2\ months. In the thicker powders
it occurred after about 4, 9, lOf and 12f months, respectively.

The lower stability of smokeless powder in comparison with that of nitrocellulose
is accounted for by its content of residual solvent and of the oxidation products
of this solvent. Since in a finer powder the ratio of the surface to weight is high,
the oxidation processes are more intense. A larger amount of decomposition
products of residual solvent is formed by oxidation, and their destructive effect

552

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

on nitric esters (nitrocellulose, nitroglycerine) in the powder is more severe.

The following findings point to the harmful effect of air on the stability of "green"
nitrocellulose powder, i.e. freshly pressed and containing a considerable quantity
of solvent (alcohol and ether) (Table 171).

*J0

%48

5 36
c

Z>24

8 12V

3 6 9 12 15 18 21 24 27 30
Storage time, months

Fig. 193. Stability of tubular nitrocellulose powder at 75-80°C as a function of the web
thickness of the tubes, according to Brunswig [17].

— i r

1 1 —

— ' ' ' i

m/iy

y^^

- /

- / /

-//

...J 1 1 I

Table 171
The effect of air on the stability of nitrocellulose powder

Stability, min

Method of drying

at a temperature of

the powder

110°C

(for reddening of

litmus paper)

135°C
(for appearance of
first brown fumes)

Dried in the air

285

5 days in absence of air

570

110

7.5 days in absence of air

825

180

10 days in absence of air, then in

the

open

1020

190

When the powder is freed from most of the solvent in absence of air, subsequent
drying in the open is not detrimental, since the amount of solvent is too insignificant
to form oxidation products in a quantity which might impair stability.

The experiments of Swietoslawski, T. Urbariski, Calus and Rosinski [52] showed
that "green" flake powder containing about 15% solvent emits a certain amount
of heat which is very small but which may be detected in 3wietoslawski's [53] micro-
calorimeter. The heat effect fades after the green powder has been stored for a
certain time in a sealed calorimetric vessel but reappears if oxygen is added, hence
it may be assumed that this heat effect is caused by oxidation reactions between the
residual and atmospheric oxygen. A powder containing very little solvent (soaked
and dried) gives no such heat effect.

This means that the presence of a substance of very low volatility such as amyl
alcohol may under certain circumstances prove detrimental to the powder's stability.

SMOKELESS POWDER . 553

Indeed, it was found that amyl alcohol is converted into amyl nitrite and nitrate by
the action of nitric oxides resulting from the decomposition of nitrocellulose. These
substances are then oxidized to form valeric acid and amyl valerate products distin-
guishable by their characteristic smell.

There are many factors which affect the decomposition of powder during storage,
the most important of which is temperature.

The decomposition of powder at an elevated temperature does not differ greatly
from that of the nitric esters themselves, i.e. nitrocellulose and nitroglycerine (the
decomposition of nitrocellulose at various temperatures was discussed earlier in Vol.
II, p. 310). The higher the temperature, the more actively the decomposition of the
powder proceeds, with total loss of nitrogen, as NO and N0 2 , and carbon as CO
and C0 2 . Hydrogen is evolved chiefly as water, the amount of water decreasing
with increase in the temperature of decomposition of the powder (Sapozhnikov's
investigations).

The course of the decomposition reaction is different when the powder is stored
at or just above room temperature. A marked oxidation then occurs. It is chiefly
an internal oxidation with the ON0 2 groups acting as an oxidizer, but also involves
atmospheric oxygen, and the residual solvent.

Powder for use in a tropical climate requires higher stability than that in a tem-
perate climate. Powder for the navy must also have a high stability, since it may
be transported into a tropical zone.

Atmospheric humidity has a deleterious effect on the stability of powder. Storm
[54] reports that a good nitrocellulose powder, which withstood heating at a tem-
perature of 65.5°C for 400 days without marked decomposition, showed evident
decomposition in 175 days when stored at the same temperature in an atmosphere
saturated with water vapour. Powder which passed the first test subsequently with-
stood heating for 5 hr at a temperature of 135°C without explosion, whereas powder
from the second test exploded at the same temperature after 10 min.

The ease of decomposition by moisture is particularly pronounced in nitrocellu-
lose powders, which are hygroscopic but is much less in those with a lower hygro-
scopicity i.e. in nitroglycerine powders. According to Brunswig [17] nitroglycerine
tubular powder lost the same amount of weight at a temperature of 46°C in both
a dry and a damp atmosphere, when stored for a period of 6 months. Nitroglycerine
powder of the cordite type, containing vaseline, i.e. a moisture-proofing agent,
is also noteworthy, its stability in a damp atmosphere being the same as that in a
dry one. In Brunswig's opinion the addition of 3% vaseline to nitroglycerine pow-
der protects it against the harmful effect of moisture. This powder when stored
in a damp atmosphere, at a temperature of 46°C showed no loss of weight over
6 months, while common nitrocellulose powder, stored under the same conditions,
for comparison, showed a loss of weight of 7.1%. The nitrocellulose powder with
added vaseline is however unacceptable ballistically, so that proofing of this kind
has no practical significance for nitrocellulose powders.

Nitric oxides exert a very detrimental effect upon the stability of powder, indue-

554

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ing rapid denitration. Vieille [11] gives the following figures for nitrocellulose powder

maintained at a temperature of 40°C, in an atmosphere containing nitrogen dioxide

(Table 172).

Table 172

The effect of nitric oxides on the stability of
nitrocellulose powder

Duration

of test,

days

Powder without
stabilizer

Powder with an

addition of 7%

amyl alcohol

Powder with an
addition of 1%
diphenylamine

Nitrogen content of powder (%)

12.10

11.05

10.68

9.18

12.40
11.0
9.85
9.40

12.50

11.20

9.60

9.41

On the basis of extensive investigations into the stability of powder de Bruin
and de Pauw [55] proved that the gaseous products of the decomposition of smokeless
powder, in particular nitric oxides and water, increase the rate of the decomposition
of powder at a temperature of 110°C. This is confirmed by the fact that when ab-
sorbents that take up the gaseous products are introduced into the test room, the
duration of the test is extended. Figure 194 shows a number of curves for the stability
of nitrocellulose powder at a temperature of 110°C in the presence of various sub-
stances absorbing the gaseous products of decomposition. The sample of powder
tested without an absorbent decomposed most rapidly (curve VIII). The substances
which inhibit the decomposition of powder most efficiently are : phosphoric anhy-
dride (combining with water) (curve /), calcium oxide (combining with water and N0 2 )
(curve //), carbamite (centralite) (combining with N0 2 ) (curve ///), activated carbon
(absorbing water and N0 2 ) (curve IV). Sodium carbonate (combining with water
and N0 2 but evolving C0 2 ) (curve V), vaseline, absorbing nitric oxides (curve
VI) and anhydrous cupric sulphate combining with water (curve VII) were less
efficient. Thus, when the decomposition of powder involved the isolation of nitric
oxides its denitration occurred rapidly both in powders with and without a stabilizer.

Other acid substances, e.g. hydrogen chloride or vapours of sulphuric acid,
affect the powder in a similar deleterious way.

If a powder which decomposes and forms acid products is mixed with a "healthy"
powder it causes the latter to decompose. The products of decompostion of smoke-
less powder were found to contain formic acid, hydroxypyruvic acid CH 2 OH-
•COCOOH, hydroxyisobutyric acid (CH 3 ) 2 C(OH)COOH and oxalic acid. All
these acids except oxalic acid are hygroscopic, hence by increasing the moisture
content in the powder, they hasten its decomposition. Acid products react with
alkaline stabilizing components.

Brunswig [17] describes the following experiment. 30 g of flake powder was
poured into a flat glass vessel. In the middle of this charge was placed a piece of

SMOKELESS POWDER

555

tubular powder in a state of decomposition, containing acid products. The vessel was
covered with a watchglass and put into an atmosphere saturated with water vapour.
In a few days the tubular powder had changed into a greasy substance. In a few
weeks the flake powder gradually underwent a similar transformation with pro-
nounced decomposition progressing from the tubular powder in concentric rings. The
flake powder became soft so that it could be rubbed through the fingers.

It is a practice in some countries to add a small amount of sodium hydrogen
carbonate to the powder dough, to neutralize dinitrogen tetroxide evolved during
the decomposition of the powder. This would lead to formation of sodium nitrate

160

u 140

1 120

u 100

e 80
h

■£ 60

^ 40

I 20

150hr

Fig. 194. Stability of nitrocellulose powder
at 1 1 °C (measured as a reduction of weight)
in presence of substances absorbing gaseous
products of decomposition (H2O, NO2).

12 3 4 5 6 7
Layers (-^towards inside)

Fig. 195. Quantity of sodium nitrite in

various layers of Italian powder (Polvere C)

at various times of manufacture [57].

and nitrite. Angeli and Jolles [56] established a colorimetric method for deter-
mining the quantity of sodium nitrite in powder (by extracting it with water, diazo-
tizing aniline and coupling with dimethylaniline). They suggested it as a method for
estimating the degree of decomposition of the powder. The quantity of nitrites
(calculated as nitrous acid) varied form to 145 mg in 100 g of the powder.

Jolles and Socci [57] also examined the distribution of sodium nitrite in various
layers of an Italian tubular double base powder (Polvere C). According to their
estimation the distribution of sodium nitrite along the tubes is practically the same
whether on the surface or in an internal layer. The distribution varies from the surface
to the inside of the tube. In "young" powders the quantity of nitrite on the surface
is higher than inside. With time this may be reversed and in "old" powders the
internal layers may be richer in nitrite. This is shown in the diagram (Fig. 195)

556 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

which gives the results of an examination of "young" (e.g. 1931) and "old" (e.g. 1915)
powders.

Decomposing powder which on heating emits visible nitric oxides (N 2 4 ?±2N0 2 )
also emits a small amount of heat at room temperature. This was ascertained (Swieto-
slawski, T. Urbanski, Calus and Rosinski [52]) in Swietoslawski's microcalorimeter.
Immediately before this stage of decomposition, however, powder which has not
yet started to emit nitric oxides abundantly shows no positive heat effect at room
temperature.

These results do not conform to those of the "silvered vessel test". This test
determines the time necessary to produce pronounced exothermic reaction in a
sample of powder heated to a temperature of 80°C. This reaction causes a rise of
temperature by 2°C above the ambient temperature. The relatively high tempera-
ture (80°C) may account for the different results. It may well be that the exothermic
reaction recorded here is a reaction between the acid products of the decomposition
of the powder and the basic ingredients of the glass of the vessel at the relatively '
high temperature.

A good powder should withstand this test for at least 500 hr. Immediately
before the rise of temperature up to 82°C, or just after this rise, nitric and nitrous
acids are evolved from the powder. Sometimes, shortly after a temperature of 82°C
has been achieved, explosion follows.

On heating nitroglycerine powder of a ballistite type to a temperature of 95°C in
a Dewar vessel without a stabilizer, de Bruin [58] obtained the following data on
temperature rise :

after a lapse of 48 hr 95°C

after a lapse of 72 hr 95.2°C

after a lapse of 96 hr 95.2°C

after a lapse of 108 hr 96.4°C

after a lapse of 119 hr 97.9°C

after a lapse of 120 hr 98.6°C

after a lapse of 120^ hr 102.6°C

after a lapse of 120f hr explosion

When the powder used for this test was mixed with Centralite II, the tempera-
ture first rose to 97.2°C during 26 days of heating (an increment by 2.2°C) followed
by a gradual drop.

Strong bases have an adverse effect on the stability of smokeless powder as
described above. Moreover, Angeli [59] found that pyridine and its homologues
cause decomposition of nitrocellulose. (On the action of pyridine on other nitric
esters see Vol. II.) At an elevated temperature (e.g. 110°C) pyridine can produce
an intense denitration of esters which may even lead to an explosion.

After World War I the influence of sea water on the stability of smokeless powder
was examined. It was found that nitrocellulose powder submerged in the sea during
military activities did not suffer any perceptible deterioration as a result of immersion
in sea water for several years, neither in its colloidal properties nor in its stability.

SMOKELESS POWDER 557

In all probability this may be partly accounted for by the low temperature of sea
water at a certain depth.

Sunlight is a factor which hastens the decomposition of smokeless powder.
Experiments carried out by D. Berthelot and Gaudechon [60] showed that powders
containing various types of stabilizers behave in different ways towards light. E.g.
powder stabilized with amyl alcohol proved to be more resistant to sunlight than
that containing diphenylamine. The latter darkens very rapidly under the influence
of light, which no doubt accelerates the decomposition of diphenylamine.

Whatever its behaviour in sunlight, powder should not be exposed to direct
sunshine at any stage of manufacture; all the windows of the plant should face
north and any that do not should be covered with a layer of blue or yellow varnish,
to cut off the rays of shorter wavelength.

STABILITY TESTS

The majority of the stability tests for smokeless powder are much the same as
the methods used to determine the stability of nitric esters, in particular nitro-
cellulose (Vol. II). They are based on heating samples of the powder, thus starting
decomposition processes or hastening processes already initiated within the pow-
der. The value of such methods is comparative, since at an elevated temperature
different reactions occur than those which would arise under normal conditions
of storage. Nevertheless experiments over many years have shown that certain
interrelations may be established for the stability of powder at various tempera-
tures. Vieille [61] reports that the heating of a sample of powder for 1 hr at a tem-
perature of 110°C involves approximately the same decomposition as:

24 hr of heating at a temperature of 75°C
7 days of heating at a temperature of 60°C
30 days of heating at a temperature of 40°C

In addition to the testing methods common to nitric esters (nitrocellulose, nitro-
glycerine) and smokeless powder there are also methods used exclusively for testing
the stability of smokeless powders.

One of the inspection methods employed in smokeless powder factories consists
in taking samples (about 500 g) from each lot of powder and placing them in hermet-
ically closed jars. The jars are kept in thermostatic premises where a temperature
of 30-50°C is maintained. A methyl violet reagent-paper (a paper tinted with
crystal violet and rosaniline) is placed in each jar above the surface of the powder.
If oxides of nitrogen are evolved by the powder, the paper gradually loses colour.
A change in the colour of paper is therefore a sign that particular attention should
be paid to this lot of powder, which should be subjected to detailed testing. This
method may also be used in powder magazines.

Another method of checking stability consists of placing the powder into boxes
of special design connected with an exhaust valve and a tube that leads any gases

558 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

formed inside the box to a vessel containing a solution of potassium iodide and
starch. Any traces of nitrogen dioxide given off by the powder are detected immediate-
ly since they produce a coloration of the solution.

In addition to the above methods which are designed to detect active decomposition
periodical sampling and stability testing by one of the known methods is absolutely
indispensable. The simplest of these methods are :

(1) The Abel heat test (heating at a temperature of 75-80°C in the presence
of standard potassium iodide-starch paper).

(2) The stability test for nitrocellulose powder at a temperature of 134.5°C; the
sample will not be considered serviceable unless the time necessary to produce the
evolution of nitrogen oxides is at least 45 min and the powder withstands this tem-
perature without exploding for 5 hr. The test can be combined with methyl violet
test : decoloration of the test paper should not occur before 30 min or longer accord-
ing to particular specifications.

(3) The test for nitroglycerine powders at a temperature of 120°C (nitroglycerine
powders cannot withstand higher temperatures); here the same conditions apply
as in the test for nitrocellulose powders, viz. appearance of the nitrogen oxides
after a lapse of at least 45 min and no explosion for 5 hr. The same test can be carried
out in the presence of methyl violet test papers. Decoloration of the paper should
not occur before 30 min.

(4) The reduced heating test at a temperature of 110°C in the presence of a
blue litmus paper, which should not redden in less than 10 hr.

(5) The Vieille test at a temperature of 110°C (Vieille [61]). For this a sample
of powder is heated daily for 10 hr or until the litmus paper has assumed a standard
red tint. The sample is then aerated for 14 hr and the procedure is repeated each
day until the litmus paper reddens in one hour or less. The course of the test can
be seen from a following example :

1st day 10 hrj together the first reddening

2nd day 9 hr] after 19 hr

3rd day 7.5 hr

4th day 7.5 hr

«th day 1 hr (or less)

Together x hr

The value of x for good powders should not be less than 70 hr.

Quantitative tests are rarely conducted in magazines and for production inspec-
tion purposes except for a test devised by Bergmann and Junk [62] in which the
quantity of acid products (calculated as NO) ^evolved by the powder is determined
by titration. The quantity of NO evolved on heating for 2 hr at a temperature of
132°C should not exceed 2.5 cm' NO per 1 g of powder. (For more details see Vol.
II p. 26.) Other quantitative tests are usually employed in research.

According to Jolles et al. [56, 57] the colorimetric determination of the
quantity of sodium nitrite in powder (p. 555) can be an auxiliary method of esti-
mating the stability of powders. Results agree well with the results of the heat test.

SMOKELESS POWDER

559

Tonegutti and Debenedetti [63] found that this test also agrees well with the me-
thyl violet test at 120°C.

Jolles and Socci [57] give the following figures confirming the findings of Tone-
gutti and Debenedetti (Table 173).

Table 173

Comparison of sodium nitrite content and the methyl
violet test of italian double base powder (polvere c)

Sample

Date of
manufacture

1915
1919
1924
1929
1931

NaN0 2
in mg %

Decoloration

of methyl violet

test paper at 120°C

after min

150.0

116.0

80.0

16.0

7.0

50-55

70-75

100-105

105-110

STABILIZATION OF SMOKELESS POWDER

STABILIZATION WITH DIPHENYLAMINE

An important advance in the stabilization of nitrocellulose powder was the
addition of diphenylamine to the powder cake, suggested by Alfred Nobel in 1889
[64] and introduced into practice in Germany. According to Gorst [65] diphenyl-
amine was also used in Russia at the end of the nineteenth century on the suggestion
of Nikolskii. The application of diphenylamine to the powder in Germany was
kept a profound secret, but as early as 1896 it was known in France that German
nitrocellulose powder contained 2% diphenylamine. However it was believed in
France that diphenylamine was too basic, and liable to hydrolyse nitrocellulose.
Nevertheless, in view of the disaster on the batHeship Liberte, the use of diphenyl-
amine as a stabilizer for nitrocellulose powder was approved in 1911. Comparison
of the stability of powders without stabilizer, with amyl alcohol and with diphenyl-
amine gave inter alia the following results :

At a temperature of 75°C in a dry atmosphere, a powder with an admixture
of 2% diphenylamine gave off only about one quarter the amount of gases evolved
by a powder with an admixture of 8% amyl alcohol heated for the same period
of time. At a temperature of 1 10°C the stability of the powder containing diphenyl-
amine proved to be 2.5 times greater than that containing amyl alcohol.

A powder with an admixture of 1.5% diphenylamine, when heated at a tem-
perature of 75 C C in a dry atmosphere showed signs of decomposition only after
512 days, whereas powder containing 2% amyl alcohol began to decompose after
122 days. In a damp atmosphere at a temperature of 75°C a powder containing

560

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

2% diphenylamine decomposed much more slowly (at least four times more slowly)
than the powder containing 8 % amyl alcohol.

Berger [66] defined the degree of decomposition of powder at temperatures of
40-1 10°C, by determining the amounts of heat emitted by the fresh powder on
burning and by partly decomposed powder. From this difference he calculated the
amount of heat emitted on decomposition. Berger thus examined the behaviour
of powder without a stabilizer (AT), powder with amyl alcohol (AM) and D powder
with diphenylamine; he obtained curves (Fig. 196), the shapes of which are charac-

Time of heating

Fig. 196. Effect of the addition of amyl alcohol and diphenylamine on the stability of
nitrocellulose powder, according to Berger [66],

teristic for the transition from slow to fast decomposition of powder. In the author's
opinion violent decomposition coincides with the sharp inflection in the curve that
occurs after an initial period characteristic of a resistance to decomposition that
varies with the type of stabilizer added.

Investigations have shown that the basic properties of diphenylamine are so
weak that it cannot hydrolyse nitrocellulose, but they are sufficiently strong to
neutralize any acid product arising either from the decomposition of impurities in
the nitrocellulose, from the oxidation of residual solvent or even from decomposi-
tion of the nitrocellulose itself. It was also demonstrated that the basic properties
of diphenylamine may have a deleterious effect on the powder if the diphenylamine
content exceeds 5%. The best stabilizing results are achieved by using 1.0-2.5%
diphenylamine.

The results of extensive investigations into the influence of the content of diphenyl-
amine and other stabilizers on the stability of powder are tabulated below.

More recently Demougin and Landon [67] examined the stability of nitro-
cellulose powder containing 1.02-7.8% diphenylamine at a temperature of 110°C.
After 160 hr of heating they determined the nitrogen content in nitrocellulose iso-
lated from powder (Table 174.). The initial content of diphenylamine in the sample
was 7.8%; on heating for 180 hr at a temperature of 110°C it was reduced to 1%.

SMOKELESS POWDER 555

Table 174

The influence of diphenylamine content on the

stability of nitrocellulose powder

Time of heating at

a temperature of

100°C

Content of diphenylamine

1.02 2.2 3.75 | 7.8

Content of nitrogen in nitrocellulose,

hr (fresh powder)
160 hr

12.82
11.89

12.89
13.01

12.76
11.14

12.58
10.13

A large content of diphenylamine may be particularly detrimental to the ballistic
properties in fine-grain (rifle) powders. That is why 0.5-1.0% diphenylamine is
used in these powders whereas in slower burning cannon powders 1.5-2.0% is used.

Diphenylamine is not used for stabilizing powders containing nitroglycerine
since it hydrolyses this ester. Diphenylamine also causes the decomposition of
higher nitrated aromatic compounds and therefore should not be used in powders
containing such compounds.

In wartime the content of diphenylamine was reduced to 0.5% and even to
0.25% in expectation of the rapid utilization of the powders. Such powders should
be labelled, e.g. by the addition of red dye, so that after hostilities they can be
re-checked and used up quickly or destroyed. After World War I the storage of
such powders caused a number of catastrophes; e.g. in Poland there were explosions
of the magazines in the Warsaw citadel (1924) and at Witkowice (1927) and in
France at Bergerac (1928).

Diphenylamine behaves not only as a stabilizer, but also as an indicator of the
oxidation and decomposition processes which occur in powders. It was noticed
long ago that powders containing diphenylamine assume under certain conditions
colours ranging from greenish or bluish shades to dark blue or almost black and
sometimes to yellow or brown, i.e. :

(1) The powder turns dark blue if it contains a large quantity of solvent and
if it is exposed to the action of hot air, e.g. on drying at a temperature of 50-60°C.
The investigations of Desmaroux [68], Marqueyrol and Muraour [69] and of Mar-
queyrol and Loriette [70] showed that this is due to the oxidation of diphenylamine
produced by peroxides formed from residual ether and atmospheric oxygen.

(2) The powder turns blue if it contains traces of metals or if it is in contact
with such metals as iron, copper, zinc, etc. Traces of metals present in powder
sometimes cause the formation of blue stains around the metallic particles. A blue
coloration appears sometimes at the points where the powder is in contact with
metal on the inner surface of cases lined with metal.

Dark coloration of the powder need not necessarily signify decomposition, but
it proves that some of the diphenylamine has undergone changes and has been
consumed to form new products. This might reduce the stability of the powder
and, as a rule, it is no longer serviceable for military purposes.

562

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Much research has been devoted to finding the mechanism of the changes which
diphenylamine undergoes in powder. Marqueyrol and Loriette [70] report that
under the action of oxidizing agents, particularly of peroxide formed from ether,
diphenylamine undergoes the following reactions to produce tetraphenylhydrazine
(I) and diphenylphenazine (II):

2(C 6 H 5 ) 2 NH -> (C 6 H 5 ) 2 N— N(C 6 H 5 ) 2
I

C 6 H 5

S\/

In these authors' opinion phenazine (II) causes powders to darken.

Davis and Ashdown [71] report that diphenylamine in powder undergoes the
following reactions :

V_ NH— /~

N0 2

VIII

N-Nitrosodiphenylamine (III) which is formed first, has been proved experiment-
ally to be nearly as good a stabilizer as diphenylamine itself. Neither of the above
substances gives powders an intense coloration.

Davis and Ashdown devised methods for detecting these substances in powder.
If the alcoholic extract from powder contains diphenylamine it forms a blue colour
with ammonium persulphate; N-nitrosodiphenylamine gives no coloration with
this reagent. The intensity of the colour developed by a mixture of both substances
evidently depends upon the concentration of diphenylamine. On the other hand
N-nitrosodiphenylamine gives an intense blue coloration to concentrated sulphuric
acid.

Another test for N-nitrosodiphenylamine is based on the treatment of an
alcoholic solution of this substance with a 1 % alcoholic solution of a-naphthylamine
followed by heating, when it turns orange.

•k SMOKELESS POWDER 563

With mineral acids, including nitric acid produced by the decomposition of powder,
the N-nitroso compound undergoes rearrangement to form />-nitrosodiphenylamine
(IV). This substance is readily oxidized to />-nitrodiphenylamine (V). Furthermore,
the higher nitrated products, i.e. dinitro derivatives (VI) and (VII) and trinitro
derivatives (VIII) may be formed in powder. Davis and Ashdown isolated 2,4,4'-tri-
nitrodiphenylamine from American pyrocollodion powder by heating a sample in a
closed vessel for 240 days at a temperature of 65°C. At the end of the heating period
brown nitric oxides were given off by the powder.

It is generally agreed that powder in which diphenylamine is completely converted
into N-nitrosodiphenylamine is suitable for storage. However, on the disappearance
of N-nitrosodiphenylamine and its conversion into nitro compounds, the powder
should be considered unsuitable for storage since it then lacks a stabilizer.

Powder in which nitro derivatives of diphenylamine have been formed is coloured
reddish-yellow or brown.

According to Schroeder et al. [72] the chromatographic analysis on silica of the
conversion products of diphenylamine in nitrocellulose powder indicates that di-
phenylamine may be converted into hexanitrodiphenylamine. From \ to f of diphenyl-
amine is converted into nitro derivatives if the powder is kept at a temperature of
71°C for 258 days.

Stabilizers which had or still have practical application are classified into inorganic
and organic.

INORGANIC STABILIZERS

As early as 1867 Abel [73] realized that nitrocellulose tends to decompose in an
acid medium, and suggested that sodium carbonate should be added to it to neutralize
the acid products of the decomposition of the impurities in the powder or of nitro-
cellulose and nitroglycerine per se. However more than 2% sodium carbonate in the
powder proved detrimental— due to its strongly alkaline reaction it impairs the
stability of the powder.

Accordingly, attempts were made to use the less alkaline sodium hydrogen
carbonate (NaHC0 3 ). It was shown that 1 % of the latter has no marked effect, either
positive or negative, on the stability of nitrocellulose powder, but with nitroglycerine
powder containing vaseline, its influence is decidedly helpful. Brunswig [17] des-
cribes cases in which nitroglycerine powder with vaseline and NaHC0 3 withstood
storage for 20 years without any signs of decomposition whereas the same powder
without NaHC0 3 showed a pronounced decomposition, several times ending in
self-ignition after a lapse of 5 years.

Calcium carbonate, often added to nitrocellulose at the end of stabilization,
affects the stability of powder beneficially, its influence, however, becomes strongly
marked only when used in large quantities. Brunswig reports that powder con-
taining 0.1% of CaC0 3 withstands heating at a temperature of 94°C in a closed
vessel for 4.5 hr; nitric oxides are then given off and after 20 hr its loss of weight
amounts to 19.7%. If the same powder contains 6% of CaC0 3 the loss of weight on

564 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

heating at a temperature of 94°C for 200 hr does not exceed 0.4%. Nevertheless,
it is inadvisable to introduce large amounts of a non-explosive substance to the
powder mass lest its ballistic properties should be impaired. Magnesium carbonate
acts similarly to calcium carbonate.

Magnesium oxide was added to the newest German nitroglycerine powders used
during World War II. A content of 0.25% MgO considerably improved their stability
and, in addition, facilitated pressing the powder paste. In all probability this is the
most efficient inorganic stabilizer.

ORGANIC STABILIZERS

Apart from diphenylamine a number of other organic bases were tested for use
as stabilizers. Some of them, e.g. aniline, were used only temporarily, chiefly during
World War I when diphenylamine was in short supply. The basic properties of aniline
are too marked and this is detrimental to stability. On the other hand, relatively good
results have been obtained with carbazole which resembles diphenylamine in its
structure:

Nshf

Marqueyrol [74] reported experiments carried out in France over a period of
15 years to compare the efficiency of various stabilizers. The results are shown in Table
175. In addition to amyl alcohol and diphenylamine, the action of N-nitrosodiphenyl-
amine (diphenylnitrosamine), carbazole, diphenylbenzamide, nitronaphthalene and
naphthalene was also tested. The powder was stored at temperatures of 40, 60 or 75°C.
The experiments were stopped when the powder showed signs of intense decomposi-
tion, giving off nitric oxides. This was also manifested by a sudden faiyrf the nitrogen
content in nitrocellulose isolated from the samples.

Non-volatile solvents ("gelatinizers") such as camphor and butyl phthalate have
also a definite stabilizing effect on powder. This applies especially to those which
contain nitrogen: urea substitution derivatives (centralites or carbamites, acardit)
and urethane substitution derivatives (see p. 645). It was found that on decomposi-
tion of the powder the centralites are nitrated and their nitro derivatives are formed.
Nitro groups are obviously introduced into aromatic rings. Lecorche and Jovinet [75]
ascertained that the centralite (carbamite) in solventless nitroglycerine powder is
converted into substances volatile and non-volatile with steam.

The volatile fraction consists chiefly of p-nitrophenylethylnitrosamine (I) and
the non- volatile fraction of dinitrocentralite (II) :

<C2 H 5

.N<

° 2N \Z/ N ~" NO co C6H4N ° 2

C 2 H 5 \ /C 6 H 4 N0 2

x: 2 h 5
i ii

smokeless powder

Table 175
the efficiency of various organic stabilizers

565

Time (days) of storage at
a temperature of 40°C

387

843

1174

2991

3945

4016

Stabilizer

Nitrogen content in the nitrocellulose, %

No stabilizer
2% amyl alcohol
8% amyl alcohol
1 % diphenylamine
2% diphenylamine
5% diphenylamine

10% diphenylamine

12.63
12.65
12.60
12.60
12.48
12.52
12.52

12.59
12.45

12.48
12.43
12.46
12.48
12.40

12.40

9.25
12.55
12.57
12.58
12.46

12.46
12.44
12.57
12.47

~~ 1

12.36

10.81
12.40
12.58
12.52

Time (days) of storage at
a temperature of 60°C

146

295

347

1059

2267

3935

Stabilizer

Nitrogen content in the nitrocellulose, %

No stabilizer
2% amyl alcohol
8% amyl alcohol
1 % diphenylamine
2% diphenylamine
5% diphenylamine

10% diphenylamine

12.65
12.65
12.60
12.60
12.48
12.52
12.52

9.15
12.35
12.34
12.36
12.27

12.26
12.03

9.2
12.41
12.51
12.41

10.0

11.62
10.82

Time (days) of storage at
a temperature of 75°C

231

312

516

652

667

Stabilizer

Nitrogen content in the nitrocellulose, %

2% amyl alcohol
1 % diphenylamine
2% diphenylamine
5% diphenylamine
10% diphenylamine

12.71
12.60
12.48
12.52
12.52

11.97
12.26
12.18

12.06
11.52

12.39
12.40

11.94

_
12.02

11.65
11.00

Time (days) of storage at
a temperature of 75°C

146

312

419

493

—

Stabilizer

Nitrogen content in the nitrocellulose, %

1 % diphenylamine

2% diphenylamine

10% diphenylamine

12.54

12.38

12.41

12.46

12.36

12.40

12.51

12.61

12.41

12.42

12.36

10.73

12.61

12.22

12.15

12.07

11.93

11.72

12.14

11.53

566

chemistry and technology of explosives
Table 175(contd.)

Time (days) of storage at
a temperature of 75°C

108

197

377

633

Stabilizer

Nitrogen content in the nitrocellulose, %

2% amyl alcohol
1.25%carbazole
10% carbazole

12.57
12.55
12.53

12.44
12.47
12.43

12.31

12.46
12.37

11.44
12.40

12.07

11.90

Time (days) of storage at
a temperature of 75°C

227

556

Stabilizer

Nitrogen content in the nitrocellulose, %

1.5% diphenylbenzamide
10% diphenylbenzamide
1.5% mononitronaphthalene
10% mononitronaphthalene
1.5% naphthalene
10% naphthalene

12.52
12.52
12.66
12.64
12.66
12.63

12.47
12.40
12.50
12.40
12.52
12.46

11.65
12.19

12.07

12.53

12.46

12.52

Such substances as vaseline (added to nitroglycerine powders of the cordite type),

castor oil and rosin are also capable of stabilizing powder.

Brunswig [17] gives the following comparative figures based on the weight loss

Am
coefficient, i.e. the loss of the weight of powder in a unit of time — • , as a criterion of

stability. The values of the weight loss coefficient have been found for powder con-
sisting of 10 parts of guncotton and 8 parts of nitroglycerine, obtained with the use
of acetone as a solvent. They are summarized in Table 176.

Table 176

Type of stabilizer

Amount of
stabilizer

Am
A/

No stabilizer

0.77-1.05

Vaseline

0.8 pt.

0.31

Rosin

0.8 pt.

0.30

Centralite

0.8 pt.

0.28

The systematic studies of T. Urbafiski, Kwiatkowski and Miladowski [76] proved
that the addition of an aromatic nitro compound distinctly enhances the stability of
nitrocellulose and nitrocellulose powder. Thus, nitrocellulose containing 13.4% N
which on heating for 5 hr at 120°C had pH=2.28 showed pH=2.89 on addition of
9.1%^-nitrotoluene, pH = 3.17 on addition of 9.1% 2,4-dinitrotoluene and pH = 3.34
on addition of the same amount of a-trinitrotoluene. The same samples when heated
in a constant volume (Tagliani test) gave at 134.5°C a pressure of decomposition

SMOKELESS POWDER 567

products of 109 mm Hg after 32.5 min with pure nitrocellulose, after 44.5 min with
nitrocellulose and the addition of 9.1% ^-nitro toluene, after 48.5 min with nitro-
cellulose and the addition of 9.1% 2,4-dinitrotoluene and after 52.5 min with nitro-
cellulose and the addition of 9.1 % a-trinitrotoluene. The same nitro compounds do
not influence the stability of nitroglycerine.

The experiments summarized in Table 175 show the stabilizing action of nitro-
naphthalene. Dinitro- and trinitronaphthalene also act in a stabilizing manner. On
the other hand mixed nitramines such as tetryl are detrimental to stability.

APPARENT STABILIZERS

There are substances which appear to stabilize a powder by masking the results
of stability tests. One of them is mercuric chloride (corrosive sublimate). Sell [77]
originally suggested the addition of sublimate to nitrocellulose to prevent the devel-
opment of mould on moist nitrocellulose. A test of the purity and accuracy of the
stabilization of nitrocellulose containing stfblimate led to an unexpected result, i.e.
even the worst stabilized nitrocellulose gave no coloration to the standard test paper
in the Abel heat test. For a certain time sublimate was added to the powder mass in
an amount of 0.02-0.03% and to powder earmarked for tropical countries in an
amount of 0.05%. This greatly hindered research on genuine stabilizers. It was later
shown that mercuric chloride is partly reduced under the influence of nitrocellulose
and during the heat test mercury volatilizes (usually at 65-82°C) and combines on the
test paper with iodide formed by nitric oxides, to form colourless mercuric iodide.

Experiments showed that minute amounts of mercury vapour in the atmosphere
are sufficient to discolour a blue iodide-starch paper, e.g. on heating a mixture of
0.2 g of barium nitrate with 0.8 mg of mercury to a temperature of 80°C there is
immediate discoloration of the paper.

LITERATURE

1. C. F. SchOnbein, Sitzungsber. Naturforsch. Ges. Basel 7, 27 (1846).

2. J. H. Pelouze, Compt. rend. 23, 809, 837, 861, 892 (1846).

3. Lenk von Wolfsburg, according to S. J. Romocki, Geschichte der Explosivstoffe, Bd. II,
Oppenheim, Berlin, 1896.

4. E. Schultze, Deutsche Industrie-Ztg. 10, III, 1865; Dai neue chemische Schiesspuher, Berlin,
1865.

5. Hartig, Untersuchungen iiber den Bestand und die Wirkungen der explosiven Baumwolle, Braun-
schweig, 1874.

6. F. Volkmann, Austrian Pat. 21/208, 21/257 (1871); also according to O. Guttmann, Zwanzig
Jahre Fortschritte in Explosivstoffen, Berlin, 1909; Z. ges. Schiess- u. Sprengstoffw. 4, 16 (1909).

7. D. Spill, Brit. Pat. 1739 (1879).

8. W. F. Reid, Brit. Pat. 619 (1882).

9. W. Wolf and F. FOrster, Ger. Pat. 23808 (1883).

56g CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

10. M. von Duttenhofer, according to H. Brunswig, Das rauchlose Pulver, de Gruyter, Berlin-
Leipzig, 1926; Brit. Pat. 6022 (1887); 8776 (1902).

11. P. Vieille, Mem. poudres 3, 9, 177 (1890); 4, 256 (1891); 7, 19, 30 (1894); 11, 157 (1901);
15, 61 (1909-1910); Z. ges. Schiess- u. Sprengstoffw. 6, 181, 303, 327, 441, 464 (1911).

12. D. I. Mendeleyev, Sochinenya, Vol. 9, Izd. Akad. Nauk SSSR, Leningrad-Moskva, 1949;
Morskoy Sbornik 268, 38 (1895); 272, 39 (1896); Bull. Soc. d'encour. Sci. [4], 10, 1100 (1893);
Engineering 63, 180 (1897).

13. A. Nobel, Brit. Pat. 1471 (1888).

14. F. A. Abel and Dewar, Brit. Pat. 5614 (1889).

15. C. Claessen, Ger. Pat. 256572 (1910-1913).

16. H. Gallwitz, Die Geschiitzladung, Heereswaffenamt, 1944, according to Technical Report PB
925, U.S. Dept. of Commerce, Washington.

17. H. Brunswig, Das rauchlose Pulver, de Gruyter, Berlin-Leipzig, 1926.

18. O. Poppenberg and Stephan, Z. ges. Schiess- u. Sprengstoffw. 4, 281, 305, 388 (1909).

19. Andrew Noble, Proc. Roy. Soc. {London) A 76, 381, 512 (1905); 78, 218 (1906).

20. H. C. Knight and D. C. Walton, Ind. Eng. Chem. 18, 287 (1926).

21. According to M. E. Serebryakov, Vnutremtaya ballistika, Oborongiz, Moskva, 1962.

22. A. F. Belayev and Ya. B. Zeldovich, according to M. E. Serebryakov [21].

23. E. K. Rideal and A. J. B. Robertson, Illrd Symposium on Combustion, p. 536, Williams
& Wilkins, Baltimore, 1949.

24. J. D. Hufftngton, Nature 165, 840 (1950); Trans. Faraday Soc. 47, 864 (1951).

25. D. R. Cameron, in Encyclopedia of Chemical Technology, Ed. R. E. Kirk and D. F. Othmer,
Vol. 6, Interscience, New York, 1951.

26. R. Klein, M. Menster, H. v. Elbe and B. Lewis, /. Phys. Chem. 54, 877 (1950).

27. J. Taylor, Solid Propellants and Exothermic Compositions, Newnes, London, 1959.

28. H. Kast, Z. ges. Schiess- u. Sprengstoffw. 15, 196 (1920).

29. T. UrbaAski and T. Galas, Wiad. Techn. Uzbr. 34, 501 (1939); Z. ges. Schiess- u. Spreng-
stoffw. 34, 103 (1949).

30. L. Medard, Mem. artill. franc. 18, 277 (1939).

31. E. Burlot, Mem. artill. franc. 23, 183 (1949).

32. H. Nash, Army Ordnance 11, 293 (1931); Mem. artill. franc. 12, 765 (1931).

33. A. Langevin and P. Biquard, Mem. poudres 26, 255 (1934-1935).

34. F. W. Brown, D. J. Kusler and F. C. Gibson, U.S. Bureau of Mines Report 3852 (1946).

35. F. Morris, Engineering 49, 73 (1947).

36. Comit6 Scientifique des Poudres.

37. J. Roszkowski, Z. physik. Chem. 7, 485 (1891).

38. H. F. Coward and F. J. Hartwell, Safety in Mines Research Board, London 19 (1926).

39. H. Dautriche, Compt. rend. 146, 535 (1908).

40. 3. Fauveau and le Paire, Mem. poudres 25, 142 (1932-1933).

41. M. Prettre, Mem. poudres 25, 160, 169, 531 (1932-1933).

42. R. N. Pease, J. Am. Chem. Soc. 52, 5106 (1930).

43. K. F. Bonhoeffer and H. Reichardt, Z. physik. Chem. A139, 75 (1928).

44. L. J. Avramenko and V. N. Kondratyev, Zh. eksp. teor.fiz. 7, 842 (1937).

45. V. N. Kondratyev, Kinetika Khimicheskikh gazovykh reaktsii, Izd. Akad. Nauk SSSR,
Moskva, 1958.

46. J. Dwyer and O. Oldenberg, /. Chem. Phys. 12, 351 (1944).

47. R. G. W. Norrish and G. Porter, Nature 164, 658 (1949); Proc. Roy. Soc. {London) A210,
439 (1952).

48. R. G. W. Norrish, G. Porter and B. A. Trush, Proc. Roy. Soc. {London) A216, 165 (1953);
A227, 423 (1955).

SMOKELESS POWDER 569

49. R. G. W. Norrish, XVI Congres de Chimie Pure et Appliqud, Paris, 1957, Experientia Suppl.
VII, 87 (1957).

50. According to V. Recci, Z. ges. Schiess- u. Sprengstoffw. 1, 285 (1906).

51. A. Buisson, Le probleme des poudres, Dunod & Pinat, Paris, 1913.

52. W. Swietoslawski, T. Urbanski, H. Calus and M. Rosinski, Roczniki Chem. 17, 444
(1937).

53. W. Swietoslawski and J. Salcewicz, Roczniki Chem. 14, 621 (1934).

54. C. Storm, Army Ordnance 9, 230 (1929).

55. G. de Bruin and P. F. M. de Pauw, N. V. Koninklijke Nederlandsche Springstoffenfabrieken
3, 4 (1926); 6 (1927); 8 (1928); 9 (1929).

56. A. Angeli and Z. E. Jolles, G. Chim. Ind. ed Appl. 14, 65 (1932).

57. Z. E. Jolles and M. Socci, G. Chim. Ind. ed Appl. 15, 113 (1933).

58. G. de Bruin, N.V. Koninklijke Nederlandsche Springstoffenfabrieken 5 (1927).

59. A. Angeli, Atti reale accad. Linzei, Roma 23, 20 (1919).

60. D. Berthelot and Gaudechon, Compt. rend. 153, 1220 (1911).

61. P. Vieille, Mem. poudres 5, 81 (1892).

62. E. Bergmann and A. Junk, Z. angew. Chem. 17, 1022 (1904).

63. M. Tonegutti and B. Debenedetti, Ann. Chim. Applicata 22, 627 (1932).

64. A. Nobel, Ger. Pat. 51471 (1889).

65. A. G. Gorst, Porokha i vzryvchatyye veshchestva, Oborongiz, Moskva, 1949.

66. E. Berger, Bull. soc. chim. France [4], 11, 1 (1912).

67. P. Demougin and M. Landon, Mem. poudres 26, 273 (1934-1935).

68. J. Desmaroux, Mem. poudres 21, 238 (1924).

69. M. Marqueyrol and H. Muraour, ibid. 21, 259 (1924).

70. M. Marqueyrol and P. Loriette, Mem. poudres 21, 277 (1924).

71. T. L. Davis and Ashdown, Ind. Eng. Chem. 7, 674 (1915).

72. W. A. Schroeder, E. W. Malmberg, L. L. Fong, K. N. Trueblood, J. D. Landerl and
E. Hoerger, Ind. Eng. Chem. 41, 2818 (1949).

73. F. A. Abel, Trans. Roy. Soc. 157, 181 (1867).

74. M. Marquerol, Mem. poudres 23, 158 (1928).

75. H. Lecorche and P. L. Jovinet, Mem. poudres 23, 147 (1928); Compt. rend. 187, 1147 (1928).

76. T. UrbaNski, B. Kwiatkowski and W. Miladowski, Przemysl Chem. 19, 22 (1935); Z. ges.
Schiess- u. Sprengstoffw. 32, 1 (1937).

