You are on page 1of 822

proceedings of

New Trends in Research


of Energetic Materials

Pardubice, April 1921 2006


University of Pardubice, Faculty of Chemical Technology
Institute of Energetic Materials

Sponsors

University of Pardubice, Faculty of Chemical Technology


Institute of Energetic Materials

proceedings of the 9th Seminar

New Trends in Research


of Energetic Materials

held at the University of Pardubice


Pardubice, Czech Republic
April 1921 2006

intended as a meeting of students, postgraduate students, university teachers,


and young research and development workers, concerned from the whole world

This publication has not been submitted to language corrections and contributions have not
been reviewed.
Contributions of the Proceedings will be quoted in the Chemical Abstracts.
The only distributor of the present publication is the
Institute of Energetic Materials
University of Pardubice
532 10 Pardubice
Czech Republic
where the publication can be ordered or gained by exchange of similar pulications.

NTREM 06
Proceedings of New Trends in Research of Energetic Materials
Jan Ottis, Miloslav Krupka (editors).
Published by University of Pardubice, publication number UP 06-19.
Czech Republic, April, 2006.
828 pages, 150 copies.
Printed by Press Center of University of Pardubice.
Typeset by authors, checked by editor, and completed by Vt Zka
(www.typokvitek.com).
C Copyright to all papers are retained by the authors.

ISBN 80-7194-849-7

9th Seminar of the New Trends in Research of Energetic Materials


Seminar is supported by:
Schlumberger Technology Corporation, Rosharon, Texas,
European Office of Aerospace Research and Development of the USAF,
Austin Detonator, Inc., Vsetn,
Indet Safety Systems, Inc., Vsetn, a member of Nippon Kayaku group,
Explosia, Ltd., Pardubice,
BORGATA, Ltd., Praha 5,
Faculty of Chemical Technology, University of Pardubice,
OZM Research, Hrochuv Tnec.
Chairman: Prof. Svatopluk Zeman, D.Sc.
Scientific Committee:
Chairman
Dr. Adam Cumming
Members
Prof. Ang How-Ghee
Dr. Anthony J. Bellamy
Dr. Yurii M. Burov
Dr. Stanislaw Cudzilo
Prof. Manfred Held
Dr. David Jones
Prof. Mikhail Ilyushin
Prof. Thomas Klapoetke
Prof. Michel Lefebvre
Prof. Frantiek Ludvk
Prof. Andrzej Maranda
Dr. Henric stmark
Prof. Peter Politzer
Dr. William G. Proud
Prof. Valerii P. Sinditskii
Prof. Igor Tselinski
Prof. Yuanjie Shu
Dr. Muhamed Sueska
Prof. Waldemar A. Trzciski
Assoc. Prof. Pavel Vvra
Dr. Woodward Waesche
Organizing Committee:
Chairman
Assoc. Prof. Betislav Janovsk
Members
Assoc. Prof. Ladislav Lehk
Dr. Jan Jakubko
Dr. Marcela Jungov
Dr. Pavel Valenta
Dr. Iva Ulbrichov

DSTL, Sevenoaks, U.K.


National Univesity of Singapore, Singapore
Cranfield University, U.K.
Russian Academy of Sciences, Chernogolovka, Russia
Military Univ. Technol., Warsaw, Poland
EADS/TDW, Schrobenhausen, Germany
CERL, Ottawa, Canada
St. Petersburg State Inst. of Technol., Russia
Ludwig-Maximilians-Universitt, Mnchen
Royal Military Academy, Belgium
Univ. of Defence, Brno, Czech Republic
Military Univ. Technol., Warsaw, Poland
FOI, Tumba, Sweden
Univ. of New Orleans, USA
Cavendish Laboratory, Cambridge, U.K.
Mendeleev Univ. of Chem. Technol., Moscow
St. Petersburg State Inst. of Technol., Russia
Inst. of Chem. Materials, CAEP, Sichuan, China
Brodarski Inst., Zagreb, Croatia
Military Univ. Technol., Warsaw, Poland
Univ. of Pardubice, Czech Republic
SAIC, Gaineswille, USA

IEM, Univ. of Pardubice, Czech Republic


Explosia, Ltd., Pardubice, Czech Republic
Indet Safety Systems, Vsetn, Czech Republic
Univ. of Pardubice, Czech Republic
Austin Detonator, Vsetn, Czech Republic
University of Pardubice, Czech Republic
1