77. E. Sell, Arbeiten Raised. Gesundh. Amt (1888).

CHAPTER VIII

THE MANUFACTURE OF SMOKELESS POWDER

INTRODUCTION

The present chapter does not deal comprehensively with all the existing methods of
powder manufacture. The manufacturing processes for smokeless powders described
below should be regarded only as typical of some of the methods which were or still
are used but which are often liable to very considerable variation. It should be borne
in mind that some of the more recent methods are kept secret. The author has ac-
centuated, as far as possible, the differences existing between the methods adopted
in various countries.

Nitrocellulose powders can be classified into semicolloidal powders made of two
kinds of nitrocellulose (insoluble and soluble in the solvent— ether and alcohol), almost
fully colloidal, made of pyrocollodion cotton (highly soluble in ether-alcohol) and
Schultze type powders with a very low content of colloidal nitrocellulose and con-
taining inorganic salts. The Schultze nitrocellulose powders are now of little signifi-
cance and very little used, so they will be discussed only briefly.

Nitrocellulose powders completely gelatinized with such solvents as acetone,
ethyl acetate etc. were also known and manufactured for some time in various coun-
tries. They were of passing interest only because they possessed a number of dis-
advantages : the high cost of the solvents, the difficulty of igniting the powder and
a number of difficulties in the course of manufacture produced by the high
viscosity of the dough and its ready adhesion to metallic surface which made extrusion
or rolling difficult.

Ball-grain powder is an example of a modern powder approaching fully colloidal
structure through the use of ethyl acetate.

Another type of nitrocellulose powder used for some time which eventually
disappeared from the market was a semi-colloidal nitrocellulose powder containing
inorganic salts such as potassium or barium nitrate (e.g. Poudre T in France, with
2% potassium nitrate) or with dichromates (e.g. Poudre J in France with 14% am-
monium dichromate and 3% potassium dichromate). The dichromate powder was
very sensitive to friction and its dust contained toxic dichromates.

THE MANUFACTURE OF SMOKELESS POWDER 571

Nitroglycerine powders can be classified into two groups: with and without a
volatile solvent. Semi-colloidal nitrocellulose powders and nitroglycerine solventless
powders are the most important types of smokeless powder.

NITROCELLULOSE POWDER

NOMENCLATURE

The semi-colloidal powders are designated in every country according to the
purpose for which they are intended. In the U.S.S.R. (Gorst [1]) rifle powders were
formerly given the letter B (Russian V from the word "vintovka" i.e. rifle) followed
by another letter denoting the type of projectile to be used, thus BA (Russian VL)
for powder for light projectiles, BT (Russian VT), for powder for heavy projectiles.
Tubular and multiperf orated powders for field artillery are, in addition, designated
by a fraction, in which the numerator shows the web thickness in tenths of millimetres,
and the denominator the number of perforations. The letters TP (Russian TR)
following these figures mean that the powder in question is a tubular one. Thus y TP
denotes a single-perforated tubular powder with a web thickness of 1 mm. Similarly
■j TP denotes a multiperforated tabular powder (7 perforations) with a web thickness
of 0.9 mm.

Naval and coastal artillery powders are also designated by fractions, but for these
the numerator signifies the calibre of the gun in millimetres, while the denominator
shows the length of bore in calibres. Thus 75/50 signifies a powder for 75 mm guns,
50 calibres long.

The number of the lot of the powder, and the date and place of manufacture are
given at the end, e.g. 2/49<D is a powder forming part of the 2nd lot of 1949 produced
at <D (Russian F).

In Anglo-Saxon countries nitrocellulose powders are known as single base
powders, i.e. made of only one explosive component.

In France nitrocellulose flake (strip) powder is shown by the letter B followed by
further letters indicating its purpose. E.g. BF denotes rifle powder (fusil), BnF— newer
(nouveau) rifle powder, BFP— progressive rifle powder, BC— powder for field
(campagne) guns, BSP— powder for siege howitzers (siege et place) used mainly in
75 mm field guns. Recently, powders for the larger military guns were given letters
BGC (gros calibre) with a subscript showing the calibre, e.g. BGC 4 , BGC 5 etc.
or simply BG 4 , BG 5 etc. Powders for naval ordnance have the letters BM (marine)
also with a subscript denoting the calibre :

Powders Calibres

BM 5 to BM 7 100 and 138.6 mm

BM 5 to BM 9 164.7 mm

BM 9 to BMio 194 mm

BM 7 toBMi3 240 and 274.4 mm

BM13 to BM17 305 mm

BM13 to BMi» 340 mm

572 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Other letters and subscripts show the stabilizer used and its amount, thus BM 7 AM 8
means a powder BM 7 stabilized with an addition of 8% amyl alcohol. BSP D1.5
is a powder for siege howitzers stabilized with an addition of 1.5% diphenylamine.
BM 5 powder is thin since it is extruded through a die of about 2.3 mm, and BM 17
powder through a die of about 7.5 mm.

In Germany, the following nomenclature of nitrocellulose powder was in use:

SP— flake powder for rifle;

Gesch. Bl. P. (Geschutz Blattchenpulver)— flake powder for 8.8 cm guns;

Gr. Bl. P. 03 (Grobes Blattchenpulver)— large flake powder for 15 cm and

21 cm guns;
RG96I
RP 051 tu ' : ' u ' ar powders for 96 n/a (127 mm long);

RP 97 and 99 (Rohrenpulver)— tubular powders for 10 cm guns;

RP 07— tubular powder for 13 cm guns;

Man RP (Man over-Ringpulver)— blank ring powder for field guns.

According to Gallwitz [2] more recent German coding used during World War II
also indicated dimensions, composition etc. The dimensions of the dies used for the
extfusion of tubular or strip powder were given in millimeters. They differ from the
actual grain dimensions. All powders containing volatile solvents shrink in the
course of drying and their dimensions are therefore smaller than indicated by the
code name.

All the German nitrocellulose powders are marked with the letters Nz. This
designation is followed by letters indicating the shape of the grain and its dimen-
sions.

RP.— (Rohrenpulver) tubular powder followed by figures (in brackets) indicating
the length of the tubes, and the outer and inner diameter, e.g. RP (150-2- 1) means a
tubular powder 150 mm long outer diameter 2 mm, inner diameter 1 mm.

St. P.— (Streifenpulver) indicates strip powder: e.g. St. P. 150-15-1 means strip
150 mm long, 15 mm wide, 1 mm thick.

Bl. P. (Blattchenpulver) indicates square plate powder, followed by dimensions
as above, e.g. Bl. P. (4-4- 1).

Rg. P. (Ringpulver) is ring powder— and the figures indicate its dimensions:
(thickness, outer diameter, inner diameter) e.g. Rg. P. (3.25/5).

PI. P. (Plattenpulver) is plate or disk having diameter, thickness as indicated,
e.g. PI. P. (50-0,2).

N. P. (Nudelpulver) is cylindrical (macaroni) powder of the length and diameter,
indicated e.g. N. P. (1,5-1,5).

German double base powders are also marked with figures or letters indicating
their calorific values. Nitrocellulose powders had no such marks, as their calorific
value is approximately the same in all types of powder.

The German nomenclature for double base powder is given in the appropriate
chapter — p. 660.

THE MANUFACTURE OF SMOKELESS POWDER 573

MANUFACTURE OF NITROCELLULOSE POWDER
THE DEHYDRATION OF NITROCELLULOSE

For safety purposes nitrocellulose is delivered to the factory in a wet state and
before it is partially dissolved in a mixture of alcohol and ether the water must be
removed since this prevents the process of swelling and dissolution.

Formerly the water was removed by drying the nitrocellulose. This operation is
dangerous due to the high sensitiveness of the dry nitrocellulose to friction, impact
and static electricity. The dust of dry nitrocellulose either suspended in the air or
spilled on the floor, or on radiators etc. is particularly dangerous, hence the drying
of wet nitrocellulose has caused many accidents.

In the manufacture of nitrocellulose powders the water is displaced with alcohol.
This method was proposed by Lundholm and Sayers [3] and widely used in many
countries [4, 5]. Despite the simplicity of the idea the dehydration process is rather
complicated. It is influenced by such factors as the solubility of nitrocellulose in
alcohol and the ability of nitrocellulose to swell under the influence of alcohol:
the lower the solubility of nitrocellulose in alcohol, the more easy dehydrated
with alcohol. Since, however, the solubility of nitrocellulose depends primarily on
its nitrogen content dehydration is easier with the higher nitrated types of nitrocellu-
lose.

The advantage of dehydrating nitrocellulose with alcohol lies in the fact that any
residual alcohol may be subsequently included in the solvent. The amount of residual
alcohol in nitrocellulose depends not only upon the pressure applied in the dehydra-
tion press, but also on the type of nitrocellulose, i.e. it is somewhat larger in more
highly nitrated nitrocellulose. Nitrocellulose made from wood cellulose swells in
alcohol more readily than that made from cotton which is why the former retains
more alcohol and more water.

The dehydration process is based primarily on the ability of alcohol to displace
water. Since, however, the water is not always perfectly displaced the alcohol becomes
partly mixed with water. In addition, some of the water is adsorbed by the nitrocel-
lulose and cannot readily be removed, which causes a further dilution of the alcohol.
The subsequent portions of fresh alcohol displace the dilute alcohol, the residual
alcohol adsorbed by nitrocellulose is mixed with concentrated alcohol, the latter is
displaced by fresh alcohol etc. This course of the operation is illustrated by variations
in the concentration of alcohol in the liquid flowing out of the dehydration appa-
ratus (Figs. 203-205).

It has been shown that washing out nitrocellulose with alcohol also serves another
purpose; it dissolves and removes from the nitrocellulose degradation products which
for the most part are known to be of low stability (Berl and Delpy [6]). Thus, the re-
moval of water from nitrocellulose by displacement may be considered as an ad-
ditional stabilization process.

The accuracy of the process, its duration, and the variation in the concentration

574

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

of alcohol, depend to a great extent on the apparatus used for dehydration. The yield
of the recovered alcohol also depends upon the apparatus and the particular process
used.

Dehydration with centrifuges

The simplest method of dehydration is by centrifuging. Various types may be
used e.g. power driven centrifuges usually with a capacity of 30 kg of nitrocellulose
(calculated on the weight of the dry substance). Power drive, however, is somewhat
risky if the nitrocellulose is moistened only with alcohol, since it very quickly dries
up and can penetrate into the driving gear. For this reason a special centrifuge driven
by means of a water turbine was designed (Figs. 197 and 198). The basket of this

Fig. 197. Centrifuge for dehydration of nitro-
cellulose with alcohol, according to Brunswig

Fig. 198. Basket of a dehydration

centrifuge, according to Brunswig

Ul.

centrifuge can rotate at two speeds: 1000-1200 r.p.m. and 500-600 r.p.m. The rate
of rotation should be reduced when introducing alcohol, after which it is increased
to remove the liquid. Some centrifuge baskets have a double wall of perforated sheet
(Fig. 198), the annular space between the two walls being lined with a strong filter
cloth. The basket is loaded with 60 kg (dry substance) of wet nitrocellulose. The
cloth acts as a filter and retains the nitrocellulose in the basket.

THE MANUFACTURE OF SMOKELESS POWDER 575

After the centrifuge has been loaded it is closed with a lid fixed with clamps. The
basket of the centrifuge is set in slow motion and a pipe that injects alcohol into the
space between the shaft and the outer wall of the basket is introduced through a
special hole in the lid. The pipe has numerous delivery nozzles facing the outer wall
of the basket. At first 80% alcohol from the previous dehydration is introduced and
the motion of the basket is speeded up to drain away a part of the water and alcohol.
Further alcohol is added in several portions, the centrifuge being slowed down and
then speeded up after each injection. When the content of alcohol in the drained
liquid reaches about 60%, fresh 96% alcohol is introduced. This operation is repeated
2-3 times as described above until the concentration of alcohol flowing out of the
centrifuge reaches 92%. The amount of alcohol supplied each time is established
experimentally for a given centrifuge.

When operating the centrifuge described above with a load of 60 kg of nitro-
cellulose the alcohol is usually injected in four portions:

I ca. 40 1. of 80% alcohol
II ca. 40 1. of 80% alcohol

III ca. 30 1. of 95 % alcohol : \

IV ca. 301. of 95% alcohol

The nitrocellulose is then centrifuged to a definite alcohol content. The dehydration
' of one batch of nitrocellulose requires approximately 1 hr.

The best method for recovering the alcohol is to collect the 60-70% alcohol and
rectify it and to collect the 80% alcohol separately and use it for the dehydration of the
next batch of nitrocellulose. 125 1. of 95% alcohol is consumed per 100 kg of dehydrat-
ed nitrocellulose; of this 30-35 1. remain in the nitrocellulose and about 90 1. are
recovered by rectification.

After dehydration 30-35% alcohol remains in the nitrocellulose and is subse-
quently utilized as a part of the solvent. The final centrifuging should last until the
content of residual alcohol remains constant. It is found that with certain forms of
nitrocellulose (e.g. those made from wood cellulose) the removal of excess alcohol
by centrifuging is difficult. In such cases centrifuging should last somewhat longer.
If the content of alcohol in the nitrocellulose is too high, more solvent will be put at a
later stage of dilution, or its composition is altered to increase the content of alcohol
in it.

The centrifuged nitrocellulose is taken out of the centrifuge together with the
cloth. Since the cloth and its charge adheres strongly to the wall of the basket, it may
be necessary to use aluminium spades to separate it. The unloaded nitrocellulose is
weighed. This gives the content of alcohol, the weight of the dry substance being
known. The alcohol-damp nitrocellulose is broken up into lumps and loaded into
cylindrical cans made of strong galvanized iron, hermetically sealed. Each can con-
tains approximately 20 kg of nitrocellulose (calculated as dry substance).

The nitrocellulose to be loaded into the centrifuge may be in lump form or com-
pressed (taken directly from the transportation cases). To attain uniformity in the

576

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

operation of the centrifuge it is recommended that before loading the nitrocellulose
is rubbed through a 1-2 cm mesh sieve either manually or mechanically. Manual rub-
bing may be carried out either by hand or with wooden paddles. For mechanical
rubbing Nussbaumer's [8] apparatus (Fig. 199) may be used, in which bronze scrapers
fixed on horizontal shafts (/) rotating around a vertical shaft are clamped down to
screen (2) made of brass or aluminium wire. Swivel dampers (3) are used to load the
rubbed nitrocellulose into sacks.

Fig. 199. Schematic view of Nussbaumer's apparatus for rubbing nitrocellulose before

dehydration.

Dehydration in presses

Another method for dehydrating nitrocellulose is based on the use of pneumatic
or hydraulic presses. Dehydration in pneumatic presses (diffusers) (Fig. 200) consists
of filling the cylinder of the press with wet guncotton squeezed out of the tank under

i «>

1 Alcohol

T rfii

S" ^v

— . •»— .

<u a

Fig. 200. Schematic view of a pneumatic press (a pressure diffuser) for dehydration of
nitrocellulose with alcohol.

THE MANUFACTURE OF SMOKELESS POWDER

577

a pressure of 5-8 atm. The cylinder is then closed and the alcohol is forced into it
from above by compressed air (5-8) atm. The water, followed by dilute alcohol
and finally by concentrated alcohol flows out from below.

After dehydration in a diffuser, the nitrocellulose contains about 50% of alcohol.
This is obviously too much and on removal from the cylinder the nitrocellulose is
again subjected to pressure in a hydraulic press (100-200 kg/cm 2 ) to reduce the
content of alcohol to 30-35%. The method described above requires an expensive
installation, more numerous staff and takes longer than dehydration in centrifuges
or in hydraulic presses. It does not seem to be in use any more.

' Draining
of alcohol

Dehydration

HT Trn

Pushing out
the nitrocellulose
cake

Fig. 201. Champigneul hydraulic press for dehydration of nitrocellulose.

Dehydration in hydraulic presses is the most general method now in use. The
Champigneul press (Fig. 201) consists of four cylinders rotating around a vertical
shaft and of four pistons taking positions accomodating to the movements of the
pots. The pistons can move in a vertical direction and always perform the same
successive operations, whereas the set of cylinders rotates, each one taking four posi-
tions in turn, corresponding to four seperate operations. The functioning of the
press is illustrated by the diagram in Fig. 202. The arrangement of the cylinders as
seen from above is shown in section (7). Section (77) represents all four cylinders as
developed on one plane along AOB'OA'OB. As can be seen, the diameters of each
particular piston are different. This is important since the pressure on the whole

578

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

piston is proportional to r 2 (r= radius of the piston). The bottoms of the cylin-
ders constitute a lower immobile section (supports) non-rotatable around shaft O,
with tubular conduits for the outflow of alcohol.

The operation proceeds as follows: a brass mesh and a linen disk are put on
"the bottom" of a cylinder at position /, and then 20 kg (dry substance) of wet
nitrocellulose is poured in from above. Commonly used types of nitrocellulose
(guncotton with collodion cotton) are packed alternately in layers, which in fact

Non-rotatable
sections

Rotatable
section

Non-rotatable
sections

AOB'A'OB section

Fig. 202. Diagrammatic view of the operation of the four cylinders of a dehydrating

press.

constitutes a preliminary mixing. The upper and lower layers are made of the
type of nitrocellulose used in the larger quantity. The nitrocellulose is held in place
by the lower piston. The upper piston rises and falls, thus ramming the load as it
is poured, causing the brass mesh and linen disk to adhere to the lower surface of
the load. The pressure of the piston may range from 25-50 kg/cm 2 .

After the cylinder has been loaded at position (/), the set of cylinders is turned
through 90° so that cylinder (7) moves to position (2). Here the bottom of the cylinder
is formed by a lower piston in which there are furrows and conduits to drain away
alcohol and water. About 20 1. of 95-96% refined alcohol is now poured onto the
layer of compressed nitrocellulose and forced through this layer by the upper piston
at pressures of 50-100 kg/cm 2 . The water, dilute alcohol and finally less dilute alcohol
flow out through the conduits in the lower piston.

After the upper piston has touched the layer of nitrocellulose, i.e. after the
alcohol has been forced through the dehydrated material, the upper piston rises,
while the lower one moves down and the set of cylinders again turns through 90°, so
that the cylinder which started from position (/) through (2) now takes position (5).
Here the lower piston is again fitted with conduits to drain away the alcohol. The upper

THB MANUFACTURE OP SMOKELESS POWDER

579

piston at position (3) has the largest diameter and may develop pressures of 200-300
kg/cm 2 . It reduces the alcohol content to 30-35%. After the upper piston has been
raised and the lower one lowered, the set of cylinders once again moves through 90°.
The cylinder passes from position (3) to position (4), where the upper piston pushes the
cake of dehydrated nitrocellulose down. The brass mesh and linen disk are removed
and the cake is broken up with a wooden mallet and quickly rubbed through a coarse
(1-2 cm mesh) brass sieve. The screened nitrocellulose is weighed to determine its
alcohol content. The mesh and linen disk are re-placed on the bottom of the cylinder
at position (/), before starting the cycle again with another load of nitrocellulose.

Dehydration using this type of press precludes the recovery of dilute alcohol. Only
92-96% alcfhol is used for dehydration. The waste alcohol below a certain con-
centration (usually 50%) is discarded. Any of a higher concentration is sent to recti-
fication.

Ponchon [9] studied the course of dehydration of nitrocellulose in a Champigneul
press, and drew the graphs indicating how water and alcohol are displaced in each
position of the cylinder. The graphs in Fig. 203 denote the process of dehydration

^J"

-c

htl

- ^100

- ie 20

- -2 50

- ° 70

$200

$ 100

• v^T.

s **" f

S / S" jv

f>t

1 2

Time, win

Fig. 203. Characteristics of the dehydration process of nitrocellulose in a Champigneul
press (position (2) of the cylinder).

when the cylinder is at position (2). Curve / shows the change in volume, curve //
the change in concentration of waste alcohol, curve /// the change in water content
of the nitrocellulose and curve IV the pressure indicated by the manometer. Changes
in all these values are expressed as a function of time. A graph illustrating the process
of dehydration when the cylinder is at position (3) (separation of alcohol) is shown
in Fig. 204. The curves I-IV denote the same relationships as in Fig. 203 and in
addition curve V shows the decrease in alcohol content of the nitrocellulose.

Ponchon also produced a graph (Fig. 205) showing the difference between
the dehydration of CP t nitrocellulose (containing 1.5% of soluble in alcohol)

580

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

and CP 2 nitrocellulose (containing 5.2% of soluble in alcohol). The graph indi-
cates for both forms of nitrocellulose the volume of waste alcohol (broken lines)
and the pressure within the press (continuous lines) against time. The graph shows
that the pressure necessary to displace the water is lower with guncotton (CP t nitro-
cellulose) than with collodion cotton (CP 2 nitrocellulose). In CP t the alcohol flows

■c

700

- a 700

- £ 20
o

<•-

- .| 50

- <^ 70

. <o

-1=200

- §700

1 1 1

-//

-IV

-I

-V

-III

S- ■■ -r-

■ '

Time, min

Fig. 204. Characteristics of the dehydration process of nitrocellulose in a Champigneul
press (position (J) of the cylinder).

20 r 200

o
o

3 70 - £ 700

1
CP,

"^\--^?

/**'

^•*"

^.*~~ m

jr //

*."

/ / /

S ' /

/ Ch > Zs

/ S/Wi "'

1 / ' s^CP„

/ / ' *

1/ / S

IS / '

// ''''

// "

SS /y l

Time, min
Fig. 205. Comparison of the course of dehydration of CPi and CP 2 nitrocelluloses.

out more rapidly. These findings show that CP X is dehydrated more readily than CP 2 .

Ponchon's experiment led him to conclude that the most advantageous arrange-
ment lies in placing the CP 2 layer on the bottom of the cylinder and the CP t layer
in the upper part.

As shown in Ponchon's graphs the use of a hydraulic press leads to very distinct
changes in the concentration of alcohol on dehydration and permits the collection

THE MANUFACTURE OF SMOKELESS POWDER 581

of alcohol of a definite concentration at a given time. Alcohol of approximately 50%
concentration is usually collected and rectified.

Svetlov [10] quotes the following figures characteristic of the output of the Cham-
pigneul press. In one working cycle of the press 20 kg of nitrocellulose containing
about 6 kg of absolute alcohol (23% of the total weight) and about 0.4 kg of water
(1.5% of the total weight) are obtained. In the first stage of the operation (positions
1 and 2) 12-14 kg of dilute 40-50% alcohol flows out, and in the second stage (position
3) about 7 kg of 93% alcohol. One working cycle lasts up to 2 min. In one hour
approximately 30 charges of nitrocellulose of 20 kg each are dehydrated, e.g 600
kg/hr.

According to Ponchon 110 kg of 93% alcohol are used for the dehydration of
100 kg of nitrocellulose (dry weight). In French factories (1917) an average of 33 kg
of 42-52% waste alcohol was recovered and returned for rectification.

The consumption of energy per 1000 kg of nitrocellulose (dry) is 13.5 kWh.

The Becker and van Hiillen (Krefeld) press is of another design. In this press
the wet non-centrifuged nitrocellulose is first dehydrated by removal of up to 20-30%
of its water content, and then subjected to further dehydration with alcohol. Partial
dehydration of the nitrocellulose takes place in the feeding screw used with this press.

Rectification of alcohol from dehydration

Alcohol from dehydration contains a certain quantity of nitrocellulose, i.e. its
soluble fractions, mostly degraded, and a certain amount in suspension. Experiments
have shown that about 2.2 g of dissolved nitrocellulose and 1.3 g of nitrocellulose
in suspension— a total 3.5 g— occur in 1 1. of 70%alcohol from centrifuges. Sometimes
however, the content of nitrocellulose in the alcohol may reach 10-12 g/1.

The nitrocellulose present in suspension readily passes through filters, and should
therefore be separated by decantating the alcohol out of the tank in which the waste
alcohol is stored. In this manner the nitrocellulose content in the waste alcohol is
reduced to 2-3 g/1. In some factories nitrocellulose is removed more completely by
diluting the alcohol to about 40% with water (from washing the powder, and therefore
containing a small amount of alcohol). By this means a certain quantity of the nitro-
cellulose is precipitated as sludge. After settling, the alcohol is decanted. This method,
however, is troublesome and not commonly used, because the residual nitrocellulose
in solution and traces of its suspension in alcohol are liable to decomposition in the
course of prolonged heating during distillation. The decomposition of residual nitro-
cellulose in a distillation still (vat) has several times caused explosions. Moreover due
to the decomposition of nitrocellulose there are traces of nitrites, nitrates and even
nitric oxides in the distillate. To avoid this, it is advisable to add calcium oxide to
the distillation vat, using 1 kg per 100 1. of alcohol. This causes hydrolysis of the
nitrocellulose and neutralizes the products of hydrolysis. In addition the vat should
be freed from solid residue as often as possible. Losses on rectification amount to
approximately 1 . 5 % alcohol.

582 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

THE PREPARATION OF NITROCELLULOSE MIXTURES

Nitrocellulose for the manufacture of powder must meet requirements of chemical
stability, nitrogen content and solubility in a mixture of alcohol and ether according
to the regulations discussed in the chapter on nitrocellulose.

For the manufacture of most kinds of semi-colloidal nitrocellulose powders
with a volatile solvent, and recently also for the manufacture of solventless nitro-
glycerine powders, a mixture of two forms of guncotton is used : high nitrated guncot-
ton, insoluble in a mixture of alcohol and ether, and low nitrated collodion cotton
soluble in a mixture of alcohol and ether. In some countries (U.S.A. and U.S.S.R.)
pyrocollodion cotton is used and for powders requiring a high nitrogen content, a
mixture of pyrocollodion and guncotton.

Since nitrocellulose is stored and weighed in a wet condition (25-30% water),
the moisture content should be determined accurately before batching. The powder
plant usually receives nitrocellulose in semi-sealed containers. Since these are not
uniformly tight the moisture content is liable to some variation. This makes batching
more difficult, less accurate and even completely faulty. It is therefore advisable to
equalize the moistnessof the nitrocellulose by unloading it into concrete pits, holding
5000 to 20,000 kg of nitrocellulose (dry weight) hermetically closed with a sheet
iron cover fitted with a rubber seal and storing it there for a few days. The nitro-
cellulose is then analysed for its water content, nitrogen content and its solubility
(dry weight) in a mixture of alcohol and ether. Samples are taken from several places
to make the analysis more reliable.

After analysis each form of nitrocellulose (guncotton, collodion cotton, pyro 7
collodion cotton) is batched separately into a linen bag which for convenience and
safety is stored in an air-tight iron vessel (to protect the nitrocellulose from drying up
and becoming dusty). Nitrocellulose is batched by charges, the size of which depends
on the dimensions and the type of apparatus used. In France, for instance, a total
charge of nitrocellulose (CP t and CP 2 ) is 20 kg when dehydrated in a hydraulic press
or 30 kg when dehydrated in a centrifuge.

The more modern approach to the problem of mixing nitrocellulose consists of
mixing defined types of nitrocellulose in the nitrocellulose factory itself. In this case
the two forms of nitrocellulose are mixed under water in mixers as described in Vol.
II, p. 374. The water is then centrifuged and the mixture dehydrated with alcohol.
This method, however, creates certain inconvenience to the powder factory which
loses the possibility of changing (within certain limits) the composition of the mixtures,
i.e. nitrogen content and total solubility of nitrocellulose. The powder factory is
therefore compelled to limit the number of factors which can be varied to obtain
the powder of required ballistic properties.

When treating the nitrocellulose mixture with solvent, only collodion cotton is
dissolved and converted into a colloidal state. Guncotton is incorporated into the
colloidal mass in the form of fibres. Thus by the solubility of a mixture of nitrocellulo-
ses in a mixture of alcohol and ether, we mean the total solubility of the mixture.

THE MANUFACTURE OF SMOKELESS POWDER 583

The batching of nitrocellulose consists of weighing the guncotton and collodion
cotton in a ratio which gives a mixture of suitable nitrogen content with the required
total solubility.

The following rule is helpful in calculating the composition of a mixture. Suppose
that a mixture of 30% total solubility is needed consisting of guncotton (CP t ) of 10%
solubility and collodion cotton (CP 2 ) of 95% solubility. To prepare it 65 parts by
weight of CPi and 20 parts by weight of CP 2 have to be mixed.
CP X 10 65 parts by weight of CP,, i.e. 76 5°/

30
/\
CP 2 95 20 parts by weight of CP 2 , i.e. 23.5%

Total 85 parts- 100%

After the ingredients have been mixed the total nitrogen content in the mixture
so obtained must be checked.

The nitrogen content and solubility of the nitrocellulose mixture must be maintain-
ed within the following limits according to its intended use (Table 177).

Table 177

Total solubility

Nitrogen content

Rifle powder

18-25%

13.1-13.3%

Powder for smaller

calibre cannon

(below 90 mm)
Powder for heavier

25-35%

12.8-13.1%

calibre cannon

35-45%

12.5-12.9%

The lower the nitrogen content and the higher the solubility of the nitrocellulose
in a given solvent, the more slowly the powder made of it will burn. Thus in matching
the composition of a mixture of guncotton and collodion cotton to obtain the reouired
total solubility the burning rate of the powder can be regulated within certain limits.
E.g. in France, for modern rifle powders, which are faster burning than cannon
powders, the mixture of CP t and CP 2 is so balanced that the total solubility is 15-20%
while for the earlier type of powder Bn 3 F (Le Bel rifle) it is 25-30%. For small calibre
field guns (75 mm) powder, a total solubility of 30-35% is chosen. For 105 mm field
guns powder and for small calibre naval guns it is 40% and for heavy calibre guns
40-50%.

If the powder is made of pyrocellulose, only part of it dissolves in the solvent
(usually 60-70%) while a certain amount of the substance remains in fibre form.

PARTIAL DISSOLUTION OF NITROCELLULOSE

For the manufacture of nitrocellulose powder a volatile solvent is used, i.e. a
mixture of alcohol with ether in a weight ratio of about 2:1.

584 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The alcohol (usually refined to 95-96%) must not be acid. To make a 100 cm'
sample of alcohol neutral towards phenolphthalein no more than 1.6 cm 3 of 0.1 N
NaOH is required. A larger quantity of acid is detrimental since it may adversely
influence its stability and gives a dark powder. The alcohol should not contain nitrites.
A nitrocellulose powder plant should be equipped with a plant for the distillation
of alcohol.

Ether, in its common commercial form, is sufficiently pure to be used in admixture
with alcohol. Its acid content should not exceed 40 mg (calculated as H 2 S0 4 ) per
litre, nor should it contain nitrites. A nitrocellulose powder plant must include a plant
for the manufacture of ether.

In France, two types of an alcohol-ether mixtures are used for the manufacture
of nitrocellulose powder specified in terms of Baume degrees :

(1) 56° ether consisting of 64 parts by weight of ether and 36 parts by weight of
alcohol. v

(2) 54° ether consisting of 56 parts by weight of ether and 44 parts by weight of
alcohol.

Nitrocellulose dissolves more readily in 56° ether which is therefore used more
often than 54° ether. If, however, the nitrocellulose is very soluble and gives a powder
with too slow a rate of burning, the use of 54° ether is recommended, i.e. a solvent
which leaves more insoluble material, thus giving a more fibrous powder.

The volatile solvent (alcohol-ether mixture) is used in such an amount that the
dough is fairly loose but sticks when pressed between the fingers. The actual amount
depends to a great extent on the properties of the nitrocellulose, since some forms
require more solvent than others. The amount of solvent is usually determined by
experiment. It ranges from 70-150 parts per 100 parts by weight of a mixture of nitro-
cellulose (dry substance).

With wood nitrocellulose smaller amounts of solvent are used (70-90%). This is
accounted for by the lower degree of polymerization of wood cellulose in comparison
with that of cotton. With its lower degree of polymerization wood nitrocellulose
swells and dissolves more readily, producing solutions of a relatively low vis-
cosity.

With pyrocellulose which contains 60-70% of material soluble in an alcohol-ether
mixture, the use of 70-80 parts by weight of solvent per 100 parts of nitrocellulose
is sufficient.

A volatile solvent is not used for the manufacture of completely colloidal powders
containing nitroglycerine, with the exception of British cordite. For this, acetone is
used, which is a solvent well fitted for both high nitrated cellulose (guncotton of
13% N) and nitroglycerine. Small amounts of acetone are also sometimes employed
as an auxiliary solvent for the manufacture of ballistite. During World War I, when
there was a shortage of acetone an alcohol-ether mixture was adopted as a substitute
for the manufacture of cordite. It then became necessary to use a lower nitrated
cellulose soluble in this solvent (RDB cordite). This is discussed more fully in the
chapter on nitroglycerine powders.

THE MANUFACTURE OF SMOKELESS POWDER 585

Nitrocellulose is gelatinized (dissolved) and the powder dough is prepared in
kneaders so that the dough-like mass is thoroughly mixed. Werner-Pfleiderer tvoe
kneaders (Figs. 206, 207, 208 and 209) are most commonly used. They consist of a
trough made of bronze (surrounded by a cooling jacket) in which two powerful
bronze stirrers in the form of worm-shaped blades rotate in opposite directions one
twice as fast as the other. Effective kneading combined with simultaneous mixing is ■
attained by the movement of the stirrers downwards in the centre and upwards by
the walls. The angular speed of the slower stirrer is 20-30 r.p.m. and of the faster
one 40-60 r.p.m. The trough is fitted with an outer cooling jacket

The kneaders in use are of varying capacity, and usually hold a charge of 60 kg
of dehydrated nitrocellulose (dry weight) containing alcohol. After the kneader has
been loaded its lid is closed (until then the lid is slung on chains from a block attached
to the ceiling) and screwed down to the trough as tightly as possible. The stirrers are
then set in motion; the ether is fed through a conduit in the IM as is the additional
quantity of alcohol. Simultaneously the stabilizer is introduced in such' a manner
that diphenylamine weighed out in a silk bag is placed into a small container attached
to the lid The latter is connected with the conduit feeding the ether so that the ether
flowing down from the batcher to the kneader, on passing through the container
With diphenylamine, dissolves it and thus introduces it into the powder dough To
facilitate checking whether or not diphenylamine is present in all the containers it
is advisable to fit an inspection window.

Kneading requires 2.5-3 hr, although in exceptional cases if rapid manufacture
is necessary, this period may be shortened to 1-1.5 hr.

Since the mass is heated up during kneading, due to friction, cold water is fed
to the cooling jacket during the whole time of kneading so that the temperature
does not exceed 30°C, otherwise the ether evaporates. Dissolution may be in-
complete and excess pressure may be produced within the kneader which may blow
oil the hd when it is unscrewed.

When the kneading is finished, the lid is unscrewed and lifted with the block, the
stirrers are set to rotate in the opposite direction (upwards in the centre and down-
wards by the walls) and the trough is tilted up by a special mechanism driven manu-
ally or mechanically (Fig. 210) so that the dough falls from the trough into two
containers previously placed below. To prevent spilling the dough the containers
are covered with a protective hopper made of sheet brass or leather. As a container
is Med, the dough is rammed with bronze rammers mounted on wooden handles.
Ramming gives a more uniform dough and removes the air. This at a later stage,
facilitates pressing.

The containers loaded with the dough are hermetically closed. When unloading,
a sample is taken from each kneader to be sent to the testing laboratory to check
the presence of diphenylamine. A drop of concentrated sulphuric acid on a sample
ot the dough gives a blue coloration if diphenylamine is present. If the test is negative
the dough is returned for kneading once again with the addition of the required
amount of diphenylamine.

Fig. 206. Werner-Pfleiderer kneader with electric motor shown behind a wall (Cour-
tesy Werner & Pfleiderer, Maschinenfabriken und Ofenbau, Stuttgart).

Fig. 207. Werner-Pfleiderer kneader at work (Courtesy Werner & Pfleiderer, Maschi-
nenfabriken und Ofenbau, Stuttgart).

Fig. 208. Werner-Pfleiderer kneader in unloading position (open) (Courtesy Werner &
Pfleiderer, Maschinenfabriken und Ofenbau, Stuttgart).

Fig. 209. View of paddles (Courtesy Werner & Pfleiderer, Maschinenfabriken und

Ofenbau, Stuttgart).

588

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

*"/

i \

y \

""-

\
^ \

--'"M

//'*'

\ \

'<?r\

/ 1 ^

Fig. 210. Mechanism for tilting the kneader.

>,n, nrjK it

XXrrXX 1 "

■ •>>».;,•

Fig. 211. A diagram of a Chaudel-Page kneader.

Fig. 212. A diagram of the mechanism for moving the stirrer of a Chaudel-Page
kneader in translatory and angular motion, according to Yegorov [11].

THE MANUFACTURE OF SMOKELESS POWDER 589

In France Chaudel-Page kneaders (Fig. 211) are in general use. These take the
form of an iron cylinder in which an iron bar rotates about a horizontal shaft [11].
The bar performs an angular-translatory motion, driven by the mechanism sketched
in Fig. 212. A spur-gear rotates in a box with interior left-hand and right-hand
threads. The spur-gear is fastened to the horizontal shaft. If turns forwards in the
grooves of one of the threads with a translatory motion of the shaft in one direction
(e.g. to the right). When the spur-gear reaches the end of the box further movement
in this direction becomes impossible. As it turns it falls into the threads imparting
a translatory motion to the shaft in the opposite direction. Thus it travels to and
fro, and kneads and mixes the contents of the cylindrical kneader throughout its
full length.

60 kg of dehydrated nitrocellulose containing alcohol is loaded into the kneader,
through a manhole in the upper section of the cylinder. It is unloaded through the
lateral cover. Kneading takes as long as that in a Werner-Pfleiderer kneader.

The advantage of kneaders of the Chaudel-Page type is that they take up con-
siderably less space than Werner-Pfleiderer kneaders. Premises which provide accom-
modation for six kneaders of the Chaudel-Page type can hold at most three
Werner-Pfleiderer type kneaders.

Safety in operating kneaders is, in general, fairly high. This stage of production
is considered to be one of the safest in the manufacture of nitrocellulose powders,
now that it is known how to avoid the danger created by electrification of the solvent,
especially of ether. It was observed that numerous explosions and inflammations of
ether were caused by discharges of static electricity, e.g. there were accidents caused
by the ignition of ether when the valve feeding it to the kneader was opened. Such
accidents ceased when all containers and pipes containing solvents were carefully
earthed. The kneaders must also be earthed. The equipment driving them— engines,
transmission gears etc.— should be situated in a separate room.

The moment of unloading is particularly dangerous since large quantities of
ether vapours mix with the air to form an explosive mixture. At this moment any
impact of metal against metal which may produce sparks is dangerous (e.g. allowing
the heavy lid to strike against the rim of the kneader in the Werner-Pfleiderer sys-
tem).

Dough containing a considerable amount of solvent is non-flammable and
almost non-explosive. Only the solvent burns easily and only if there is an access
of air. Indeed, the solvent strongly "phlegmatizes" the nitrocellulose, considerably
reducing its explosiveness.

In some countries (e.g. the U.S.A.) the nitrocellulose-alcohol-ether dough is
subjected to an additional treatment intended to improving its homogeneity.

This consists of "blocking" which involves pressing the mass for several minutes
at a pressure of ca. 200 kg/cm 2 (3000 lb/in 2 ). The block thus formed is transferred
to another press of a type much similar to the extrusion press (Fig. 213). Here the
dough is forced through a series of screens and perforated plates. This is referred
to as "macaroni" pressing owing to the shape of the extruded threads.

590

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The shredded, "macaronied" dough is introduced into the extrusion press, as
described in the next paragraph.

Fig. 213. Blocking press at E. I. du Pont de Nemours and Co. factory, according

to Davis [12].

SHAPING THE DOUGH

' The dough is usually pressed into the desired shape by extruding it from a hy-
draulic press. Formerly, the dough was rolled into sheets by passing it between rolls,
predried and cut into flakes or strips. This method is retained for the manufacture
of some types of nitrocellulose powder, i.e. very fine ones with a very high rate
of burning ("coefficient of vivacity"*), used as sporting, pistol, and blank powders.

* According to French nomenclature "coefficient of vivacity" (coefficient de vivacity is the
rate of increase of pressure during the burning of the main part of the grain of smokeless powders
in a closed vessel (p. 530).

THE MANUFACTURE OF SMOKELESS POWDER 591

Rolling is the most desirable method for obtaining flakes of very low web thickness
since it is difficult to obtain strips of a thickness below 0.40 mm by pressing and
there is much waste, due to the clogging of the die slots and the folding and tearing
of the extruded strips (rolls are shown on Fig. 260).

For rolling the dough should contain a relatively small amount of solvent (should
be "dry") otherwise it adheres to the surface of the rolls and is then difficult to remove.
On an average, dough to be rolled should contain less solvent (about 50%) than
that subjected to pressing. Rolls especially adapted for the manufacture of nitro-
glycerine powder are most widely used, but not heated.

The use of hydraulic presses makes it possible not only to shape the powders
into flakes or strips but also into cords or tubes or even into more complicated
profiles. The operation of the press consists of the filling the cylinder so that the
highest possible density is obtained with all air spaces eliminated. The piston of
the press forces the mass first through a sieve that removes mechanical impurities
and then through the die giving it a required shape. It is particularly important
to remove all impurities from the dough, otherwise they may interfere with pressing
and may distort the shape of the product— troubles that obviously reduce the efficiency
of operation and increase operating expenses. If the dough is "macaroni" pressed
(see p. 589) it is already filtered and homogenized and extrusion to the final shape
is much facilitated. ^

The extruded strips, cords or tubes are forced from the die directly on to a linen
conveyer or other receptacle from which they are passed to the drying room.

The die of press is the most essential structural component of the system since
it imparts a definite shape to the dough. If the powder is to be shaped into flakes
or strips, a die with a mouth in the shape of a slot or several slots is used. E.g.
German rifle powder was extruded in the form of strips 2 mm wide and 0.45-0.55 mm
thick. The thickness of strips was modified by the change of die depending on the
required time of burning of the powder. On changing the lot of nitrocellulose used
for the manufacture of the powder, the experimental batch may indicate that it is
more, or less, "vivacious" than the standard powder. To obtain a faster burning pow-
der than that of the experimental batch made of strips, e.g. 0.50 mm thick, a die giving
thinner strips, e.g. 0.45-0.47 mm thick is used. If, on the other hand, the experi-
mental batch has too high a "coefficient of vivacity", the die should be replaced by one
which gives thicker strips, e.g. 0.53-0.55 mm.

The influence of the web thickness on the ballistic properties of cannon powder
is illustrated by the following example of strip powder used for 105-120 mm field
guns. The total solubility of this powder is 35%.

Web thickness Muzzle velocity Pressure p

obtained

*>o

2.7 mm 733.5 m/sec 2563 kg/cm2

2.8 mm 724.2 m/sec 2323 kg/cm*

2.9 mm 719.0 m/sec 2282 kg/cm*

592

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

With thinner powders the influence of changes in web thickness is greater. E.g.
for the powder used for 75 mm field guns :

b thickness

1.2 mm

547.5 m/sec

3015 kg/cm*

1.3 mm

538 m/sec

2587 kg/cm2

The final web thickness offtake and tubular powders, after the removal of solvent,
amounts to 53-65% of the web thickness of the strip at the moment of extrusion.
The extent of shrinkage depends on such factors as total solubility, the viscosity of
nitrocellulose, the quantity of solvent and the composition of the dough at the
moment of extrusion. A greater solubility and a higher content of solvent lead to
greater shrinkage of the strip.

The die in which tubular powder is extruded is shown in Fig. 214 (according to
Yegorov [11]). It consists of two sections (7) and (2) inserted one in the other. The

Fig. 214. Design for a die for shaping
tubular powder.

Fig. 215. Design for a die for shaping multi-
tubular powder.

lower channel of section (/) of the die gives the outer diameter to the tube while
section (2) of the die contains a centrally-fixed steel wire, the thickness of which
determines the inner diameter of the single-perforated tube. Multiperforated dies
are of similar design (Figs. 215 and 216, according to Yegorov).

The presses are usually operated with two cylinders working alternately. When
the dough is extruded from one of them the other is loaded with material from the
kneaders and vice versa (Fig. 217). The dough is loaded in batches, using a brass
shovel, together with a ramming piston working under a pressure of about 50 kg/cm 2 ,
A cylinder may hold 15-25 kg of dough (calculated as dry nitrocellulose) depending
on the dimensions of the press. When the cylinder is loaded, a dried cake of dough
and a leather disk are put over the charge as a seal to prevent the "leakage" of the
mass through the gap near the pressing piston. As a cylinder is loaded it is turned
through 180° to take up the position at which the piston, working under a pressure
of 75-125 kg/cm 2 , extrudes the mass through the die.

THE MANUFACTURE OF SMOKELESS FOWDER

593

There are presses in which a cylinder fitted with a die and a system of filters is
filled with the dough and turned through by 180°. In other presses, e.g. those of
the Champigneul type, the bottom of the cylinder, when filled with the rammer
is formed by a separate piston. When, however, the cylinder takes up the position

Fig. 216. A view of a multiperforated die.

for pressing the bottom section is formed of a set of filters together with the die
In turning the cylinder from the loading to the pressing position, only the cylinder
is moved, while the bottom stays in position. A general view of a press of the Cham-

Fig. 217. Schematic view of the functioning of a press for extruding strips and tubes

of powder.

pigneul type used in France and Werner-Pfleiderer used in Germany is shown in
Figs. 218 and 219, respectively. Figure 220 shows a hydraulic press used in Sweden
(Bofors Nobelkrut).

594

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 218. General view of a Champigneul press used in France for extruding strips
and tubes of powder, according to [13].

THE MANUFACTURE OF SMOKELESS POWDER

595

A horizontal extrusion press used in the U.S.A. is shown on Fig. 221.

The yield of a press depends not only on its design, applied pressure and the
dimensions of the extruded strip or tube, but also on the plasticity of the dough.
The effect of the dimensions of a strip or tube on the yield of the press is illustrated
by the following examples. In a press of the Champigneul type, when pressing a
strip 0.70 mm thick for rifle powder (total solubility 30%, amount of solvent 120%),
the yield is 90 kg/hr. When pressing a strip 2.8 mm thick (total solubility 40%,

a) b)

Fig. 219. General view of Werner-Pfleiderer powder press (Courtesy Werner & Pflei-
derer, Maschinenfabriken und Ofenbau, Stuttgart).

amount of solvent 125%) the yield is 175 kg/hr (calculated as dry nitrocellulose).
The same powder mass, when pressed into strips 5.0 mm thick, gives a yield of
200 kg/hr.

Pressing is not a particularly dangerous operation. Its safety is secured to a
great extent by care in ramming the dough in the cylinder, so as not to leave free
spaces filled with air. If this is not done an explosive mixture of solvent vapour in
air is formed in these spaces and violent (adiabatic) compression of such a gaseous
mixtures under the pressure of the piston may occur together, elevating the tem-
perature above that of initiation. This could lead to an explosion. As previously
stated, nitrocellulose containing solvent explodes with difficulty, thus there is no

596

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

danger that the explosion of a vapour-air mixture may bring about an explosion
of the whole nitrocellulose mass. Nevertheless it does mean that there may be an
explosion of a mixture of air with ether and alcohol present in the pressing room.
This would have disastrous consequences.

Fig. 220. Extrusion press used in Sweden (Bofors Nobelkrut).

Predrying

After the dough has been shaped, it is subjected to predrying which reduces
the solvent content to 20-30%. The mass so dried becomes mechanically resistant
and may therefore be cut without being deformed. Due to the presence of residual

THE MANUFACTURE OF SMOKELESS POWDER 597

solvent the mass also retains a certain plasticity and elasticity, preventing crumbling
and dusting during cutting.

Fig. 221. Extrusion of powder in the form of perforated cylinders at E. I. du Pont de
Nemours and Co. Factory, according to Davis [12].

Predrying should be conducted at temperatures as low as possible, i.e. within
the range 15-25°C. Higher drying temperatures may cause the powder to swell
or crack, due to intensive evaporation of solvent. It is therefore advisable, especially
at the beginning, to maintain a low, but not too low temperature -in the latter case
water vapour may condense on the powder, leading to the partial precipitation of
nitrocellulose from the colloidal sol. This causes bright stains to appear on the
surface of the powder. Since the heat of evaporation of the solvent is hi 2 h, the
temperature of the air coming into the drying room should be higher than the~drying
temperature. The composition of the residual solvent gradually changes. Ether,
being more volatile, evaporates more quickly, so that the residual solvent is enriched
in alcohol. A small quantity of ether remains in the alcohol, however, and is not
removed by simple drying.

Since the rate of volatilization decreases as the ether is removed, the temperature
should be raised gradually during drying.

An important factor influencing the dimensions of the predried powder is the
viscosity of the nitrocellulose solution. The lower the viscosity of the nitrocellulose,
the greater is the shrinkage of the nitrocellulose gel on drying. This produces flakes
or tubes of considerably smaller size than those from nitrocellulose of higher vis-
c osity.

598

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Solvent recovery is extremely important from the economic point of view. The
solvents are recovered by various methods, dealt with in a separate section (p. 599).
From the time the strips or tubes are extruded until they are loaded into the predriers
a certain period of time elapses depending upon the type of plant, the method of
operation etc.

The amount of solvent escaping into the air depends on this period of time and
on the season, the temperature of the air etc. Solvent losses range from 10-25%,
calculated as the solvent introduced into the powder mass in the kneaders.

Originally, predrying was conducted in ordinary cabinet driers fitted with racks.
Strips or tubes intended for drying were thrown over the racks. This type of drier

i^zm^rzzzzzrz:

Cooler

Radiators ,. ,. . , ,
Liquefied solvent

Fig. 222. A cabinet drier for predrying powder and solvent recovery.

is however unsuitable for the recovery of solvents since the relatively large surface
of the door allows considerable solvent loss on loading and unloading. The cabinet
drier used in the U.S.A. is shown in Fig. 222. Here the powder was placed in the

/TX.

Fig. 223. A cabinet drier for predrying powder : 1 — hole for loading, 2— hole for unload-
ing the powder strips and tubes, i— rails inclined towards the hole unloading, 4— air

supply, 5— air exit.

drier already cut up since the method of manufacture required that thicker powder
tubes must be cut up directly after pressing (before predrying). In some driers of
more modern design used in France the surface of the loading and unloading holes
is very small in comparison with the drier's capacity (Fig. 223). There are slightly

THE MANUFACTURE OF SMOKELESS POWDER

599

inclined rails inside the cabinet, on which the racks in the form of rods fitted with
rolls, are moved. The powder strips, tubes or cords are suspended on the racks.
Fresh air or air with a reduced content of solvent (depending on the recovery meth-
od) enters from the bottom, becomes saturated with the solvent, and escapes
through the top. It is most important that the powder is dried as uniformly as possible.
Driers of this design should not be too long as this prevents uniformity of drying.
The best length is approximately 5 m.

Solvent recovery

Solvent escapes from nitrocellulose powder at various stages of its manufacture.
The largest amount volatilizes during the predrying of the freshly shaped strips or
tubes in cabinet driers specially constructed for this purpose. Installations for the
recovery of this solvent have therefore long been in use in factories.

At other stages of the manufacture of the powder the solvent escapes into the
atmosphere.

The losses of solvent in various stages of manufacture, for each 100 kg of nitro-
cellulose used (based on French findings from World War I) are tabulated below.
Losses of alcohol and ether are calculated as alcohol (1 kg of ether equivalent
to 1.4 kg of alcohol).

Table 178
Losses of alcohol and ether in the manufacture of nitrocellulose

POWDER

No.

Stage of manufacture

Losses of alcohol
kg

Dehydration

Kneading

Loading of presses

Pressing (for strips 1.8 mm thick on the average)

Loading and unloading of predriers

Cutting

Screening

10.8

Soaking

Refining of alcohol from dehydration

Refining of alcohol and ether recovered on

predrying

Storing of alcohol and ether

1.2

Total |

76.0 kg

This summary gives the amounts of solvent that escape to the atmosphere of
the factory premises at various stages of manufacture. This solvent is partly recovered
by means of an additional installation which will be discussed later.

There are two essential methods for the recovery of solvent:

600

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(1) Methods that rely on condensation of the solvent vapours by cooling (or
compression).

(2) Methods relying on adsorption or absorption of the solvent.
Condensation methods are successful only when there is a high concentration

of solvent vapours in the air. In practice, however, the concentration of the solvent
in the air is usually not high enough, so that the efficiency of the installation is
relatively low. When the vapours of solvent are considerably diluted in the air,
the condensation method cannot be applied and the solvent should therefore be
recovered by absorption.

Installations for the recovery of solvent usually consist of main and ancillary
sections. The main installation comprises equipment for the recovery of the solvent
which escapes from the powder on predrying. The ancillary installation recovers
solvent from various production departments. Solvent recovery ducting is installed
in rooms containing high solvent concentration. This improves the working condi-
tions by removing ether-alcohol vapours from the atmosphere.

The main installation may be used for the recovery of solvent either by condensa-
tion or by absorption. The ancillary installation usually deals with considerably
dilute solvent vapours so that the best results are attained by absorption.

Condensation or absorption methods may also be applied to the recovery of
the solvent removed from the powder on drying under reduced pressure ("p. 614).

The recovery of solvent from the water after soaking the powder and from the
vapours evolved on soaking is a separate problem and will be dealt in a section
on p. 620.

With more recent solvent recovery methods the consumption of alcohol and

Table 179
Recovery and consumption of solvent per 100 kg of

POWDER READY FOR USE

Solvent (kg)

Used | Recovered

2
3

Ether

Recovered in installations for
predrying

Recovered after the soaking of powder
Recovered with cresol or activated coal

92-116

44-63
1-9

4 | Consumption

39-56

3
4

Alcohol
Dehydration

Recovered in installations for
predrying

Recovered after the soaking of powder
Recovered with cresol or activated coal

97-148

26-75

29-35
15
1-3

Consumption

18-^0

THE MANUFACTURE OF SMOKELESS POWDER

601

ether m powder manufacture is considerably reduced. In Table 179 figures based on
statistical data from various French factories during World War I are summarized
The consumption of alcohol (1 kg of ether equivalent to 1.4 kg of alcohoH fnr
the production of 100 kg of powder is 72-1 1 7 kg. '

In choosing a method for the recovery of solvent operational safety is a primarv
factor to be taken into account. While the machinery is in operation the greatest
danger is associated with the mixture of air and alcohol-ether vapour which flows
through the pipelines and appears in various parts of the plant.

As moderate concentrations of solvent in the air, i.e. 4-9% are the most danger
ous, danger is minimized either by very high concentrations (over 9°/) or bv verv
low ones (under 4%). Very high concentrations are only likely to occur in the main
installation. There is always a danger, however, that with the predrying of the powder
and the condensation of the solvent, the concentration of solvent vapours may fall
to a limit at which the gaseous mixture becomes dangerous. The mixture of solvent
vapours may also become diluted to the danger limit due to a plant leakage For
these reasons installations which require high concentrations should be considered
relatively dangerous. Those which can work on low concentrations of solvent va-
pours, ,.e. considerably below the lower limit of dangerous concentrations (ie in-
of safety 8 ^ rCCOVery ° f SOlV6nt ^ absorption ) P ermit a Skater margin
Installations for the additional recovery of solvent are an additional hazard in
a powder factory, because separate buildings, often some distant apart, are connect-
ed by primes through which a mixture of air and alcohol-ether vapour flows
Th.s contradicts the basic rule that all buildings in a powder factory should be
separate and at a safe distance from each other. There are devices for breaking a
flame moving along a pipeline, but their effectiveness is limited, e.g. when weak
explosions occur in the gaseous mixture (the design of these fire breakers is dis-
cussed on p. 607).

_ Instances are known of the destruction of nearly whole factories simply due to
the fact that the separate buildings were connected with a net of solvent recovery
pipelines, e.g. the large explosion at- Hasloch, Germany, in 1926.

A method which might ensure safety in the recovery system is based on a partial
or total replacement of the air within the pipelines by a gas containing no oxygen
that does not form explosive mixtures with alcohol^ther, e.g. by nitrogen or carbon
dioxide from exhaust gases. This method, however, proved too expensive, and was
not carried beyond small scale tests.

Recovery of solvent by the condensation of vapours. This is one of the oldest
methods which once was widely used in many countries (France, U.S. A) and
well checked in practice.

The recovery devices used in France (Fig. 224) function in the following way.
Fan (2) forces the air, heated in heater (7), into the predrying chamber (5). The air
in the chamber is saturated with the solvent up to a content of approximately
700 g/ m 3 ( m percentage by volume this amounts to about 30% for alcohol and to

602

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

about 20% for ether). From there it passes through heat exchanger (4) into cooling
chamber (5), where a temperature of — 5°C is maintained. Here part of the solvent
is condensed and flows down along the sloping bottom of the chamber. The air
leaving the chamber contains about 300 g/m 3 of solvent. It then passes through
the heat exchanger in which it is preheated to a temperature of 15-30°C (depending
on the method used for predrying the powder) and returned to the predrying chamber.
Thus the air circulation cycle is a closed one.

^ i

4 Z

v J

Powder
3

Air \^_/ 2

Fig. 224. Schematic view of an installation for predrying powder and solvent recovery,

according to [13].

Recovery stops when the content of solvent in the air in the predrying chamber
drops to 10%.

In the installation described above 80-85 kg of liquid solvent is recovered per
100 kg of powder (calculated as dry nitrocellulose). Accepting that before predrying
the powder contains 110-115% of solvent and after it, 15-25%, the amount of
recovered liquid represents 80-90% of the solvent lost by the powder on predrying.
This accounts for 60-70% of the solvent used for the manufacture of the powder.
Since the recovered liquid contains about 5% of water, its amount is somewhat
lower.

The average composition of the recovered liquid is :

Ether 62%
Alcohol 33%
Water 5%

The composition of the liquid varies depending on the temperature inside the
predrier (at the beginning it is 10°, by the end 30°C). At a high temperature (e.g.
50°C) the content of water in the liquid is less than 3%.

The composition of the recovered liquid also varies in time. At the beginning
the liquid is rich in ether (60-70%) and relatively poor in alcohol (about 30%).
By the end alcohol predominates (about 70%), and the content of ether falls to
about 25%. The content of water remains almost constant.

THE MANUFACTURE OF SMOKELESS POWDER 603

The operational safety of this recovery installation is far from perfect. Explosions
in driers caused by the impact of steel or even brass tools against iron, by the friction
of powder strips against the iron edge of driers or by the electrification of powder
strips, are well known.

Recovery of solvent by isothermal compression. This method was proposed by
Claude [14]. It was applied to the recovery of alcohol containing camphor which
escapes during the manufacture of celluloid. With alcohol and ether this process
entails compressing the vapours to 7 atm, thus causing the condensation of the
alcohol and after that rapidly expanding them. Ether is condensed by intensive
cooling. The necessary plant was very expensive and there was risk of explosion
when the mixture of the air with alcohol and ether was compressed too rapidly. It
never attained wide application.

Absorption of solvent with sulphuric acid. This is another of the oldest methods
for the recovery of solvent. It was first used for the recovery of alcohol and ether
in the manufacture of artificial silk by the old Chardonnet process and was then
widely applied in the manufacture of powder in Germany and Austria before and
during World War I. The air containing alcohol and ether entered the tanks filled
with sulphuric acid. The tanks were cooled from outside by spraying with water.

At the Troisdorf powder factory near Cologne, lead towers sprayed inside with
sulphuric acid were used for the absorption of the solvent. Air containing alcohol
and ether entered from below in counter-current to the sulphuric acid.

The solvents were distilled from the sulphuric acid by heating to 120°C. Alcohol
is then partly converted into ether. The yield was low: only 10-12% of solvent
used for the manufacture of the powder was recovered. The disadvantages of this
method were numerous and the method is no more in use.

Absorption of solvent with cresol. During World War I Bregeat [15] in France
suggested the recovery of alcohol and ether with cresol. After successful tests in
1917, installations for solvent recovery by this method were erected in all the powder
factories in France. The installations for ducting the air containing solvent were so
arranged that they were in operation during unloading of the kneaders, loading of
the presses, extrusion of the powder strips in the presses, and the loading and un-
loading of predriers.

The method is based on the fact that with alcohol and ether cresol forms a mole-
cular compound which may then be decomposed by heating to a temperature of 130—
135°C. The absorption is conducted in towers sprayed with cresol. The air and
solvent vapour enters the towers from below. The towers are filled with ceramic
rings. A diagrammatic view of the installation is shown in Fig. 225. Air plus solvent
is introduced from below into tower (7), sprayed with cresol containing alcohol
and ether. Partly freed from solvent, the air enters tower (2), sprayed with fresh
cresol pumped from container (4). Tower (3) serves for recovering the drops of
cresol entrained by the air. Cresol containing alcohol and ether flowing down from
tower (2) is pumped by pump (J) into tower (/), from which, through pump (6)
and container (7), it is passed to the heat exchanger (8) and the retort (9), in which

«M

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

alcohol and ether are distilled off. When free from solvent it flows down through
the heat exchanger (8) and is returned to container (4).

Solvent absorption using cresol gives a good recovery yield, but large quantities
of cresol are required. This is a drawback, and so is the costly installation required.

The recovery of alcohol and ether by the Bregeat method was used in Great
Britain at Gretna during World War I, and was later adopted in Belgium at Caulilles
and in Germany at Hasloch.

Air circulation
Cresol circulation

X to rectification

Fig. 225. Schematic view of an installation for the absorption of solvent with cresol

(Bregeat method).

Absorption of solvent using water and aqueous solutions. This method, though of
no value for alcohol and ether, was used for the recovery of acetone in cordite
manufacture in Great Britain. Acetone was also recovered by using a solution of
sodium hydrogen sulphite.

Adsorption of solvent on silica gel. Attempts to use this adsorbent failed since
activated silica adsorbs moisture from air more strongly than alcohol and ether
vapour.

Adsorption of solvent with activated charcoal. This is the most modern method of
solvent recovery. It was introduced after World War I and immediately attracted
attention by the exceptionally high recovery obtained, amounting to approximately
98% of the solvent entering the plant. The first plants using activated charcoal were
used extensively in the oil industry for separating methane from heavier fractions
in natural gas.

Initially, adsorption with charcoal was not efficient when used as a main installa-
tion in nitrocellulose powder factories. With the high concentration of alcohol and

THE MANUFACTURE OF SMOKELESS POWDER

60S

ether in the air the charcoal became very hot during adsorption, and this often led
to an outbreak a£fire in the adsorber. However, the safety of the operation was con-
siderably higher when this method was applied to ancillary recovery. When working
with dilute mixtures of alcohol and ether in air, there was less risk of fire and the
yield of solvent recovered was increased. A concentration of solvent of less than
15 g/m 3 of the air (below 1 %) is now common.

This method is applicable both for ancillary and main recovery. For the latter
however, a closed cycle is avoided by introducing large amounts of fresh air into
the driers, as a diluent.

The simplest device for the adsorption of solvent on charcoal is shown in Fig.
226. The air containing alcohol and ether is introduced from below through valve
(/) into a cylindrical container (adsorber) filled with activated charcoal. Alcohol
and ether are adsorbed on the charcoal and the solvent-free air escapes through

Steamy

I Solvent
T and steam

Fig. 226. Schematic view of an activated charcoal adsorber.

valve (2). The heat of adsorption raises the temperature of the charcoal by 30-40°C
above ambient temperature. When the activated charcoal approaches saturation,
the air-solvent mixture is directed to another adsorber. (It takes 5-7 hr to saturate
an adsorber, depending on the concentration of solvent in the air.) Valves (1) and
(2) are then closed, valve (4) is opened, and steam is introduced through valve (5) to
remove the adsorbed solvent. Solvent and steam escape from the container through
aperture (4). After the solvent has been removed, the inflow of steam is stopped,
valve (4) is closed and hot air (110°C) is introduced through valve (5). This desiccat-
es the layer of charcoal and air charged with water vapour escapes through valve
(2). The evaporation of water from the charcoal during drying leads to a drop in
temperature of the charcoal to 100°C or lower. The end of drying is indicated by

606 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

a sudden rise of temperature to 1 10°C. This is a moment of great danger when the
charcoal may ignite. To prevent this, carbon dioxide or nitrogen may be introduced
into the container towards the end of drying, which may take 6-8 hr (the time is
variable and depends on the operating conditions). When drying is completed the
inflow of hot air is replaced by one of cold air. Cooling is completed when the tem-
perature falls to 15-20°C, which takes 1-2 hr depending on the season. After the
adsorbent has been dried and cooled, the container is ready for the next adsorption
cycle.

One adsorber holds a charge of charcoal ranging from 500 to 2000 kg, depending
on the dimensions of the charcoal grains. A complete adsorption cycle comprising
adsorption, distillation, drying and cooling lasts 14-18 hr.

3000 kg of steam, 0.08 m^ of water and 0.2 kWh of current are required pSt
100 kg of solvent recovered. The consumption of activated charcoal (to replace
losses due to collapse of the grains or loss of activity) is calculated at 0.05 kg per
100 kg of the solvent recovered.

The large steam consumption acted as an incentive to the development of another
and more economic process. In the "Acticarbone" method developed in France
between 1925-1926, the heater-cooler was located inside the adsorber, so that heating
and cooling the charge was much more economical. In this installation 300 kg of
steam was consumed for each 100 kg of the solvent recovered.

However, this system was abandoned because location of the steam heater inside
the adsorber was not sufficiently safe. The modified system now most widely used
is described below.

Efficiency and safety were improved by changes in design. A schematic view of
a more modern "Acticarbone" installation is shown in Fig. 227. The time for each
operation in this installation is considerably less than in earlier designs, i.e. :

Adsorption

2hr

Steam distillation

of solvent

fhr

Drying the char-

coal with hot air

fhr

Cooling

Ihr

In addition to the charge of activated charcoal A, a "thermal" layer T of broken
rock is laid in the adsorber to absorb heat should the charcoal layer ignite. The
air plus solvent passes through fan V, valve (7), layers T, A and valve (2). When
the charcoal is saturated with solvent both valves are closed, the steam is introduced
through valve (3), valve (4) is opened and the alcohol and ether are distilled off and
passed to the condenser (5). The condensed solvent and water is collected
in the lower section (6), and from there conveyed by pump (7) for rectification.
After the solvent has been distilled the inflow of steam is stopped and hot air is
passed through the adsorber, with valves (i) and (2) open. When the charcoal is
dry the air heater is shut off (it is not shown in the figure) and the charge is cooled
by means of cold water, afterwhich the adsorber is ready for another adsorption cycle.

THE MANUFACTURE OF SMOKELESS POWDER

607

Fig. 227. Schematic view of an installation for the adsorption of solvent vapours
(alcohol and ether) with activated charcoal by the "Acticarbone" method.

Fire breakers are used to protect the installation against flame that may arise
in the pipelines carrying an inflammable mixture of air and ether-alcohol vapour
and to cut off the drier from the pipelines that serve it. Breakers of the Sudlitz
[16] type (Fig. 228) have proved the most efficient. In Sudlitz's opinion, the efficiency
of the breaker is high if there is a partition in box (7), permitting easy passage of

Fig. 228. Schematic view of a Sudlitz fire breaker.

the gas and filled with heat-conducting metal. Copper mesh is used to form the
partition and spirals, rings or copper shot are used as the filling. The box and pipe-
line outlets are closed with lids of thin sheet (2) so that they break if an explosion
occurs in the pipeline. Any excess pressure leading to detonation of the gaseous
mixture is thus avoided.

608 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Cutting

In some countries (e.g. in the U.S.S.R. and in the U.S.A.) thick strips (for
cannon powders) may be cut immediately after the extrusion of the tubes. This is
possible only when pyrocollodion cotton is used for the manufacture of the powder
since it is easily soluble and requires only a small amount of solvent (about 80% by
weight of nitrocellulose). After cutting the powder is subjected to predrying. When

Fig. 229. A guillotine for cutting powder strips into flakes [7].

dried to a solvent content of 15-30%, the powder strips or tubes are usually suffi-
ciently plastic and elastic to be cut without breaking or crumbling, caking or
deforming.

The cutters are designed according to the required shape of the powder. For
cutting the powder strips into square flakes "guillotines" are often used (Fig. 229),
in which a bundle of strips is moved forward intermittently, driven by a mechanism
synchronized with the knife of the guillotine. The length of the stroke of the driving
mechanism can be varied within certain limits. The knives must be very sharp,
and are therefore changed frequently (every 0.5 hr). This greatly influences the
uniformity of the dimensions of the flakes which, in turn, affects the uniformity of
the ballistic properties of the powder.

Similar guillotines are used for cutting powder tubes. For short tubes a guillotine
of the design described above is used. Slightly different ones are used for cutting
of long tubes which are usually moved under the knife by hand.

Cutters for cutting wide powder strips into square or rectangular flakes are of
quite different design, the principle of which is shown in Fig. 230. They consist
of two systems of knives. The first is composed of two rollers with rectangular
knives of a width corresponding to the required width of the powder flakes. The
strip is thus cut lengthwise into ribbons (Fig. 231).

The second system cuts the strips crosswise, i.e., at right angles to the first one,
and consists of a fixed knife-edge against which two knife blades rotate.

THE MANUFACTURE OF SMOKELESS POWDER

609

In this system, which is used extensively in France, a frequent change of knife
blades and repeated sharpening are indispensable in order to ensure uniform proper-
ties in the powder.

In some factories cutting is combined with rolling (smoothing) the powder
strips, if the powder is to be shaped in strips or flakes. Special cutters are used for

Fig. 230. A French design for cutters for wide

powder strip, according to Vennin et al. [13];

1— rollers cutting lengthwise; 2— rotational knife

cutting crosswise; 3— fixed knife.

Fig. 231. A design for rollers cutting

lengthwise, according to Vennin

et al. [13]; a— cutting knives.

this purpose (Fig. 232). They are fitted with smoothing rollers (7) and cutting rol-
lers (2). The rollers are clamped with screws (5). Smoothing considerably reduces
the quantity of waste resulting from deformation of the powder strips on predrying.
Cutting is a fairly safe operation if the cutters are carefully earthed to prevent
them from building up static electricity, which has been a frequent cause of accidents.
Care also should be taken to prevent overheating of the cutter component, includ-
ing the knives. If overheating occurs the cutter must be stopped for some time until
the knives are cool.

Grading

After being cut the powder should be graded to remove irregularly-shaped flakes,
strips or tubes. The quantity of waste depends on the shape of the powder, its dimen-
swns and on the condition of the cutters. With flake rifle powders, for instance, the

610

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

smaller the dimensions of the powder flakes, the larger is the quantity of waste.
The thicker the strips or tubes, as extruded from the press, the smaller the quantity
of waste discarded during grading.

Fig. 232. A cutter for powder strips with smoothing and cutting rollers, according

to Yegorov [11].

The method of grading depends on whether a fine-grained powder is to be dealt
with or whether it is in the form of strips or long tubes. The former is graded mechanic-
ally, the latter manually.

Mechanical grading of fine-grained powder is carried out using vibrating or rotary
screens of various designs (Figs. 233 and 234). Dust and oversized or irregular
grains are collected in separate receivers together with undersized grains. The screen
dimensions (usually made of brass gauze) are chosen in accordance with the type
of powder being graded and the required dimensions of the grains. Waste and dust
are recycled for processing (this will be discussed in the section devoted to the proces-
sing of waste— p. 631).

Manual grading consists of inspecting the strips or tubes and discarding those
which are ill-shaped or discoloured (dark-blue stains). Grading is usuaUy carried
out on tables from which dust and waste are discarded into special boxes. It is
advisable to use tables made of clouded glass lighted from below by electric light.
This facilitates the recognition of defects in the powder (stains, wrong colour, etc.).
Powder with dark stains is rejected into a separate box so as not to mix it with powder
discarded because of faulty shape. The latter is recycled. Dark coloured powder is

THE MANUFACTURE OF SMOKELESS POWDER

611

considered unstable and is either destroyed or made into training or sporting powders
which can be used without delay.

The output of manual grading depends on the dimensions of the strips and tubes
i.e. it is higher for larger dimensions of powder. When grading one of the finest

Fig. 233. Schematic view of an installation for the mechanical grading of fine-grain
powder (vibrating screen), according to Yegorov [1 1].

Fig. 234. Schematic view of a Marot installation for grading fine-grain powder (rotary
screen), according to Yegorov [11].

types of strip powders (powder for 75 mm field gun), of dimensions 0.7 x 20 x 144 mm,
output is approximately 10 kg/hr per one worker.

Powder which is cut but neither air- nor water-dried is known as "green" powder.

612

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Final removal of solvent

A predried powder burns very slowly since the residual solvent acts as an inhibitor.
It is not until the content of residual solvent is reduced below a certain limit
that the powder attains a usable rate of burning. It also acquires ballistic stability,
so that its ballistic properties are unchanged during storage, since it contains no
volatile constituents capable of volatilizing spontaneously and thus of modifying the
ballistic properties.

The final removal of solvent from nitrocellulose powder may be carried out by
two methods:

(1) By drying at reduced pressure at a temperature of 80°C or at atmospheric
pressure at a temperature of 50-60°C.

(2) By soaking the powder in cold or hot water followed by drying.

Drying at reduced pressure. It was widely practised in Germany and often made
it possible to avoid the soaking of powder. Driers used for this purpose (Fig. 235)

Fig. 235. A drier for drying nitrocellulose powder [7].

consist of a cylinder made of boiler iron and shelves heated with hot water to a tem-
perature of 60-80°C. The cylinder is tightly closed on both sides with lids clamped
down by means of screws.

The powder is spread on cotton muslin stretched over wooden frames and placed
on the shelves. The drier is then closed and the vacuum pump set in motion. The
pump usually operates a' a pressure of 100-150 mm Hg. When the pressure becomes
steady, the screws clamping the lids are released, but the lids continue to adhere
hermetically to the drier, owing to the reduced pressure created inside. As a safety
precaution the driers work in this manner until drying has been completed. Thus
should ignition occur the lids are freely detached. This prevents the powder from burning
in a confined space, and exploding, and so destroying the drier. The capacity of the

THE MANUFACTURE OF SMOKELESS POWDER

613

9&l-

3
T3

a
o

o
o

614 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

drier is 1000-1200 kg of powder (calculated as the weight of dry nitrocellulose).
Driers of this type are commonly used for the manufacture of fine-grained powder
(e.g. rifle flake powder). Such a powder takes about 24 hr to dry at a temperature
of 80° C. Powder dried at a reduced pressure has the advantage of being less porous
than that from which the solvent is removed chiefly by soaking. After drying at
reduced pressure, the powder may be subjected to soaking to remove residual
solvent and then dried again. This treatment does not lead to the formation of ad-
ditional pores in the grain. A more modern design of vacuum-drier is shown on
Fig. 236.

An installation for the recovery of solvent (chiefly alcohol) by condensation
is connected with a drier operating at a reduced pressure. The installation works at
atmospheric pressure. It consists of two condensers cooled with cold water in which
ether and alcohol (containing some water and ether) are condensed. From 10 to 25 kg
of 95% alcohol is recovered per 100 kg of powder according to the season (more
in winter).

Drying at atmospheric pressure. Various types of driers are used to dry powder
at atmospheric pressure. In France drying houses with natural air circulation were
often used (Fig. 237). They are reinforced concrete buildings with three stout
walls and one— the blow-out wall— made of wood. The building is divided into
several compartments each of which holds 600-1000 kg of powder for drying. The
partition walls should be thick enough to withstand a high pressure in the event of
ignition of the powder, so that only the light blow-out wall is wrecked. A door in
the light wall is fitted with a reinforced concrete canopy to protect the drying house
from rain and to direct the flame in the required direction should ignition of the
powder occur. The canopy walls are of double thickness. There are openings with
brass grids in the lower section of the light wall, above the canopy and in the door of
the drying house. Air enters through the lower openings and escapes through the
upper ones. The air is circulated as indicated by arrows through radiators heated
with hot water [17]. They are located under iron shelves on which the trays of powder
are set. The trays are made of wire netting stretched over wooden frames. When fine-
grained powder is dried they are also covered with muslin. A separate chamber con-
tains an automatic system for keeping the temperature of the water which feeds the
radiators constant. This should be higher by 5°C than that inside the drying house
and is usually 60°C (inside it is 55°C).

Drying at a temperature of 55°C requires at least 24 hr, but the actual duration
depends primarily on the web thickness of the powder. As a guide the following
interrelations may be assumed between the time of the drying of strip powder at
55°C and its web thickness:

) thickness

Approximate time of drying

0.8-1.5

2.3-2.5

3.2-7.5

THE MANUFACTURE OF SMOKELESS POWDER

615

When drying has been completed the inflow of hot water to the radiators is
stopped and the door is opened for 3 hr to cool the powder before unloading.

The drying house described above is very economical since it does not consume any
mechanical energy. Its disadvantage lies in its rather poor air circulation which
prolongs the drying process and therefore exposes the powder to the action of a high
temperature for a relatively long period of time.

Shelves

Radiators

E^:

;

y '.'/.■/■'////','////»,

—_^^

/'</////, ■■■■■ ■■ ■■ '

-..•■.V-Ww -V-:- X.

Temperature
control room

Fig. 237. Schematic view of a standard

drying house for nitrocellulose powder

used in France.

Fig. 238. Schematic view of a drying
house with mechanical ventilation.

There exist a wide variety of drying houses with mechanical ventilation. Cabinet
driers (Fig. 238) are one of the simplest types. In these the powder is spread on
shelves through which there is a free passage of air. The air forced by means of a
fan passes through a heater and is then supplied from below by pipelines to each
partition of the drying house. The door opens from outside. The outer corridors
(not shown in the figure) should be wide enough to allow for the passage of the
trucks bringing the powder to be dried or removing it after drying.

There is another method (de Quinan method, Fig. 239) in which the powder is
placed in shallow cylindrical (or otherwise shaped) tubs of brass sheet, lined internally
with cloth. The bottom of the tubs are perforated. A fan forces heated air through the
bottom; the air passes and escapes through an outlet in the roof of the building.
When dry, the powder is cooled by cold compressed air. Drying in drying houses
with mechanical ventilation takes approximately half the time needed when a natural

616

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

air stream is used. Drying with mechanical ventilation has the disadvantage that a
strong stream of air favours the oxidation of the diphenylamine present in the
powder, and thus may easily lead to dark coloration (dark-blue or dark-green).
The drying of powder was formerly one of the most hazardous operations, but
since the causes of danger have been eliminated by new methods it is now no more
dangerous than any other of the processes for manufacturing nitrocellulose powder.
The following rules must always be observed :

-Lid

'Silken net

,- Powder

p -- Perforated
bottom

Wire net

. Drier

. Concrete
walls

Set-up of driers

v~7q&

Cold compressed Heated air

air for cooling

Fig. 239. Schematic view of a de Quinan drier.

(1) All shelves and other metallic parts of drying houses should be earthed.

(2) Doors in drying houses should be shut with clamps made of wood or of non-
sparking metal.

(3) The premises must be kept clean with particular attention to the removal
of spilt powder. Powder lying scattered on shelves or the floor for long periods under-
goes spontaneous decomposition with the passage of years and has often been the
cause of the ignition of powder in drying houses. Keeping the radiators clean is
particularly important.

(4) The powder must not be removed from the drier until it has been cooled to a
temperature of about 30°C.

Drying with infra-red radiation. The studies of Brun and Ratouis [18] show that
drying with infra-red rays may reduce the time necessary for the removal of solvent
by 20 or even 100 times, although for two reasons it is more expensive:

(.1) The heat supplied by infra-red radiation is more expensive (at least three times)
than steam heating.

(2) Only thin layers of powder may be dried since in thick layers the surface may
acquire a dry skin, leaving a damp layer underneath. This may upset the uniformity
of burning.

THE MANUFACTURE OF SMOKELESS POWDER 617

For the latter reason infra-red radiation is not used for drying nitrocellulose
powder.

The design of an infra-red drier will be discussed later in the text (p. 636 and
Figs. 254 and 255).

Soaking the powder. It appears that the final removal of solvent by soaking
("bathing") was first applied by Sukhinskii [19] in Russia, in 1892. By this method
the solvent is removed from the powder much more rapidly than by any other.
Soaking was widely practised in the manufacture of nitrocellulose powders during
World War I, since it considerably accelerated production. The soaking temperature
was then 80 or even 90°C which was inadvisable for three reasons:

(1) Due to the high temperature of the water some of the diphenylamine dissolves,
thus reducing the content of stabilizer in the powder; in addition, the high temperature
leads to more vigorous oxidation reactions of the diphenylamine.

(2) Due to the high temperature of the water the nitrocellulose partly coagulates
from the colloidal state particularly on the surface.

(3) Powders become porous which is not always desirable.

A soaked powder may be easily distinguished from a non-soaked one. The latter
retains a certain transparency while the former is dull, opaque (due to partial coa-
gulation) and usually dark, of a grey-greenish colour, on account of the reactions
undergone by the diphenylamine at a high temperature in the presence of water.

Saoking is most efficient if the powder has been only predried, but not dried at an
elevated temperature. On drying at an elevated temperature (50-80°C) a dry skin forms
on the surface of the powder, and this prevents the solvent inside from diffusing into
the water. For these reasons it is most desirable to soak the predried powder at a
low temperature (15-30°C) which is then gradually elevated to 50-60°C.

Vigorous prolonged soaking at a high temperature, like drying for a long period,
may reduce the stability of the powder. It is believed that the decrease of stability
is particularly pronounced when a non-dried powder, which contains a large amount
of solvent, is subject to soaking. This may be attributed partly to the fact that on
soaking a non-dried powder loses more diphenylamine than a dried powder containing
little solvent.

Long detailed investigations have shown that if the powder is predried immediately
after extrusion in a damp atmosphere facilitating moisture penetration its stability
is impaired. If however drying takes place in a confined space by the action of a stream
of dry air, without access of moisture, the stability is not reduced. In the latter case
it makes no difference whether or not the powder is soaked after drying. With modern
predrying equipment there is no fear of decreasing stability on soaking unless this
is done too vigorously (soaking at a temperature of 90°C and over should be consider-
ed harmful, at a temperature of 80°C soaking should not exceed 72 hr).

Sometimes, when a highly porous powder is wanted (e.g. a sporting or training
powder of high vivacity) the powder strips or cords are soaked in hot water immedi-
ately after extrusion from the press. A violent evaporation of solvent then ensues, the
strips or cords swell, and holes are formed inside (Bazylewicz-Kniazykowski and

618

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Partyka's method [20]). Powder may also be made porous by the previous introduc-
tion into the powder mass of salts soluble in water which are extracted by soaking.
Potassium nitrate is generally used for this purpose so that a certain amount of it
remains in the powder and enhances its vivacity.

The soaking of nitrocellulose powder in which alcohol and ether were used as a
solvent should be carried out so that the solvent (chiefly alcohol) which passes into
aqueous solution may be recovered. Nitroglycerine powder manufactured with a
volatile solvent such as acetone should never be soaked since water would remove
some of the nitroglycerine from the powder.

Soaking may often be replaced by drying at a reduced pressure, as previously
discussed (Passburg [21]). This, however, requires expensive installations and
is therefore relatively rarely used or only on a small scale.

Soaking is carried out either in cold, non-heated water at room temperature
(15-30°C) or at an elevated temperature (50-80°C).

Cold water "

' ' Steam

Hot water
Overflow

Fig. 240. A concrete basin for soaking tubular powders.

In the first case large concrete basins are used. It is advisable to line them with
glazed wall tiles. Fine-grained powder is packed in cloth sacks, although this has the
disadvantage that it hinders the circulation of water between the powder grains.
Strip and tubular powders may be soaked either in baskets made of brass mesh,
perforated brass sheets or in wooden boxes with openings through which the water
circulates immersed in a vat. To increase the water circulation and thus accelerate the
soaking process, pumps may be installed to circulate the water. The pump delivery
must be adapted to the capacity of the vats (e.g. in vats with a capacity of about
3.5 m3, water is circulated at about 300 l./hr). To facilitate the loading and unloading
of the powder an electric lift with a non-spark mechanism may be installed.

Soaking in hot water is carried out in a very similar manner, although it is more
convenient to use vats of a smaller capacity. The device for heating the water should
also serve for the circulation of the water in the tub.

A powder soaking vat of the type used in France is shown in Fig. 240. It is made
of concrete and divided by a partition into two parts, each 2.3 m long, 1.0 m wide

THE MANUFACTURE OF SMOKELESS POWDER 619

and 1.4 m deep. The outer walls are 10 cm thick and it takes a charge of 300-700 kg.
At some distance from the bottom a tough perforated sheet of brass is laid. Under
the sheet there are pipes that supply hot water and drain the water to be heated from
the lower section. Each vat has a heating system of its own with a steam injector and
is fitted with an overflow for draining the excess water into a special container. It
is advisable to use large containers for heating the water to a temperature of 50°C
and to draw it from these containers for soaking the powder. The consumption of
steam in the vats described above is:

for heating the water from 50 to 80°C— ca. 580 kg/hr
for maintaining a temperature of 80°C— 650-850 kg/hr

The soaking time depends on the temperature and the type of the powder, i.e.
the lower the temperature of the water and the larger the web thickness of powder', the
longer the soaking time. In some factories, as a guide it is assumed that decrease of the
soaking temperature for coarse-grained powders (a thickness of ca. 1 mm) by 10°C
requires a threefold longer soaking time, e.g. : at a temperature of 80°C soaking lasts
for 32 hr, at a temperature of 70°C it lasts for 96 hr, and at a temperature of
60° C it lasts 288 hr.

Other data show that finer powders (of a thickness of 0.5-2 mm) require an
increase in processing time by 50% if the soaking temperature is decreased
by 10%.

The soaking time also depends on whether the powder has been previously dried
(to 1.5-2% of volatile matter) or only predried (to 10-15% of volatile matter). In
the latter case soaking is much more effective although there is a danger that stability
may be reduced.

Soaking greatly increases the vivacity of powder to an extent which depends on
the method applied. The following example may illustrate the efficacy of soaking
preceded or not by drying. One sample of powder was first subjected to drying for
24 hr at a temperature of 55°C, then to soaking for 12 hr at a temperature of 80°C
and finally to drying for 24 hr at a temperature of 55°C. Another portion of the same
powder was first soaked for 8 hr at a temperature of 50°C, then for 10 hr at a tempera-
ture of 80°C and finally dried for 24 hr at a temperature of 55°C.

It was found that the powder soaked without previous drying was faster burning
and in the 75 mm gun gave a muzzle velocity higher by 9 m/sec and a pressure
higher by 37 kg/cm2 than the powder that was dried before soaking.

Table 180 gives data showing the effect of additional soaking (after polishing— see
p. 627) on the increase of vivacity of rifle flake powder.

Vigorous soaking may distort the powder strips or tubes, especially thick ones.
This is a further reason for recommending that soaking should be conducted at a
gradually rising temperature, e.g. : 12 hr at a temperature of 30°C, 24 hr at a tempera-
ture of 45°C and 48 hr at a temperature of 60°C. To prevent distortion on soaking,
especially at higher temperatures, the strips or tubes can be clipped into bunches
with brass collars fitted with brass springs.

620

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

After soaking the powder is dried to remove moisture together with some of the
residual solvent. Drying at a temperature of 55° requires 4-160 hr depending upon
the web thickness of the powder (the duration of drying is given for a drying house
with natural air circulation).

Table 180

Additional soaking

Increase by a value of

Av j Ap

Time
hr

Temperature
°C

m/sec

kg/cm 2

24
24
24

80
85
90

36^14
45-60
60-70

300-450
350-550
500-600

After drying, cannon powders, i.e. strips or tubes, are subjected to a second grading
as they may have been distorted or discoloured on soaking.

, , Fresh
water

To rectification

Fig. 241. A battery of powder soaking vats (the content of alcohol in various parts

of the vat is indicated).

The recovery of solvent from the water after soaking

The soaking of nitrocellulose powder involves recovery of the solvent (chiefly
alcohol) which passes into the water. It is worth while recovering this solvent if the
content of alcohol in the soaking water is not less than 5%. By applying the battery
counter-current soaking system the alcohol content in the water may be increased to
15-20% and the recovery of solvent thus made more profitable.

Soaking together with solvent recovery was adopted in France during World
War I. A battery consisting of three vats was used (Fig. 241). Fresh water heated by
steam injection was introduced into vat (7). From there it passed through an overflow
and injector to the upper section of vat (2) and then, to the lower section of vat (5).
The un-soaked powder was loaded in reverse order: first into vat (5), from there
after soaking into vat (2), and, finally, into vat (7). By soaking at a temperature of
50°C, the following concentrations of alcohol were obtained in the vats:

Vat 1 2.5% in lower section

11% in upper section

THE MANUFACTURE OF SMOKELESS POWDER 621

Vat 2 11% in lower section

13% in upper section

Vat 3 15% in lower section

20% in upper section

The water from the vat containing an average of 18.5% alcohol was recycled for
rectification.

It was thought that this method might lead to much greater losses of diphenyl-
amine than with single stage soaking in water (when the concentration of alcohol is
less than 6% by the end of bathing). This however was not so. The losses of diphenyl-
amine on soaking by the battery method are less than 0.01 %.

Besides recovering alcohol from the water after soaking it is also necessary to
recover the ether and alcohol escaping from the vat as vapour, together with steam,
when soaking with hot water. E.g. a powder containing 25% volatile matter (thus only
predried and not dried) gives the following composition for the liquid formed by the
condensation of vapours evolved from the soaking vat (according to Bonneaud [22],
Table 181).