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Contents
Part 1

Invited lectures
1

European Govermental Collaboration in Energetics Research The Future


Adam Cumming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

New Diagnostic Techniques in Blast Waves


Manfred Held . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

The Thermal Behaviour of BTAW, a High Nitrogen Fuel


David Jones, K. Armstrong, T. Parekunnel, Q. Kwok . . . . . . . . . . . . . . . . . . . . . . . . . 42

New Trends in Detection of Explosives


Petr Mostak, Miroslav Stancl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Molecular Surface Electrostatic Potentials in the Computational


Characterization of Energetic Compounds
Peter Politzer, Jane Murray . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

Lectures

The Influence of Geometry and Material Properties on an Explosives


Gurney Velocity and Energy
Joseph Backofen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Computer Simulation of Thermal Decomposition Mechanism for


Compounds with Nitroguanidine Fragment
Eugeniya Bakhmatova, Vyacheslav Korolev, Aleksey Porollo, Tatyana Pivina . . . . . 90

Thermal Decomposition of Solid Energetic Materials


Yurii Burov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

Use of Compressed Gas Gun to Investigate Response of Explosives to a


Range of Initiation Stimuli
Michael Cartwright, Andrew Lewis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107

10

The Relationship between Shock Sensitivity and Morphology in Granular


RDX
Helen Czerski, William Proud, John Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

11

Guanidinium Nitroformate Salts: Possible New Oxidizers for High


Performance, Halogen Free Solid Propellants
Michael Gbel, Thomas Klaptke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

12

Preliminary Study of Superfine Metal Powders Effect on the Decomposition


of Energetic Materials
Alexander Gromov, Elena Popenko, Alexey Sergienko, Young-Soon Kwon,
Ulrich Teipel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

13

Preparation, Properties and Application of Energetic Complex Perchlorates


of d-Metals
Mikhail Ilyushin, Igor Tselinsky, Irina Bachurina, Yurii Gruzdev,
Valery Sinditskii, Valery Serushkin, Vyacheslav Egorshin . . . . . . . . . . . . . . . . . . . . . 143

14

Balistic Stability of a Spherical Propellant: Comparison with a Flattened


Spherical Propellant
Laurence Jeunieau, Michel Lefebvre, Pierre Guillaume . . . . . . . . . . . . . . . . . . . . . . . 150

15

Synthesis and Characterization of 1,2-Dinitroguanidine (DNG) and its


Derivatives
Nikolaj Latypov, Martin Johansson, Larisa Yudina-Wahlstrm, Stefan Ek,
Carina Eldster, Patrick Goede . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160

16

Energetic Characteristics of Solid Composite Propellants and Ways for


Energy Increasing
David Lempert, G. Nechiporenko, G. Manelis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

17

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From


the Point of View of the Concept of Spatial-Periodic
Alexander Lukin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181

18

Stable Salts of Methylated 5-Aminotetrazoles


Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke,
Jan Welch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202

19

Applicability of Dynamic Mechanical and Thermal Methods in Investigation


of Ageing Processes of Double Based Propellants
Sanja Muani, Muhamed Sueska, Sanko Baka . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214

20

High Energy Density Materials (HEDM): Overview, Theory and Synthetic


Efforts at FOI
Henric stmark . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

21

Aerobic Biodegradation of Mononitrotoluenes in Different Reactor


Configuration
Jan Pca, Martin Halecky, Rakesh Bajpai . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251

22

Polyfunctional N-nitramines Structure Differentiation as a Basis for


Simulation of Their Decomposition Mechanism
Tatyana Petukhova, Victor Ivshin, Vyacheslav Korolev, Tatyana Pivina . . . . . . . . 260

23

Process Hazard Evaluation for Explosives Manufacturing Using


TEMCLEV-EX Procedures
Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak . . . . . . . . . . . . . . . . . . . . . . 268

24

Determination of the Thermal Resistance of the Emulsion Explosives


Tomasz Saaciski, Waldemar Witkowski, Andrzej Maranda,
Johann Kasperski, Barbara Gobek, Henryka Boniuk . . . . . . . . . . . . . . . . . . . . . . . . 279
3

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

25

Reduction of Nitroaromatic Explosives by Oxygen-insensitive


NAD(P)H:Nitroreductases: Implications for their Cytotoxicity
Jonas Sarlauskas, Lina Miseviciene, Henrikas Nivinskas,
Zilvinas Anusevicius, Ausra Nemeikaite-Ceniene, Evaldas Maldutis,
Narimantas Cenas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285