Table 181

Components of
gaseous phase

Soaking at 50°C
Start | End

Soaking
Start

at 80°C
End

Ether (%)
Alcohol (%)
Water (%)

92
5
3

52
32
16

In these experiments the quantity of liquids condensed from the vapours evolved
during soaking at a temperature of 80°C amounted to 2-3%, calculated on the
quantity of powder used.

To recover the solvent vapours evolved during soaking in the hot, the vat is
fitted with an air-tight lid and a pipe through which vapours for condensation are
conducted. 30 kg of liquid containing 96% of ether were obtained per 700 kg of
powder in the vat.

At another plant 4 kg of liquid containing 71 % ether and 16% alcohol per 100 kg
of the powder soaked at temperatures of 60 and 80°C were obtained by condensation
of the solvent vapours.

The content of residual solvent and moisture in the powder

Powder produced with a volatile solvent should be freed from it as thoroughly as
possible since too large a content of residual solvent is detrimental to the ballistic
stability of the powder. In nitrocellulose powders the content of residual solvent
should be lower than 1 % ; in coarser powders (thicker flakes, strips or tubes) its
content may be relatively higher, while it is relatively lower in finer ones. Powders

622 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

gelatinized on the surface with camphor also contain a certain amount of camphor
unbound with nitrocellulose which may volatilize on heating or storing.

Apart from residual solvent, nitrocellulose powder contains a certain quantity of
water, either in the form of residual moisture, incompletely removed by drying, or
in the form of moisture introduced to the powder.

Nitrocellulose powder is slightly hygroscopic since it partly consists of non-colloidal
fibrous nitrocellulose which is hygroscopic. The colloidal part of the powder is less
hygroscopic but retains a certain amount of residual solvent, chiefly alcohol, which
as a hydrophilic substance increases the hygroscopicity of the whole mass. The
hygroscopicity of nitrocellulose powder mainly depends on the total nitrogen content
in the nitrocellulose, the latter being less hygroscopic the higher its nitrogen content
(Vol. II, p. 283). Powder made of nitrocellulose with a total nitrogen content of
13% absorbs 1.0-1.5% of moisture. With a nitrogen content of 12.5-13.0% the mois-
ture content of the powder may increase to 1.5-2.0%. These figures are characteristic
for the climate of Central Europe. They are lower in a dry climate and higher in a
humid one. According to Hansen [23] for a flake or tubular powder of web thickness
over 0.8 mm the hygroscopicity may be expressed by the formula:

where a is a constant, x is the web thickness of the powder flake or tube, and y the
moisture content as determined by drying at a temperature of 80°C for 3 hr.

Nitrocellulose powders gelatinized on the surface with centralite, camphor or
nitro compounds are less hygroscopic since the layer of gel on the surface constitutes
a non-hygroscopic coating which prevents the powder inside from attracting moisture.
Nitrocellulose powders containing aromatic nitro compounds, e.g. dinitrotoluene
(DNT) or dinitroxylene (DNX) are less hygroscopic.

Davis [12] reports the following figures characteristic of the hygroscopicity of
pyrocollodion cotton containing various substances :

Pyrocollodion cotton with

Hygroscopicity

5 % hydrocellulose

2.79%

10% crystalline DNX

2.09%

10% DNX oil

1.99%

10% crystalline DNT

1.92%

20% crystalline DNT

1.23%

25% crystalline DNT

1.06%

Nitroglycerine and nitrodiglycol powders are the least hygroscopic (practically
non-hygroscopic) since nitroglycerine-nitrocellulose or nitrodiglycol-nitrocellulose
gels are virtually non-hygroscopic. For this reason even those nitroglycerine and nitro-
diglycol powders which contain highly nitrated nitrocellulose, partly in a fibrous state,
are non-hygroscopic.

Analytical methods for the determination of moisture content and residual
solvent content are of the utmost importance in view of the effect exerted by these pro-

THE MANUFACTURE OF SMOKELESS POWDER

623

perties on the ballistic properties of the powder. Various methods are used in dif-
ferent countries. During World War I the moisture content in nitrocellulose powder
was determined as follows :

in Russia
in France
in the U.S.A.

6 hr of drying at 100°C
4 hr of drying at 60°C
6 hr of drying at 60°C,
under reduced pressure.

Vieille investigated the loss in the weight of nitrocellulose powder on heating at
temperatures from 40 to 130°C. He obtained a typical curve (Fig. 242) which

1.8

i i i i '■-

ii iii

1 1 ■■

1.6

^JL-°— * °

^-°— a ~~ t

* 1.4

%1.2

* 1.0

°0.8

3 0.6

0.4

0.2

i i i i

1 1 1 1 1 1 1

1 1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Days of drying at 60°C

Fig. 242. Loss of weight of nitrocellulose powder at a temperature of

60°C: /—evolution of moisture ("elimination"), //—evolution of

solvent ("coefficient of emission") [24].

consists of two parts: / is very steep and // is gently inclined. Part /of the curve
corresponds mainly to the evolution of moisture and part // to that of the residual
solvent.

By comparing the losses of residual solvent at temperatures of 110, 75 and 40°C
Vieille found that identical losses are obtained on heating the powder for an identical
number of

hours at 110°C
days at 75°C
months at 40°C

In France Vieille's experiments provided a basis for the determination of the
loss of weight during the first 4 hr of drying at a temperature of 110°C, i.e. the so-
called "elimination" (period /) and of the loss of weight during the next 16 hr, i.e.
the so-called "coefficient of emission" (period II).

624 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The coefficient of emission is a function of the content of residual solvent in a pow-
der. It usually amounts to 0.3-1 .00 %, while the content of residual solvent ranges from
1.5 to 3.5%. The determination of the coefficient of emission is a convenient and
rapid method for acquiring an idea of the content of residual solvent.

The total content of residual solvent is determined by partially dissolving a weighed
sample of the powder in a solvent (a mixture of alcohol and ether) and adding water
so as to precipitate the nitrocellulose from the solution in flocculant form. The
weight of this nitrocellulose is determined by evaporation to dryness, repeated dis-
solution and precipitation with water, and by final drying. The difference between the
weight of the powder sample and that of the nitrocellulose is the weight of the residual
solvent.

Davis [25] distinguishes :

(1) Total volatile content (TV)

(2) External moisture (EM) (determined by drying at 100°C, for 1 hr)

(3) Residual solvent (RS) (calculated from the difference of RS = TV-EM)
Increase in the moisture content leads to a decrease in the vivacity of the powder.

According to Vieille the coefficient of the vivacity of powder (p. 530) decreases by
13% when the moisture content increases by 1%.

In small arms and small calibre cannon increase of the moisture content by 0. 1 %
reduces the muzzle velocity by 4-5 m/sec and the pressure by 50-70 kg/cm 2 .

In the climate of Central Europe the moisture content, as determined by drying
at a temperature of 60°C for 4 hr, should be constant within the limits of 1-2%.
If the moisture content is determined under more drastic conditions the figure is
correspondingly higher.

Surface gelatinization

Attempts have been made to improve the ballistic properties of fine-grained
(flake or tubular) powders for rifles by coating the grains with a layer which would
burn more slowly than the inner part of a grain. Such grains would give a better
ballistic effect (would be more "progressive") since after a certain time of burning
the surface of the flakes would decrease and the higher rate of burning would thus
facilitate maintaining a steady pressure.

At first, attempts were made to coat the powder grains with gum arabic or gelatine.
An aqueous solution of these substances left a layer of phlegmatizing substance on
drying. In practice, powders of this kind gave a somewhat better ballistic effect
(were more "progressive") although the most favourable results were obtained by the
gelatinization of the surface of the grains with centralite, camphor or DNT. Since
nitrocellulose powder contains a considerable quantity of fibrous, non-gelatinized
guncotton, coating the grains with a non-volatile solvent gives a totally gelatinized
slow burning surface layer. The concentration of the solvent gradually decreases
towards the inside of the grains, so that the rate of burning of the powder increases
with the combustion of the outer layers.

THE MANUFACTURE OF SMOKELESS POWDER

625

Surface gelatinization is almost always accompanied by polishing the grains
with a small amount of graphite. The duration and temperature of gelatinization
depends to a great extent on the type of gelatinizing agent. With a strongly gelatinizing
agent the process is shortened and the temperature may be lower. E.g. when using
camphor (the strongest employed gelatinizing agent) the temperature may be kept at
30-35°C, while with centralite it should be 50-55°C and with DNT (the weakest
gelatinizing agent used) 80-90°C.

The polishing of rifle powder

Flake or tubular rifle powder should be subjected to surface gelatinization and
coating with graphite. This is usually called "polishing" and constitutes the most
delicate operation in the manufacture of rifle powder (for calibres from 7.6-20 mm).

Polishing is usually carried out in copper or brass drums. These may be cylindrical,
with a horizontal axle, and fitted with an opening for loading and unloading and with
another for supplying the solutions and predrying the powder (Fig. 243). Drums for

1 1

Fig. 243. Schematic view of a design for a drum for polishing fine-grained nitrocellulose

powder.

the manufacture of dragees used in the confectionery and pharmaceutical industries
("sweet barrels") may also be adopted (Fig. 244). Some designs of dragee apparatus
provide for heating jackets which surround the polishing vessel, thus making it
possible to obtain a higher temperature (50-80°C) inside. The inside of the drums is
ribbed which enhances the polishing effect by making the powder rise higher as the
drum rotates. The capacity of the drum ranges from 100-300 kg of powder. The
rotational speed is approximately 30 r.p.m.

Polishing is usually carried out as follows : the soaked and dried powder is placed
into the drum together with the graphite (0.01 % by weight of powder). The presence
of the graphite from the beginning of the operation increases the conductivity of the
powder and prevents it accumulating static electricity, thus reducing the possibility
of accidents during polishing.

After a preliminary "dry" polishing for several minutes, an alcoholic solution of a
gelatinizing substance— centralite (p. 645) (3% in relation to the weight of powder)
or camphor (1.5-2%)— is introduced through an opening in the side wall of the drum.
These substances are usually injected in the form of a 10-20% solution sprayed with
compressed air. The solution should have a temperature of about 50°C when centralite

626

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

is used, or 30-35 c C when camphor is used. With centralite the temperature of the
powder in the drum should not exceed 40-50°C and with camphor 25-30°C. These tem-
peratures may be attained by maintaining a sufficiently high temperature in the
premises where the polishing takes place. The ldtter temperature should be lower by
some degrees than that which is required within the drum since the temperature of
the drum contents is somewhat higher than ambient temperature.

Fig. 244. A dragee apparatus ("sweetie barrel") for polishing fine-grained nitro-
cellulose powder [7].

The solution is introduced into the drum, which is then closed with a lid and set
in motion for one or more hours. A side lid is then opened permitting alcohol to
evaporate thereby gradually predrying the powder. During the latter operation a
small amount of graphite is added so that its total content amounts to 0.25-0.3 %.
The evaporation of alcohol from the drum may be combined with the recovery of
this solvent. For this purpose the lid is not removed but connected with a duct
producing a reduced pressure within the drum. The solvent thus drawn off is then
recovered either by condensation or adsorption.

The whole polishing operation lasts for 4-20 hr depending on the method em-
ployed. The efficiency of polishing may be checked by measuring the apparent den-
sity, which amounts to about 0.700 before polishing and to 0.800-0.900 afterwards. In
some factories a charge of wooden balls equal to £ of the charge of the powder is
added when polishing flake powder. The balls hasten the surface smoothing of the
powder flakes and their acquisition of the required apparent density. They also
prevent caking of the powder grains by the gelatinized solution.

There are various modifications of the polishing process, e.g. the introduction
of centralite or camphor without a solvent or in the form of a hot (e.g. 80°C) aqueous

THE MANUFACTURE OF SMOKELESS POWDER 627

emulsion. This is the last stage of polishing, based on the volatilization of the water.
The use of a tilting drum with a heating jacket hastens evaporation of the water or
solvent.

After polishing, the powder 4S soaked for a short period (4-8 hr) at temperatures
from 50-80°C, and then dried. This is an operation of great importance designed
to remove from the powder any excess of gelatinizing agent unbonded with nitro-
cellulose. The content of the gelatinizing agent in the powder is thus stabilized. If
the powder is not soaked after polishing, but only dried, it may change its ballistic
properties on storage. In particular there may be a decrease of muzzle velocity and
an increase of pressure, i.e. the powder becomes less progressive, because the excess
of the gelatinizer unbonded with nitrocellulose present in the non-soaked powder
penetrates into the deeper layers of the powder on storage.

In some countries (U.S.A.) DNT is used for the surface coating of tubular rifle
powder, instead of centralite or camphor. The polishing is carried out in the presence
of water at a temperature of 80°C, i.e. above the melting point of DNT. As the
polishing proceeds, the water is removed by evaporation under reduced pressure,
the DNT remaining on the surface of the powder tubes.

Polishing was a hazardous operation until the drums and all the pipelines supply-
ing alcoholic solutions were carefully earthed. Before such precautionary measures
were taken there were frequent serious accidents of ignition during polishing.

The polished powder, bathed and dried, should undergo final screening to separate
dust and powder grains.

Finishing the powder

In the finishing of powder there are two operations : blending and damping.

Blending. This operation aims at obtaining a large, uniform lot of the product,
amounting to 5000-50,000 kg. Powder is produced in small batches (in a kneader
60 kg; from a press still smaller quantities are extruded), so that it is difficult to
obtain an identical product each time, due to slight deviations in weighing CP t and
CP 2 , in adding the solvent, diphenylamine etc. At later stages of processing, uniform
batches increase in size and reach a quantity corresponding to the capacity of the
soaking vat or the drier. Lots amounting to 1000 kg are obtained in this way. In
rifle powders a polishing drum charge, which amounts to about 300 kg of powder, is
considered to be acceptably uniform.

For powders which require a high precision in manufacture (polished rifle pow-
ders) the ballistic properties are determined for the contents of each polishing drum
charge and the contents of all the drums are blended on the principle that powders
of various ballistic properties give a mixture with properties corresponding to the
arithmetic mean.

The ballistic properties of ordinary cannon powders are more easily controlled,
therefore those of individual batches intended for blending are not examined. Excep-
tions are made when manufacture is restarted after a long interruption or is commen-

628

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ced with a new type of nitrocellulose, or when there are doubts as to the quality of
production for some reason or other. In these cases every manufactured batch (e.g.
the charge of every drier) is subjected to ballistic tests and then blended according
to the rule expounded below.

The blending method depends on whether the powder is a grain (flake or short-
tube) or a strip powder (or long-tube).

In the first case (cannon or rifle grain powder) blending is easy and simple funnel
equipment, similar to that for blackpowder blending (p. 359) may be used. In the
U.S.A. a continuously operating plant, consisting of towers, is used for blending.
The towers contain a series of funnels placed one above the other (Fig. 245). The

Fig. 245. Schematic view of equipment for blending powder by continuous working.

powder from several drying houses is conveyed to the upper funnel. It flows succes-
sively through all the funnels, is partially blended, and falls onto a conveyer which
carries it to the upper section of another system of funnels, where the procedure is
repeated. Again the powder flows down, is further blended, falls onto the conveyer
and is carried to the upper section of the first system of funnels and so on. After
several passages the powder is sufficiently blended. A general view of a mixing house
is given in Figs. 246 and 247.

Strip powder or long tubular powder can be blended by hand, in the following man-
ner. Sacks with powder from different drying houses are placed around a cloth onto
which the workers pour out the strips or tubes and blend them by holding the four
corners of the cloth. The powder so blended is poured back into the sacks and the
whole process is repeated.

Figure 248 shows an arrangement for blending operations that permits them to be
carried out meticulously. Diagram / represents the order in which the sacks from
6 units (driers) are placed for blending. Diagram II shows the final position of the

THE MANUFACTURE OF SMOKELESS POWDER

629

Fig. 246. Mixing house for nitrocellulose powder. Each day's output is hoisted to

the upper floor where it passes through a mixing apparatus to form uniform batches

of 10-50 tons (Bofors Nobelkrut Factory).

Fig. 247. Smokeless powder blending tower, according to Zaehringer [26].

630

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

sacks. A worker walks along each side of the rectangle, as indicated by arrows and
empties out the sacks of tubes in turn. After blending on the cloth the contents
are dumped into the sacks denoted by letter A in the figure. The next stage of blending
on the cloth gives a new, more or less uniform batch denoted by letter B. After the
contents of all the sacks placed around the cloth have been blended, the blending
operation with sacks A, B etc. is repeated several times until a sufficiently uniform
product is obtained.

5 13 5 13
OOOOOO

40
20
50

20
SO

L 6 C L 6 C
OOOOOO

i«o

[do

o«f

TWO

OOOOOO

15 3 15 3

A E I A f /

Fig. 248. An arrangement for blending larger grains of powder.

Damping. After soaking and drying the powder usually contains less moisture
than that required by standard specifications and should therefore be damped.
Methods of damping, like those of blending, depend on whether the powder is in
grains, strips or long tubes.

The simplest method of damping granular powder consists of loading it in a
rotary drum and injecting the calculated amount of water into it, in accordance
with an analytical determination of the powder's moisture content. After rotating
the drum with the lid closed for about 30 min the moisture content in the powder
becomes uniform.

Strip or tubular powder is usually damped by placing it in a room, either on
shelves or on grids standing on the floor, on which vessels of water are placed or
water is spilled. A sample of powder is taken every few hours and its moisture content
is determined. After one or more days, when the moisture content reaches the standard
specification, the powder is taken out of the damping machine and poured into sealed
boxes.

The standard specifications permit the moisture content to range within certain
limits. This enables the producer to regulate the moisture content to obtain the
required ballistic properties. If the powder is too vivacious it should be moistened
to the upper limit, while if it is less vivacious, the lower limit is preferable. In deter-
mining the moisture content the following practical rule should be observed. In rifle
powder an increase of the moisture content by 0.01 % lowers the muzzle velocity
v o by 4-5 m/sec and the pressure by 50-70 kg/cm 2 . In cannon powder the variations
of ballistic properties are smaller and depend on the type of powder and on the
calibre of the gun.

THE MANUFACTURE OF SMOKELESS POWDER 631

The processing of waste products

There are two principal types of waste products:

A. The waste products of non-soaked powder,

B. The waste products of soaked powder.

The utilization of these waste materials differs in principle. The former are re-
turned to the kneaders to be reworked while, as a rule, the latter are never recycled
in the normal manufacture of powder. It is generally recognized that a soaked
powder must not be added to the powder dough since it reduces the chemical
stability of the powder, except when it is intended for immediate use.

Waste products A derive from various stages in the manufacture of the powder,
prior to soaking. They are :

(1) Scrap comprising remnants of dough from the kneaders and conveyers.

(2) Shapeless strips or tubes extruded from the press.

(3) Cakes remaining in the press between the bottom of the press piston and
the die.

(4) Flakes, strips or tubes with non-standard dimensions or shapes, graded after
cutting, non-dried or sometimes dried (when graded after drying).

More of materials (2) and (4) is produced from powder of smaller dimensions.

These products, are processed by dissolving them in a mixture of alcohol and
ether, kneading into a dough, filtering in a hydraulic press to remove mechanical
impurities and adding to the fresh powder mass in the kneaders.

Since products (1) and (2) contain a considerable amount of solvent (60-100%),
they are loaded directly into the kneader together with the solvent, kneaded for
1-2 hr and filtered in a press in which the die is replaced by a steel plate with circular
openings, approximately 1 mm dia. Since there is less ether in the residual solvent
in the powder than is primarily used for its manufacture, the solvent added to the
waste products should be richer in ether (approximately 70% by weight of ether
and 30% by weight of alcohol).

Waste products (4) take longer to dissolve since they contain less residual solvent
(20-30%). They are poured into hermetically closed tins (in batches of 30-35 kg)
and flooded with solvent, so that its total amount ranges from 100-150%. The solvent
should be rich in ether (70-75% by weight of ether and 25-30% by weight of alcohol)
as for waste products (1) and (2). After the solvent has been poured in, the tins are
sealed, and tumbled through 90° every 15 or 30 min for 2-4 hr. The thicker the
powder strips or tubes, the longer they take to dissolve. The tins are then turned
upside down and a few hours later returned to the normal position.

With fine-grained powders (web thickness of 0.5-1.5 mm) the contents of the
tins may be transferred after a few hours to the kneaders to produce a uniform mass.
With coarse-grained powders the dissolution of the waste material requires from
four to ten days or more; e.g., with a web thickness of 3 mm the time is 4 days,
while with a web thickness of 7 mm it increases to twenty days. After being mixed
in the kneaders the dough is filtered and added to the normal dough in the kneaders.

632 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In principle waste products (3) are processed in the same way as waste products
(4), although they contain a large amount of solvent. The reason for this is that
the cakes are thick and therefore dissolve slowly. To hasten the process the cakes
may be cut into several pieces before being placed in the tins.

The quantity of waste products added to the fresh dough in the kneaders varies
depending on the amount of factory waste arising during manufacture. From 20 to 90 kg
of waste products may be added per 100 kg of fresh nitrocellulose. Naturally the
waste products added to fresh powder should have a suitable nitrogen content and
total solubility.

Waste materials from soaked powder are used chiefly for preliminary ballistic
tests. They can also be converted into less valuable types of powder, earmarked for
rapid consumption (practice or sporting powders). Such powders are treated with
solvent in the same way as waste products (4), although they require a longer time
to dissolve. Thus, during the World Wars, when powders were consumed rapidly,
soaked waste products were utilized for this purpose.

THE STABILIZATION OF AN UNSTABLE POWDER

If the stability of the powder does not deviate greatly from the required standard,
it may be improved by soaking it in an alcoholic solution of diphenylamine.

For this purpose the powder is put into cylinders, filled with a diphenylamine
solution, sealed and left to stand for several days. The concentration of diphenylamine
must be such that its total amount is no higher than 0.5-1.0% of the weight of the
powder.

Some methods recommend the use of an alcoholic solution with a small addition
of ether, e.g. one part of ether per five parts of alcohol. The ether causes part of
the nitrocellulose to swell and facilitates the penetration of the diphenylamine solution
into the strips or tubes. It should not be used in such an amount as to cause the
nitrocellulose to dissolve.

BALL-GRAIN POWDER

Between 1936 and 1940 Olsen et al. [27] designed a process for the manufacture of
nitrocellulose powder in the form of uniformly shaped balls (Fig. 250). The manu-
facturing process at Western Cartridge Co., Division of Olin Industries, Inc. at
East Alton, Illinois, U.S.A. has been described by Olive [28].

A diagrammatic presentation of the process is shown in Fig. 249. The nitrocellulose
containing about 13.45% N is stabilized in kiers, cut in beaters and introduced under
water into still (7) fitted with a stirrer, containing ethyl acetate with a small amount
of diphenylamine (Fig. 251). Calcium carbonate is also added. The presence of water
is not harmful when dissolving nitrocellulose in ethyl acetate. Diphenylamine remains
in the organic solvent phase (ethyl acetate), thus neutralizing the acid products
dissolved in the ethyl acetate, while the calcium carbonate remains suspended in the

THE MANUFACTURE OF SMOKELESS POWDER

633

aqueous phase and neutralizes the acid derived from nitrocellulose and passing from
the organic to the aqueous phase. The neutralization of the acid products in the
nitrocellulose and the final stabilization of the nitrocellulose are thus completed,
Olsen claims that for this reason there is no need to use a completely stabilized nitro-
cellulose since the same process may be performed more quickly and equally success-
fully with a nitrocellulose which has not been stabilized in boilers and poachers.

Fig. 250. Ball-grain powder, according to Olin Industries, Inc. [29].

After the contents have been mixed for 0.5 hr a solution of a protective colloid
(gum arabic or starch) is added to secure uniformity of the suspension. The material
is then stirred vigorously until small balls are formed. From that moment the proce-
dure varies depending on whether porous, fast burning or dense, slow burning balls
are required. In the first case ethyl acetate is distilled off rapidly by reducing the
pressure in the still. During distillation the balls harden, but should retain their
shape. If the solvent is distilled too quickly, the grains become elongated. The dis-
tillation rate should be such that the solvent is evaporated from the surface of the
grains no faster than it moves from the interior to the surface. The distillation should
therefore be slow at first and more rapid towards the end when the hard surface of
the balls is already shaped. Since they contain much water, drying, which is the next
operation after grading, makes them porous.

If, however, the balls are dehydrated before hardening, i.e. before distillation of
the solvent, they have a high density. To dehydrate them some sodium sulphate is

634

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

introduced into the apparatus. Due to the high osmotic pressure of the aqueous
solution surrounding the balls water moves from them to the solution, so that they
are dehydrated. The degree of dehydration is controlled by altering the time, the
temperature and the concentration of the sodium sulphate.

The distillation of ethyl acetate is carried out either by reducing the pressure in
the apparatus or by raising the temperature to 77°C (the boiling point of ethyl acetate
under atmospheric pressure).

Fig. 251. General view of the large still (Courtesy Olin Industries, Inc.) [29].

The distillate (in receptacle (2)) consists of a layer of ethyl acetate plus water
and a layer of water in which 8% ethyl acetate is dissolved. The aqueous layer is
rectified to recover the solvent. In all, 95 % of the ethyl acetate used is recovered from
the distillate and recycled for manufacture.

After the balls have cooled in the slurry they may be coated with nitroglycerine.
For this purpose they are treated with an aqueous emulsion of nitroglycerine dissolved
in toluene, using nitroglycerine up to 15% of the weight of the powder. By heating
at a reduced pressure the toluene is distilled off, leaving a solid solution of nitrocellu-
lose in the nitroglycerine deposited as a layer on the surface of the balls. This surface
may then be coated with centralite, applied in a similar manner, i.e. in the form of

THE MANUFACTURE OF SMOKELESS POWDER

635

an emulsion dissolved in a solvent immiscible with water, in which the nitrocellulose
is insoluble, i.e. in toluene.

If the ball-grain powder is to be a true double base powder with a relatively high
content of nitroglycerine, the nitroglycerine is usually introduced at the first stage
of manufacture, together with the ethyl acetate solvent. This forms a solution of
nitrocellulose and diphenylamine in nitroglycerine and ethyl acetate. Subsequent
procedure is similar to that described above, i.e. ethyl acetate is rapidly mixed in r
the balls are dehydrated and solvent distilled off. The whole operation, from dissolving
to the end of graining, takes approximately 16 hr for a batch of 3000 kg.

The aqueous slurry of grains produced in this way is conveyed to storage tank
(5) and held there pending a laboratory report on the suitability of the product.
If this shows that balls have incorrect dimensions, they are separated from the water
and returned to be re-dissolved in ethyl acetate and recycled.

If the laboratory report is satisfactory, the balls in the slurry are graded by
pumping the slurry over a succession of water-sprayed, rotating, drum-type screens.
Fig. 252 gives an idea of their design. The graded balls are directed to storage tanks

Fig. 252. Diagrammatic presentation of rotating drum screens, according to P. Brown

[30].

(5), from which the slurry is pumped over into a still (6~) in which the balls are coated
with a gelatinizing agent— usually dibutyl or diphenyl phthalate, with or without
the addition of carbamite (centralite). Formerly centralite alone was used for coating.
The gelatinizer is introduced into the still as an emulsion. The proportion of
water to powder in the still is 3 1. of water per 1 kg of powder. Steam is introduced
into the heating jacket and the charge is slowly agitated until the gelatinizer has
penetrated to the desired depth. To coat a batch of 5000 kg takes 4-5 hr, after which
the coated grains are separated from the water on a vacuum filter (7). A cross-section
of a coated grain is given in Fig. 253. After filtering about 8 % of water remains in the
powder. The damp powder is dried on a conveyer in a tunnel drier equipped with
infra-red lamps (<$). The conveyer is a travelling belt of rubber and canvas 48 in. wide

636

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

and 80 ft long (Figs. 254 and 255). The tunnel walls consist of light metal panels,
insulated and hinged to serve as explosion hatches. One hundred and forty infra-red
lamps, each of 250 W, are arranged in twenty-eight rows of five lamps each, more closely

Fig. 253. Cross-section of a ball powder grain magnified 300 times. The light ring
on the outside indicates the location of the gelatinizer (phlegmatizer). (Courtesy Olin

Industries, Inc. [29].)

Fig. 254. Diagrammatic presentation of infra-red drier, according to P. Brown [30].

spaced at the wet than at the dry end. They are placed about 18 in. above the bed
of powder. A relatively weak current of air flows counter to the direction of movement
of the powder.

.__.,-„„._,._

THE MANUFACTURE OF SMOKELESS POWDER

637

The damp powder is fed directly to the belt from a hopper containing an agitator,
and drying is accomplished at a rate of 140 lb (53 kg) per drier per hour, at a tem-
perature of 72°C. The air supplied is heated to about 50°C by steam coils, and the
temperature in the tunnel rises due to the heat from the lamps.

Fig. 255. General view of infra-red drier (Courtesy Olin Industries, Inc. [28]).

The temperature of the powder is controlled by thermocouples on the surface
of the moving bed from which three banks of lamps are automatically controlled.
In addition six other thermocouples per drier actuate recorders but do not control.

Power consumption of the driers averages 0.196 kWh per pound of dry powder
and drying time is approximately 60 min. The operation is safe. The only fire which
occurred between 1942 and 1946 was one that was started deliberately to see what
the result would be. It was brought under control so quickly that production was
resumed in less than two hours, with no damage to the lamps, belt or housing (Olive
(28]).

Since the shape of the balls is ballistically unfavourable, they may be flattened
between rolls before the final drying so that they acquire a more favourable form.

The balls are conveyed to the rolls in a slurry from apparatus (6) and pass through
a. feed tank, and then between the rolls (9). A general view of the rolls is given in

638

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Fig. 256. Larger balls may be flattened to reduce their dimensions and burning time.
E.g. U.S. 0.45 ammunition requires balls of 0.006-0.010 in. dia. They may be pro-
duced by rolling balls to a web (minimum) thickness of 0.004 in. Rolled powder
grains are shown in Fig. 257.

Fig. 256. General view of rolling operations (Courtesy Olin Industries, Inc. [29D.

After flattening, the slurry passes to the centrifuge (10) (Fig. 249) for dehydration
and then to the drier (8), as described above.

The ball-grain powder is then processed in the conventional manner, i.e. being
weighed (11), polished and coated with graphite (12), blended in towers (13), similar
to those described above (pp. 628-629), and stored (14).

The gravimetric density of ball-grain powder varies from 0.400 to 0.975.

Olsen suggested using old nitrocellulose powder, withdrawn from use due to
insufficient chemical stability, for the manufacture of ball-grain powder, instead of
nitrocellulose. The powder is milled in disintegrators under water containing a sus-
pension of calcium carbonate to neutralize the acid decomposition products. Coarse
grains are thus formed which after the excess water has been removed in a centrifuge
are introduced into ethyl acetate containing diphenylamine. The substance is then
stabilized and traces of acid are removed. Further processing is as described above.

The advantage of this method of powder manufacture is that it is quicker and

THE MANUFACTURE OF SMOKELESS POWDER 639

much safer than the usual methods, since kneading, pressing and cutting are elimi-
nated and all operations are carried out under water.

Olin Industries [29] report that the safety is further enhanced by the fact that
at any moment a relatively small amount" of material is being processed.

A third important safety feature is the continuous drying by infra-red radiation.
No "drying" to remove solvent is required and so no dangerous mixtures of organic

Fig. 257. Powder rolled from 0.020-0.025" dia. to a thickness of 0.018" (Courtesy Olin

Industries, Inc. [29]).

vapours in air are formed as in the conventional process. Approximately 100 lb
of powder is dried at a time. This amount is within the safety limits for this type of
equipment.

Olin Industries report that no fatal accident attributable to ball powder has
occurred in twelve years of manufacture, during which time more than 60,000,000 lb
of ball powder have been produced. Labour requirements are low : ball powder can
be manufactured by using fifty man hours per 1000 lb of finished powder as compared
with 125-200 man hours per 1000 lb required by the conventional process for making
extruded powder.

640

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

NITROCELLULOSE BULK POWDER (SCHULTZE POWDER)

These propellants containing mainly fibrous non-colloidal nitrocellulose are
usually referred to as Schultze powder, after the inventor— see p. 528.

In Great Britain they are frequently referred to as E.C. powder (this stands for
Explosives Company, one of the early manufacturers in England). In the U.S.A. they
are usually referred to as bulk powder, probably because they were loaded in bulk,
volume for volume, with blackpowder.

The production of nitrocellulose bulk powder is based on the same principle as
that of blackpowder. Nitrocellulose together with water, potassium or barium nitrate
and a binding substance (e.g. gum arabic, gelatine, agar-agar, starch) is mixed in
edge runners, granulated, screened and dried. On drying the grains harden like those
of blackpowder. The mixture may also include vaseline, which facilitates the adhesion
of the grains, or camphor, acting as a binder by gelatinizing the nitrocellulose. Pow-
ders of this kind are extremely fast burning and are used in sporting or practice
ammunition. The composition of some of them is tabulated below (Table 182).

Table 182
Composition of some Schultze sporting powders

Ingredients

Amberite

Schultze

Kynoch

Chasse M

CPi (guncotton)

CP2 (collodion cotton)

KNO3

4.5

1.5

Ba(N0 3 ) 2

7.5

22.5

Vaseline

—

Starch

—

Woodmeal

1.5

—

Agar-agar

—

DNT

—

Camphor

Heat of explosion (kcal/kg)

762

745

786

807

According to the method adopted in France the manufacturing process consists
of the following operations:

(1) Mixing the ingredients in an edge runner with the addition of about 10%
water which takes approximately 1 hr. Sometimes, in addition to water a certain
amount of alcohol-ether mixture is added (e.g. 50-60% of the weight of nitrocellulose).
Mixing should then be commenced in a hermetically sealed kneader to avoid losing
ether and alcohol. After mixing and the partial dissolution of nitrocellulose the mass
is reloaded into edge runners, and after kneading again is further processed.

(2) Rubbing the mass through a 1.5-2 mm mesh sieve to form grains.

(3) Granulating, which is carried out by placing the grains in a rotating wooden
drum, in which they are rounded off, forming granules of different sizes.

THE MANUFACTURE OF SMOKELESS POWDER 641

(4) Drying. The granules obtained in the wooden drums are dried at a temperature
of 45°C. Volatilization of the water makes the grains harden.

(5) Grading. The dried grains are graded by passing them through 0.5-1.5 mm
mesh sieves.

(6) Polishing. The graded grains are placed in small rotating drums made of
brass. An alcoholic solution of camphor, or acetone with alcohol, or an alcohol-
ether mixture is injected into the powder charge in the drums. The introduction of
solvent causes the formation of a coating on the grains which makes them harder
and makes the powder burn more progressively, i.e. the outer layer burns more
slowly than the interior of the grain. The gravimetric density of the grains in-
creases slightly on polishing. E.g., with Poudre Chasse M it increases from 380 to
0.400-0.430.

(7) Drying and final grading are the same as (4) and (5).

There are designs which make it possible to carry out operations (3), (4) and (5)
in one apparatus comprising a drum that acts as a granulator, a drier and a grading
machine. This is a long, inclined rotating drum with a system of screens inside. Warm,
drying air is passed through the drum. As the drum rotates the grains are polished,
dried in the air stream and graded between the screens.

According to U.S. Army Specification No. 50-13-8B E.C. powder should pass
two tests [31].

When the powder is to be used in blank ammunition the wad from the round
loaded with a specified weight of powder when fired in a 0.30 calibre rifle shall not
penetrate a craft screen placed at a specified distance in front of the muzzle of the
rifle. Only 1 % of variation is permitted.

The same powder can be used as a high explosive to fill hand grenades. On deto-
nation it should give 40 ± 10 fragments large enough to be held on a 2-mm mesh
screen.

DOUBLE BASE POWDERS

As with nitrocellulose powders the manufacturing procew*$of nitroglycerine
powders described below should be regarded as examples of typical methods which
may vary from one country to another. As before, the author has endeavoured, as
far as possible, to draw attention to the differences in methods adopted in various
countries.

Smokeless powders containing nitroglycerine are classified into two types: those
produced with the use of a volatile solvent and those produced without such a solvent.
In both types nitrocellulose is in a completely colloidal form.

Powders with a volatile solvent are becoming obsolete and are now produced
in only a few countries (Great Britain) whereas those without a volatile solvent are
being used increasingly. Nitroglycerine powders are designated as follows: France—
SD; Germany— R PC; Great Britain— Cordite SC (solventless cordite); Poland —
BR; the U.S.S.R.— letter H (Russian "N") following the figures showing the dimen-

£42 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

sions of tubes. In the U.S.A. nitroglycerine powders are called "double base powders"
i e those composed of two explosive ingredients: nitroglycerine and nitrocellulose.
Double base powders, i.e. nitroglycerine powders without a volatile solvent also
include modern flashless powders containing nitroguanidine.

NITROGLYCERINE POWDERS WITH A VOLATILE SOLVENT

Cordite Mk I and Cordite MD

British cordite is a typical powder of this kind. There are two types: the obsolete
Mark I and the more modern modified cordite MD. They differ in composition
(Table 183).

Table
Composition of

183
cordites (%)

Ingredient

Mark I
(abbr. Mk I)

Nitrocellulose (12.9-13.1% N)

Nitroglycerine

Vaseline

65
30

Cordite MD is an improved powder. It contains less nitroglycerine than Mk I
and is, therefore, less erosive. Cordite owes it name to the fact that it is made of
cords, the size (diameter) of which is denoted by a figure giving the diameter
of the die in hundredths of an inch. Thus, for instance, Cordite 50 denotes a powder
in the form of a cord extruded from a die of \ in. (50/100 in.) dia.

Cordite MD is manufactured by a method very similar to that used for the manu-
facture of nitrocellulose powder.

Drying the nitrocellulose. Dehydration with alcohol is not possible in this case
since in powders of this kind there is no alcohol in the solvent. To improve operational
safety in the drying house, and especially to prevent the formation of dust, the damp
nitrocellulose is pressed into cylindrical blocks. Natural draught drying houses are
used. Drying takes several days at a temperature of 43°C. The operation is dangerous
because of the sensitiveness to friction of dry nitrocellulose and its inflammability.
Precautionary measures should therefore be taken, especially when unloading the
nitrocellulose from the drying house. It should be unloaded only after cooling, with
great care to avoid friction. The regular removal of nitrocellulose dust by sweeping
the floor and dusting the radiators, shelves etc. is also of great importance.

Premixing nitrocellulose with nitroglycerine and incorporation. For this purpose
a special lead table in the form of a trough is used. One of its ends is slightly raised
and perforated with holes 1-1.5 mm dia. which form a kind of screen. Blocks of
nitrocellulose are placed in the trough and a weighed amount of nitroglycerine is
poured into it. A worker wearing rubber gloves first mixes the ingredients, and then

THE MANUFACTURE OF SMOKELESS POWDER 643

rubs the mixture carefully through the lead screen, from which it falls into a bag
attached below.

The cordite paste so obtained is next loaded into Werner-Pfleiderer kneaders
where it is incorporated with vaseline and acetone. The quantity of acetone in relation
to the nitrocellulose is 50-60%. Kneading lasts for 2.5-3.5 hr.

Next the dough is formed into cords by pressing it in hydraulic presses with
suitably shaped dies.

The pressing of nitroglycerine powder is quite a hazardous operation. The cor-
dite extruded from the press often ignites although this is not dangerous since the
presses are small with a cylinder of low capacity. The presses should be arranged so
that the workers operating them have a speedy exit from the building if a fire breaks
out.

Drying. The solvent is removed from the cordite by drying at temperatures of
38-43°C for several days to a volatile matter content of 0.4-0.6%. The drying time
depends on the thickness of the powder cords. Cordite MD loses its solvent more
easily than cordite Mk I and therefore dries more quickly.

Mater and solvent
Clamp^~~~ Nitroglycerine
Fig. 258. A device for trapping nitroglycerine in pipelines.

The drying houses employed for cordite consist of rooms with shelves heated
from beneath by radiators. Natural draught or forced circulation may be used. Air
containing acetone |mm the drying houses is passed into recovery towers where it
is sprayed with water. A dilute aqueous solution of acetone is thus recovered and
then rectified. There is better recovery of acetone if the towers are sprayed with
sodium hydrogen sulphite which reacts with the acetone (R. Robertson and Rintoul
method [32]). The solution is then concentrated and after acidification the acetone
is distilled off. The air exhaust ducting (especially with forced air circulation) should
be fitted with equipment for trapping the entrained nitroglycerine (Fig. 258) so as
to prevent the penetration of nitroglycerine to the machinery (e.g. to the fan).

Blending and packaging. Individual lots of cordite are blended to obtain a uni-
form product by the usual methods. It is then packed into wooden semi-hermetic
boxes since it is non-hygroscopic (p. 532).

Cordite RDB

Since there was a great shortage of acetone for the manufacture of cordite in
Great Britain during World War I, a solvent composed of alcohol and ether in a
weight ratio of 2 : 3 was used. Since the nitrocellulose usually employed for the manu-

644 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

facture of cordite is insoluble in this mixture, nitrocellulose containing 12.9-13^2% N
was replaced by a much lower nitrated one which contained 12.2% N and was soluble
in the alcohol-ether mixture. The composition of cordite RDB was:

Nitrocellulose (12.2% N) 52%
Nitroglycerine 42%

Vaseline 6%

About 90% solvent was used, consisting of 58% ether and 42% alcohol.

In principle the manufacture of cordite RDB was similar to that of ordinary
cordite with the difference that instead of drying the damp nitrocellulose it was
dehydrated with alcohol and the required amount of ether and additional alcohol
were introduced in the kneader. The manufacture of cordite RDB was discontinued
after the wa^ since its ballistic properties were inferior to those of common cordite.

Powders of the cordite type were never widely used outside Great Britain. For a
certain time before World War I cordite was in use in the U.S.A. Navy. Before and
during World War I a tubular nitroglycerine powder, similar to cordite in its com-
position, was manufactured in Austria-Hungary. It contained 30-40% nitroglycerine
and 60-70% nitrocellulose (13.2-13.4% N). Acetone was used as a solvent. In Germa-
ny a similar powder containing 25-30% nitroglycerine was manufactured up to 1912.

SOLVENTLESS NITROGLYCERINE POWDERS

In completely colloidal nitroglycerine and nitrodiglycol powders these nitric
esters are non-volatile, explosively active solvents. Similarly, triethylene glycol dini-
trate (nitrotriethylene glycol) is used as an ingredient in some powders without a
volatile solvent.

Aromatic nitro compounds may also be used as active solvents. During World
War I they were adopted to compensate for the shortage of nitroglycerine. They
have also been used more recently to reduce the heat of explosion and flash. In World
War II they were used partly for this purpose and partly to make up for the lack of
such non-volatile, explosively inert solvents as carbamite (centralite), acardite and
urethanes. When they are included in nitroglycerine and nitrodiglycol powders, it
is possible to reduce the content of these nitric esters by increasing the content of
nitrocellulose. This gives a powder of a lower calorific value and erosiveness. Non-
volatile solvents are also used in the surface gelatinization of semi-colloidal nitro-
cellulose rifle powder (see above, p. 625). They produce an outer layer that is com-
pletely colloidal, which burns more slowly and thus improves the progressiveness
of the powder. '

Camphor was also used for the same purpose. It has the advantage over centralite
of being a better solvent for nitrocellulose so that it can be used in smaller quantities
and can be gelatinized at lower temperatures.

Centralite (called in Great Britain carbamite) is 5ym-diethyldiphenylurea. It was
obtained by Zentralstelle fiir wissenschaftlich-technische Untersuchungen in Neu-

THE MANUFACTURE OF SMOKELESS POWDER

645

babelsberg and first used in powder manufacture in 1906. Shortly afterwards a
homologue of this substance— jjm-dimethyldiphenylurea— was prepared and put to
use. The diethyl derivative was named Centralite I and the dimethyl derivative
Centralite II. Other urea derivatives have also been developed and used, e.g. Acard-
ite — iMJjw-diprienylurea.

During World War II N-arylurethanes were extensively used for the manufacture
of nitroglycerine powder, e.g. ethylphenylurethane, diphenylurethane, o-tolylure-
thane, also, to a lesser extent, N-arylsubstituted amides of aliphatic acids, e.g. formyl-
diphenylamine were used. The latter was used in Japan [33] as gelatinizer and
stabilizer.