26

Complex Research of Al Nanopowders Produced by the Exploding Wire


Method
Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov,
Ludmila Chemezova, Alexander Alekseev . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293

27

Ab Initio MD Computation of the Vibrational Relaxation Time in HE


Molecules
Alexander Selezenev, Aleksey Aleynikov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 304

28

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong


Bonds or Other Reasons Exist?
Valerii Sinditski, Sergei Smirnov, Valerii Egorshev . . . . . . . . . . . . . . . . . . . . . . . . . . 314

29

The principal approach to the estimation of cookoff temperature using FEM


Jakub eleovsk, Miloslav Krupka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329

30

Low Temperature Activation Energy of Propellants


Stephan Wilker, Uldis Ticmanis, Gabriele Pantel . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338

31

Theoretical Study of the Tetrazines C-N Heterocyclic Derivatives


Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min

32

Self-aggregation of Thermal Decomposition Products


Tomasz Zielenkiewicz, Teresa Ksiczak, Andrzej Ksiczak

. . . . . . . . . . . 356

. . . . . . . . . . . . . . . . . . 368

Only in Proceedings

33

Study of the Routes for Producing Nitroguanidine(NQ) in Aqueous Nitric


Acid Solutions. Development of the Efficient Meth
Alexandr Astratev, Dmitry Dashko, Andrei Stepanov . . . . . . . . . . . . . . . . . . . . . . . . 377

34

Fuel Dispersion in Air by Detonation Wave


Esmaeil Bazrafshan, Seyed Pourmortazavi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 393

35

Preliminary Study of the Behavier of Al Nano Powders Water Slurry


Compositions at Near Room Temperatures
Mikhail Laritchev, Olga Laricheva, Ilia Leipunsky, Pavel Pshechenkov,
Alexey Jigatch, Michael Kuskov, Valentin Sedoi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404

36

Carbon Nanotubes as Support for Co/TiO2 Electrocatalysts for Hydrogen


Evolution
Orce Popovski, P. Paunovi, A. Dimitrov, D. Slavkov, D. Smileski,
M. Tasev, S. Hadi Jordanov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 416

37

Sliding Shock Wave Causing Deformation of the Metal Plate Attached to


the Cover
Huiying Zhao, Zhaowu Shen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 423

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Part 2

Posters
38

Simulation of Coffee Mill Explosion Accident Underatex 137 Directive


Vladislav Adamk, Betislav Janovsk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451

39

Explosion Hazard of Water Gel-like Systems on a Base of High Explosives


Alexei Apolenis, Ekaterina Aleshkina, Vladimir Annikov, Georgii Kozak . . . . . . . . 464

40

Heat of Melting of Nitrocompounds Measuring by Means of the Differential


Scanning Calorimetry Method
Svetlana Arinina, Georgii Kozak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473

41

Characterization and some Properties of 1-Nitroamidino-3-amino-5nitrimino-1,2,4-triazole


Alexander Astachov, Vitaliy Revenko, Ludmila Kruglyakova,
Andrew Nefedov, Eduard Buka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 480

42

The Use of a Kenics Static Mixer for the Reaction of TNT with NaOCl
Anthony Bellamy, Victoria Lomax, Kate Read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491

43

Investigation of Stabilizing Properties of Triphenylamine (TPA)


Jan Bladek, Stanislaw Cudzilo, Sylwia Pietrzyk, Stephan Wilker . . . . . . . . . . . . . . . 500

44

Some Properties of a Composite Propellant Premix


Daniel Buczkowski, Bogdan Florczak, Tomasz Saaciski . . . . . . . . . . . . . . . . . . . . . 507

45

Thermophysical Properties of Thermoplastic Composite Propellants Based


on Modified PVC Binder
Vladica Bozic, Marko Milos, Djordje Blagojevic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 513

46

Thermal Decomposition of the trans-1,4,5,8-Tetranitro-1,4,5,8tetraazadecalin


Yurii Burov, Irina Kucherova . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526

47

Observation of Changes in Chosen Properties of Plastic Explosives During


Artificial Ageing
Martina Chovancov, Peter Oko, Jozef Lopch, Albta Pechov . . . . . . . . . . . . . 529

48

Hexanitrostilbene - Density Dependend Sensitivity


Dominik Clment, Karl Rudolf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 540

49

Self-sustaining reductive defluorination of (CF)n and characterization of the


reaction products
Stanislaw Cudzilo, Andrzej Huczko, Mateusz Szala, Michal Bystrzejewski . . . . . . . 544