The formulae of the most important non-volatile solvents are given below. All
of them are characterized by the presence of carbonyl groups and with the exception
of camphor by the presence of the amido group — CO— N<

CH,

Camphor

C 2 H 5
6 H 5

CO ^

\3C 2 H 5
Ethylphenylurethane

C 2 H 5

C 6 H 5

\ '/C 6 H 5

X C 2 H 5
Centralite I
(Carbamite)

/CH 3

X C 6 H 5

\ /C 6 H 5

X CH 3
Centralite II

C6H5
C 6 H 5

X NH 2

Acardite
NH— C 6 H 4 — (o)CH 3

\)C 2 H 5
Diphenylurethane

C 6 H 5

/
CO

X OC 2 H 5
o-Totylurethane

S C 6 H 5

N — Formyldiphenylamine

Non-volatile phthahc esters, (e.g. butyl phthalate) are also used for the manufac-
ture of some nitrocellulose and nitroglycerine powders.

Individual non-volatile solvents vary in their capacity for dissolving nitrocellulose.
The type of nitrocellulose dissolved is also a factor to be considered. Marqueyrol
and Florentin [34] report the following figures which denote the amount of solvent
required to obtain a uniform gelatinous film (Table 184).

Table 184

Solvent

CPi

CP 2

Dimethylphenyl-o-tolylurea

260

Dimethyldiphenylurea

—

Diethyl sebacate

320

Diethyl phthalate

360

646

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Davis [25] found that non-volatile solvents vary in their capacity for gelatinizing
nitrocellulose, depending on the liquid in which they are dissolved. The following
data show the gelatinizing power of pyrocollodion cotton (Table 185). The figures
indicate the parts of non-volatile solvent required for the complete gelatinization of
100 parts of pyrocollodion cotton.

Table 185

Ability to completely gelatinize pyrocollodion cotton
of various non-volatile solvents, according to davis [25]

Solvent

In alcohol

In benzene

Methylurea

100

Ethylurea

100

Sym-dimethylurea

Sy/n-diethylurea

^5ym-dimethylurea

100

Tetramethylurea

S.y/M-diphenylurea

100

^.yy/n-diphenylurea

100

Triphenylurea

Tetraphenylurea

160

Centralite II

Centralite I

/4s;y/n-dirnethyldiphenylurea

—

Carbamic acid ethyl ester

140

Methylcarbamic acid ethyl ester

Ethylcarbamic acid ethyl ester

Phenylcarbamic acid ethyl ester

Phenylcarbamic acid phenyl ester

200

Diphenylcarbamic acid phenyl ester

Methyl phthalate

Ethyl phthalate

Isoamyl phthalate

DNX oil

120

130

TNT

300

Symmetrically substituted, especially tetra-substituted, urea derivatives have
particularly good gelatinizing properties.

The manufacture of solventless nitroglycerine (also nitrodiglycol) powders (i.e.
powders without a volatile solvent) differs from that of powders with a volatile
solvent at the stage when the nitrocellulose is converted into a colloidal state. This
highly important stage is not accomplished in a kneader, but between rolls heated
to a high temperature (80-90°C). An elevated temperature is needed during pressing
to obtain the required plasticity.

Nitroglycerine powders are produced in the form of flake powder, which is easy
to manufacture (ballistite). They are also produced as tubular powders.

THE MANUFACTURE OF SMOKELESS POWDER 647

Huffington's [35] investigations showed that nitroglycerine powder manufactured
without a solvent is not completely uniform, the nitroglycerine being present in
the form of small drops. This means that nitroglycerine powder burns rather irreg-
ularly at relatively low pressures (e.g. 27 atm) common in rockets. The irregularity
manifests itself as successive periods of slow- and fast-burning due to the explosions
of the nitroglycerine drops.

Huffington carried out his experiment with a powder containing 29% of nitro-
glycerine and 10% of dinitrotoluene, as mentioned before (pp. 644, 650 and 653).

Ballistites

Ballistites initially consisted of equal amounts by weight of nitroglycerine and
soluble nitrocellulose CP 2 with the addition of aniline or diphenylamine as stabiliz-
ers. It was found, however, that the presence of aniline and diphenylamine is detri-
mental to the stability of the powder, and they were therefore omitted. The valuable
properties of centralite as a solvent of low basicity were then recognized and it
was used both for its ability to dissolve the nitrocellulose and for its stabilizing
action.

This led to the development of ballistite 50/50 which is still in use It consists
of:

Collodion cotton 49-49.5%

Nitroglycerine 49-49.5%

Centralite 1-2%

Ballistite 40/60 has a reduced content of nitroglycerine. It consists of 60% collod-
ion cotton and 40% nitroglycerine. To this mass 1-2% centralite is added.

The manufacture of ballistite is divided into the following stages.

The incorporation of nitroglycerine and nitrocellulose. The two ingredients and
carbamite are simply mixed in hot water by stirring with compressed air. Water is
heated to a temperature of 60°C. Nitrocellulose is suspended and nitroglycerine
poured into the slurry of nitrocellulose. The required amount of centralite may
be dissolved in nitroglycerine. On mixing, the nitrocellulose absorbs the nitroglycer-
ine. After 15-30 min, when the nitrocellulose is uniformly mixed with the nitro-
glycerine, the contents of the vat are poured into a cloth filter. The vat was usually
emptied either by tilting (convertor principle) or by lowering a flap clamped with
a lever to an outlet in the bottom.

The water may be removed from the nitrocellulose-nitroglycerine "paste" by
centrifuging. The system described is primitive and virtually obsolete, but it is
quite adequate for the manufacture of ballistite and is therefore still in use. A more
modern system, giving a uniform paste with more evenly incorporated nitroglycerine
is based upon the use of an aqueous emulsion of nitroglycerine (see Fig. 268).

Rolling (for drying). The damp mass is passed between rolls heated to a temper-
ature of 50-60°C. Most of the water is removed and at the same time dissolution

648 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

of the nitrocellulose in the nitroglycerine is promoted. This causes the mass to form
into lumps which are here and there transparent. For this operation horizontal
rolls are commonly used. (Fig. 259). Under the roll there is a tin tray to catch the
water pressed out of the mass, and the pieces of paste that drop during rolling.

Fig. 259. Horizontal drying rolls [7].

Drying is complete when no more water is squeezed out of the paste. The amount
of moisture is reduced to ca. 5%.

Very often the "differential rolling mills" are used here: two horizontal cylindrical
rolls rotate at different circumferential speeds (the ratio of the speeds being 1.5-2
to 1). This uneven rotation produces a shearing action which facilitates mixing.
The further processes of manufacture may vary. Usually the paste is subjected to
a final rolling.

Final rolling. This operation aims at obtaining a uniform translucent, completely
colloidal mass in the form of a flat sheet. It is conducted at an elevated temperature
(70-95°C), and consists of repeatedly passing a sheet folded in various directions
between the rolls. Calanders are usually used for the final rolling (Fig. 260). They
have a highly polished surface and the spacing between them may be regulated
with a high accuracy (to 0.05 mm).

Rolling is considered to be complete when the sheet is quite uniform to the eye,
translucent and without streaks or stains. In some factories during the final rolling
a weighed amount of centralite is added, spread over the sheets during rolling.

Both cylinders should run at the same speed. This increases the safety of the
process by minimizing the friction. In spite of this the operation of final rolling is
considered relatively dangerous (see p. 65 1).

Cutting. Warm sheets (at about 50°C) are cut on a guillotine into squares, by
first cutting the sheet into strips and afterwards, crosswise, into squares. The sheets

THE MANUFACTURE OF SMOKELESS POWDER 649

must be warm in order to keep them plastic, so that they are easy to cut. Cold sheets
which are hard and brittle, soon blunt the knives, the cut flakes are not suffi-
ciently regular in shape, and more dust is produced.

Ballistite is usually cut into "squares" with sides ten times larger than their
web thickness (e.g. flakes of 1x10x10 mm or 3x30x30 mm are obtained). The

Fig. 260. Finishing rolls (calanders) [7].

dimension ratio may, however, be varied, e.g. for sporting ballistite, requiring a high
vivacity, the dimensions are 0.1 x 1.5 x 1.5 mm.

Grading. After they are cut, the powder flakes are graded on vibrating screens
to separate out flakes of the right size from dust and coarse or irregularly shaped
flakes. The waste material is returned to the dried mass prior to final rolling.

Graphite glazing. The ballistite flakes are glazed with graphite in drums. For
this purpose 0.1% graphite is added to the powder and the drum is rotated for
15-30 mm. The powder is then regraded, mainly to remove any graphite adhering
to the surface of the flakes and the dust formed when the sharp edges of the
flakes are rounded off, and is then ready for use. Its ballistic properties are tested
and the mass is blended in hoppers or drums.

The method for the manufacture of ballistite outlined above gives good results
only when totally soluble nitrocellulose is used, the solutions of which are not too
viscous. A high viscosity hampers gelatinization while a low viscosity gives solid
solutions that are too brittle. This makes the sheets brittle, causes a larger amount
of wastage on cutting and impairs the strength of the flakes, which may be damaged

65 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

during transport. It is of great importance, therefore, that a plant manufacturing
ballistite should always use nitrocellulose of the same properties, especially as far
as solubility and viscosity are concerned.

Some factories (e.g. in Austria) which before and during World War I encount-
ered many difficulties in manufacturing ballistite by the method described above
modified the process: after the mass had been rolled to dry it, it was transferred to
kneaders and a small amount (2-8%) of acetone was added. This greatly facilitated
the next, final rolling which could be then conducted at a lower temperature (70°C).

The powder so obtained contained some acetone which did not volatilize on
final rolling. After cutting, grading, glazing and final grading, therefore, it had to
be dried at a temperature of 40-43°C, which took up to two days.

Ballistite is virtually non-hygroscopic and is therefore transported and stored
in tightly closed wooden boxes, lined with cloth or waxed paper with a lid fixed
with brass screws.

Attenuated ballistites

In some countries, especially during World War I when nitroglycerine was in
short supply, it was partly replaced by aromatic nitro compounds, e.g. liquid DNT
(a liquid mixture of DNT and TNT isomers). Partial replacement of nitroglycerine
by nitro compounds also reduced the erosive effect of the powder by lowering the
heat of explosion and the flame temperature during explosive decomposition.

The composition of such flake ballistite WP (Wurfelpulver) was:

Collodion cotton

61%

Nitroglycerine

20%

TNT

15.25%

DNT

3.50%

Centralite

0.25%

Another type of attenuated ballistite (WP) containing less nitroglycerine and
an increased content of nitrocellulose was also produced in Germany. It consisted

of:

Nitrocellulose (12.6-12.7% N and 50-70% solubility) 60%
Nitroglycerine 40%

To this mass 0.5-1.0% centralite was added.

During World War I and later attenuated ballistite (Ballistite ATT) in which
the nitroglycerine was partly replaced by DNT was used in France. To prevent
excessive attenuation the collodion cotton was partly replaced by high nitrated

nitrocellulose (CPi):

Nitrocellulose (CPi) 30%

Collodion cotton (CP 2 ) 30%

Nitroglycerine 25%

DNT 15%

In this case DNT also acted as a stabilizer.

If**'

THE MANUFACTURE OF SMOKELESS POWDER 651

Progressive baUistite

To obtain baUistite with a more progressive rate of burning, attempts were made
to produce laminated flakes, with the two outer layers made of attenuated baUistite
and an inner one, sandwiched between, made of ordinary baUistite. In the attenuated
baUistite DNT was substituted for part of the nitroglycerine. However a powder
of this type retained its ballistic characteristics for only a few months, since, due
to diffusion, the composition of all three layers gradually became equal.

Safety in the manufacture of ballistites

Final rolling is a dangerous stage of the manufacture, since the powder may
ignite, especially when it is rolled to a low web thickness (below 1 mm). Ignition
may be caused in various ways. Sometimes a foreign body (e.g. a grain of sand)
can increase friction, or a pocket of air confined in a fold of the sheet may be viol-
ently compressed when the latter is introduced between rolls.

A hot sheet of powder burns very fast. Special automatic installations are used
to extinguish fire as soon as possible with a strong jet of water. The simplest system
consists of suspending above the rolls a vessel filled with water which is balanced
with a strip of smokeless powder. The strip is burned out at once, by the flame that
shoots upwards when a sheet ignites, and the vessel immediately falls over and pours
its contents onto the fire.

This equipment, however, may not apply the water quickly enough if the sheet
does not burst into flame immediately below the strip of powder.

A more modern installation comprises a system in which a photoelectric cell
detects the first flash of flame emitted by the sheet of powder. The electric current
generated by the photoelectric cell passes through an amplifier and ignites a charge
of 2 g of smokeless powder which a quarter of a second after the accident opens
a 9 in. dia. water release valve. This method has proved reliable and very useful in
industrial practice [36].

All workers must wear special clothes of heavy wool, leather or asbestos to protect
them from brief but very hot flame. Hands must be protected with gloves that
leave the fingertips bare so that the workers retain the sense of touch in the fingers,
otherwise the hand may be drawn between the rolls.

In the course of their work the staff are exposed to the risk of inhaling nitro-
glycerine vapours but most people grow accustomed to this without detriment.
The centralite vapours present in the air also exert an irritating effect upon the upper
respiratory tract and it is advisable, therefore, to provide the staff with light respi-
rators containing cotton wool or an adsorptive layer.

652 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

SOLVENTLESS POWDERS WITH A LOW CONTENT OF NITROGLYCERINE

Powder with a low content of nitroglycerine is simply called solventless powder
to distinguish it from ballistite.

Solventless powder is usually tubular. It was first produced at the Diineberg
factory, in 1912, under the name of RPC/12 (Rohrenpulver mit Centralit 1912).
Initially the manufacture of this powder was based upon the use of easily soluble
nitrocellulose (11 %N), plus guncotton to bring the total nitrogen content up to
11.7%. About 25% nitroglycerine was used plus a large quantity of centralite (4-5%)
as a non-volatile solvent.

The manufacturing process was divided into the following stages:

(1) Preparation of the mixture of ingredients ;

(2) Rolling, to promote drying and gelatinizing;

(3) Hot pressing (at 90°C) into tubes.

Since the powder contains no volatile solvent it does not require long drying
(cf. the manufacture of cordite). Drying lasts only a few hours and its purpose is
solely to equalize the moisture content in the powder tubes.

The rapid manufacture of RPC/12 powder was one of reasons for the protracted
resistance of the Central Powers during World War I. The lack of acetone suffered
by the Central Powers at that time had no effect on the production capacity of
this powder in German factories.

Another of its advantages was that various substances, e.g. flash-reducing com-
pounds, could be introduced into the powder mass. During World War I potassium
oxalate or potassium tartrate were used to reduce or suppress flash. These salts
cannot be introduced into nitrocellulose powder since they would be washed out
with the water during soaking.

Potassium hydrogen carbonate and vaseline were also tested as flash reducing
agents and stabilizers.

Powder of the RPC/12 type proved not very erosive, no more so than common
nitrocellulose powders. The only disadvantage of this powder is that it is hazardous
to manufacture. Rolling and hot pressing often leads to ignition. This has caused
a large number of explosions. The manufacture of this powder was kept secret and
not disclosed until after World War I, when the essential features of its production
were revealed. The type of nitrocellulose used for the manufacture of the powder
is of great importance. Mixed nitrocellulose is required since more uniform tubes
are then obtained, and the presence of non-gelatinized guncotton facilitates ignition
and increases vivacity. A powder made from one type of nitrocellulose containing
11.7% N is irregular, less vivacious and ignites with greater difficulty, hence its
ballistics are less uniform. The uniform ballistic properties of RPC/12 powder
were achieved by accuracy in production and particularly by strict production
control in terms of the heat of explosion. This property was therefore kept within
narrow limits.

THE MANUFACTURE OF SMOKELESS POWDER 653

After World War I the manufacture of solventless powder was started in other
countries: France under the name of powder SD (sans dissolvent), Great Britain as
Cordite SC (solventless cordite) and the U.S.S.R.

According to Wheeler, Whittaker and Pike [37] British solventless powder
consisted of:

Nitroglycerine 41 %

Nitrocellulose 50%

Diethyldiphenylurea 9%

At that time the composition was improved by the addition of a small amount
of graphite which acts as a lubricant and so facilitates extrusion. In many cases
centralite was replaced either by Acardite or by phenylethylurethane or diphenyl-
urethane. In Japan formyldiphenylamine was applied.

In the U.S.S.R. a type of solventless powder was introduced in which organic
nitro compounds, e.g. DNT, partly replaced the nitroglycerine. DNT acts as a
non-volatile solvent and as a stabilizer. Since it also reduces the heat of explosion
these powders are either fiashless or partly so.

During World War II large quantities of fiashless powder containing nitro-
diethyleneglycol and nitroguanidine were produced. Nitroguanidine has the advan-
tage of considerably reducing the heat of explosion, although it cannot dissolve
nitrocellulose and is, therefore, only mechanically incorporated into the colloidal
mass.

Two operations in the manufacture of solventless powder are of particular
importance :

(1) Careful mixing of the ingredients.

(2) Uniform, hot pressing of the powder tubes.

Mixing. Mixing is conducted first by making a slurry of the nitrocellulose in
water and then stirring in the nitroglycerine in which the centralite is dissolved.
When, however, solventless powder is prepared with a small amount of nitroglycer-
ine, the "paste" obtained is not sufficiently uniform in spite of continuous stirring.
Uniformity is adequate, if the mixture is to be made into ballistite, i.e. if the content
of nitroglycerine is relatively large (40-50%). The manufacture of the powder is
then concluded by rolling and cutting the sheets. If however, the nitroglycerine
content is relatively low (20-30%) and the sheets are to be extruded into tubes,
the primitive method described above gives insufficient uniformity and this is not
improved by mixing in kneaders. The most effective method is to allow the paste
to stand in a wet condition (after removing the water by filtering or centrifuging)
for some time, usually two weeks. In this "ripening" period, the composition of
the mass is partially equalized by the diffusion of liquid nitroglycerine, the swelling
of the nitrocellulose and its partial dissolution in the nitroglycerine.

It is obvious that this method has some serious disadvantages:

(1) The manufacturing process is considerably prolonged.

(2) Large concrete pits must be installed for storing the ripening paste.

654

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The difficulties in obtaining a unifoim paste were completely eliminated when
nitroglycerine or dinitrodiethyleneglycol was used in the form of an aqueous emulsion
for mixing with nitrocellulose under water (see pp. 647 and 661). Nitrocellulose adsorbs
nitroglycerine very uniformly from an aqueous medium, so the paste thus obtained
may be utilized for further processing immediately after the removal of water.
This is best achieved in centrifuges. They reduce the water content to about 30%,
while for further reduction (to 8-10%) screw presses are preferable.

Rolling. The paste intended for the manufacture of solventless powder is first
rolled between drying rollers (at 50-60°C) and then between gelatinizing rollers, as
in the manufacture of ballistite. This operation must be very carefully controlled.

Fig. 261. A press for the extrusion of sol-
ventless powder tubes.

Fig. 262. A press for the extrusion of sol-
ventless powder tubes.

The powder paste should always contain the same proportion of water. The rollers
used for a given powder paste should always be heated to the same temperature
and their peripheral speed should also be constant (the diameter of the rollers should
be directly proportional to their angular speed). The distance between the rollers,
the amount of material to be rolled and the number of passages between the rollers
must not be changed during the operation. Rolling of the paste should be more
prolonged than with ballistite, since this improves the uniformity of the tubes ex-
truded by the press, and gives them a smoother surface. An increase in the number
of passages, however, may lead to difficulties in pressing, primarily in an increase
of the pressure required for pressing since the paste becomes more gelatinized.

Powder containing dinitrodiethyleneglycol is much safer to roll than that con-
taining nitroglycerine powder.

Pressing. This is an operation of the greatest importance since the quality of
the powder tubes and their ballistic properties depend to a high degree on its proper
execution. It should be carried out under a uniform but moderate pressure to give
tubes with an even, smooth surface, and identical dimensions.

The tubes are extruded from the hot material, under high pressure (200-700 kg/cm2)
using hydraulic presses, with the cylinders heated with hot water to a temperature
of 90°C. Since the pressing operation is rather dangerous and sometimes leads to
explosion, the cylinders are usually fairly small and hold at most 10-25 kg of paste.
As a safety precaution the press should be separated by a wall from the conveyor

THE MANUFACTURE OF SMOKELESS POWDER 655

receiving the extruded tubes (Figs. 261 and 262). In addition, the press itself may
be fitted with safety devices protecting it from explosion. There are presses, for
example, in which the die is joined to the cylinder by crocodile clips. If the pressure
inside the cylinder increases to above 1200 kg/c m 2 it exceeds the mechanical strength
of the clips so that they are broken, the die is thrown out and the gases inside the
cylinder can expand freely.

The causes of explosions during the pressing of hot, solventless powder have
not yet been fully explained. They may include: discharge of electricity accumulated

kg/cm 7

120

2 4 6 8 10 12 14 16 win
Fig. 263. Piston speed during pressing.

2 4 6 8 10 12 14 16 min

Fig. 264. Hydraulic pressure in terms of
time.

by intensive friction, friction itself, and compression of hot air present between the
layers of paste. Fleury [38] suggests that almost all the piston work is converted
into heat energy which may lead to local overheating, up to the initiation temperature
of the mass.

In fact, the material is pressed at such a high temperature that explosive ingre-
dients particularly sensitive to friction and impact such as nitroglycerine and nitrocel-
lulose may be exploded by a minute initiating thermal or mechanical impulse.

The mechanism of the pressing process should be known in great detail in order
to recognize the dangerous moments. Fleury [38] investigated the pressing of sol-
ventless powder, and recorded the following:

(1) A diagram of the piston speed (Fig. 263).

(2) A diagram of the variation of hydraulic pressure with time (Fig. 264).

(3) Diagrams of piston acceleration (curve V in Fig. 265) and of piston work
(curve P in Fig. 265).

Two distinct sections ab and cd of different piston speed can be seen in Fig.
263. The speed for the first section ab is distinctly high. This is followed by a short,
transistory section be, that passes into the third section erf in which the speed becomes
steady. The diagram in Fig. 264 shows that from the point b a sudden rise of pressure
begins and persists for the first two minutes, while curve P in Fig. 265 shows that

656

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

during the first 30 sec the piston work increases violently with intense heat emission
and then falls almost to zero, soon increasing again. These changes correspond
to the following stages: first the kneading of the mass continues for half a minute,

n/cm

2
P

146

125-

100-

75-

lf\

50-

25-

I V

I I I I

2 4 6 8 10 12 14 16 win

Fig. 265. Piston work in terms of time P; V— acceleration of piston motion.

when the piston is slowed down, since the pressure is still too low to cause the outflow
of the mass through the die. Next the pressure increases and the mass commences
to flow out.

Fleury used the above diagrams to explain some explosions which occurred in
an experimental factory at Sevran-Livry, in September and October, 1929.

In all these accidents the explosions occurred during the first stage of pressing
(points b on diagrams in Figs. 263 and 264), when the pressure (ca. 40 kg/cm 2 )
was considerably lower than the maximal one, but when the piston speed was fairly
high (section a-b in Fig. 263) and the amount of heat from the transformation of
its work into heat energy was very large (curve P in Fig. 265).

Fleury concluded that it is necessary to determine experimentally the most
appropriate piston speed for a given press and a given powder mass. During the
whole process the speed of the piston should be so adjusted as not to produce an
excessively large quantity of heat that might lead to an explosion. The danger is
particularly great during the initial period of pressing since the quantity of heat
than produced may be exceptionally large.

The most important factors influencing the pressing process are listed below.

THE MANUFACTURE OF SMOKELESS POWDER 657

Uniformity of the paste is a prerequisite for good pressing. This depends chiefly
on the uniformity with which the ingredients are mixed under water and on the
uniform gelatimzation of the mass between the rollers. Insufficient accuracy or
defective execution of either operation can be highly detrimental to the pressing
process, sometimes rendering it impossible.

In addition the results of pressing also depend upon filling the cylinder with
the sheets of gelatinized substance as completely as possible and upon the main
tenance of a uniform temperature inside the cylinder (for nitroglycerine pastes
about 90°C, for nitrodiethyleneglycol pastes about 70°C). The cylinder is pronerlv
filled by competent loading. There are two methods which differ in economy of
materials. In the first method disks with a diameter equal to that of the inner dia
meter of the cylinder are cut from a sheet and are then piled onto each other ("sand-
wich loading"). In the other method the sheet is coiled into a rouleau (roll) with a
diameter as above ("carpet loading"). The first of these methods is less economical
since cutting the disks from the sheet leaves a considerable amount of waste clip
pings. Although they are recycled for rolling, this leads to an increase of running costs"
Carpe loading gives only an insignificant quantity of clippings, when the edges
are cut from a big sheet, and this method is therefore the more widely used

of I°?T am a f teady aDd Unif ° rm tem P erature ^thin the cylinder the surface
of the disks or rouleaux must not be allowed to cool during cutting or other opera-
tions preceding loading. To prevent such cooling, in many factories the powder

tZV^°1 Z f ; lled ° n taWeS hCated Wkh h0t Water and the —ll then
immediately loaded into the cylinders of the press. If the disks or rouleaux should

JST Vu "' ^ tUbCS CXtrUded by the die are uneven > often shredded,

and should be rejected. '

Uniformity of pressure depends upon the uniformity of the charge in the press
The amount of pressure applied depends on a number of factors. For example
a higher temperature of the paste enhances its plasticity thus facilitating extrusion
and making it possible to carry out the pressing under a lower pressure

On the other hand there are a number of factors which lead to an increase in
pressure, i.e.: a reduction in the dimensions (diameter) of the openings of the die,
an elongation of tfie outlets of the die or a large number of passages between the
rollers as described above.

The pressure is also greatly influenced by the composition of the powder mass,
n particular by the nitrogen content of the nitrocellulose. The highest pressure
should be .applied to nitrocellulose with a 12.0-12.3% nitrogen content. With a
lower or higher nitrogen content a lower pressure is required. An increase of the
content of gelatinizing agent (e.g. over 9% centralite) may also involve increased
pressure during the pressing operation. Similarly the pressure increases when a
paste with a larger content of nitrocellulose is used.

However, the pressure may be considerably reduced by the addition to the powder
mass of substances such as graphite, or magnesium oxide, or both, that act as a
lubricant by reducing the internal friction. The presence of graphite also diminishes

658

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

the danger of electrification of the powder tubes during extrusion. Pressing is also
facilitated if a certain percentage of water is left in the sheets coming from the rolls.

The uniformity (smoothness) of the surface of the tubes extruded by the press
is favourably influenced by the following factors:

An increase in the number of passages in rolling;

An increase in the length of the die.

The use of nitrocellulose with a nitrogen content below 12% also increases the
smoothness of the surface of the tubes. The addition of higher nitrated cellulose is
not detrimental in this respect if the production process is well managed.

Zones:

Fig. 266. Diagram of a screw press, according to Schenkel [40]. 1 -heating and
cooling jacket, 2— heating elements.

The surface becomes less smooth due to :

(1) An increase in the extrusion speed of the tubes;

(2) A rise in temperature of the mass in the cylinder.

The surface of the tubes often looses smoothness if the centralite content in
the powder is increased above 9%, while it is improved when the content of nitrocellu-
lose is increased above 50%. It was ascertained experimentally that a well gelatinized
paste, containing 50% nitrocellulose of 12% N content may give tubes with a rough,
uneven surface, while a paste containing nitrocellulose with the same nitrogen
content but composed of a mixture of lower nitrated (about 11% N) and higher
nitrated (about 13% N) nitrocellulose gives a smoother surface.

The outer diameter of the tubes extruded by the die is somewhat larger and
the inner diameter somewhat smaller than the corresponding dimensions of the
die. This is due to the fact that the paste is to a certain extent elastic and therefore
expands after the pressure is discontinued. The shorter the die, the greater the ex-
pansion, but a longer die requires a higher presure. Thus, under a pressure of about
230 kg/cm 2 , with a die about 1 mm long, the tubes may expand by about 12%.
When using a considerably longer die -about 25 mm -the pressure must be raised
to approximately 400 kg/cm 2 for the same paste, to obtain the same speed of extru-
sion of tubes. Under these conditions, the expansion barely amounts to 5% (the
above figures relate to a definite type of powder only— they may differ for other types).

THE MANUFACTURE OF SMOKELESS POWDER

659

The expansion of the tubes also depends on other factors. It increases with an
increase in the speed of extrusion, a rise of temperature of the mass in the cylinder,
and an increase in its centralite content. An increase in the number of passages
between the rollers may also lead to a greater expansion of the tubes. The expansion
also depends on the nitrogen content in the nitrocellulose, being nearly inversely pro-

-**?%

Fig. 267. "Guillotine" cutter for powder tubes [7].

portional to it. Thus, when using nitrocellulose which contains approximately 1 1 % N,
the expansion is much greater than with ritrocellulose which contains 12.3% N.

The uniformity of the diameter of the powder tubes is also influenced by different
factors. Greater uniformity is obtained with longer dies and an increased number
of passages between the rollers. A decrease in the nitrocellulose content to 50%
reduces the uniformity of the tube dimensions as does an increase in the centralite
content.

Recently a new technique of extrusion of solventless powder dough was introdu-
ced, in which extrusion is effected by a worm screw extruder instead of a hydraulic
press.

A few patents have been issued covering the use of screw presses [39], but they
do not include details of the design of the presses. However these do not differ from
the screw presses widely used for extrusion of plastic rods, tubes and other shapes.
A detailed description of the screw presses used in plastics technology can be found
in monographs [40, 41]. A diagrammatic representation of a screw press is given in
Fig. 266.

Cutting. Tubes extruded as described above are cut on a "guillotine" (Fig. 267).
The powder should be warm when cut (about 50°C), and the trough of the guillotine
is therefore heated externally with warm water. This makes it possible to maintain
an adequate temperature and elasticity of the powder and prevents it from cooling
down and cracking when being cut.

660 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

THE MANUFACTURE OF SOLVENTLESS POWDER IN GERMAN FACTORIES [42]

Notation

German nitroglycerine powders were all marked with letters Ngl. at the beginning
of the code name. This was followed by letters denoting the form, e.g. R.P. for
tubular powder, Bl.P. for square plates powder etc., as described earlier (nitrocellu-
lose powder, p. 572). After that a figure denoting the calorific value in hundredth
of kcal/kg was given, followed by the dimensions in brackets. E.g. Ngl. B1.P.-12,5-
(4-4- 1) meant: nitroglycerine powder in square plates wittya calorific value 1250
kcal/kg and the dimensions 4 mmx4 mmxl mm; Ngl. R.P.-8,2-(100.2/l) meant:
nitroglycerine tubular powder of 820 kcal/kg, 100 mm long, with external and internal
diameters 2 and 1 mm, respectively.

Diglycoldinitrate powders had the same notation as nitroglycerine powders
with only one difference, the calorific value was marked with letters for the sake
of secrecy.

The letter A denoted calorific value 930 ±25 kcal/kg

E denoted calorific value 740+10 kcal/kg

K denoted calorific value 690!?o kcal/kg

G denoted calorific value 690*10 kcal/kg

E.g. Digl. R.P.-G. followed by dimensions in brackets meant diglycoldinitrate tubular
powder of calorific value 690+i" kcal/kg.

DGDN powder usually contained flash reducing salts. Their quantity was indi-
cated by a figure following the letter denoting the calorific value, e.g. G.0. meant
no salts are present, G 1,5 denoted 1.5% salts.

Flashless powder containing nitroguanidine was called "Gudol Pulver" and was
marked with an abbreviation Gu. This was followed by the usual notation giving
the form of the powder grain. Next was the letter giving the calorific value (usually
A in nitroguanidine powders) followed by a figure indicating the amount of flash
reducing salts. The absence of such salts was indicated by O; e.g. Gu.Bl.P.AO
meant Gudol square plates powder, calorific value 930 ± 25 kcal/kg, without flash
reducing salts; Gu.R.P.A. 1,2 meant Gudol tubular powder, of 930 ±25 kcal/kg
with 1.2% flash reducing salt.

Manufacture

The following method has been described as that used in Krummel and Diineberg
during World War II. First the crude powder paste was prepared. The ingredients
comprised (weight of dry substance) :

Nitrocellulose 70%

Explosive oil (nitroglycerine
or nitrodiethyleneglycol or
nitrotriethyleneglycol or
nitrometriol) 30%

THE MANUFACTURE OF SMOKELESS POWDER

661

The nitrocellulose used for the manufacture of the mass consisted of a mixture
of two qualities: high-nitrated nitrocellulose S (13.15-13.25% N), with a solubility
of about 10 % and a degree of fineness of about 85 cm * and low-nitrated nitrocellulose
EH (11.3-11.45% N), with a solubility of 100% and a degree of fineness of about
90 cm.

For nitrocellulose supplied in hermetically-sealed containers (35-10% of water)
the water content was determined by drying for 12 hr at a temperature of 45°C. Both

Explosive oil

Nitrocellulose
Water

to a centrifuge

Fig. 268. Schematic view showing an arrangement for mixing powder with an in-
jector for producing the explosive oil-in-water emulsion.

types of nitrocellulose were then mixed to obtain the required nitrogen content. In
Krummel, mixtures were used with the following nitrogen contents:

I 11.5% N

II 12.0% N

III 12.2% N

IV 12.6% N

V 12.75% N

VI 13.0% N

Mixtures II, IV and VI were used most frequently.

The nitrocellulose was weighed in rubberized bags and conveyed by electric
trucks to the mixer.

The explosive oil (usually dinitrodiglycol) was used in the form of an aqueous
emulsion. It was drawn from a wooden container lined with lead and introduced
through a feeder into a water injector in which the aqueous emulsion formed. This
emulsion was introduced into the mixer, filled with an aqueous suspension of nitro-
cellulose (Fig. 268).

Incorporating. Approximately 1.6 m3 of water and 280 kg of nitrocellulose
(weight of dry substance) were poured into a mixer 3 m 3 in capacity and a mechanical
stirrer was set into vigorous motion. After 10 min of stirring, when both modifications
of nitrocellulose were uniformly incorporated and a slurry was formed in water,

* The figures characterize the volume of a layer of nitrocellulose settling from a suspension
of 10 g in 250 ml of water. The finer the nitrocellulose the smaller the figure.

€62

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

120 kg of the aqueous emulsion of explosive oil was introduced. The whole was
stirred for a further 10 min and then poured into a centrifuge. After centrifuging
the paste contained about 35% of water. The centrifuge discharged its contents
from below into rubber bags which were placed into larger cloth sacks and transported
by electric trucks to the store where the crude paste was kept.

The water flowing from the centrifuge was collected in large tanks in which
particles of crude paste carried out with the water settled on the bottom. The clear

Fig. 269. Worm drying and mixing machine: A — inlet for introduction of dough
with a 30% water content, B— water discharge outlet, C— exit for the mass containing

8% water.

water was returned to the mixer and the sediment from the bottom of the tanks
was removed from time to time, when larger amounts had collected.

Storage. The crude, paste was emptied out of the bags in thin layers into large
wooden boxes of 10,000 kg capacity. To attain a uniform sample, the mass for pro-
cessing into powder was taken in vertical layers.

The crude powder paste was conveyed to the adjacent factory in Duneberg, where
it was processed into charges ready for use in the following way:

To make the crude paste uniform in composition it was mixed in a large wooden
drum and loaded into Werner-Pfleiderer kneaders heated to about 50°C. The remain-
ing ingredients of the powder (Centralite or Acardite, graphite, magnesium oxide
etc., depending on the type of powder) were then added. From the kneaders the
mixture was placed in air-tight cans. To ensure good results in subsequent operations,
the paste was allowed to ripen for about a week. After ripening the mixture may
be rolled, but its moisture content should first be reduced from 30 to 8% by passing
it through a worm press (Fig. 269) to facilitate subsequent rolling operations.

Rolling. It was carried out using horizontal rollers 40 cm dia. and 100-120 cm
in length, rotating with a speed of 1 1 r.p.m. For paste made with dinitroglycol it
was sufficient to maintain a temperature of 70-80°C. A charge of about 15 kg was
rolled for 18-30 min. The sheets issued from the rollers completely gelatinized. In
cannon powders 3-5% water was left in the sheets since this facilitated the formation
of tubes.

THE MANUFACTURE OF SMOKELESS POWDER

663

To load the press one or two sheets, with the edges evened-off, were coiled around
a brass rod approximately 4 cm dia. This gave charges of 15-30 kg which were placed
into the cylinders of a hydraulic press. The diameter of the cylinders ranged from
17-24 cm, and the temperature when pressing dinitrodiglycol powder was 70-80°C

For pressing powder of large dimensions the Mamut press was used. This had
a cylinder with 52 cm dia., a charge height of 65 cm and a charge weight of 210 kg.
The pressure applied was 60-70 kg/cm2. If the powder extruded was damp (3-5 y
of moisture) it had to be dried to 1.0-1.2% moisture content.

SOLVENTLESS POWDER IN JAPAN

The manufacture of solventless double base powder began in Japan in 1924
It appears that the Japanese Army and Navy used two types of double base
powders [33] (Table 186).

Table 186

Ingredients

Names

G OTSU Mk I

| G OTSU Mk II

Nitrocellulose

63.9-64.3

58.9-59.3

(11.85% N)

(12.79% N)

Nitroglycerine

27.0

35.0

Centralite

4.0

2.5

Formyldiphenylamine

4.0

2.5

Inert compounds in proportion:

Ammonium oxalate

50,

Sodium bicarbonate

1.1-0.7

Graphite

iol

Properties

Heat of explosion (kcal/kg)

726-734

960-967

Vq a/kg)

979-980

892-893

t (°Q

2410-2427

3006-3025

/ <m)

9946-10,016

11,077-11,148

Powder G OTSU Mk I was designed for general ordnance use. It was characterized
by low corrosion.

Powder G OTSU Mk II was mainly used in naval revolving turret guns.

FLASHLESS CHARGES AND FLASHLESS POWDERS

The first flashless charges were made during World War I. They were developed
from an observation of Dautriche [43] that addition of blackpowder reduces and
attenuates flash or even entirely prevents the formation of a secondary flame. The
French therefore added blackpowder to nitrocellulose powder and during World
War I regularly loaded part of their machine gun ammunition with a mixture con-

664 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

sisting of nine parts of smokeless powder and one part of blackpowder. In their
cannon they used silk anti-flash bags filled with potassium hydrogen tartrate. Since
this substance reduced the muzzle velocity, an extra charge of smokeless powder
was used. E.g. in 155 mm naval guns, 305 g of BM 7 powder were added to restore
the muzzle velocity to a charge of 10 kg of BM 7 powder with a priming of 115 g of
blackpowder, when three bags of 500 g of potassium hydrogen tartrate were used.

Another method used in France was to add anti-flash pellets, consisting of four
parts of potassium nitrate and one part of DNT, to propellant charges. The pellets
weighed one gramme each and were about 2 mm thick and 15 mm dia. They were
sewn in silk bags, in numbers depending on the calibre e.g. 200-300 were used for a
155 mm gun. Such pellets behaved as a propellant charge and did not reduce the
muzzle velocity.

The Germans used anti-flash charges containing potassium chloride in their
cannon propellants. The charges, in bags of artificial silk or cotton cloth, were loaded
between the base of the projectile and the propellant. Obviously, all the additions
described above increased the smoke formed when the rounds were fired.

After World War I FNH powder was produced in the U.S.A. It was flashless,
and non-hygroscopic and according to one of the relevant patents [44] consisted of:

Nitrocellulose (13.15% N)

76-79%

DNT

21-24%

Diphenylamine

Nitrocellulose

84%

DNT

10%

Butyl phthalate

Diphenylamine

Flashlessness was attained by reducing the heat of explosion with an addition of
DNT. At the same time, however, smoke was increased.

Other patents of the interwar period include several that specified addition of
substances rich in carbon, e.g. of powdered hydrocellulose, to obtain flashless charges.
In the U.S.S.R. nitroglycerine powder was used in which a part of the nitroglycerine
was replaced by aromatic nitro compounds. During World War II the most widely
used flashless powder contained nitroguanidine (in Germany called "Gudol" powder).

The idea of adding nitroguanidine to smokeless powder had been already con-
sidered by Vieille [45]. He suggested adding nitroguanidine to reduce the erosiveness
of the powder (see p. 548)

The idea was revived by various authors (e.g. Recchi [46]).

It was difficult to manufacture since the nitroguanidine had to be introduced
into the powder mass in a state of fine subdivision, otherwise the powder was not
uniform. In some factories, therefore, methods were worked out to obtain nitroguani-
dine in the form of fine dust.

The introduction of a large amount of nitroguanidine would be very difficult,
were it not for the replacement of the nitroglycerine by dinitrodiethyleneglycol which

in
si
2

D
O

-I

as
w

a
u

w
a!

a!
o

ei
W

I
O

THE MANUFACTURE OF SMOKELESS POWDER

665

(8>(/[B0>|)
3ll[BA 3gU0[B3

tSjuaipajaui Jsqjo

3}«qd[ns lunissBjOji

§§8§§§§§§§§

«"> in in v~i

rn n m io ^ ^ ^

o © o o © © o

I I I I

I I

spixo uimsauSBjAj

SJUJClBJO

3}ipjB3V

I I I I

■n in in in m

in in in

o o o
odd

I I I

o
d

I O rt o

odd

3}IIBJ)U93

I I I I I I I

in in in m

d d d -.'

ampiuBnSojjit^

[OoXiSipojjif^

3UU33X(gOJ}I)sI

^ ^ ^ en ci in m
" " —' d d d d

I I I I

I I I I I I I I I

I I I

°. P. P

ON OS ON o
m m ro ■*)■

it —I I

% iraSoijiN

o o o

555

I I I I I

qqqqqqooooo

m rn m ro r»i r<i r<i ni n' r<i r^'

ininmrNcNfNrt^ortqrN!

S 3 S 8 8 8 8 S 3 3 55

C-4

00 CN

d
x

„• X •*

j3 .5 .5 rt =tf =tf G

n, c* t* E E E 5

t; so

o ^

60 £

a a

c22 , o

c u
o

»n »n

»n vi in

a g

o .5

J3 _

oo t^ K S

22a

f*l 't "O lO h. M

666

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

(85(/lBO>i) anjBA oguopo

m o
oo t—

V"!

■ r4

saoistraima

»n

i -

rn
X
O

^J"

suips^A

i 2

**'

>>
>

sjmidjns uinissBjoj

1 1

a
a

»n

■<

v-i in

«->

H
<

spixo umis3u3Bj^

d d

sjiqdBjQ

d d

ajBiBqjijd i/Cjng

1 »5

t- "S

«->

00 A 1

• o

O tN

0^5

3)IJBI}U33

S "0

"3 O m

00 On

.& £

(S °

S fi >/"i

T3 "3 5 U

S +

l~ IH

C «4— 1 >> C

ea O C (S

i-I

looX[8ipoj}iis[

(N CS

l-H

% usSojjiM

o
c4

'-i

*— 1 »— 1

t— 1

vo 3

<£ «S

a! m

'e3

.3 k> .3

1 §

1
e

u g

C 3 c
3 M ™

E - £ S

a -o £ 3 e

°°. 9 i «j 1

E 1

•*; §i

^ a

00 00

*— 1

<N en

TJ"

THE MANUFACTURE OF SMOKELESS POWDER

667

a
n

r 5

(8>[/tB0>() anjBA oguopo

o o

suoisnaimg

C^' c^«

ajBqdjns umtssEjoj

i 5

apixo uinissuSBj^

•— N

ajiqdEjr)

T— 1 '"I

ajipjBoy

in
© 1

ft
S
o

3UBV[j3iri]/Caaqdia

m in

streqjamiXugqdi^qjg

t~; q

en >n

stnpiirenSojjijsj

o ©

«i

IOdXiSipojjim

u
o

% uaSojjiM

■— i t— t

J3
1

"1 "1

<*i On
^- en

C4-I *i-

p p

.fci .fa
as a

3 3

S E

00 00

-H CI

668

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

,0
'«>

o
a

E
o
O

Wl^l)
snjBA Mjuon33

690-730

730
720

suoisuauiiQ

as"
o

ajBqdjns umissBjoj

| i 3

spixo uinissuSejAj

"ft in vi
d do'

sjiqdBio

d do

asoinijsooipXjj

1 °-

1 m *t

suBijjsjniXusxjdiQ

1 1 2

P
«

3UBH}amiifri3udiKu,}g

1 1 2

ajipjBoy

1 1 ^

' ' o

3JIIEIJU83

"ft
^ 1

3U9iBqiqdBU0JJ!is[

"ft o

XNa

O "1

ampiuenSoflif^

1 1 ©

Ioo^i8ipojjifv[

14.8-15.0

25.95
18.85

o
_o

1
o

% nsgojjijst

tN »S (S

OS
C\
SO

I >ft
^ "ft o

SO so #

3
i

E
o
oo

•a

(3
3
O
(-.

guns
All guns, commonly

used (Universal

powder)
Various guns

»—i

<s m

THE MANUFACTURE OF SMOKELESS POWDER

669

» ^ t- » » i-i

SU0ISU3UIIQ c~. e- e- e- c- c-

ajBqdpis umissEjoj

III ^ •=?

apixo mnissuSej^

N ^ (N N N (M

d o o d o ©

sjrqdBiQ

3}Breqjqd [Xjng

m
1 ^ 1 1 1 1

suBqpinj^usqdia

i i i i 55

z .