50

Insensitive Environmentally-compatible and Non-toxic Replacements for


RDX in EMs New Method Of Synthesis Of DAT
Chaza Darwich, Thomas Klaptke . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 551

51

The Nitration Reaction of Isowurtzitane Acyl Derivatives


Maciej Duda, Pawe Maksimowski, Wincenty Skupiski . . . . . . . . . . . . . . . . . . . . . . 561
5

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

52

Initiation of Deflagration and Detonation in the Metal Dust/Air Systems


Irina Egorova, Boris Kondrikov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 569

53

JET Velocity and the Effects of the Shaped Charge Perforator for Use in
the Oil and Gas Well
Zvonimir Ester, Mario Dobrilovi, Branimir Jankovi, Zvonimir Tomii,
Mile Odak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 581

54

Thermochemical Calculations and Experimantal Study of Composite


Propellants Containing DADNE or NTO
Bogdan Florczak, Andrzej Orzechowski, Tomasz Saaciski . . . . . . . . . . . . . . . . . . . 587

55

Implementation of the Absolute Method in the Ballistic Analisys of


Gunpowder
Tomasz Gawor, Tomasz Wolszakiewicz, D. Korsze . . . . . . . . . . . . . . . . . . . . . . . . . . 598

56

Green Polymer-Bound Explosive (PBX) for Laser


Mikhail Ilyushin, Igor Tselinsky, Irina Shugalei, Anatoliy Chernay,
Valeria Toftunova . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602

57

Synthesis of 2,4-dinitro-2,4-diazapentane from methylnitramine and


dihalogenmethanes
Zdenk Jalov, Pavel Mareek, Kamil Dudek, Mat Ondrejkovi, Frantiek Lika

608

58

Application of Supercritical Fluid Extraction for the Isolation of


Nitrocompounds
Vra Jeov, Tom Fojtk, Ale Eisner, Martin Adam, Karel Ventura . . . . . . . . . 613

59

Possibilities of Degradations Process of Nitroaromatic Compounds


Vra Jeov, Tom Fojtk, Karel Ventura, Vt Matj, Eva Kyslkov,
Simona Voshlov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 618

60

Preparation of Low-melting Compositions Based on 1,3,3-Trinitroazetidine


Dmitrii Katorov, N. Vyalova, G. Rudakov, A. Ladonin, V. Zhilin, E. Veselova . . 624

61

Salts OF 5-Nitrotetrazolate
Thomas Klapoetke, Peter Mayer, Kurt Polborn, Jan Wlech . . . . . . . . . . . . . . . . . . . 631

62

5,5-Bis-(1H-tetrazolyl)amine (H2bta): A Promising Ligand in Novel


Copper Based Priming Charges (PC)
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand . . 641

63

Emulsive Explosive Sensitized by Inorganic Peroxides


Igor L. Kovalenko, Alexandr V. Kuprin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 652

64

Detonability of Nitromethane Mixtures with Azideethanol


George Kozak, Vlada Raikova . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657

65

Effect of Rubbing Surfaces Velocity on the Results of Friction Sensitivity


Testing of Explosives
Miloslav Krupka, Jan Horkel, Jakub Selesovsky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

66

Thermal Effects of a Slow Thermal Decomposition and Inflammation of


Nitramines and Its Mixtures with Al, B and Zr Powder
Valeriy Lashkov, A. Strikanov, A. Tikhonova . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 673

67

Sensitivity Research of Explosive Mixtures Containing Nitroguanidine and


3-Nitro-1,2,4-triazol-5-one
Andrzej Maranda, Jerzy Nowaczewski, Andrzej Orzechowski, Dorota Powaa . . . . 680

68

Cytotoxicity of Novel Nitrotriazole and Trinitroaniline Explosives: The Role


of Oxidative Stress
Ausra Nemeikaite-Ceniene, Jonas Sarlauskas, Henrikas Nivinskas,
Zilvinas Anusevicius, Narimantas Cenas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689

69

Optimalization of Crystallization of 1,1-Diamino-2,2-dinitroethylene


Andrzej Orzechowski, Dorota Powaa, Bogdan Florczak, Andrzej Maranda,
Jerzy Nowaczewski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 693

70

Selective Modifiers of Thermal Decomposition and Combustion of


Composite Solid Rocket Propellants Containing HMX
E. Popenko, A. Sergienko, E. Sokolov, D. Petrov, V. Shandakov,
Y. Zharinov, A. Lobanova . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701