3 .«

3 3UBqj3Jni^u3qd[Xqjg

' i i i 53

LE 191

FOR NAVAL G

n
5

3- sjipjBoy

' | 2 i 2 i

3}I[BJ}U33

»n

!fl C\ N 2 1 |

Tab

DWDER

suaiBqjqdBuojjijvj

1 1 "- 1 1 1

<
p

3UipiUBngOJ}I«[

o o
1 1 1 1 d ri

D
H

[03^[glJ}OHI[<[

1 1 1 £ 1 1

Ioo/tySipoajiN

25.27

23.9

23.95

21.75
18.2

M N N M N O

oo in to

n ^j r^ in in tj-

on "o in oo in ri

\o ^o id in n tf

Purpose

All guns, commonly used
AH guns, commonly used
All guns, commonly used
Signal rescue charges
37cmgunSKC/32
Most important guns

h m n ^ in io

670

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

imparts a higher elasticity to the powder. This enabled the "foreign" substance to
be incorporated into the powder mass while retaining its crystal form.

Powder containing nitroguanidine has certain disadvantages. It is more fragile
than the ordinary solventless powder, due to the presence of nitroguanidine, i.e. a
substance which is not included in the powder colloid. On the other hand it has the
advantage that it does not increase the smoke to a marked degree. Small amounts
of other anti-flash substances such as potassium nitrate or sometimes powdered
hydrocellulose are added to this powder. The compositions of powder of this type
and of other solventless powders manufactured in Germany during World War II,
are summarized in Tables 187-191.

In Great Britain small amounts of sodium or potassium cryolite (potassium
aluminium fluoride) were added to nitroglycerine powder containing nitroguanidine.
According to Pring [47] a powder with 55% nitroguanidine and cryolite gives no
flash in all calibres of cannon up to 16 in. (40 cm). Charges of this powder, however,
are harder to ignite and thus require a larger quantity of blackpowder as a primer.
This, in turn, increases the smokiness, since the majority of smoke in firing with
smokeless powder derives from the blackpowder primer.

Other substances suggested for flashless powder instead of nitroguanidine included
aminotetrazole which was, however, rejected due to its hygroscopicity, and mefhylene-
urea.

Recently it was found that caesium salts prevent the formation of secondary-
flame.

SMOKELESS POWDER WITH PENTHRITE

During World War II, the German tried using penthrite as an ingredient of smoke-
less powder. Penthrite, in spite of being a nitric ester, does not dissolve nitrocellu-
lose and thus retains its crystal structure in the powder mass, and causes brittleness.
(like nitroguanidine). The introduction of a large amount of penthrite into the powder
mass was possible only when dinitrodiethyleneglycol was used, which, as is known,
gives a more elastic and mechanically resistant powder. Powder containing penthrite
(Nipolit) has been manufactured on a small scale. The manufacture never passed
beyond pilot plant scale. The production process was the same as in other nitroglyc-
erine powders. The composition of Nipolit is given in Table 192.

Table 192
The composition of Nipolit

Tubes

Cords

Ingredients

80x27/9.1

50x9.1

Nitrocellulose (12.6-12.7% N)

34.1

29.1

Dinitrodiethyleneglycol

Penthrite

Magnesium oxide

0.05

Graphite

0.1

Centralite or urethane

0.75

THE MANUFACTURE OF SMOKELESS POWDER 67 j

SMOKELESS POWDERS CONTAINING NITROALIPHATIC COMPOUNDS

Hexanitroethane was recommended as an ingredient of smokeless powder [481
It is a good solvent of nitrocellulose and due to its high energy of explosion it could
produce a high energy propellant.

It was believed that the German Army used smokeless powder containing hexa-
nitroethane for the long range artillery that bombarded Paris in 1915.

The complicated and costly manufacture of hexanitroethane prevented further
use of this substance as an ingredient of smokeless powder.

There is now a trend towards the use of nitroaliphatic compounds more readily
available from simple nitroparaffins. Thus Aerojet-General in Sacramento California
[49] suggested using a 50/50 mixture of bis-(2,2-dinitropropyl)-formal (I) and bis-
(2,2-dmitropropyl)-acetal (II) as a "nitroplasticizer". The mixture of the two com-
pounds is liquid. The products are obtained from nitroethane through the following
sequence of reactions :

CH 3 CH 2 N0 2 _*__> CHrf^-J™ CH 3 cf N ° 2 5!_>

\ K2CO3 NX

\N0 2 ^NO©K®

Hr wn ° 2N N °2

n 3 <\ /No 2 CH20 n^

no 2 S55J-* CH3 ~ c - CI * a - \c Ha

CH 3 -C-CH 2 -0
CH 2 o *L

CH3CH0 N x x NO

(BF 3 ) t 2

2 N N0 2

\/
CH 3 — C— CH 2 — (X

"* >CH— CH

CH 3 — C— CH 2 — <y

2 N N0 2
II

The composition of the propellant was not revealed. It is only known that the
propellant contains ammonium perchlorate as an oxidizer and has a very high
density.

SMOKELESS POWDERS FOR ROCKETS

In the interwar period, after 1930, a number of countries (e.g. Germany, Great
Britain) began to manufacture double base powder for rockets. In Great Britain
(according to Wheeler, Whittaker and Pike [37]) it was manufactured in the form
of tubes 2 in., 3 in. and 4.3 in. dia. The composition of German rocket powders
is given in Table 193. Tubes of larger diameter were also made.

672 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Due to the great web thickness the burning time of the powder tubes was consi-
derable: about 3-4 sec under a pressure of about 10 kg/cm 2 .

In composition, nitroglycerine powder (with a relatively low content of NG
or DEGN without solvent) is the same as that commonly used without a volatile
solvent for firearms ; it has, however, a characteristic which calls for an examination
of its use for rocket propulsion.

(1) The gases evolved on burning are intensely luminous. Non-transparent dark-
coloured particles, often not visible to the naked eye, e.g. undissolved nitrocellulose,
fibres or various impurities which are always present in the powder grains, adsorb
this radiation more readily than the colourless or light-coloured powder mass. The
temperature in the neighbourhood of these dark spots is therefore higher than that
in the whole mass and even may be high enough to produce ignition of the surrounding
powder. This creates the danger of irregular burning at many points inside the powder
mass, causing the powder grains to crack. It also causes a sudden increase in the
surface of the burning grains and a rise in pressure that may lead to the explosion of
the rocket (blowing out the steel rocket case). Accidents of this kind have been noted
with nitroglycerine smokeless powders of a web thickness of over 1 5 mm. After a
certain time it was found that the addition of a darkening agent prevents the prema-
ture burning of the powder mass below the surface. Nigrosine and lamp black were
used for this purpose. The darkening agent finally adopted was carbon black in the
extremely fine form used in the rubber industry with a grain diameter from 0.025-0.5
fi. It is added to the powder in an amount of 0.01-0.2% by weight. It is also advisable
to add some graphite, as was used formerly to facilitate the extrusion of a powder
mass without a volatile solvent.

(2) Powder for rockets is usually in the form of perforated grains with a large
diameter, considerably larger than that of the tubes used for cannon charges. This
makes the production process for extruding the powder mass very complicated.

(3) The powder in rockets burns under much lower pressure than that in firearms
(usually below 2000 lb/in 2 ) and its burning rate is also considerably lower. The low
pressure and the low burning rate, together with the shape of the powder and of
the combustion chamber, specific for rockets, lead to a reaction, totally unknown in
firearms, called "resonance burning", which is characterized by certain periodic
pressure oscillations. It may well be that this is caused by unevenness of the powder
and the presence in it of droplets of nitroglycerine, which was discussed earlier (p. 647).
Irregular burning may be alleviated by drilling radial holes in a spiral pattern along
the length of the powder tube, or a non-combustible rod may be inserted into the
perforation. This "stabilizing rod" is used in many current rockets.

Experience in using large charges of nitroglycerine powder has shown that powder
tubes with a large diameter are unsafe to use due to the internal stresses which arise
in them during cooling. While the powder is burning the tube may crack due to the
local weakening of the walls and the pressure of the gases. The burning surface of
the powder then increases, the pressure rises and the rocket may be blown up.

THE MANUFACTURE OF SMOKELESS POWDER 675

In small Soviet and German rockets a powder charge made of a bundle of small
diameter powder tubes (e.g. to 20 mm) has been used. The internal stresses in these
charges were not sufficiently strong to be dangerous.

In Great Britain and the U.S.A. cruciform, non-perforated powder grains have
been used for rockets of heavy calibre. In these charges the mechanical stresses are
far less important than in tubular charges. Another modification, consisting in the
very slow cooling of the charges after extrusion and cutting up, necessitated new
installations and prolonged the flow process, so a new method was introduced in
which the charges were cast. This is discussed further on.

The composition of nitroglycerine powder for rocket propulsion is similar to
that of conventional powder.

The composition of German rocket propellants manufactured during World War
II are given in Table 193.

rABL

E 193

^OCK

ET POWDER [42]

Composition (%)

Nitrocellulose

c
a

No.

Purpose

n
n

I-

"3
o
>.

•3
o

■3

u
3

">.

C
<D

-C
Q,
">■

_3
">>
C
<D

a.
S

~o
o
o

•a

"53

"2
'J?
o

3
'tri
U

o
ao

"3
<»

"53
to
t3

<*>

c
o

E
Q

Calorific value
(kcal/kg)

"Universal"
powder for all

rocket launchers

60.0-60.2

12.6

35.3

—

1.4

1.0

1.5

n?5

0^5

900

For 30 cm rocket

launcher

59.05

12.6

34.8

0.5

1.9

3.0

0.5

0.25

865

Serebryakov [50] reported the following composition for rocket powders:
J.P.N. (U.S.A.) rocket powder consists of:

Nitrocellulose (12.2% N)

Nitroglycerine

Diethyl phthalate

Carbamite

Potassium sulphate

Wax

Carbon black

Moisture

M.R.N. (U.S.A.) rocket powder consists of:

Nitrocellulose (12.2% N)
Nitroglycerine

51.5%
43.0%
3.25%
1.0%
1.25%
0.08%
0.2%
0.6%

56.5%
28.0%

674 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

Diethyl phthalate

3.0%

Aromatic nitro compounds

11.0%

Carbamite

4.5%

Wax

0.08%

Moisture

0.6%

German powder:

Nitrocellulose (12.2% N)

64.5%

DGDN

29.0%

DNT

11.0%

Centralite

3.5%

Candelilla wax

1.0%

Carbon black

0.35%

Moisture

0.5%

According to Ley [51] British smokeless powder for 2

Nitroglycerine

50%

Nitrocellulose

41%

Centralite

The powder was in the form of tubes about 4 cm dia. and about 1 m long.
Japanese 10-20 cm rockets were loaded with powder consisting of:

Nitroglycerine

27%

Nitronaphthalene

Nitrocellulose

60%

Centralite

Potassium sulphate

Much attention is now paid to the mechanical properties of smokeless rocket
powder [52]. This stems from the requirement for physical integrity during manufac-
ture, storage and firing. Any imperfection, such as a crack or deformation or a
stress which may lead to a crack in the course of burning can cause serious ballistic
malfunction. This defective behaviour is particularly liable to occur in large double
base powder grains, hot extruded under pressure.

According to Steinberger [53] the main problems, dependent on the physical
properties of the propellant, are :

(1) Thermal variations. A solid propellant grain may be exposed to temperature
extremes ranging from conditions of low temperature in the upper atmosphere,
to tropical heat. The grain must not be too brittle at low temperature nor too soft
at high temperature. The problem can be aggravated by "case-bonding" since differ-
ent coefficients of expansion of the metal-case and the propellant can produce a
very harmful effect.

(2) Grain collapse. This is a problem which exists in some rockets where an
appreciable pressure difference can develop between the forward and aft ends of
the rocket. The latter is subjected to compressive forces which can squeeze it. A
propellant of high tensile strength is therefore desirable.

THE MANUFACTURE OF SMOKELESS POWDER 675

(3) Grain expansion. This is the problem opposite to (2). It can be serious in
some rockets. The propellant must be sufficiently flexible to accomodate expansion
without cracking.

Steinberger [53] suggested that propellants for case-bonded units should possess
a tensile strength greater than 50 lb/in2 and elongations greater than 15%. According
to Boynton and Schowengardt [54] many applications require an ultimate tensile
strength of 120 lb/in2 and elongation of ca. 30%.

The difficulty in the manufacture of large grains may be overcome by substituting
the classical method for the extrusion of powder by a novel technique whereby cast
charges of double base propellant are produced.

CAST DOUBLE BASE PROPELLANTS

The process consists of loading a mould with granules of nitrocellulose powder
and filling the interstices between the granules with a solution of a nitroglycerine
plasticizer. This causes the nitrocellulose powder to swell and to gelatinize with the
ingredients of the liquid phase to form a horny solid on heating the mould.

This technique seems to have been developed in various countries during World
War II.

In Germany, owing to the short supply of nitroglycerine and nitrocellulose,
TNT was used as the main constituent [55] (p. 681).

The composition of various more modern cast double base propellants (accord-
ing to Sutton [56] and J. Taylor [57] is shown in Table 194. Table 195 gives the
physico-chemical properties of propellant I.

Table 194

Ingredients

Nitrocellulose

Nitroglycerine

Plasticizer (mainly dimethylphthalate)

DNT

Ethyl centralite

Diphenylamine

Carbon black or graphite

Propellant numbers
(given by the author of this book)

45-55
25-40
12-22

1-2

47.0
37.7
14.0

1.0

0.3

III "OV"

58.6 (12.6% N)
24.2

9.6

6.6

1.0

0.1

60.0

37.8

0.9
0.2
0.2

Table 195

Constants

Adiabatic flame temperature (°F)

Typical sea level specific impulse, (sec)

Burning rate of 1000 lb/in2 (kg/cm*) and 70°F (21°C) (in./sec)

Burning rate exponent (equation V=k.p n )

Figures

2600-4000
160-220
0.22-0.3
0.1-0.8

676 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The main advantage of cast propellant charges consists of the ease with which
very large charges may be made without very costly equipment.

The safety of the process is higher than that of the conventional extrusion process
and the labour cost is relatively low.

According to Sutton [56] the amount of additives in the cast double base charges
varies from 5-30%. Their object is to improve the physical properties of the powder,
facilitate manufacture, confer better stability under various storage conditions,
suppress flash and inhibit ignition caused by the radiation of energy through the
translucent grain. According to Sutton, an increase in the nitrocellulose content
of the charge usually improves its mechanical strength.

A description of the technology of the process is given below.

Technology of casting

According to published information [52, 54, 58] two methods are in use: (1) pow-
der casting process, and (2) slurry casting.

Powder casting process. This is the most versatile process for manufacturing
double base powder charges of any size and shape.

The first step consists of making a "casting powder" composed of nitrocellulose,
all the solid ingredients and a part of the non-explosive gelatinizer, such as carbamite
(centralite), phthalates, etc. Solvent (alcohol-ether) is added to this mixture and
threads of ca. 1 mm dia. are made by the usual extrusion method and cut to the
length of ca. 1 mm. The solvent is removed by drying.

This casting powder is poured into a mould with a core of the shape required
for the future charge ("grain") (Fig. 270). The mould is usually lined internally with a
cellulose acetate or ethyl cellulose sheet which forms the inhibitor for the finished grain,
if the grain is to burn from the inside outwards (as in most solid propellant grains
used in large rockets). The thickness of the sheet of inhibitor may range from 3-1 2 mm.
The bottom of the mould may also be covered with the same material. The mould
should be provided with an equipment for the evacuation and pressurization of
the assembly.

The casting powder fills ca. \ of the mould. Then the inside of the mould is
evacuated to remove air and residual solvent vapours.

The interstitial spaces between the granules of the casting powder are slowly
filled with the "casting solvent" composed of a mixture of plasticizers such as:
nitroglycerine, diglycol dinitrate and a further quantity of carbamite (centralite)
and esters. Introducing the solvent too fast can disturb the packed bed of granules.
Filling can be done from the top, from the bottom or radially from the core (if
perforated) and the advancing solvent front sweeps out the air, which is evacuated
through a vent (at the bottom on Fig. 270).

The filled mould is sealed, sometimes pressurized and placed in an oven at mo-
derate temperature (usually 60°C) to cure for a period ranging from 8 hr to several
days.

THE MANUFACTURE OF SMOKELESS POWDER

677

The curing action consists of diffusion of the plasticizers into the casting powder,
resulting in swelling and partial dissolution of the granules, thus consolidating the
whole into a reasonably homogeneous mass. This should be tested by X-ray ex-
amination before firing.

Castii

Inhibitor

Fig. 270. Mould for casting a rocket powder grain [53].

The process has the great advantage that very large charges can be made this
way. Grains up to 4 tons have been prepared.

Its main disadvantage consists in the fact that the process is slow.

Slurrying process. The main aim of this process is to avoid the necessity of making
casting powder. Granular nitrocellulose is used instead, a fibrous form being inaccept-
able because of the high rate of swelling and solution which makes the gel insuffi-
ciently uniform.

Nitrocellulose granules, partially colloided, of sizes ranging from a few microns
to a fraction of millimetres are suggested. The best known of the described methods
was developed by the Hercules Powder Company [54]. It uses "densified" nitro-
cellulose originally developed for the lacquer industry, with an average particle
size of 0.25 mm.

. The basic idea of slurry casting consists of preparing the propellant in one simple
mixing operation. The plasticizer is poured into a mixing pot and the solid ingredi-

678

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

ents, including the nitrocellulose, are added. The mixture is stirred for a few minutes
(usually 15 min), poured into a mould or a rocket case and cured. Figure 271 gives a
diagrammatic presentation of the main parts of the arrangement. When hoppers

Remote
control TV
camera

Hydraulic motor

Dry ingredient hoppers
Tachometer

Gangway

Sight f
port^'t

Mechanical seat chamber
• Fluid seal chamber

Mixing vessel

Hydraulic
cylinders

To vacuum

• Fluid seal chamber

Fig. 271. Diagram of a slurry mixing arrangement at Hercules Powder Co., Bacchus

Work [54].

are charged and the mixers are operated, the personnel are withdrawn from the build-
ing and all is operated by remote control.

This method can be used to make propellant charges of ammonium perchlorate,
RDX, nitroglycerine and triacetine.

Physical properties of cast propellants

Steiberger [53] gives a number of diagrams which illustrate the change of mechan-
ical properties: tensile strength, elongation and modulus of elasticity with nitro-
cellulose content (Figs. 272, 273 and 274). It is evident that nitrocellulose exerts a
great influence on the properties of the system.

THE MANUFACTURE OF SMOKELESS POWDER

679

3000

"1 : r

2000

WOO

500

300

200

-J 1 L_

-60 -40 -20 20 40 60 80 WO 120 140
Temperature, °F

Fig. 272. Tensile strength of cast double base powder as a function

content [53],

of nitrocellulose-

j 1 1 i_

SO -40 -20 20 40 60 80 100 120 140
Temperature, °F

Fig. 273. Elongation of cast double base powder as a function of nitrocellulose-
content [53].

680

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

3
■9

1000

-60-40-20 20 40 60 80 100 120 140
Temperature, °F

Fig. 274. Modulus of elasticity of cast double base powder as a function
of nitrocellulose content [53].

100

—I r- 1 1

y^M-1500

/ yM -1516.2

// /M-1515.3

'l/y —- — <VV- 1514. 2

< / ^^M-1499

/Grain No. XCrossllnker

M - 1499 4.4

Uu/

M -1514.2 3.1

i v/

1 f/

M- 1515.3 2.1

: ■'

M- 1516.2 1.0

M-1500 OH

1 1 .1 1

-40

40 80
Temperature, °F

120

IFig. 275. Influence of cross-linking of nitrocellulose on physical properties of cast

charges [53].

THE MANUFACTURE OF SMOKELESS POWDER 681

Variations in the viscosity of nitrocellulose have only negligible effect on the
physical properties of the propellants. On the contrary, cross-linking the nitro-
cellulose by adding polycarboxylic acid anhydrides, di-isocyanates or metallic salts
may produce a profound influence as shown on Fig. 275. Elongation is significantly
decreased especially at high temperature. The modulus is increased correspondingly
while the tensile strength is unaffected. A low degree of cross-linking may be partic'
daily beneficial, decreasing deformation or flow at high temperature while leaving
the low temperature properties unchanged.

According to Boynton and Schowengardt [54] the ultimate tensile strength of
double base rocket charges should be 120 lb/i n 2 and elongation ca 30°/ The
propellant which gives these properties contains less binder than the maximum im
posed by the castmg powder process, but it can be produced by the slurry process

Grosse [55] indicated that a typical NG double base powder on static testing
shou d produce a maximum stress of 200 kg/cm*. When the loading rate is increased
to 0.23 sec and 0.15 sec until fracture, the stress is 360 and 442 kg/c m 2, respectively.

German cast propellants for rockets

Originally this powder (Giessling Pulver) consisted of:

TNT

50-52%

Nitrocellulose

28-30%

DGDN (Diethyleneglycol

dinitrate)

17-18%

Centralite

0.5%

Diphenylamine

0.5%

The propellant was in the form of tube or plain grain, the latter having a dia-
meter up to 50 cm and a length of about 100 cm.

Finely ground nitrocellulose of low viscosity was used. It was impregnated with
DGDN under water in the usual way. After dehydration in a centrifuge to a moisture
content of 18-22%, the blend was stirred into molten TNT. The vat was evacuated
to evaporate the water and to produce a homogeneously gelatinized grain.

These propellants had rather poor physical properties produced by a high pro-
portion of TNT: they were too brittle and unsuitable for spin stabilized rockets.

A ridged, block-shaped propellant with nine perforations was later used for 21 cm
Nebelwerfer rockets. It was 12.5 cm in diameter, about 40 cm long and weighed
about 6.5 kg. Its composition was:

Nitrocellulose

63%

Diethyleneglycol dinitrate

35%

Centralite

0.5%

Wax

0.2%

Graphite

1.2%

682 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

The disadvantages of smokeless powder as a rocket propellant, the difficulty
in manufacturing very large charges and the high operating expenses led to a search
for explosives based upon another principle: plastomers or elastomers with an
oxygen carrier.

GENERAL SAFETY CONSIDERATIONS IN THE MANUFACTURE OF

SMOKELESS POWDER

The safety aspects of the manufacture of smokeless powder have already been
mentioned when describing manufacturing methods.

There are however some safety aspects which can be considered in general terms.
They can be classified into two main groups according to the origin of the danger:

(1) The use of combustible solvents ;

(2) The sensitiveness of nitrocellulose or nitroglycerine and similar esters to
thermal or mechanical factors or to electric discharge.

Solvents

A number of unexplained accidents were encountered in nitrocellulose powder
factories associated with handling alcohol and particularly ether.

Thus, the simple opening of a valve closing a pipe bringing ether from a tank
placed at a certain height may cause the ether to catch fire. Investigation of the
origin of this phenomenon at the begining of the XXth century led to the conclusion
that inflammation of the ether occurred when the static charge acquired by the
solvent flowing along the pipe was discharged on contact with an earthed object.

The first experimental work on the production of static charges in liquid during
its flow along insulated pipes were carried out at the end of the last century [59].

Freytag [60] collected the data on the subject obtained by various authors. Some
of the figures referred to solvents used in the manufacture of powders are given
in Table 196, according to Dolezalek [61].

An electric discharge from other sources can also ignite solvents.

The following accident was described in a cordite factory. A worker wearing
rubber shoes wanted to clean his hands which were soiled with powder dough.
He rubbed his hands together and held them towards a bucket filled with acetone
standing on the concrete floor. A spark between the hands and the surface of the
acetone ignited the latter.

In this instance, the worker's body became charged by the friction generated on
rubbing his hands. According to Freytag [60] this kind of friction may charge the
human body to a potential of 10,000-14,000 V.

Because of these accidents all tanks containing solvents and all pipes were earthed.
This prevented further accidents.

THE MANUFACTURE OF SMOKELESS POWDER

683

Another danger associated with solvents has already been mentioned (pp. 601, 607),
i.e. the explosibility of mixtures of combustible vapours with air.

These mixtures can be exploded by an electric discharge or by sparks from steel
tools etc. Bronze is much safer in that respect and bronze tools should be used
wherever possible. Beryllium bronze was strongly recommended at one time. How-
ever the toxic properties of beryllium limit its application.

The inflammability of gas mixtures were investigated by early workers, such
as Davy, Bunsen and particularly the French School of Le Chatelier and M. Berthelot.

Table 196

Potential of the electric charge

produced by moving liquids along

METAL TUBES AT THE RATE OF 3 Itl/seC
(DOLEZALEK [61])

Liquid

Metal

Potential
(V)

Moist ether

3100

2500

2000

Brass

1200

700

Dry ether

3100

2200

3000

Brass

1300

2200

Investigation of the limits of flammability and of self-propagation of flame
through a gaseous explosive medium became one of the main subjects of research
on the combustion of gaseous mixtures.

Early results are collected in the monograph of Bone and Townend [62].

The investigations of White [63] were particularly important. He discovered
that the limits of flammability depend not only on the chemical nature of the vapours
but also on the direction of propagation of the flame. Figures for combustion in
glass tubes 7.5 cm dia. are collected in Table 197.

Table 197

Limits of propagation of flame in air

Vapour

Upward

Downward

(%)

Methyl alcohol

7.05-36.5

7.45-26.5

Ethyl alcohol

3.56-18.0

3.75-11.5

Ethyl ether

1.71-48.0

1.85-6.40

Acetone

2.89-12.95

2.93-8.60

Benzene

1.41-7.45

1.46-5.55

684

CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

In his later experiments White [64] found that the limits vary with the diameter
of the tube, being wider in tubes of larger diameter.

Although considerable research has been carried out after these classic exper-
iments, and a number of excellent monographs has appeared dealing with explosions
of gas mixtures [65-68] little has been added to our knowledge of the limits of gas
explosions.

Recent data given by Lewis and von Elbe [66] are collected in Table 198.

It is now well established that in large tanks and large compartments the limits
of gas explosions are wider than in relatively narrow tubes.

Powder grains

Powder grains are sensitive to impact and friction. This is due to the sensitiveness
of the main ingredients, nitrocellulose and nitroglycerine. The sensitiveness of these
compounds was discussed in Vol. II.

It should be born in mind that the sensitiveness to impact and friction increases
with temperature. Hence any handling of warm powder (e.g. immediately after
drying) should be avoided.

Another danger is produced by the ease with which a powder (particularly
warm one) can be charged with static electricity. This can be produced by the fric-
tion of warm air passing through the layer of powder grains in the course of drying.
It can also be produced in the course of mixing and screening, and in the polishing
drum.

The last operation should be considered particularly dangerous because of the
high potential of the electric charge which can be created under prolonged friction
at an elevated temperature. According to the literature [60] celluloid (which possesses
properties very similar to nitrocellulose powder) can be charged to 40,000 V by
simple friction.

All machinery for powder manufacture should therefore be carefully earthed.

Table 198
Limits of vapour-air mixture inflammability [66]

Limits

Vapour

of inflammability

Lower

Upper

(%)

Methyl alcohol

6.72

36.50

Ethyl alcohol

3.28

18.95

Amyl alcohol

1.19

Isoamyl alcohol

1.20

Diethyl ether

1.85

36.50

Acetone

2.55

12.80

Ethyl acetate

2.18

11.40

THE MANUFACTURE OF SMOKELESS POWDER

Buildings and their lay-out

685

In the manufacture of smokeless powder the primary danger lies not in explosion
but in an outbreak of fire. For this reason the buildings usually have walls of a
standard thickness, with the exception of those in which a large quantity of powder
is accumulated. In drying houses, for instance, a blow-out construction, with one

□ D
D □

//\

n n n

IV\

Fig. 276. A rescue ramp in a multistoreyed building.

L *• • -Vti

*,;&

'**.•- s* . «

Fig. 277. General view of a smokeless powder factory (Western Cartridge Company,
Division of Olin Industries, Inc., East Alton, Illinois, U.S.A.).

686 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

light wall, may be adopted. Buildings in which there is a risk of explosions of a
mixture of solvent vapours with air are of similar construction. All buildings are
usually one storied, although sometimes there are installations requiring a multi-
storied design. These buildings are provided with emergency chutes (Fig. 276) for
rapid exit from the upper floors. In some countries to permit the rapid escape of
workers in case of fire, the regulations demand that no part of the premises
should be further than 4.5 metres from a door. As in all factories or laboratories
where explosives are dealt with, the doors open outwards.

Cement basins filled with water are placed by buildings in which particularly
dangerous work is carried on. The water is heated in winter. If a workman's clothes
should catch fire he can immediately jump into one of the basins. The water should
be changed frequently and kept clean to prevent the risk of infection of burns.

Electric installations should be hermetically sealed and non-sparking. The bulbs
should be enclosed in safety shields. Electric motors should not be installed in pre-
mises in which powder is manufactured.

The distance between the buildings should be large enough to prevent trans-
mission of a fire from one building to another. A typical lay-out of a powder factory
is given in Fig. 277.

LITERATURE

1. A. G. Gorst, Porokha i vzryvchatyye veshchestva, Oborongiz, Moskva, 1957.

2. U. Gallwitz, Die Geschiitzladungen, Heereswaffenamt, 1944.

3. C. Lundholm and J. Sayers, Ger. Pat. 53296 (1889).

4. I. N. Zakharov (1892), according to V. A. Boldyryev and S. A. Brouns, Kratkii kurs
tekhnologii porokha, p. 108, Gosnatekhizdat, Moskva-Leningrad, 1932.

5. D. I. Mendeleyev, Sochinenya (1890-1894), Vol. IX, Izd. Akad.' Nauk SSSR, Leningrad-
Moskva, 1949.

6. E. Berl and I. M. Delpy, Z. ges. Schiess- u. Sprengstoffw. 8, 129 (1913).

7. H. Brunswig, Das rauchlose Pulver, W. de Gruyter & Co., Berlin & Leipzig, 1926.

8. P. Pascal, Explosifs, poudres, gaz de combat, Hermann, Paris, 1924.

9. Ponchon, Mem. poudres 19, 81 (1922).

10. L. G. Svetlov, Piroksilin i bezdymnyi porokh (Ed. L. G. Svetlov and N. S Puzhai) 2, 308
(1935) (ONTI, Moskva).

11. T. S. Yegorov, Proizvodstvo bezdymnogo piroksilinovogo porokha, Moskva, 1935.

12. T. L. Davis, The Chemistry of Powder and Explosives, Wiley, New York, 1943.

13. L. Vennin, E. Burlot and H. Lecorche, Les poudres et explosifs, Beranger, Paris-Liege,

14. Claude, according to Celluloide de Oyonnax, 1910.

15. J. H. Bregeat, Fr. Pat. 502882, 502957 (1916); 503728 (1917); Technical Records of Explosives
Supply 1915-1918, No. 8, Solvent Recovery, H.M.S.O., London, 1921.

16. C. Sudlitz, unpublished results (1932).

17. Poudrerie du Ripault, Mem. poudres 21, 178 (1924).

18. R. Brun and Rotouis, Mem. poudres 36, 163 (1954).

19. A. V. Sukhtnskh (1892-1894), according to V. A. Boldyryev and S. A. Brouns, Kratkii
kurs tekhnologii porokha, Gosnatekhizdat, Moskva-Leningrad, 1932.

THE MANUFACTURE OF SMOKELESS POWDER 687

20. E. Bazylewicz-Kniazykowski and K. Partyka, Ger. Pat. 570459 (1932); Swedish Pat.
73851 (1931).

21. PASSBURG, Ger. Pat. 28971,40844 (1884); 56330 (1890).

22. Bonneaud, unpublished report (1918).

23. N. L. Hansen, Z. ges. Schiess- u. Sprengstoffw. 6, 461 (1911).

24. According to L. Vennin and G. CHESNEAU, Les poudres et explosifs et les mesures de se'curite
dans les mines de huille, Beranger, Paris-Liege, 1914.

25. T. L. DAVIS, Army Ordance 2, 9 (1921); Ind. Eng. Chem. 14, 1140 (1922).

26. J. A. Zaehringer, Solid propellant rockets, Amer. Rocket Co. Box 1112, Wyandotte, Michigan.
1955.

27. F. Olsen, G. C. Tibbitts and E. B. W. Kerone, U.S. Pat. 2027114 (1936); 2111075 (1938);
2175212 (1939); 2206916 (1940).

28. T. R. Olive, Chem. Engineering 53, 136 (1946).

29. Olin Industries Inc., East Alton, 111., U.S.A.

30. P. Brown, The American Rifleman, p. 17, December, 1952.

31. According to D. R. Cameron, in Encyclopedia of Chemical Technology, Ed. R. E. Kirk and
D. F. Othmer, Vol. 6, Interscience, New York, 1951.

32. R. Robertson and W. Rintoul, Brit. Pat. 25994 (1901).

33. BIOS/JAP/PR/1292/Report. Japanese Propellants— Research on Non-volatile Solvent Pow-
ders, H.M.S.O., London.

34. M. Marqueyrol and Florentin, Mem. poudres 18, 150 (1921).

35. J. D. Huffington, Trans. Faraday Soc. 47, 864 (1951).

36. Engineer {London) 185, 286 (1948)

37. W. H. Wheeler, H. Whittaker and H. H. M. Pike, /. Inst. Fue 20, 137 (1947).

38. G. Fleury, Mem. poudres 24, 49 (1930-1931).

39. H. Freyer and Dynamit Nobel A.G. Troisdorf, Ger. Pat. 1013555 (1957); 1053373 (1959);
1082175(1960).

40. G. Schenkel, Schneckpressen fiir Kunststoffe, Hanser, Miinchen, 1959.

41. K. WrObel and J. Luczaj, Wytlaczanie tworzyw sztucznych, PWT, Warszawa, 1961.

42. CIOS XXVII-72, Manufacture of Solventless Type Smokeless Powder and Nipolit, Deutsche
Sprengchemie G.m.b.H.; Technical Report P.B. 925, Explosive Plant D.A.G. Kriimmel, Dune-
berg, Christianstadt, U.S. Dept. of Commerce, Washington, 1945.

43. H. Dautriche, Compt. rend. 146, 535 (1908); Fr. Pat. 385769 (1907).

44. E. S. Goodyear, U.S. Pat. 2228309 (1941).

45. P. Vieille, Mem. poudres 11, 157 (1901).

46. V. Recchi, Z. ges. Schiess- u. Sprengstoffw. 1, 285 (1906).

47. J. N. Pring, Industr. Chem. Manuf. 24, 467 (1948); Chem. Trade 122, 473 (1948).

48. Koln-Rottweil Pulverfabrik, Ger. Pat. 277594 (1913).

49. Chemical & Engineering News 41, No. 8 (February 25) 29 (1963).

50. M. E. Serebryakov, Vnutrennaya ballistika, Oborongiz, Moskva, 1962.

51. W. Ley, Rockets, Missiles and Space Travel, The Viking Press, New York, 1954.

52. 20th Meeting Bulletin, Joint-Army-Navy-Air-Force-ARPA-NASA Panel on Physical Properties of
Solid Propellants, 1961, Riverside, California, Vol. I and II, John's Hopkins University, 1963.

53. R. Steinberger, Preparation and properties of double base propellants, in The Chemistry of
Propellants, AGARD Panel, Paris, 1959 (Ed. S. S. Penner and J. Ducarme), Pergamon Press,
Oxford, 1960.

54. D. E. Boynton and J. W. Schowengardt, Chemical Engineering Progress 59, 81 (1963).

55. H. GROSSEin The Chemistry of Propellants, AGARD Panel, Paris, 1951 (Ed. S. S. Penner
and J. Ducarme), p. 303, Pergamon Press, Oxford, 1960.

56. G. P. Sutton, Rocket Propulsion Elements, Wiley, New York, 1956.

57. J. Taylor, Solid Propellants and Exothermic Compositions, Newnes, London, 1959.

688 CHEMISTRY AND TECHNOLOGY OF EXPLOSIVES

58. F. A. Warren, Rocket Propellants, Reinhold, New York, 1958.

59. Richter, Angew. Chem. 6, 218, 502 (1893).

60. H. Freytag, Raumexplosionen durch statische Elektrizitdt, Verlag Chemie, Berlin, 1938.

61. Dolezalek, Chem. Ind. 33, 147 (1913).

62. W. A. Bone and D. T. A. Townend, Flame and Combustion in Gases, Longmans, Green & Co.,
London, 1927.

63. A. G. White, /. Chem. Soc. 121, 1244, 1688, 2561 (1922).

64. A. G. White, J. Chem. Soc. 125, 2387 (1924); 127, 48, 672 (1925).

65. W. JOST, Explosions- u. Verbrennungsvorgdnge in Gasen, Berlin, 1939 ;W. JoSTand H.O. Croft,
Explosion and Combustion Processes in Gases, McGraw-Hill, New York, 1946.

66. B. Lewis and G. von Elbe, Combustion, Flames and Explosions in Gases, Academic Press,
New York, 1951.

67. V. N. Kondratyev, Kinetika khimicheskikh gazovykh reaktsii, Izd. Akad. Nauk SSSR, Moskva,
1958.

68. A. S. Sokolnik, Samovosplomenenye, plamya i detonatsia v gazakh, Izd. Akad. Nauk SSSR,
Moskva, 1960.

AUTHOR INDEX

Page numbers in parentheses denote references to Literature lists

Aaronson, H. A. (243)

Abel, F. A. 335, 336, 340 (363), 402 (495),
530, 563 (568, 569)

Abernethy, C. L. 45, 49 (75)

Abrams, A. J. J. 15, 29 (38)

Adams, R. (200)

Adicoff, A. (394)

Adryanov, N. F. (526)

Ahrens, H. 421 (496, 497)

Akre, R. L. 482 (497)

Albertus Magnus 323

Alekseyev, D. 168 (198)

Allen, C. F. H. 51 (75)

van Alphen, J. 19 (38)

Alterman, Z. 380 (394)

Altman, D. 295 (320)

Alvisi, U. 280 (287)

Amble, E. 162 (197)

Amos, A. A. (39)

Amster, A. B. (394)

Andersen, W. H. 381, 391 (393)

Anderson, J. D. C. 201 (241)

Andrews, D. H. 256 (286)

Andreyev, K. K. 55 (75), 86 (125), 171 (198),

342 (364)
Andreyevs™ 130
Andrussow, L. (285), 290 (319)
Angeli, A. 222 (242), 555, 556 (569)
Angelico, F. 136 (159)
AprN, A. Ya. 85 (125)
Aristoff, E. (126)
Ashdown 562, 563 (569)
Askey, P. J. 305 (320)
Assonov, V. A. 424, 425 (496)
Atkinson, A. 121 (126)
Atzwanger, H. 313 (321)
Aubertein, P. 33 (39), 84, 121, 122 (125,

126), 268 (287)
Audibert, E. 184 (199), 398, 409, 411 (495, 496)

Audrieth, L. F. 169 (198), 307 (320)
Aunis, G. 30 (39)
Avogadro, M. 83, 84 (125)
Avramenko, L. J. 547 (568)
AzAroff, L. V. 169 (198)

Bacchus (363)

Bachmann, W. E. 18 (38), 112, 114, 118,

121 (126)
Backer, H. J. 2, 4 (14, 75)
Bacon, Roger 323
Badoche, M. 122 (127)
Badowski 493
Baeyer, A. 225 (242)
Bailey, B. M. 316 (321)
Bain, C. J. 58 (76)
Baker, B. E. (320)
Baker, L. L. 295 (319)
Baldwin, R. R. 303 (320)
Baly, E. C. C. 2 (13), 24 (39)
Bamberger, E. 5, 7, 10, 13 (14), 201 (240)
Bamford, C. H. 305 (320)
Barab, J. 6 (14)
Baranger, P. 226 (242)
Barcikowski, J. 217 (242)
Baron, L. I. (526)
Barrere, M. 294 (319)
Barrot, J. 6 (14)

Bartlett, B. E. 143 (159), 184 (199)
Barton, S. S. 26, 27 (39)
Bauer, W. 296 (320)
Bazylewicz-Kniazykowski, E. 617 (687)
Beck, W. 133 (158)
Becker, E. D. 167 (198)
Becker, F. 35 (39), 83 (125)
Beckman, A. O. 167 (197, 198), 305 (320)
Bedard, M. (126)
Bednarczyk, M. (241)

690

AUTHOR INDEX

Behrens, H. 298, 299 (320)

Belayev, A. F. 49 (75), 139, 147 (159), 204

(241), 537 (568)
Belayeva, A. E. 147 (159), 204 (241)
Bell, R. P. 15, 16 (38)
Bellamy, L. J. 3, 4 (14), 25 (39), 165 (197)
Bellinger, F. 296 (320)
Bellot 129

Bennett, G. M. 70 (76)
Benton, A. F. 224 (242)
Berchtold, J. 146 (159), 171, 182 (198)
Berg, A. 11 (14)
Berger, E. 560 (569)
Bergmann, E. 558 (569)
Berl, E. 573 (686)
Berman, L. 92 (126)
Bernard, M. L. J. 295 (320)
Bernardini 254 (286)
Berthelot, D. 557 (569)
Berthelot, M. 138, 148 (159), 190 (199, 200),

201, 224, 227, 228, 229 (240, 242, 243),

326, 336 (363), 397, 683
Berthmann, A. 137 (159)
Berthollet, C. 274 (287), 334 (363)
Beyersdorfer, P. 420 (496)
Beyling 409, 410, 412 (496)
Bhaduri, K. 228 (243)
Bichel, C. E. 281 (287), 398, 400 (495)
Bickford 130 (131)
Bills, K. W. 381, 391 (393)
Binnie, W. P. 108 (126)
Biquard, P. 543 (568)
Blackwood, J. D. 326, 338, 339, 341

(363)
Blanchard, K. C. 15, 33, 34 (38)
Blanksma, J. J. 63, 65, 66 (76)
Blinov, I. F. 276 (287)
Blochmann, R. 228 (243)
Blokhshtein, F. I. 193 (200)
Bobovich, Ya. 3 (14)
Bogdal, S. 46 (75)
Bogoyavlenskii, A. (75)
Bogush, O. F. 256 (286)
Boileau, J. (287)
Boldyryev, V. A. (686)
Bombrini-Parodi, Delfino 258 (286)
Bone, W. A. 683 (688)
Bonhoeffer, K. F. 547 (568)
Bonneaud 621 (687)
Boucard, J. 432 (497)
Boudet, J. 138 (159)
Bourjol, G. 31, 32 (39)

Bowden, F. P. 144 (159), 172, 183, 184 (198,

199), 211 (241, 242), 326, 338, 339,

341 (363)
Bowden, R. C. (126)
Bowersock, D. C. 316 (321)
Boyd, N. C. 254 (285)
Boyer, J. H. 164, 194, 196 (197, 200)
Boynton, D. E. 675, 681 (687)
Bradt, P. 167 (198)
Brattan 303 (320)
Braude, E. A. 163 (197)
Braun, E. 163 (197)
Bregeat, J. H. 603, 604 (686)
Breyre, A. (495)
Brian, R. C. 11 (14), 17 (38)
Brinkley, S. R. (497)
Brochet, A. 121 (127)
Brockmann, F. J. 81, 122 (125)
Brouns, S. A. (686)

Brown, F. W. 176 (199), 314 (321), 543 (568)
Brown, H. R. 217 (242)
Brown, P. 635, 636 (687)
Brugere 335 (363)
de Bruin, G. 554, 556 (569)
Brun, R. 616 (686)
Brunet, P. E. 25 (39)
Brunswig, H. 532, 533, 535 (536, 539, 551,

552, 553, 554, 563, 566 (568), 574 (686)
Bruson, H. A. 73 (76)
Bryan, G. J. 172 (199), 216 (242, 394)
Budnikov, M. A. 150, 152, 153 (160)
Buisson, A. 551 (569)
Bull, W. 296 (320)
Bunsen, R. 336 (363)
Buriks, R. S. (200)
Burlet, G. 44, 45 (75)
Burlot, E. 154 (160), 236, 237 (243, 527),

541 (568, 686)
Burton, O. E. 208 (241)
Bushnell, V. C. (320)
Bystrov, I. V. (160)

Caldin, E. F. 16 (38)
Calus, H. 552, 556 (569)
Cambier, R. 121 (127)
Cameron, D. R. (568, 687)
Campbell, C. 280 (287), 382 (394)
Canright, R. B. 293 (319)
Canter, F. C. 196 (200)
Carleton, L. T. 365, 366 (393)
Carlson, O. F. 280 (287), 473