71

Crystal Structures of Selected Energetic Materials Calculated by Molecular


Simulations
Miroslav Pospil, Pavel Vvra, Petr Kov . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 712

72

Comparison of Coating with Various Deterrents


Pavel Prchal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 719

73

Treatment of WastewatersContaining Nitrate Esters


Jan Skldal, Martin Kouba, Martina Mudrukov, Miloslav Slezk . . . . . . . . . . . . . 723

74

The Use of Low Pressure Closed Vessel and Rocket Motor for Measurements
of Burning Rate of Rocket Solid Propellants.
Petr Stojan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730

75

Experimental Rating of Surge Withstand Capability of Reinforced Concrete


to the Standard Concrete with Dynamic Load
Ji toller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 736

76

An Experimental Study of Detonation Transmissionfrom Hydrogen-Oxygen


to Hydrogen-Air Mixture
Andrzej Teodorczyk, A. Dabkowski, M. Bratowski . . . . . . . . . . . . . . . . . . . . . . . . . . . 742

77

On Some Methods of Determination of the Detonation Energy of Explosives


Waldemar A. Trzciski, Jzef Paszula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 748

78

Regarding the Fundamental Problem Solving of Interior Ballistics in the


Case of Flegmatised Powders
Titica Vasile, Cristian Barbu, Doru Safta . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 756

79

Numerical Description of the Effects in Gas Generator


Tomasz Wolszakiewicz, Tomasz Gawor, Zbygniew Walenta . . . . . . . . . . . . . . . . . . . . 764

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

80

The Mechanisms of Thermal Decomposition of Simple Hydronitrogen


Compounds
Ying Xiong, Yuanjie Shu, Ge Zhou, Xinfeng Wang, Anming Tian,
Ning-Bew Wong . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 769

81

Influencing Parameters on HE Projectiles with Natural Fragmentation


Berko Zecevic, Jasmin Terzic, Alan Catovic, Sabina Serdarevic-Kadic . . . . . . . . . . 780

82

New Findings in the Electric Spark Sensitivity of Nitramines. Part I.


Aspects of Molecular Structure
Svatopluk Zeman, Ji Ko, Vojtch Pelikn, Ji Majzlk, Zdenk Friedl . . . . . . . 796

83

New Findings in the Electric Spark Sensitivity of Nitramines. Part II.


A Problem of Hot Spots
Svatopluk Zeman, Ji Ko, Vojtch Pelikn, Ji Majzlk . . . . . . . . . . . . . . . . . . . . 810

84

Theoretical Study on the Decomposition of 1,1-Diamino-2,2-Dinitroethylene


Hehou Zong, Yuanjie Shu, Yigang Huang, Xinfeng Wang . . . . . . . . . . . . . . . . . . . . 816

Abstract
85

Polymorfic Transformations and Thermal Decomposition of Ammonium


Nitrate at Thermal Analysis Conditions
Valeriy Lashkov, E. Egorychev, A. Strikanov, A. Tikhonova, V. Rybakov . . . . . . . 825

Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 826

situation of the middle European countries, Balkan and former Soviet Union, no fee was and
is asked from the participants. This concept has been maintained up to now.
New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In the year that has passed from the eigtth seminar, we lost some of the well-know
figures in the international energetic community. The greatest loss for us is the demise of Dr.
Preface
Fred Volk in September 2005; he was a very familiar person at international meetings and
cooperation.
Fred, who was know by his unselfish help given young scientific-research
This is the ninth seminar New Trends in Research of Energetic Materials and the eighth inworkers ternational
in the filed
ofthe
energetic
materials
wasDepartment
actively ofengaged
in Technology
the process of
one in
series organized
by theand
former
Theory and
of
Explosives,
now
the
Institute
of
Energetic
Materials
(IEM)
at
the
University
Parinvolving the Czech workplaces of this branch into cooperation within Europeanofstructures,
dubice.
The
original
purpose
of
these
meetings
was
to
teach
young
research
workers
how
to him
contributed significantly to moulding the character of our April meetings. We will miss
present their results in front of scientific audience. With regard to economic situation of the
as not only
excellent
scientist
but
a Soviet
very Union,
good nofriend.
loss
middlean
European
countries,
Balkan
andalso
former
fee was Another
and is asked
fromfor
the us in
December
2005 wasThis
theconcept
death of
of Czech
of Blasting Techniques and
participants.
haspresident
been maintained
up toAssociation
now.
Pyrotechnics and external IEM assistant professor Dr. Vclav Tamchyna. In his case, IEM
the year that has passed from the eigtth seminar, we lost some of the well-know figures in
has lost In
direct
connection with the blaster community of the Czech Republic and Europe as
the international energetic community. The greatest loss for us is the demise of Dr. Fred Volk
well. It need
not
be emphasised
losses
areatappreciable
it will
some time to
in September
2005; he was athat
verythese
familiar
person
international and
meetings
andtake
cooperation.
know
by his unselfish
help given
young continue,
scientific-research
workers
in the
field of is its
get overFred,
themwho
at was
least
partially.
However,
life must
and our
ninth
Seminar
energetic materials and was actively engaged in the process of involving the Czech workplaces of
constituent.
this branch into cooperation within European structures, contributed significantly to moulding