AUTHOR INDEX

691

Carlton Sutton, T. 54 (75)
Carmack, M. 2 (14), 223 (242)
Carruccini, R. J. (320)
Carruthers, A. 88 (125)
Carstanjen, E. 140 (159)
Cason, J. 120 (126), 223 (242)
Cattelain, E. 228 (243)
Cave, G. A. 32 (39)
Chaiken, R. F. 391 (393)
Chalon, P. F. (363)
Chandelon 150 (160)
Chao, T. S. (197)
Chapman, F. 116 (126)
Chapman, J. C. 374 (394)
le Chatelier, H. 396, 397 (495), 683
Chaylan, E. 268 (287)
Cheltsov, I. M. 258 (286)
Chesneau, G. (687)
Chetyrktn, W. N. 93 (126)
Chevalier 149 (160)
Chevreuil, M. 336 (363)
CmcHiBAmN, A. E. 8 (14)
Chretien, A. 211 (241)
Chute, W. J. (14), 36 (39), 92 (126)
Cirulis, A. 185 (199)
Claessen, C. 530 (568)
Clark, L. V. 70, 71, 72 (76), 202, 203, 204
Clarkson, C E. 41, 42 (74)
Claude 603 (686)
Clift, G. D. (243)
Clusius, H. 168 (198)
Clusius, K. 15 (38), 162 (197)
Cohen, H. L. 108 (126)
Cohen, J. (39)
Colli, A. J. 389

Cook, M. A. 31 (39), 54 (75), 85 (125), 290
Cooper, P. D. (39)
Cope, W. C 6 (14), 55 (76), 233
Copp, J. L. 261 (286)
Corey, B. B. 2 (14)
Costain, W. 2 (13)
Courtney-Pratt, J. S. 182, 183 (199)
Coward, H. F. 545 (568)
Cox, E. G. 2 (13)
Cox, R. F. B. 70 (76)
de Crater, W. C. 48 (75, 125)
Crawford 522
Criegee, R. (243)
Croft, H. O. (688)
Cunningham, G. L. 224 (242)
Curtius, T. 161, 166, 168, 185, 190,
192 (196, 197, 200)

Cybulski, W. 407, 410, 411, 413, 417, 418,
419, 420, 461, 477 (496, 497)

von Dahmen 266 (286)

Dailey, B. P. 162 (197)

Damon, G. H. (497)

Darapsky, A. 166 (197)

Dautriche, H. (287), 398 (495), 541, 546

(569), 663 (687)
Davidson, S. H. 490 (497)
Davidson, W. B. 201 (240)
Da vies, A. G. 227 (243)
Davies, D. A. 324 (363)
Davis, N. S., Jr. 300 (320)
Davis, T. L. 15, 24, 28, 29, 33, 34 (38, 39),
48, 49, 51, 67 (75, 76), 152 (160), 174
(199), 210, 236, 238 (241, 243), 562, 563
(569), 590, 597, 622, 624, 646 (686, 687)
Davy, Humphrey 396
Deb, S. K. 164, 165, 187, 188, 196 (197, 199)
Debenedetti, B. 559 (569)
Debus, H. 337 (363)
Deffet, L. 431, 432 (497)
Deissler 266 (286)

Dekker, A. O. 2 (14), 387, 388 (394)
(241) Delay, A. 165 (197)

Delepine, M. 81, 122 (125, 127)
Delpy, I. M. 573 (686)
Demelenne, E. 489 (497)
Demtdyuk, G. P. (526)
Demougin, P. 560 (569)
Dennis, L. M. 166, 185 (197)
Deno, N. C. 121 (126)
Descartes, Rene 335 (363)
(319) Desch, C. M. 2 (13), 24 (39)
Designolle 334 (363)
Desmaroux, J. 561 (569)
Desseigne, G. 61 (76), 256 (286)
Desvergnes, L. 23 (39), 46, 51, 53 (75)
Dewar 530 (568)
Dibeler, V. H. 167 (198)
Dickey, J. B. (14)
Dickinson, R. G. 167 (197, 198)
Dickson, G. 326 (363)
Dietl 141 (159)
Dietz, W. (321)
Dimroth, O. (200), 326 (363)
Dittmar, P. 35 (39), 83 (125, 287)
Ddcon, W. J. 445 (497)
191, Dobrzynski 217 (242)

van Dolah, R. W. 446 (497), 508 (526)

692

AUTHOR INDEX

Dole, S. H. 319 (321)

Dolezalek 682 (688)

Dolezel, Z. 122 (127)

Doixfus, F. E. (14)

Domanski, T. 37 (39)

Dougherty, C. F. 389 (394)

Downing, D. C. 81, 123 (125, 126)

Dows, D. A. 164, 167 (197, 198)

van Drebbel 129

Dubnov, L. V. 413, 420, 427, 428, 429

(496)
Dubsky, J. V. 9 (14)
Ducarme, J. (321, 687)
Duden, P. 121 (126)
van Duin, C. F. 53, 64, 65 (75, 76)
Dunn, G. E. (14)
Dunning, K. W. 92 (126)
Dunning, W. J. 92, 94, 96 (125, 126)
Dutour, M. 122, 123, 124 (127)
von Duttenhofer, M. 529 (568)
Duval, C. 165 (197)
Dwyer, J. 547 (568)
Dykstra, P. R. 372, 376, 377, 378 (394)

Easterfield, T. H. 326 (363)

Eastes, J. 296 (320)

Ebele 109

Edwards, G. 78 (125)

Effenberger, E. (198)

Egg 129

Eggert, J. 145 (159), 171, 182, 183 (198)

Ehrenberg, A. 140 (159)

Ejsmund, J. (241)

von Elbe, H. 538 (568), 684 (688)

Elderfield, R. C. 28 (39)

Eldred, D. N. 134 (158)

Elgin, J. C. 305 (320)

Elliot, M. A. 314 (321)

Engelbrecht, A. 313 (321)

Ephraim, F. 230 (243)

Epstein, S. 82, 118, 119 (125, 126)

van Erp, H. 4, 6 (14)

Escales, R. 266 (286), 352 (363, 364)

Espenschied, R. 229 (243)

Euler, H. 4 (14)

Evans, B. L. 163, 164, 169, 183, 186, 191

(197, 198, 199)
Evans, J. I. 145 (159), 172 (199), 216

(242)
Evans, W. M. 261 (286)
Eyster, E. H. 162, 164 (197)

Farmer.R.C. 51, 52,53 (75), 139, 142, 146(159)

Farrel 522

Faveau, J. 546 (568)

Fedoroff, B. T. 231 (243, 287)

Fedosyev, W. I. 294, 298, 308, 309, 311 (319)

Feitknecht, W. 178 (199)

Feldhaus (131)

le Fevre, R. J. W. 201 (241)

Ficheroulle, H. 121 (126), 189, 190 (199)

Fierz, H. E. (200)

Fiedorov, N. P. 336 (363)

Fischer, C. N. (76)

Fleming, J. S. B. 432 (497)

Fleury, G. 655, 656 (687)

Florenttn 645 (687)

Flygare, H. 231 (243)

Foner, S. N. 168 (198)

Fong, L. L. (569)

Forg, R. 269 (287)

Forst, W. 303 (320)

Forster, A. (126)

Forster, F. 529 (567)

Forster, M. O. 191 (200)

Forsyth 129

France, A. D. G. (200)

Franchimont, A. P. N. 2, 5, 6, 7, 9, 10 (14),

18, 25, 34, 36 (38, 39, 75)
Franklin, E. C. 161 (197)
Franklin, J. L. 167 (198)
Franzen, H. 230 (243)
Freamo, M. 167 (198)
Freiwald, H. 253 (285)
Frevel, L. K. 161 (197)
Freyer, H. (687)
Freytag, H. 131 (131), 682 (688)
Friedman, H. 296 (320)
Friedrich, W. 230, 231 (243)
Friese, P. 224 (242)
Fripiat, J. 407 (495)

Gaens 331 (363)

Galas, T. 85 (125), 540 (568)

Galinowski, S. 68 (76)

Galloway, R. L. 401 (495)

Gallwitz, H. 531, 540 (568), 572 (686)

Ganapathi, K. 8 (14)

Gantz, E. S. C. 26 (39)

Garfield, F. M. 202 (241)

Garner, W. E. 45, 49 (75), 142, 143 (159),

165, 169, 171, 172, 177, 182, 186, 18?

(197, 198, 199), 215 (242)

AUTHOR INDEX

693

Gay, P. F. 396, 397, 401, 414, 415, 416, 419,
422, 433, 463, 464, 465, 466, 467, 468,
490 (495), 522, 524 (527)

Gay, W. A. 290 (319)

Gaydon, A. G. 167 (198)

Gay-Lussac, J. 335 (363)

Gaudechon 557 (569)

Gibson, F. C. 176 (199), 543 (568)

Gibson, J. W. 446 (497)

Giguere, P. A. 303 (320)

Gilbert, E. C. 305 (320)

Gillette, R. H. 164 (197)

GlLLIBRAND, M. I. 4, 6 (14)

Gillies, A. 116 (126)

Gilman, H. (39)

Gilpin, V. 95, 112 (126)

Girard, C. 253 (285)

Girsewald, C. 225, 226 (242)

Giua, M. 52 (75), 256 (286)

Glassman, J. 310 (320)

Glen, K. 168 (198)

Glowiak, B. 195 (200), 202, 205, 206

(241)
Gnehm, R. 73, 74 (76)
Gobel, G. 369 (394)
Goldschmidt 266 (286)
Gomberg, M. 222 (242)
Gomm, A. S. 165, 169, 172, 186 (197), 215

(242)
Goodyear, E. S. 664 (687)
Gordon, A. S. 307 (320)
Gordon, S. 280 (287), 382 (394)
Gordon, W. E. (497)
Gorst, A. G. (131), 146 (159), 234 (243, 285,

286), 324 (363), 559, 571 (596, 686)
Gowan, J. E. (200)
Graham, J. A. (126)
Grant, R. L. 212, 233 (241), 446 (497)
Gray, P. 164, 165, 167, 169, 170, 183, 186,

187, 188, 191, 196 (197, 198, 199)
Greek, B. F. 389 (394)
Greenspan, F. P. 300, 302, 304, 305

(320)
Greenspan, J. 15 (38)
Griess, P. 121 (126), 202, 214 (241)
Griffith, R. L. 224 (242)
Grigorovich, P. 139 (159)
Grimshaw, H. C. 420, 446 (496, 497)
Grodzinski, J. 280 (287), 381, 382 (394)
Grollier-Baron, R. 293, 294 (319)
Groocock, J. M. 171 (198) »

Grosse, H. 681 (687)

Grottanelli (126)
Grundman, C. 190 (200)
Gunther, P. 289 (319)
Gunther, P. L. 171, 190 (198, 199)
Guttmann, O. (364, 527)

Hackel, J. 265, 260, 262 (263, 286)

Hackspill, L. 421 (496)

Hagui, J. 256 (286)

Haid, A. 35 (39), 53 (75), 83 (125), 147 (160),

172, 195 (199), 492, 494 (497)
Hailes, H. R. 142 (159), 215, 216 (242)
Haissinsky, M. 171 (198)
Hale, G. C. 19, 20 (38), 77, 87 (125)
Hall, R. H. 26, 27 (39), 217 (242)
Hancock, J. C. 416, 490 (496, 497)
Hanna, N. E. (497)
Hannum, J. A. 299 (320)
Hansen, N. L. 622 (687)
Hantke 278 (287)
Hantzsch, A. 4, 6 (14), 34 (38, 39), 201, 222

(240, 242)
Harris, S. R. 21 (38), 212 (241)
Harrison, P. L. 173 (199)
Harrow, G. 121 (126)
Hartig 529 (567)
Hartmann, I. 217 (242)
Hartwell, F. J. 545 (568)
Haslam, R. 490 (497)
Hatton, W. G. 212 (241)
Hawkes, A. S. 173 (199)
Haycock, E. W. 142 (159), 184 (199)
Heidenreich, K. 192 (200)
Hein, F. 232 (243)
Hellhoff 290 (319)
Helmholz 522
Hendricks, L. B. 161 (197)
Henning, G. F. 77 (125)
Henri, V. 163 (197)
Henry, R. A. (39)
Herring, K. G. (14), 36 (39)
Herron, J. T. 167 (198)
Hersh, C. K. 309 (320)
Herz, E. 37, 38 (39), 70 (76), 201, 214 (240,

241)
Herz, G. C. V. 77, 80 (125)
Herzberg, G. 162 (197)
Hess 141 (159)
Hilgert, H. 305 (320)
Hino, K. 423, 435, 436 (496, 497)
Hinshelwood, C. N. 53 (75), 224 (242)

694

AUTHOR INDEX

Hirst, E. L. 17 (38), 88, 91 (125)

Hiscock 409 (496)

Hoare, D. E. 303 (320)

Hoare, F. R. I. 231 (243)

Hock, H. (241)

Hodgkinson, W. R. 231 (243)

Hodgson, H. H. 42 (74), 201 (241)

Hoerger, E. (569)

Hoffman, C. W. W. (197)

Hoffmann, K. A. 206, 207 (241), 337 (364)

Hofman, W. 327 (363)

Hofmann, A. W. 19 (38)

Holden, I. G. 41, 42 (74)

Holleman, A. F. 138 (159)

Holznagel, W. 191 (200)

Horton, W. J. (38, 126)

Horwitz, J. P. 207 (241)

Howard 129 (131)

Hrynakowski, K. 256 (286)

Huber, H. (126)

Hudson, R. L. 168 (198)

Huffington, J. D. 538 (568), 647 (687)

Huggett, C. 375 (394)

Hughes, A. M. 305 (320)

Hughes, E. D. 5 (14)

"van Hullen 581

Hultgren 78 (125)

Ilosvay, L. 228 (243)
Ingold, C. K. 5 (14)
Isham, H. 166, 185 (197)
Issoire, J. 44, 45 (75)
Izzo, A. 86 (125)

Jacobs, P. W. M. 163, 183, 187, 188, 189

(197, 199)
Jacquet, R A. 314, 315 (321)
Jaffe, I. 393 (394)
Jahn, A. 169 (198)
Jahnsen, A. 230 (243)
Jannasch, P. 169 (198)
Jarry, R. L. 313 (321)
Jeffreys, R. A. (200)
Jenkins, H. P. 295 (319)
Jenner, E. L. (38, 126)
Jolles, Z. E. 555, 558, 559 (569)
Jones 306, 307 (320)
Jones, E. 54 (75)
Jones, J. K. N. (125)
Jones, R. N. 2, 3 (13), 24 (39), 222 (242)

Jones, W. 136 (159)
Jones, W. H. 82 (125)
Jost, W. (688)
Jousselin, L. 22, 25 (38)
Jovinet, P. L. 256 (286), 564 (569)
Joyner, A. R. (197)
Juettner, B. 326 (363)
Junk, A. 558 (569)

Kahane, E. (321)

Kahovec, L. 164 (197)

Kaiinowski, P. 74 (76)

Kapuscinski, Z. 25, 30, 31, 34 (39), 256 (286)

Karnojitzki, V. (243)

Karolyi, M. 336 (363)

Karpukhin, P. P. 93 (126)

Kast, H. 49, 53, 54 (75), 85 (125, 131), 136,
137, 138, 147, 148, 151, 152 (159, 160),
172, 195 (199, 241), 248, 264, 266, 267,
277, 282 (285, 286), 289 (319), 340
(364), 456, 492, 494 (497), 540 (568)

Katchalsky, A. 380 (394)

Kaufman, J. V. R. 146 (159), 171, 176 (198),
203, 217 (241)

Kedesdy, E. (495)

Keefe, J. H., Jr. 300 (320)

Keiser, E. H. 228 (243)

Kenney, J. F. (241)

Kerkovius, B. 326 (363)

Kerone, E. B. W. (687)

Khaibashev, O. K. 52 (75), 256 (286)

Kielczewski 217 (242)

Kilpatrick, M. 295 (319)

Kirchhoff, C. 224 (242)

Kirk, R. E. (241)

Kirkwood, M. W. 26 (39)

Kirmreuther, H. (241)

Kirpal, A. 10 (14)

Kirsch, M. 97 (126)

Kirst, W. 484 (497)

Klein, R. 538 (568)

Klemenc, A. 201 (241)

Klobbie, E. A. 18 (38)

Knaggs, I. E. 162 (197)

Knight, H. C. 535 (568)

Knoff, H. (198)

Koch, A. W. 253 (285)

Kochmyerzhevskh, W. 323 (363)

Koehler, A. 54 (75)

Koenen, H. (287)

Koffler 105, 111

AUTHOR INDEX

695

Kohlrausch, K. W. F. 164, 165 (197)
Kolodziejczyk, S. 257 (286)
Komar 526

Kondratyev, V. N. 547 (568, 688)
Korczynski, A. 192 (200)
Kotowski, A. 228 (243)
Kovache, A. 121 (126), 189, 190 (199)
Kowalczyk, M. 224 (242)
KRAUSE, E. 276 (287)
Krawczyk, W. 83 (125)
Kremann, R. 256 (286)
Kreyenbuhl, A. 421 (496)
Krieger, A. (14)
Krotinger, N. J. 32 (39)
Krupko, W. 170, 171, 173, 176, 185 (198)
Kruska, E. 310 (320)
Kruszynska, K. 213, 215, 216, 220 (241)
Kuhn, W. 163 (197)
Kumler, W. D. 4, 26 (39)
Kunkel 129 (131)
Kurita, M. 256 (286)
Kusler, D. J. 176 (199), 543 (568)
Kuspert, F. 228 (243)
Kustria, B. D. 193 (200)
Kwiatkowski, B. 79 (125), 251, 256 (285),
566 (569)

Lachman, A. 15, 34 (38, 39)

Laffitte, P. 85 (125), 145 (159), 251, 259, 260

(285)
Lamberton, A. H. 4, 6, 7, 11, 12 (14), 17

(38), 88, 114 (125, 126)
Landel, R. T. 368 (394)
Landerl, J. D. (569)
Landon, M. 560 (569)
Landsteiner, K. 5 (14)
Lang, F. M. 44, 51 (75, 287)
Langevin, A. 543 (568)
Langhans, A. 140, 141, 144 (159), 331 (363)
Langseth, A. 162 (197)
Latham, A. 316 (321)
Lauffenburgier 421 (496)
Leavitt, J. J. 2 (14), 223 (242)
Lebedev, Yu. A. 85 (125)
Lebrun, F. (496)
Lecomte, J. 165 (197)
Lecorche, H. 154 (160), 236, 237 (243, 257),

564 (569, 686)
Lee, H. B. 482 (497)
Lefevre, R. 226 (242)
Legler, L. 225 (242)

Lehman, H. A. 191 (200)

Lehmstedt, K. (14)

Lehman, Ya. I. 248 (285)

Lemaire, E. 398, 407, 429 (495, 496)

Lenk von Wolfsburg 528 (567)

van Lennep, B. C. R. 64, 65 (76)

Lenze, F. 53 (75), 77 (125), 130, 202 (241)

Lepin, L. K. 171 (198)

Leulier, A. 226 (242)

Levering, D. R. 3 (14, 197)

Levkovich, N. A. (160)

Levy, R. 291 (319)

Lewis, B. 295, 306, 307 (320, 394), 538 (568),

684 (688)
Ley, W. 302 (320), 674 (687)
Lbber, E. 3, 4 (14), 164 (197), 207 (241)
Lieber, F. 28 (39)
Liebig, J. 135 (158)
Lilly, C. H. 334 (363)
Linck, J. 336 (363)
Linde, C. 288 (319), 491 (497)
Lindley, C. (14, 38, 126)
Linhard, M. 231 (243)
Linstead, R. P. 12 (14), 88
Liu, I. D. 303 (320)
Llewellyn, F. J. 2 (13), 162 (197)
Lcmonosov, M. V. 335
Loriette, P. 22, 33 (38), 561, 562 (569)
Loughran, E. D. (286)
Lukin, A. Ya. 53 (75)
Lundholm, C. 573 (686)

Laszczynski 278 (287), 475
Luczaj, J. (687)

Macar 264 (286)

Macdonald, J. Y. 224 (242)

Mackenzie, J. C. (14)

MacMullen, C. W. 73 (76)

MacNab, W. 407, 489 (496)

MacNaughton, N. W. (38, 126)

Mador, J. L. 167 (198)

Maggs, J. 171, 182, 187 (198, 199)

Mai, J. 192 (200)

Majrich, A. 79, 83 (125), 139 (159), 170 (198>

Makosky, R. C. 33 (39)

Mak6wka, O. 228 (243)

Malendowicz, W. 84 (125)

Malesky, J. S. 508 (526)

Malkin, T. 41, 42 (74)

Mallard 396 (495)

696

AUTHOR INDEX

Malmberg, E. W. 2 (14, 569)

Manueli, C. 254 (286)

Marchenko, L. N. (526)

Marcus Graecus 322

Marcus, R. A. 116 (126)

Margolis, M. A. 319 (321)

Marke, D. J. B. 187 (199), 334 (363)

Marlies, C. A. 15 (38)

Marqueyrol, M. 22, 33 (38), 561, 562, 564

(569), 645 (687)
Marsch, H. E., Jr. (394)
Marsden, E. 201 (241)
Marshall, A. 138 (159)
Marshall, J. 54 (75)

Martin, F. 55 (76), 158 (160), 185, 186 (199)
Mason, L. M. (497)
Massey, F. J. (497)
Matter, O. 201 (240)
Matyushko, N. 49 (75), 139 (159)
Maxim, H. 248 (285)
Maxwell, C. E. (38, 126)
Maxwell, W. R. (395)
Mayer, F. 121 (126)
Mayer, O. V. 230 (243)
Mayes, H. A. 227 (243)
Mayor, L. 302 (320)
Mazur, J. 221 (242)
McAuslan, J. H. 172, 183 (198, 199), 217

(242)
McCaleb, J. D. 32 (39)
McCoy 484 (497)
McCrone, W. C. 117 (126)
McCullogh, F., Jr. 295 (319)
McKay, A. F. 24, 27, 28 (39), 92 (126), 211,

212 (241)
McKenzie, R. G. (320)
McLaren, A. C. 164, 170 (197, 198)
McLarren 308 (320)
Medard, L. 67, 70, 71 (76), 84, 122, 123, 124

(125, 127), 255, 256 (286), 290, 314 (319,

321), 505 (526), 541 (568)
Meen, R. H. 20 (38), 92 (126)
Meerkamper, B. 171, 182 (198)
Meissner, J. 179, 180, 181 (199), 219, 220 (242)
Mendeleyev, D. I. 166 (197), 530 (568, 686)
Menster, M. 538 (568)
Ea Mer, V. K. 15 (38)
Merchant, M. E. 314 (321)
Mertens, K. H. 40, 42 (74)
Mesrobian, R. B. (243)
Metcalf, W. V. 6 (14)
Mettengang 398 (496)

Metz, L. 227 (243)

Meyer, H. 326 (363)

Meyer, K. F. 36 (39)

Meyer, R. 166 (197)

Meyers, J. L. (126)

Michler, W. 40, 47 (74, 75)

Mieszkis, K. 37 (39)

Miles, F. D. 136, 138, 139 (159), 169, 173,

186 (198)
Millard, B. 94, 96 (126)
Millon 331 (363)

Miladowski, W. 251 (285), 566 (569)
Mishuck, E. 365, 366, 381, 391 (395)
Mitchell, D. 54 (75)
Moe, G. 381, 391 (393)
Mohler, H. 163 (197)
Monni 549

Montagu-Pollock, H. M. 211 (241)
Mood, A. M. 445 (497)
Moran, E. C. 70 (76)
Moray, R. 324 (363)
Morgan, J. D. 228 (243)
Morris, F. 543 (568)
Morris, G. 217 (242)
Moutet, A. 294 (319)
Muraour, H. 30 (39), 195 (200, 286), 561 (569)
Murata, T. 415, 422, 425, 428, 429 (496)
Mundy, W. J. 389 (394)
Munroe, C. E. (243)
Murbach, W. J. (394)
Murdock, H. D. (76)
Myers, G. S. 11 (14), 88, 92, 120 (126)

Nagy, J. 217 (242)

Namyslowski, S. (200)

Naoum, Ph. 36 (39), 282 (287)

Napoly, C. 392 (394)

Nash, H. 542 (568)

Naukhoff, S. 281 (287)

Nef, J. U. 135, 136 (158)

Newman, S. H. (200)

Newton, I. 335

Nicolardot, P. 138 (159)

Nielsen, I. R. 162 (197)

Nikolskh 559

Nishizawa, E. (39)

Nisser 280 (287)

Nobel, A. 130 (131), 280 (287), 395 (495), 530,

559 (568, 569)
Noble, Andrew 331, 336, 337, 340 (363, 364),

535 (568)

AUTHOk INDEX

697

Noonan, E. C. 172 (199), 217 (242, 394)
Norrbin, J. H. 259 (286), 395 (495)
Nqrrish, R. G. W. 167 (198), 547 (568, 569)
Nussbaumer 576
Nutt, C. W. 94, 96 (126)

Ogg, B. S. 307 (320)

Ohlsson, C. J. 259 (286), 395 (495)

Ohse, E. 194 (200)

Oldenberg, O. 547 (568)

Olin, J. M. 169 (198)

Olive, T. R. 632, 637 (687)

Olsen, F. 632, 633 (687)

Olszewski, K. 288 (319)

Orth, H. 154, 499, 500, 502, 503

Ortigues, M. 84 (125)

Orton, K. J. P. 8 (14), 43, 44 (75)

Ostrowski, R. 327 (363)

Othmer, D. F. (241)

Ott, E. 194 (200)

Owston, P. G. (14, 126)

Oza, T. 338 (364)

Pack, D. A. 290 (319)

LE Paire 546 (568)

Pannetier, G. 167 (198)

Paprotskii (363)

Parisot, A. 85 (125), 251, 259, 260 (285, 286)

Parry, E. 139 (159), 171 (198)

Partington, J. R. 323 (363)

Partyka, K. 618 (687)

Pascal, P. 256 (286), 289 (319), 352 (364, 686)

Passburg 618 (687)

Patart, G. 30 (39)

Patat, F. (197)

Patinkin, S. H. (39), 207 (241)

Patry, M. 138, 146, 147, 148 (159)

Patterson, L. J. 3 (14, 197)

Pattison, S. 47 (75)

Pauling, L. 161 (197)

Paushkin, Ya. M. 301, 304, 305 (320)

de Pauw, P. F. M. 554 (569)

Payman, W. 398, 409, 410, 411, 412, 415 (495,

496), 523 (527)
Peacock, J. 16 (38)
Pearson, J. (287, 497)
Pease, R. N. 295 (320, 394), 547 (568)
Pelizzari, G. 25 (39)
Pelouze, J. H. 528 (567)
Penner, S. S. 295 (320, 687)

Penny, E. 227 (243)

Perrott, G. S. J. 412 (496)

Philippi, E. 326 (363)

Philips, R. 28 (39)

Phillips, L. 142, 143 (159), 203, 204 (241)

Picard, J. R. 20, 25 (38, 39)

Pietrzyk, C. 221, 223 (242)

Pietsch, E. 228 (243)

Pike, H. H. M. 653, 671 (687)

Pillich, J. 83 (125), 284 (287)

Pimentel, G. C. 164, 167 (197, 198)

Piobert, G. 335 (363)

Piskorz, M. 221, 223 (242)

Platz, G. M. 309 (320)

Plucinski, J. 202, 203, 204 (241)

Podbelskh, G. N. 526 (527)

Podgornova, A. A. (497)

Ponchon 579, 581 (686)

Ponzio, G. 205, 206 (241)

Poppenberg, O. 534 (568)

Porger, J. (199)

Porter, G. (198), 547 (568)

Pravdin 161 (196)

Prentiss, A. M. 49 (75)

Prettre, M. 428 (496), 546 (568)

Price, D. 393 (394)

Pring, J. N. 670 (687)

Prttchard, E. J. 23 (39)

Prothero, J. B. (320)

Prout, E. G. 142 (159), 224 (242)

Prout, F. S. 223 (242)

PuRGorn, A. 192 (200)

Quellet, C. (320)
Quilico, A. 206 (241)

Rabek-Gawronska, I. 80 (125), 256 (286)

Raciborski, M. 307 (320)

Ramachandra Rao, C. N. (197)

Raschig, F. 329 (363)

Rathsburg, H. 141 (159), 192 (200), 207 (241)

Ratouis 616 (686)

RXtz, R. 190 (200)

Razorenov, B. 8 (14)

Recchi, V. 22 (39, 569), 664 (687)

Reckleben, H. 228 (243)

Reese, E. F. (243)

Regad, E. D. 48 (75)

Reichardt, H. 547 (568)

Reichle, W. T. 196 (200)

Reid, W. F. 529 (567)

698

AUTHOR INDEX

Reiset 331 (363)

Reitz, A. W. 164 (197)

Rice, F. O. 167 (198)

Rice, G. S. (495)

Riche, A. 227 (242)

Richter (688)

Rideal, E. K. 538 (568)

Rie, G. (363)

Rimarski, W. 226 (243)

Rinehart, J. S. (287, 497)

Rinkenbach, W. H. 48, 49, 50, 52, 58 (75),

118 (126), 184, 195 (199, 200), 208,

215, 226 (241, 242), 256 (286), 341

(364), 484 (497)
Rintoul, W. 643 (687)
Rissom, J. 166, 185 (197)
Ritchie, M. 320
Roberts, E. E. (126)
Robertson, A. J. B. 20 (38), 53 (75), 84, 118

(125), 538 (568)
Robertson, R. 54 (75), 148 (160), 643 (687)
Rodman, T. J. 360 (364)
Rogers, G. F. 164 (197)
Rogers, G. T. 173, 182, 183 (199)
Rogers, R. N. 285 (287)
Rogers, T. A. 402 (495)
Roginskh, S. Z. 53 (75), 381
Rollet, A. P. 421 (496)
van Romburgh, P. 40, 42, 45, 46, 47, 52,

63, 65, 67, 68, 70 (74, 75, 76)
Romocki, S. J. 129 (131), 322 (363, 567)
Rosbaud, P. (199)
le Rosen, A. L. 2 (14)
Rosenberg 157 (160)
Rosenquist, E. N. 210 (241)
Rosinski, M. 552, 556 (569)
Ross, J. H. 109 (126)
Rossi, B. D. 424, 425 (496)
Roszkowski, J. 544 (568)
Roth, G. 266, 269 (286)
Roth, I. F. 134 (158)
Roth, R. 206 (241)
Roth, W. A. 305 (320)
Rottweil 529

Le Roux, A. 254 (286), 505 (526)
Ruhl, G. 393 (394)
Rupe, H. 228 (243)
Rusiecki, A. 191 (200)

Sabaneyev, A. 169 (198), 228 (243) .
SAH, P. P. T. 26 (39)

Sahli, M. 178 (199)

Salathe, F. (74)

Salcewicz, J. (569)

Salvadori, R. 230 (243)

Salyamon, G. S. 3 (14)

Sandison, R. 224 (242)

Sapozhnikov, A. 553

Sartorius, R. 273 (287), 290, 314 (319, 321),
421 (496)

Satterfield, C. N. 305 (320)

Saunders, W. H. 196 (200)

Sawkill, J. 182 (199)

Sayers, J. 573 (686)

Scharff, M. 121 (126)

Scheiber, I. 228 (243)

Schenkel, G. (687)

Schepers 46 (75)

Schiessler, R. W. 109 (126)

Schmidt, A. 53 (75)

Schmidt, F. 191 (200)

Schnurr 87, 95, 104

Scholl, R. 8 (14), 140 (159)

Schonbein, C. F. 528 (567)

Schowengardt, J. W. 675, 681 (687)

Schrader, F. 192 (200)

Schroeder, W. A. 2 (14), 563 (569)

Schuck, M. 53, 63 (75)

Schultz, G. (76)

Schultz, R. D. 381, 391 (395)

Schultze, E. 528 (567)

Schumacher, H. J. 166 (197)

Schumacher, J. C. (321)

Schumb, W. C. 305 (320)

Schwaebel, R. 168 (198)

Schwanke, C. 456 (497)

Schwartz (363)

Schweitzer, C. E. 18 (38)

Schwinn, E. 231 (243)

Schwob, R. 299 (320)

Scott, J. 306, 307, 310 (320)

Scott, L. B. (126)

Seel, F. 168 (198)

Segay 411 (496)

Seidler, P. (363)

Seleznev 432

Sell, E. 567 (569)

Selle, H. 54 (75)

Semenczuk, A. 42, 43, 72 (74, 75, 76)

Semishtn, V. I. 305 (320)

Serebryakov, M. E. (568), 673 (687)

Shah, S. 338 (363)

Shakhnovskaya 161 (196, 197)

AUTHOR INDEX

699

Shanley, E. S. 300, 302, 304, 305 (320)

Sheehan, J. C. 114 (126)

Shefield, O. E. (243)

Sheinker, Yu. N. 64, 163 (197)

Shekhter, B. I. (160)

Shepherd, W. C. F. 420 (496)

Sherman, E. (39)

Shishkov, L. 138 (159), 336 (363)

Shorygin, P. P. 43 (75)

Shpitalskii, E. (276, 287)

Siegens, S. 225, 226 (242)

Siersch 398

Sillitto, G. P. 383 (394), 421 (496)

Simecek, J. 81, 82, 122, 123 (125, 127)

Simmons, W. H. (126)

Simons, J. H. (320)

Singh, K. 89 (126), 139, 143, 144 (159), 172,

187 (198, 199)
Sinyaryev, G. B. 294, 298, 308, 309, 311 (319)
Sirotinskii, V. F. (160)
Skrzynecki, J. 24 (39), 254, 257 (286)
Smart, G. N. R. 11 (14)
Smith, L. C. (286)
Smith, P. A. S. (200)
Smith, T. L. 368 (394)
Smolenski, D. 46 (75), 202, 203, 204 (241)
Snelling, W. O. 233, 341 (364)
Socci, M. 555, 559 (569)
Sofianopoulos, A. J. 166 (197) '
Sokolnik, A. S. (688)
Sokolov, A. M. 56, 57, 59 (76)
Solonina, A. 50, 135, 137, 138, 139 (159),

170 (198)
Somlo, F. 82 (125)
Sorensen, J. U. 162 (197)
Sorm, F. 192 (200)
Soubeiran (243)
Spaeth, S. P. 227 (242)
Speakman, J. C. (14, 38, 126)
Spill, D. 529 (567)
Sprengel 278 (287), 288 (319)
Springall, H. D. (125)
Sramek 123 (127)
Starke, E. A. 335 (363)
Stassart 398
Stedman, G. 168 (198)
Stein, T. W. (320)
Steinberger, R. 674, 675, 678 (687)
Steiner, A. 140 (159)
Steiner, K. 326 (363)
Steinkopf, W. 224 (242)
Stephan 534 (568)

Stern, R. 161 (196)

Stettbacher, A. 148 (160), 214 (241) 282

(287), 311 (320)
Stevens, T. S. 7 (14)
Stewart, K. 167 (197)
Storch, L. (14)
Storm, C. 553 (569)
Strasser, E. 222 (242)
Straumanis, M. 185 (199)
Strecker, W. 231 (243)
Street (287)
Sudlitz, C. 607 (686)
Sudo, H. 172, 174, 176 (199)
Suggitt, J. W. 11 (14)
Sukhinskh, A. V. 617 (686)

SUNDERMEYER, W. 191 (200)

Sutton, G. P. 675, 676 (687)
Svetlov, L. G. 581 (686)
Swietoslawski, W. 552, 556 (569)
Syrkin, Ya. K. 163, 164 (197)
Szyc-Lewanska, K. 53, 74 (75, 76), 106 (126),
203, 225 (241, 242)

Tabouis, F. 84 (125)

Taffanel, J. 398, 425 (495, 496)

Tait, C. W. 296, 297 (320)

Tamburrini, V. 256 (286)

Tanatar, S. 169 (198)

Tate, F. G. G. 324 (363)

Taylor, C. 184, 195 (199, 200), 226 (242)

Taylor, C. A. 49, 50, 52 (75), 256 (286),

484 (497)
Taylor, G. B. 233
Taylor, G. W. C. 174, 176 (199)
Taylor, H. S. 295, 305 (320, 394)
Taylor, J. (197), 217 (242), 324, 334 (363),
383 (394), 396, 397, 401, 413, 414, 415,
416, 419, 421, 422, 432, 433, 463, 464,
465, 466, 467, 468, 490 (495, 496, 497),
522, 524, 525 (527), 539 (568), 675 (687)
Taylor, W. 55 (76)
Taylor, W. G. 212 (241)
Teodorowicz, K. 135 (158)
Terpstra, P. 78 (125)
Tesiorowski, E. 326 (363)
Thelen, R. (363)
van Thiel, M. 167 (198)
Thiele, J. 15, 25, 34(38, 39), 161, 169 (197, 198)
Thomas, A. T. 174, 176, 186 (199)
Thomas, H. 331, 340 (363)
Thomas, J. G. N. 171 (198)

700

AUTHOR INDEX

Thomas, M. 122 (127)

Thorn, G. D. 2, 3 (13), 24 (39), 222 (242)

Thrush, B. A. 167 (198)

Tibbits, G. C. (687)

Tiffany, J. E. 212, 233 (241, 243)

Tikhomirova, A. (75), 256 (286)

Titman, H. 414 (496)

Tobolsky, A. V. (243)

Todd, G. 139 (159), 171 (198)

Toggweiler, U. (200)

Tompkins, F. C. 142, 143, 145, 146 (159),
163, 164, 171, 183, 184, 187, 188, 189
(197, 198, 199), 216, 220, 224 (242)

Tonegutti, M. 83, 84 (125), 559 (569)

Toombs, L. E. (14), 36 (39)

Topchiyev, A. V. 43 (75)

Tormey, J. F. 317 (321)

Towne, E. B. (14)

Townend, D. T. A. 683 (688)

Traube, W. 2, 17 (38), 221, 222, 223 (242)

Troup, H. B. 56 (76)

Trueblood, K. N. (569)

Trush, B. A. (568)

Tschinkel, J. G. 297, 299 (320)

Turek, O. 193 (200)

Turner, J. 42 (74)

Turpin, E. 288 (319)

TURRENTINE, J. W. 169 (198)

Ubbelohde, A. R. 261 (286)

Ulrich (76)

Urbanski, J. 73 (76)

Urbanski, T. 18, 24, 25, 30, 31, 34, 35, 37
(38, 39), 42, 43, 53, 63, 66, 68, 73, 74
(74, 75, 76), 79, 80, 83, 84, 85, 98, 106
(125, 127), 191 (200), 203,213, 215, 216,
220, 221, 223, 224, 225 (241, 242), 249,
250, 251, 254, 256, 257, 258, 259, 264,
265, 278, 283, 284 (285, 286, 287), 326,
327 (363), 409, 411, 434, 435, 436, 475
(496, 497), 540, 552, 556, 566 (568, 569)

Urizar, M. J. (286)

Valentinus, Basilius 129 (131)
Vari, P. 256 (286)
Vasilevskii, V. V. 193 (200)
Vaughan, J. 142, 143 (159), 203, 204 (241)
Venkataraman, A. 8 (14)
Vennin, L. 154 (160), 236, 237 (243, 527),
609 (686, 687)

Verleger, H. (197)

Vernazza, E. 81 (125)

Verneuil 229 (243)

Vervoorst, P. 230, 231 (243)

Vieille, P. 22, 30 (39), 138, 148 (159), 190
(200), 201,229 (240, 243), 340 (364),
529, 530, 538, 541, 549, 550, 554, 557,
558 (568, 569), 623, 664 (687)

Villiger, V. 225 (242)

Violette 325 (363)

Voigt, A. 330 (363)

Volkmann, F. 529 (567)

de Vries, J. H. 26 (39)

Vroom, A. H. 88, 91, 94, 95 (126)

Vyshnegradskh, I. A. 359 (364)

Waddington, T. C. 164, 165, 167, 170, 183,
186, 187, 188 (197, 198, 199)

Wagner, J. 164, 165 (197)

Walden, P. 305 (320)

Walden, R. J. 171 (198)

Walker, M. 134 (158)

Wallbaum, R. 141 (159), 172, 177 (198), 208,
215 (241)

Walsh, A. D. (320)

Walton, D. C. 535 (568)

Ware, J. C, Jr. 196 (200)

Warren, F. A. 309, 318 (320), 373, 374, 375,
387 (393, 688)

Waterlot, L. (496)

Watt, G. W. 33 (39)

Watteyne, V. 398, 407 (496)

Weisser, H. R. 162 (197)

Wenner, R. R. 305 (320)

Wentworth, R. L. 305 (320)

Wesolowski, T. 223 (242)

Wessels, G. 293, 294 (319)

Westwater, R. (497)

Wetterholm, A. 437, 482 (497)

Wheeler, R. W. 416 (496)

Wheeler, T. S. (200)

Wheeler, W. H. 653, 671 (687)

White, A. G. 683, 684 (688)

Whttmore, F. E. 2 (13), 162 (197)

Whitmore, W. C. 120, 121 (126)

Whittaker, A. E. 296, 297 (320)

Whittaker, H. 653, 671 (687)

Whittle, E. 167 (198)

Wiberg, E. 190 (200)

Wiegner, G. 34 (39)

Wieland, H. 133, 134 (158, 159)

AUTHOR INDEX

701

Wilde, D. G. 414 (496)

Wildgoose, A. B. 468 (497)

Wilkoszewski 398 (495)

Will, W. 130, 340 (364), 398, 399 (495)

Williams 296, 297 (320)

Williams, H. L. 116 (126)

Williams, H. T. 172, 184 (199)

Williams, J. F. 28 (39)

Williams, M. C. 167 (198)

Williamson, W. O. 252 (285)

Wilson, G. L. 15 (38)

Winkler, C. A. 82, 88, 91, 94, 95, 97, 109,

112, 116, 117, 118, 119 (125, 126),

173 (199)
Wischin, A. 182 (199)
Wislicenus, W. 168 (198)
Witanowski, M. 327 (363)
Witkowski 56 (76)
Wogrinz, A. 256 (286)
Wohler, L. 134, 135, 137, 158 (158, 159, 160),

170, 171, 173, 176, 185 (198), 201 (240)
Wojciechowski, W. 25, 31, 34 (39)
Wolf, W. 529 (567)
Wolff, L. 201 (241)
Wolfram 107

Wood, W. S. 296, 302, 316 (320)
Woodbury, C. A. 484 (497)
Woodcock, D. 7 (14, 126)
Woodhead, D. W. 446 (496, 497)
Woringer, B. 211 (241)
Wright 129 (131)

Wright, G. F. 9, 10, 11 (14), 23, 25, 26, 27,
36 (39), 81, 88, 90, 91, 92, 108, 109,
114, 116, 117, 120, 123 (125, 126)

Wrobel, K. (687)

Wroblewski, S. 288 (319)

Wyler, O. 256 (286)

Yamamoto, S. 469, 473 (497)
Yaroslavskh, N. G. 3 (14)
Yefremov, N. N. 52 (75), 256 (286)
Yegorov, T. S. 588, 592, 610, 611 (686)
Yoffe, A. D. (159), 163, 164, 165, 169, 183,

186, 188, 191, 196 (197, 198, 199, 242)
Yokogawa, M. 424, 435, 436, 473 (496, 497)
Yonck 280 (287)
Young, D. A. 143 (159), 164, 184, 189 (197,

199), 216, 220 (242)
Young, G. H. S. (393)
Yuill, A. M. 139, 145 (159), 170, 171, 172,

176 (198, 199), 216 (242)

Zabudskh 538

Zacharewicz, J. (38), 221, 223 (242)
Zaehringer, A. J. 368, 393 (393), 629 (687)
Zakharov, I. N. (686)
Zeldovich, Ya. B. 537 (568)
Zimmerman, G. A. 387, 388 (394)
Zingaro, R. A. 214, 215, 216 (241)
Zumstein, O. (14)

SUBJECT INDEX

Absorption towers for nitric oxides, 101
l-Aceto-3,5-dinitro-l,3,5-triazacyclohexane, 116
l-Aceto-3,5,7-trinitro-l,3,5,7-tetrazacyclo-

octane, 116
l-Acetoxymethyl-3,5-dinitro-l,3,5-triazacyclo-

hexane, 114
l-Acetoxy-7-nitroxy-2,4,6-trinitro-2,4,6-triaza-

heptane, 91
Acetylene, 227
Acetylides, 227
Aci-form,
of nitramine, 16
primary nitramines, 4
Activation energy of thermal decomposition,
of cyclonite, 83
ethylenedinitramine, 20
lead azide," 172
lead styphnate, 216
lead styphnate dehydration, 216
metal azides, 187
octogen, 118
silver azide, 184
tetryl, 53

thallium styphnate, 220
thallous azide, 188
triphenylmethyl azide, 196
Active ingredients in mining explosives, 423
Airdox charges, 526

Aliphatic nitramines— see Nitramines, aliphatic
Amatol 40/60, 261
Amatols, 261
Amines,
in rocket propellants, 294
mechanism of nitration, 11
nitration of, by nitrolysis, 12
secondary, nitration of, 9
^-Aminoanthraquinone, nitration of, 8
Aminoguanidine derivatives, 206
Aminomethylnitramines, formation of, 7

N-Amino-N'-nitroguanidine, 28
a-Aminopyridine, nitration of, 8
Aminotetrazole, 207, 210
Aminothiazole, nitration of, 8, 9
Ammelide, 29, 34
Ammeline, 29, 34
Ammonals, 268, 269

composition of, 269

German, 270

USA, 270
Ammonites, 265, 405
Ammonium azide, 190
Ammonium chloride in mining explosives, 428
Ammonium nitrate, mixtures of, in rocket

propellants, 383, 389
Ammonium nitrate-fuel oil mixtures, 489

manufacture of, 508
Ammonium nitrate mining explosives, 447, 448,
449, 450, 451, 452, 453, 454, 457, 458,
459, 462, 465, 466, 467, 468, 470, 473,
476, 477, 478, 481, 483

manufacture of, 498
Ammonium picrate propellants, 393
Ammonium powder, 331
Aromatic nitramines— see Nitramines, aromatic
Azides,

containing silicon, 196

of "ammines", co-ordination compounds of,

231

organic, 191
Azidoguanidine, 210
5-Azidotetrazole, 208

Bachmann's reagent, 112
Ball-grain powder, 632
Ballistites, 647
attenuated, 650
manufacture of, 647

rioai

SUBJECT INDEX

703

Ballistites (cont.)

progressive, 651
Baratol, 264
Barbaryt, 41 1

Barium nitrocyanamide, 212
Barium styphnate, 220
Benzidine, nitro derivatives of, 46
Benzotrifuroxane, 193
Beryllium compounds in rocket propellants,

311
Bis (acetoxymethyl)-nitramine, 114
Bitetryl — see Hexanitrodiphenylethylenedi-

nitramine
Blackpowder, 323
composition of, 324
erosiveness of, 550
explosive properties of, 340
manufacture of, 342
blending, 359
corning, 354
finishing, 357
milling, 345
mixing, 349
pressing, 352
safety in, 361
theory of burning, 335
types of, 328
ammonium powder, 331
chlorate powder, 334
modified, 330
picrate powder, 334
sulphurless, 331
Blasting caps— see Detonators
Blasting powders, composition of, 329
Bleitrizinat— see Lead styphnate
Bobbinite, 404

Boron compounds in rocket propellants, 311
Bradyt F, 422, 434, 475
Brattice drier, 155
"BSX"-see l,7-Diacetoxy-2,4,6-trinitro-2,4,6-

triazaheptane
Burning of blackpowder, theory of, 335
Burning rate,
of composite propellants for rockets, 392
cyclonite, 86

dinitrobenzenediazo-oxide, 204
mercury fulminate, 147
mixtures of tetranitromethane, 298, 299
tetryl, 55
sec-Butylnitramine, preparation of, 1 1
Butyltetryl-see 2,4,6-Trinitrophenyl-n-butyl-
nitramine

Cadmium azide, 186
Cambrite— see Carbonit
Cannon powder, 360, 540
Carbonit, 404
Cardox charges, 522
Casting, technology of, 676
Cellulose, thermal decomposition of, 327
Centralites, 447
Centrifuges, 574
Chain reactions, 547
Charcoal, 325
chemical structure of, 326
for blackpowder manufacture, 345
Cheddites, 277, 280, 403
Chlorate powder, 334
Chlorates,
in mining explosives, 408, 422, 457, 458,

474
of complex ammines, 230
of hydrazinometals, 231
Chlorine trifluoride, 312

in rocket propellants, 312
Chloroformoxime, 1 3 3
Coal-dust, explosibility of, 420, 421
Coefficient of the vivacity, 530
Collodion cotton, 511
CP 1( 530
CP 2 , 530
Combustibles in mining explosives, 423
Complex ammines, 230
chlorates of, 230
perchlorates of, 230
Complex compounds of fulminic acid, 134
Complex metal ammines, 231
Complex salts,
of cupric azide, 185
nitroguanidine, 31
Composite explosives— see Explosives, com-
posite
Composition Exploding, CE— see Tetryl
Compositions for explosive rivets, 240
Cordite MD, manufacture of, 642
Cordite Mk I, manufacture of, 642
Cordite RDB, manufacture of, 643
Cordites, 540
Corning mills, 345, 355
Cupric acetylide, 228
Cupric azide, 185
basic, 185

complex salts of, 185
Cuprous acetylide, 227
Cyamelide, 134

704

SUBJECT INDEX

Cyanamide salts, 211
Cyanuric acid, 120, 132

salts of, 133
Cyanuric triazide, 194
Gyclonite, 77
manufacture of, 87

British method, 98

E-method, 109

German method, 104

K-method, 105

KA-method, 105

W-method, 107
toxicity of, 86
Cyclotetramethylenetetramine — see Octogen
Cyclotrimethy lenetrinitramine — see Cyclonite
Cyclotrimethylenetrinitrosamine, 80

DBX, 272

DDNP— see Dinitrobenzenediazo-oxide

Decomposition,

of cyclonite, 81, 82
kinetics of, 82
hydrazine, 306
hydrazoic acid, 167
hydrogen peroxide, 301
lead azide, 171
lead styphnate, 215

activation energy of, 216
mercury fulminate, 142, 143

rate of, 141
methylenedinitramine, 17
nitramines, 16
nitroguanidine, 27

primary nitramines, with alkalis, 4, 5, 6
silver azide, 183
kinetics of, 184
tetryl, rate of, 52
Desensitization — see Phlegmatization
Detonation rate,
of acetylene, 227
Amatols, 261
T-Ammonal, 269
cupric azide, 185
cyanuric triazide, 195
cyclonite, 85
Czechoslovakian mining explosives, 450,

451
dinitrodi-(/?-hydroxyethyl)-oxamide dini-

trate, 37
dinitrodimethyldiamide of tartaric dinitrate,
37

dinitrodimethyloxamide, 35

ethylenedinitramine, 20

French mining explosives, 552, 553, 554

Gelatine-Donarit 1, 456

hexamethylenediamine peroxide, 226

hexanitrodiphenylethylenedinitramine, 70

Hungarian mining explosives, 469, 470

Hydrobel, 490

Japanese mining explosives, 472, 473, 474

lead azide, 172

lead styphnate, 218

mercury fulminate, 147, 148

nickel ammino perchlorate, 230

nitroguanidine, 31

oxyliquits, 492, 494

Polish mining explosives, 476, 477, 478,

479
silver azide, 184
tetryl, 54

trimethylenetrinitrosamine, 124
trinitrophenyl-/?-hydroxynitraminoethyl

nitrate, 71
trinitrotriazidobenzene, 194
USSR mining explosives, 485, 486, 487,

488, 489
Wetter-Astralit, 460
Wetter-Carbonit, 460
Detonator TAT-1, 233
Detonators, 218, 232

manufacture of, 236
1 ,5-Diaceto-3,7-dinitro-l ,3,5,7-tetrazacyclo-

octane, 116
1 ,5-Diaceto-3,7-endomethylene-l ,3,5,7-tetraza-

cyclo-octane, 116
l,5-Diacetoxy-2,4-dinitro-2,4-diazapentane,

115
l,9-Diacetoxy-2,4,6,8-tetranitro-2,4,6,8-tetraza-

nonane, 115
l,7-Diacetoxy-2,4,6-trinitro-2,4,6-triazaheptane,

91, 114
Diazide of carbonic acid, 192
Diazo compounds, 201

oxidation of, 2, 13
Diazomethane, reaction with primary nitr-
amines, 7
Diazonium nitrate, formation of, 7
4-Diazo-l -oxide, 205

nitro derivatives of, 205
2-Diazo-l-oxide, 205

nitro derivatives of, 205
Diethanolnitramine dinitrate— see Nitrodietha-
nolamine dinitrate

SUBJECT INDEX

705

l-Di(hydroxymethyl)-arninomethyI-3,5-dinitro-

1,3,5-triazacyclo-hexane, 90
Diluter, 100

"Dimazine"-see 1,1-Dimethylhydrazine
Dimethylaniline,
nitration of, 41, 56
continuous method, 58
Dimethylaniline sulphate, production of, 57
1,1-Dimethylhydrazine, in rocket propellants,
308

2,4-Di-(methylnitraminomethyl)-4-nitrophenol,
72

lead block expansion test of, 73
DINA-see Nitrodiethanolamine dinitrate
2,6-Dinitro-4-amino-m-cresol, 205

diazo compound from, 205
2,6-Dinitro-2-amino-m-hydroxybenzoic acid,
205

diazo compound from, 205
2,4-Dinitro-2-aminoresorcinol, 205

diazo compound from, 205
N,N'-Dinitroammeline, 120
Dinitroazidophenol, 195

plumbous salt of, 196
Dinitrobenzenediazo-oxide, 201
3,7-Dinitro-3,7-diaza-l ,5-dioxacyclo-octane, 90
2,5-Dinitro-2,5-diazahexane, 3
3,5-Dinitro-3,5-diazapiperidinium nitrate, 91
Dinitrodiazophenol -see Dinitrobenzenediazo-
oxide
Dinitrodiethyloxamide, 35

Dinitrodi-(jff-hydroxyethyl)-oxamide dinitrate,

Dinitrodi-G?-hydroxyethyl)-sulphamide dini-
trate, 38

Dinitrodimethyldiamide of tartaric dinitrate, 37
Dinitrodimethyloxamide, 34
Dinitrodimethylsulphamide, 36
2,4-Dinitromethylaniline, nitration of, 44, 61
Dinitrophenylmethylnitramine, rearrangement

of, 5
Dinitrosopentamethylenetetramine, 121, 124
3,5-Dinitroso-l,3,5,7-tetrazabicyclo[3,3,l] no-

nane — see Dinitrosopentamethylene-

tetramine

3,5-Dinitro-l,3,5,7-tetrazabicyclo[3,3,l]nonane,
90

2,4-Dinitro-6[tetrazene-(l-)]-phenylhydrazine
salt, 206

l,9-Dinitroxy-2,4,6,8-tetranitro-2,4,6,8-tetraza-

nonane, 89
Dinol — see Dinitrobenzenediazo-oxide

Diphenylamine, as stabilizing agent in smokeless

powder, 559
Diphenylphenazine, 562
Disk mill, 108, 346
1,3-Ditetrazyltriazine, 210

Ditetryl-see2,4,6-Trinitro-l,3-di(methylnitr-

amino)-benzene or Hexanitrodiphenyl-
ethylenedinitramine
DNPT - see Dinitrosopentamethylenetetramine
Donarit, 263

Double base cast propellants, 675
Double base powder, 641
Double base powder for rockets, 671
DPT-see 3,5-Dinitro-l,3,5,7-tetrazabicyclo

[3,3,l]nonane
Driers, 499, 500, 501, 502, 503, 504, 612, 613, 615
infrared, 638, 639
Mamut, 663
Dynamites, 454, 457, 458, 459, 461, 469, 471
472, 476, 479, 480, 481
manufacture of, 511
Dynobel, 405

Edge runners, 350, 351
EDNA— see Ethylenedinitramine
Elastomers in rocket propellants, 368, 380
1 ,5-Endomethylene-3,7-dinitro-l ,3,5,7-tetraza-
cyclo-octane— see Dinitrosopenta-
methylenetetramine
Energy of activation— see Activation energy
Erosion bomb, 549
Erosiveness,
of blackpowder, 550
blasting gelatine, 550
nitrocellulose powders, 550
nitroglycerine powders, 550
nitroguanidine, 550
smokeless powder, 548
Ethylene-bis-acetamide, 19
Ethylenediamine, reaction with nitroguanidine,

28
Ethylenedinitramine, 18

homologues of, 21
Ethylene oxamide, 19
Ethylene oxide in rocket propellants, 310
Ethyltetryl — see 2,4,6-Trinitrophenylethyl-

nitramine
Eutectic mixtures,
with cyclonite, 80
dinitrodimethyloxamide, 35
nitroguanidine, 24

706

SUBJECT INDEX

Expansion in the lead block test— see Lead block

expansion test
Explosibility of coal-dust, 420
Explosion, heat of— see Heat of explosion
Explosions in gas mixtures, 545
Explosions of lead azide, 174
Explosive complex compounds, 230
Explosive mixtures, incompatibility of, 283
Explosive properties,
of blackpowder, 340
composite propellants, 393
cyclonite with TNT, 250
dinitrobenzenediazo-oxide, 202
dinitrodi-(/?-hydroxyethyl)-sulphamidedini-

trate, 38
ethylenedinitramine, 20
hydrogen peroxide, 304
mercury fulminate, 146
metal azides, 186
nitroguanidine, 29
PETN with TNT, 252
smokeless powder, 532
substances containing nitroguanidine, 30
trimetylenetrinitrosamine, 123
Explosives,
composite, 245
fusible, 247
high, 247
infusible, 258
liquid, 286
semifusible, 258

Favier explosives, 403

Flash, 544

Flashless charges, 663

Flashless powders, 663

Fluorine, in rocket propellants, 312

Fluorine compounds, in rocket propellants,

312
Fluorine oxide, in rocket propellants, 312
Formhydroxamic acid, 158
Fp 60/40 -see Amatol 60/40
Free atoms, reaction of, 316
Free radicals, reaction with hydrogen, specific

impulse of, 318
Fuels, mixtures with, in rocket propellants, 293
Fiillpulver 60/40 -see Amatol 60/40
Fulminate ion, 133

structure of, 133
Fulminic acid, 132

complex compounds of, 134

polymerization of, 134

salts of, 133, 157
Furfuryl alcohol in rocket propellants, 294
Furoxanedialdoxime, 133
Fusible explosives, 247
Fusible mixtures,

manufacture of, 255

phlegmatization of, 257

Galleries for testing -see Testing galleries
Gallery testing of explosives, application of sta-
tistics in, 445
Gelatine-Donarit 1, 655
Gelignites, 461
Guanyl azide, 207

Haleite— see Ethylenedinitramine
Halogen azides, 191
HBX, 272
Heat of combustion,
of cyclonite, 78

1,1-dimethylhydrazine, 308

hydrazine, 305

tetryl, 49

trimethylenetrinitrosamine, 122
Heat of decomposition,
of mercury fulminate, 148

metal azides, 189
Heat of detonation,
of cyclonite, 84, 85

mining explosives, 429, 469, 470, 485, 486,
487, 488, 489

trimethylenetrinitrosamine, 122

Wetter-Astralit, 460

Wetter-Carbonit B, 460
Heat of explosion,
of ethylenedinitramine, 20

Gelatine-Donarit 1, 456

mercury fulminate, 148

oxyliquits, 492

smokeless powders, 536

tetrazene, 208

tetryl, 54
Heat of formation,
of acetylene, 227

ammonium azide, 190

chlorine trifluoride, 312

cyanuric triazide, 195

cyclonite, 78

dinitrobenzenediazo-oxide, 204

fluorine oxide, 312

SUBJECT INDEX

707

Heat of formation {cont.)
of hydrazine, 305
hydrogen peroxide, 300
mercury fulminate, 148
metal azides, 187
nitrogen selenide, 229
nitrogen sulphide, 229
perchloric acid, 313
perchloryl fluoride, 313
silver acetylide, 229
tetryl, 49

trimethylenetrinitrosarnine, 122
Heat of nitration,
of hexamethylenetetramine, 95
hexamethylenetetramine dinitrate, 95
hexamethylenetetramine nitrate, 95
Heat of vaporization of hydrogen peroxide,

300
Heat, specific,
of cyclonite, 78
hydrogen peroxide, 300
mercury fulminate, 139
tetryl, 49
Heterocyclic nitramines— see Nitramines, het-
erocyclic
Hexamethylenediamine peroxide, 225

chemotherapeutic action of, 226
Hexamethylenetetramine,
action of nitric acid on, 87
nitration of, 13
rate of, 94
Hexamine— see Hexamethylenetetramine
Hexanitrodiphenylethylenedinitramine, 69
Hexanitrodiphenyl-j8-hydroxynitraminoethyl

nitrate, 72
N-2,3,4,5,6-Hexanitro-N-methylaniline— see 2,3,
4,5,6-Tetranitrophenyl-N-methylnitr-
amine
Hexanitrosobenzene — see Benzotrif tiroxane
3,5,3',5',3",5"-Hexanitro-4,4',4"-tri(methyl-

nitramino)-triphenylcarbinol, 68
Hexogen— see Cyclonite
High explosives, 247
mixtures,
with aluminium and ferro-silicon, 272
chlorates and perchlorates, 274
metals, 266
mixtures of nitrocompounds, 247
HMTD— see Hexamethylenediamine peroxide
HMX— see Octogen
Homocyclonite, 119
Homohexogen— see Homocyclonite

Hydrazine, 306
decomposition of, 306
in liquid explosives, 295
in rocket propellants, 305
oxidation of, 307

Hydrazoic acid, 161
manufacture of, 168

Hydrobel, 490

Hydrogen peroxide, 300
in liquid explosives, 290
in rocket propellants, 299
thermochemical properties,

Hydrox charges, 523

303

Ignitability of smokeless powder, 541
Ignition temperature,
of charcoal, 325
cyanuric triazide, 195
cyclonite, 83

dinitrodi-(/?-hydroxyethyl)-oxarnide dini-
trate, 37
dinitrodimethyldiamide of tartaric dini-
trate, 37
ethylenedinitramine, 20
hexanitrodiphenyl-/?-hydroxynitramino-

ethyl nitrate, 72
lead styphnate, 215
metal azides, 189
methylenedinitramine salts, 221
nitrogen selenide, 229
octogen, 118
silver azide, 184
thallous azide, 188
tetryl, 53
2,4,6-trinitro-l,3-di(methylnitramino)-ben-

zene, 66
2,4,6-trinitro-3-methylnitraminoanisole, 65
2,4,6-trinitro-3-methylnitraminophenetole, 55
2,4,6-trinitro-3-methyInitraminophenol, 65
Incompatibility of explosive mixtures, 283
Inert ingredients in mining explosives, 427
Inert neutralizing agents in mining explosives,

433
Inflammability of solvents, limits of, 684
Infusible explosives, 258
Initiating ability of priming explosives, 158
Initiating compositions, 232
Initiating properties of dinitrobenzenediazo-

oxide, 204
Initiation temperature,
of basic cupric azide, 185

708

SUBJECT INDEX

Initiation temperature (cont.)
of fulminates, 157
lead azide, 172
metal azides, 186
Initiators, 129
Ionic structure of phenyldiazonium nitroform-

ate, 205
Isocyanuric acid, 133
"Isonitramines" — see Nitrosohydroxylamines

KM A mixtures, 271
Kneaders, 586, 587, 588, 589

"L-alloy", 248
Lead azide, 169
basic, 178
manufacture of, 178

continuous method, 179
neutral, 169
allotropic forms of, 169
Lead block expansion test, 85
of acetylene, 227
Amatols, 261
2,4-di-(methylnitraminomethyl)-4-nitrophe-

nol, 73
dinitrodiethyloxamide, 35
dinitrodi-(/?-hydroxyethyl)-oxamide dini-

trate, 37
dinitrodimethyldiamide of tartaric dini-

trate, 37
dinitrodimethyloxamide, 35
dinitrodimethylsulphamide, 36
Gelatine-Donarit 1, 456
hexanitrodiphenylethylenedinitramine, 70
mercury fulminate, 148
metal azides, 189
metazonic acid, 224
mining explosives, 429, 450, 451 , 452, 453,

454, 469, 470, 472, 473, 474, 476, 477,

478, 479, 485, 486, 487, 488, 489
nitrourea, 34
octogen, 119
oxyliquits, 492
Perdit, 264
tetryl, 54

trimethylenetrinitrosamine, 124
2,4,6-trinitrophenylethylnitramine, 67
trinitrophenyl-/?-hydroxynitraminoethyl

nitrate, 71
trinitrotriazidobenzene, 194

2, 4, 6 - trinitro -1,3, 5-tri(methylnitramino)-
benzene, 67

Wetter-Astralit, 460

Wetter-Carbonit B, 460
Lead dinitroresorcinate, 220
Lead picrate, 212
Lead rhodanate, 230
Lead styphnate, 213

Lead trinitroresorcinate— see Lead styphnate
Lignin, thermal decomposition of, 327
Liquid explosives, 286
mixtures with

hydrogen peroxide, 290

nitric acid, 290

nitrogen dioxide, 289

tetranitromethane, 290
Liquid oxygen explosives— see Oxyliquits

Macarit, 264
Manufacture,
of ammonium nitrate-fuel oil mixtures, 508
ammonium nitrate mining explosives, 498
ammonium nitrate powder for rockets, 386
ballistites, 647

safety in, 651
barium azide, 190
blackpowder, 342

chlorate and perchlorate explosives, 520
Cordite MD, 643
Cordite Mk I, 643
Cordite RDB, 644
cyclonite, 87

E-method, 109

K-method, 105

KA-method, 111

W-method, 107
detonators, 236
dynamites, 511
fusible mixtures, 255
hydrazoic acid, 168
lead azide, 178

continuous method, 179
lead styphnate, 218

continuous method, 219
mercury fulminate, 149
mining explosives, 498
nitrocellulose powder, 573
nitroguanidine, 32, 33
rocket charges, 373
salts of hydrazoic acid, 168
smokeless powder, 570

SUBJECT INDEX

709

Manufacture (cont.)
of smokeless powder, safety in, 682
solventless powder,
in German factories, 660
with low content of nitroglycerine, 652
tetrazene, 209
tetryl, 56
Mechanism,
of amines nitration, 11
cyclonite decomposition, 81, 82
dimethylaniline nitration, 41
N-methyl-N'-nitroguanidine formation, 27
nitramines nitration, 11
phenyl azide formation, 162
triphenylmethyl azide decomposition, 196
Melamine nitro derivatives, 120
Mercaptans in rocket propellants, 295
Mercuric azide, 186
Mercuric oxalate, 224
Mercurous azide, 186
Mercury fulminate, 135
manufacture of, 149
reactions with metals, 140
sensitiveness to sunlight, 146
storage of, 153
toxicity of, 153
Metafulminuric acid, 134
Metal azides, 186

Metals, powdered, in rocket propellants, 311
Metanits, 410, 450
Metazonic acid, 224

salts of, 224
Methylamine, nitration of, 10
Methylenedi-isonitramine, 17, 18
sodium salt of, 221
thallous salt of, 223
Methylenedinitramine, 17
homologues of, 17
salts of, 221
O-Methyl-methylnitramine, 4

hydrolysis of, 4
Methylnitramine, 16, 51
N-Methyl-N'-nitroguanidine, mechanism of for-
mation, 27, 28
N-Methylpicramide, 41, 51, 52
Methyl polymethacrylate in rocket propellants,

381
Methyltetryl-see 2,4,6-Trinitro-3-methylphenyl-

methylnitramine
Miedziankit, 403, 408, 422, 475
Minimum ignition temperature of mining explo-
sives, 428

Minimum initiating charges,
of lead azide, 177, 233
lead azide-Iead styphnate mixtures, 233
lead styphnate, 177, 233
mercury fulminate, 177
metal azides, 186
picric acid, 72
silver azide, 177
tetrazene, 177
tetryl, 72

trinitrophenyl-/?-hydroxynitraminoethyl ni-
trate, 72

trinitrotoluene, 72
Mining explosives, 395
Belgian, 404, 447, 449, 450
British, 405, 461, 462, 463, 464, 465, 466,

467, 468
chlorate and perchlorate, 408, 457, 458, 474

manufacture of, 520
components of, 420
Czechoslovakian, 448, 450
French, 403, 451, 452, 453, 454
German, 406, 408, 455, 456, 457, 458, 459, 460
Hungarian, 468, 469

Japanese, 468, 469, 470, 471, 472, 473, 474
inert neutralizing agents in, 433
manufacture of, 498
non-permitted, 461
permitted, 461

Polish, 475, 476, 477, 478, 479
sheathed, 461
stability of, 446
tests for, 433

U.S.A. 480, 481, 482, 483, 484
U.S.S.R. 484, 485, 486, 487, 488, 489
Mixtures,
fusible,

gaseous, with hydrazine, 307

manufacture of, 255

phlegmatization of, 257
of Amatols with cyclonite, 263

cyclonite with TNT, 249

PETN with TNT, 251
S-type, 271
MNO— see Dinitrodimethyloxamide
Monobel, 404
Muzzle flame, 544

NENO — see Dinitrodi-(/8-hydroxyethyi)-oxam-

ide dinitrate
Nickel ammino perchlorate, 230

710

SUBJECT INDEX

Nitramide-see Nitramine
Nitramine salts, 221
Nitramines, 1
aliphatic, 15
aromatic, 40

aromatic-aliphatic, denization of, 5
bond angles in, 2
heterocyclic, 77
interatomic distances in, 2
preparation of, through chloramines, 11

mechanism of, 11
reduction of, 6
Nitramines, primary,
acidic properties of, 4, 16
aci-form of, 4

decomposition in sulphuric acid, 4, 5
decomposition with alkalis, 6
reaction with diazonu thane, 7
Nitramines, secondary,
decomposition in sulphuric acid, 4
decomposition with alkalis, 6
Nitramino-azoxy compounds, 73
Nitramino-esters of nitric acid, 70
Nitramino group, structure of, 2
Nitraminonitrophenols, 72
Nitrates in mining explosives, 421
Nitration,
of amines

by nitrolysis, 12
mechanism of, 11
/?-aminoanthraquinone, 8
a-aminopyridine, 8
aminothiazole, 8, 9
dimethylaniline, 41, 56

continuous method of, 58
2,4-dinitromethylaniline, 44, 61
hexamethylenetetramine, 13, 94
methylamine, 10
primary amines, by acylation, 10
secondary amines, 9
2,4,6-tribromoaniline, 9
Nitrator, continuous, 99
Nitrator control panel, 100
Nitric acid, mixtures with,
in liquid explosives, 290
in rocket propellants, 292
Nitric esters in rocket propellants, 309
Nitroaliphatic compounds on smokeless powder,

671
5-Nitro-3-aminosalicylic acid, 205

diazo compound from, 205
Nitroammelide, 120

5-Nitro-3-azidosalicylic acid, 195

plumbous salt of, 196
Nitrocellulose,

dehydration of, 573, 574, 576

dissolution of, 583
Nitrocellulose bulk powder, 640
Nitrocellulose mixtures preparation, 582
Nitrocellulose powder, 541

manufacture of, 573

stability of, 555
Nitrocompounds, mixtures of, 247
Nitrocyanamide, 21

salts of, 211
N-Nitro-N',N"-diacetyl melamine, 121
Nitroethane in rocket propellants, 296
Nitroformoxime, 158
Nitrogen dioxide, mixtures with,

in liquid explosives, 289

in rocket propellants, 291
Nitrogen selenide, 229
Nitrogen sulphide, 229
Nitroglycerine powders, 541

solventless, 544

stability of, 556

with a volatile solvent, 642
Nitroguanidine, 22

complex salts of, 31

manufacture of, 32, 33

reaction with ethylenediamine, 28

tautomeric forms of, 22
Nitroguanyl azide, 28
Nitrolysis, 12

Nitromethane in rocket propellants, 296
Nitro Methylene Blue, 73
Nitromethylisonitramine salts, 223
Nitronium ion, 12, 26
Nitroparaffins in rocket propellants, 296
Nitrophenol salts, 212
2-Nitropropane in rocket propellants, 297
Nitrosamines, 121
N-Nitrosodiphenylamine, 562
Nitrosoguanidine, 210
Nitrosohydroxylamine derivatives, 1
Nitrosohydroxylamine salts, 221
Nitrosophenol salts, 221
m-Nitrotetryl — see 2,3,4,6- Tetranitrophenyl-

methylnitramine
Nitrourea, 33

salts of, 34
Non-permitted mining explosives— see Mining

explosives, non-permitted
Non-volatile solvents, 645

SUBJECT INDEX

711

Octogen, 90, 109, 117

Octyl - see Hexanitrodiphenylethylenedinitr-

amine
Organic azides— see Azides, organic
Organometallic compounds, in rocket propel -

lants, 312
Oxalic acid, 224

salts of, 224
Oxamidoazide, 192

Oxygen balance in mining explosives, 423
Oxygen carriers in mining explosives, 421
Oxygen, liquid, in rocket propellants, 309
Oxyliquits, 290, 491
Ozone, in rocket propellants, 309

2,3,4,5,6-Pentanitrophenyl-N-methylnitramine,
65

Penthrite, in smokeless powder, 670

Pentry 1 — see Trinitrophenyl-/? -hydroxynitr-
aminoethyl nitrate

Pentyl— seeTrinitrophenyl-^-hydroxynitramino-
ethyl nitrate

Perchlorates,
in mining explosives, 422
of complex ammines, 230

Perchloric acid, in rocket propellants, 313

Perchloryl fluoride, in rocket propellants, 313

Perdit, 264

Permitted mining explosives— see Mining explo-
sives, permitted

Peroxides, 225

Phenyl azide, 162

Phenyldiazonium nitroformate, 205

Phenyl-O-methylnitramine, 7

Phenylnitramine, 2

Phenylnitrosohydroxylamine, 2

Phlegmatization of cyclonite, 113

Phlegmatizing substances, 257, 276

"Phosphorous" azide, 191

Picrate powder, 334

Picratol, 266

Picryl azide, 192

Picrylmethylnitrarnine— see Tetryl

Plastic explosives, 281

Plastics, in rocket propellants, 381

Polymerization of fulminic acid, 134

Potassium chlorate, 274

Potassium chloride in mining explosives, 428

Potassium nitrocyanamide, 211

Power of explosives, 439

Preparation,
of sec-butylnitramine, 11
cyclonite, 13

ethylenedinitramine, 18, 19
methylenedinitramine, 17
nitramine, 16
nitramines, 8

through chloramines, mechanism of, 11
nitrodiefhanolamine dinitrate, 36
primer compositions, 235
tetryl, 47
general rules of, 47
Presses, 576, 577, 578, 581, 590, 593, 594, 595

596, 658
Primary amines— see Amines, primary
Primary explosives -see Initiators
Production,
of ethylenedinitramine, 19
dimethylaniline sulphate, 57
Propergols-see Rocket propellants, liquid
Pyronite-see Tetryl

Rate of detonation— see Detonation rate
Rate of nitration of hexamethylene tetramine,

94
X-Ray analysis,
of dimethylnitramine, 2

ethylenedinitramine, 2
X-Ray investigation of azides structure, 161
RDX— see Cyclonite
Reduction of nitramines, 6
Rocket propellants,
cast, 675

German, 681
composite, 365, 392

mixing of ingredients in, 389

mixtures of perchlorates with plastics, 380

mixtures with ammonium nitrate, 383, 389

thiokol propellants, 369
liquid, 291

composition of, 317

cryogenic liquids, 319

properties of, 296

storable liquidis, 319

with amines, 294

with beryllium compounds, 311

with boron compounds, 311

with ethylene oxide, 310

with fluorine compounds, 312

with fuels, 293

with fufuryl alcohol, 294

712

SUBJECT INDEX

Rocket propellants, liquid (cont.)
with hydrazine, 295, 305
with hydrogen peroxide, 299
with liquid oxygen, 309
with mercaptans, 295
with nitric esters, 309
with nitrogen dioxide, 291
with nitroparaffins, 296
with organometallic compounds, 312
with ozone, 309
with perchloric acid, 313
with powdered metals, 311
with silicon compounds, 311
with surface-active substances, 295

Safety explosives,
in the U.S.A. 484
U.S.S.R. 489
Safety,
in ballistites manufacture, 651
in blackpowder factories, 361
in smokeless powder manufacture, 682
Safety increasing ingredients, in mining explo-
sives, 427
Safety of mining explosives, 396
Safety tests for mining explosives, 439

in Japan, 472
Safety testing,
in the presence of coal-dust, 445
in the presence of methane, 441
Saltpetre, for blackpowder manufacture, 342
Saltpetre, Chilian, for blackpowder manufac-
ture, 344
"SH-Salz", 105
Saxonite, 404
Schneiderite, 260

Schultze powder— see Nitrocellulose bulk pow-
der
Semifusible explosives— see Explosives, semi-
fusible
Sensitiveness to detonation,
of smokeless powder, 540
test for, 434
Sensitiveness to impact,
of barium azide, 189
basic cupric azide, 185
cupric azide, 185
cyclonite, 86

dinitrobenzenediazo-oxide, 204
fulminates, 157
lead azide, 172, 176

Sensitiveness to impact (cont.)
of mercury fulminate, 148
metal azides, 186

mixtures with ammonium nitrate, 262
octogen, 118

phlegmatizing substances, 276
silver azide, 184
smokeless powders, 541
tetryl, 54, 55

trimethylenetrinitrosamine, 124
2,4,6-trinitro-3-methylnitraminoaniline, 65
2,4,6-trinitro-3-methylnitraminoanisole, 65
2,4,6-trinitro-3-methylnitraminophenetole,

65
2,4,6-trinitro-3-methylnitraminophenol, 65
trinitrophenyl-/?-hydroxynitraminoethyl ni-
trate, 71
Sensitiveness to y-radiation of mercury fulmi-
nate, 146
Sensitiveness to sunlight of mercury fulminate,

146
Sensitiveness to ultra-violet light of mercury

fulminate, 146
Sheathed mining explosives— see Mining explo-
sives, sheathed
Silicon compounds in rocket propellants, 311
Silver acetylide, 229
Silver azide, 182, 184
decomposition of, 183
kinetics of, 184
Silver cyanamide, 211
Silver fulminate, 157
Silver nitrocyanamide, 212
Silver oxalate, 224
Silver perchlorate, 232
Silily-azides, 191
"Sinoxyd", 218, 235
Smoke formation, 548
Smokeless powder, 528
erosiveness of, 548
for rockets, 671
manufacture of, 570

waste products in, 631
stability of, 550
stabilization of, 559
with nitroaliphatic compounds, 671
with penthrite, 670
Sodium nitrate in mining explosives, 428
Solubility,
of cyclonite, 79, 80
ethylenedinitramine, 20
mercury fulminate, 138

■r*S/gff*ft

SUBJECT INDEX

713

Solubity (cont.)
of tetryl, 50
trimethylenetrinitrosamine, 122
Solventless powder,
in Japan, 663

manufacture of, in German factories, 660
with low content of nitroglycerine, manu-
facture of, 652
Solvent recovery in smokeless powder manu-
facture, 599
Specification for cyclonite, German, 105
Specific heat -see Heat, specific
Specific impulse, 312
of composite propellants for rockets, 392
mixtures for rocket propellants, 381
Specific pressure,
of ammonium azide, 190
cyclonite, 84
mercury fulminate, 148
nitroguanidine, 30
Spectra,
electronic,

of hydrazoic acid derivatives, 163
infrared,
of azides, 162
charcoal, 327
hydrazoic acid, 164
hydrazoic acid derivatives, 164
metal azides, 165
Raman,
of azides, 162
hydrazoic acid derivatives, 164
metal azides, 165
rotational,

of azides, 162
ultraviolet,
of ethylenedinitramine, 3
2,5-dinitro-2,5-diazahexane, 3
methylenedi-isonitramine, 222
nitramines, 2, 3
nitroguanidine, 25
Spectrographic analysis of hydrazoic acid de-
rivatives, 163
Spent acid,
in cyclonite manufacture, 108
dinitromethylaniline nitration, 62
Stability,
of mining explosives, 446
nitrocellulose powder, 555
nitroglycerine powder, 556
nitroguanidine, 27
smokeless powder, 550

Stability tests for smokeless powder, 557
Stabilization,
of nitrocellulose powder, 632
smokeless powder, 559
with diphenylamine, 559
with inorganic stabilizers, 563
with organic stabilizers, 564
Stabilizers, for smokeless powder,
apparent, 567
inorganic, 563
organic, 564
Storage of mercury fulminate, 153
Sulphur, for blackpowder manufacture, 344
Sulphurless powder, 331
Surface-active substances in rocket propellants

295
Surface gelatinization of the nitrocellulose
powder, 624

Temperature of explosion,
of acetylene, 227
ammonium azide, 190
cyclonite, 84

dinitrobenzenediazo-oxide, 204
mercury fulminate, 148
nitroguanidine, 30
Temperature of ignition -see Ignition temper-
ature
Temperature of initiation -see Initiation tem-
perature
Teneres-see Lead styphnate
Testing galleries, 409, 413, 415, 416, 417, 418,
419, 439, 440, 441, 442, 443, 444, 463
for mining explosives, 397
Tetra-azido quinone, 192
Tetralita— see Tetryl
Tetralite— see Tetryl

Tetramethylenediperoxidodicarbamide, 226
3,3',5,5'-Tetranitro-bis(4,4'-nitromethyIamino)-

azoxybenzene, 73
3,5,3',5'-Tetranitro-4,4'-di(methylnitramino)-

benzophenone, 68
Tetranitromethane,
as oxidant, 297
in liquid explosives, 290
in rocket propellants, 297
N-2,4,6-Tetranitro-N-methylaniline-see Tetryl

2,3,4,6-Tetranitrophenylmethylnitramine, 63
derivatives of, 63

1 ,3,5,7-Tetranitro-l ,3,5,7-tetrazacyclo-octane -

see Octogen
Tetraphenylhydrazine, 562

714

SUBJECT INDEX

Tetrazene, 206

manufacture of, 209
Tetrazolylguanyltetrazenehydrate— seeTetrazene
Tetryl, 42

homologues of, 62

polycyclic analogues of, 68

preparation of, 42
general rules for, 47
Tetrylit— see Tetryl
Thallium styphnate, 220
Thallous azide, 188
Thiokol, liquid, 369

manufacture of rocket charges with, 373
Thiokol propellants, 368
TMTN — see Trimethylenetrinitrosamine
Tolyltetryl — see 2,4,6-Trinitro-3-methylphenyl-

methylnitramine
Torpex, 271
Toxicity,

of cyclonite, 86
hydrazine, 307
mercury fulminate, 149
tetryl, 56
Transmission of detonation test, 434
Trialen, 271

1,3,5-Triazacyclohexanetrisulphonic acid, 108
Triazide of trimesic acid, 192
Triazidonitrosaminoguanidine, 207
Triazoethanol nitrate, 191
2,4,6-Tribromoaniline, nitration of, 9
N,N',N"-Trichloro-cyclotrimethylenetriamine,

81
2,4,6-Tri-(dimethylaminomethyl)-phenol, 72
Trimethylenetriamine sulphate, 122
Trimethylenetrinitrosamine, 121
Trinitrobenzoyl azide, 193
2,4,6-Trinitro-l,3-di(methylnitramino)-benzene,

65
2,4,6-Trinitrodiphenylamine, 51
1,3,5-Trinitrohexahydro-sym-triazine— see Cy-
clonite ; trimethylenetrinitrosamine
Trinitro-N-methylaniline, 5

2,4,6-Trinitro-3-methyInitraminoaniline, 64
2,4,6-Trinitro-3-methylnitraminoanisole, 64
2,4,6-Trinitro-3-methylnitramino-N-methylani-

line, 64
2,4,6-Trinitro-3-methylnitraminophenetoIe, 64
2,4,6-Trinitro-3-methylnitraminophenol, 64
2,4,6-Trinitro-3-methylphenylmethylnitramine,

62
2,4,6-Trinitrophenyl-n-butylnitramine, 67
2,4,6-Trinitrophenylethylnitramine, 67
Trinitrophenyl-^-hydroxynitraminoethyl nitrate,

70
2,4,6-Trinitrophenylmethylnitramine — see Tetryl
l,3,5-Trinitro-l,3,5-triazacyclohexane— see Cy-
clonite
Trinitrotriazidobenzene, 193
2,4,6-Trinitro-l,3,5-tri(methylnitrarnino)-ben-

zene, 66
Triphenylmethyl azide, 196
Tritetryl-see2,4,6-Trinitro-l,3,5-tri(methyl-

nitramino)-benzene
Tritonal, 272

Trizinat— see Lead styphnate
T4— see Cyclonite

UDMH— see 1,1-Dimethylhydrazine

Vacuum driers, for mercury fulminate, 153,

154
Vacuum filters, continuous, 101
Viking Powder, 405, 410

Waste substances,
in mercury fulminate manufacture, 156
in smokeless powder manufacture, 631

Yonckites, 447

715

ERRATA TO VOLUMES I AND II

The author regrets that a few errors occurred in Volumes I and II.

Volume I

p. 15, eqn. (20) should read:

N 2 4 +3H 2 S0 4 ->NOf+H 3 0® +3HS0 9 +NO®
p. 41: the formulas should be replaced by the following:

NH 2
2 N |
N0 2 „ \y\

NHCOCH3 NCOCH3 / \/\

NO, I / NO

■^V > <?\ \ an d in some instances

NH 2

no, - X ) NH "

N0 2 V .

N0 2 N0 2

p. 77: the bottom formula should be replaced by the following:

X— C 6 H 4 — N=0

1 i

! :
N N

\/v

O O

p. 82, lines 11-12 should read:

Wieland [13] of the formation of fulminic acid ...
p. 97, line 7 from the bottom should read:

According to Wieland [79, 79a], the reaction ...
p. 97, last line of the page should read:

1,4-diphenylbutadiene is converted to l,4-diphenyl-l,4-dinitrobutylene-2 (IX)
[79b]:

p. 98, line 8 should read:

Subsequently Wieland [79c] found ...
p. 98, paragraph (3) should read:

(3). The phenyl group when treated with N 2 4 underwent nitration:

2 N— C 6 H 4 — CH— CH— COR

I I
N0 2 N0 2

XVI

716 ERRATA TO VOLUMES I AND II

In such a way ...

p. 99, line 13 from the bottom should read:
Shechter and Conrad [49] ...

p. 109, line 13 should read:

On the basis of eqns. (44) and (45) ...

p. 133:

reference 12a should be corrected to 13, reference 13 should be corrected to 14,
reference 14, H. Wieland and L. Semper, Ber. 39, 2522 (1906) should be can-
celled.

p. 135, reference 79 should read:

H. Wieland and E. Blumich, Ann. 424, 75 (1921).

New references should be inserted:

79a. H. Wieland, F. Rahm and F. Reindel, Ber. 54, 1770 (1921).

79b. H. Wieland and H. Stenzel, Ber. 40, 4825 (1907); Ann. 360, 299 (1908).

79c. H. Wieland, Ann. 328, 154 (1903).

p. 176, lines 6-4 from the bottom should read:

CC1 4 CC1 3 CN

asymmetric vibr. 1506 cm -1 1496 cm -1

symmetric vibr. 1332 cm -1 1320 cm -1

p. 179, line 14 from the bottom should read:

Holder and Kline [38], Schmidt, Brown and D. Williams [40] ...

p. 618 (Author Index)

Schechter H. should be cancelled and on p. 619 items 92, 99 (134, 600) should
be added to Shechter H.

Volume II

p. 165, lines 10-14 should read:

The boiling point of isopropyl nitrate is 101-102°C. It was originally believed
that the substance can only be obtained from isopropyl iodide and silver nitrate
[1]. Direct nitration is difficult due to oxidation at the carbon atom carrying
the secondary hydroxyl group.

However a process for the continuous nitration of isopropyl alcohol in
the presence of urea has been described in Imperial Chemical Industries patents
[20]. Isopropyl alcohol and urea was introduced into nitric acid (over 40%
HN0 3 ) at its boiling temperature and current of air removed unstable products.
According to Desseigne [21] the method gave ca. 80% yield. He used nitric
acid of over 50% HN0 3 at 108-1 10°C.

Isopropyl nitrate is becoming important as an engine starter fuel [22].

p. 165, new references should be inserted:

20. W. G. Allen and T. J. Tobin (to Imperial Chemical Industries Ltd.), Brit.
Pat. 696489 (1953); see also Brit. Pats. 749734 and 749844 (1956).

ERRATA TO VOLUMES I AND II 717

21. G. Desseigne, Mem. poudres 37, 97 (1955).

22. Engineer, London 200, 269 (1955).

180, lines 16-15 from the bottom should read:
This is the result of the symmetrical structure of PETN. PETN is completely ...

429, last line on the page should read:
(a) in weight %, (b) in mole %.

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Chemistry and Technology of Explosives - Volume III"

See other formats