the character
of ourseminars
April meetings.
will miss
as not only
an excellent
scientist but
also
Realization
of our
wouldWenever
behim
possible
without
the generous
support
of
a very good friend. Another loss for us in December 2005 was the death of president of Czech
many sponsoring institution and individuals. In the case of this ninth Seminar a very
Association of Blasting Techniques and Pyrotechnics and external IEM assistant professor Dr.
significant
financial
support
provided
by with
the the
Schlumberger
Reservoir
Vclav
Tamchyna.
In hishas
case,been
IEM regularly
has lost direct
connection
blaster community
Completions
theItEuropean
of Aerospace
Research
&
of the from
Czech Rosharon
Republic and(Texas),
Europe asby
well.
need not beOffice
emphasised
that these losses
are
appreciable
and
it will in
take
some time
get overDetonator
them at least
partially.
However,
mustSafety
Development
of the
USAF
London,
bytoAustin
Comp.
(Vsetn),
bylife
Indet
ninth Seminar is its constituent.
Systems continue,
Comp.and
(a our
member
of Nippon Kayaku group in Vsetn), by Explosia, Ltd.,
Pardubice,
by the ofCzech
company
OZM
in Hrochv
and by
the of
BORGATA,
Realization
our seminars
would
never Ltd.
be possible
without Tnec
the generous
support
many
sponsoring
institution
and individuals.
In the casesupport
of this ninth
a very significant
fiLtd., Prague.
Almost
traditionally,
some financial
has Seminar
been provided
by Dr. Oldrich
nancial
support
has
been
regularly
provided
by
the
Schlumberger
Technology
Corporation
from
Machacek, president of the Universal Tech. Corporation, Dallas, who (as chance would have
Rosharon (Texas), by the European Office of Aerospace Research & Development of the USAF
it) is onein of
the first
PhD Detonator
graduatesComp.
from (Vsetn),
IEM. Also
ourSafety
Faculty
of Chemical
Technology
London,
by Austin
by Indet
Systems
Comp. (a member
markedlyof Nippon
contributed
the costs
of thisLtd.,
Seminar.
The
efficient
help inOZM
ensuring
Kayakutowards
group in Vsetn),
by Explosia,
Pardubice,
by the
Czech company
Hrochv Tnec
and by
Prague.from
Almost
some fi- and
smooth Ltd.
and insuccessful
course
of the
theBORGATA,
meeting Ltd.,
obtained
all traditionally,
these institutions
nancial support has been provided by Dr. Oldrich Machacek, president of the Universal Tech.
individuals
is gratefully acknowledged. We greatly appreciate it that thanks to this
Corporation, Dallas, who (as chance would have it) is one of the first Ph.D. graduates from
sponsoring
all
theour
above-mentioned
specifics
of the
previous
seminars
canthe
be costs
maintained.
IEM. Also
Faculty of Chemical
Technology
markedly
contributed
towards
of this
Seminar. The efficient help in ensuring smooth and successful course of the meeting obtained

Finally,
I wish to thank the members of the Scientific Committee, the authors of all the
from all these institutions and individuals is gratefully acknowledged. We greatly appreciate it
seminar that
papers
and,
but not least,
the participants
seminar,seminars
for itscan
success
thanks
to last
this sponsoring
all theyou,
above-mentioned
specificsofofthis
the previous
maintained.
and its be
influence
on the continued success and growth of all future meetings at our
University of young people and university teachers working in the field of all kinds of
Finally, I wish to thank the members of the Scientific Committee, the authors of all the seminar
energeticpapers
materials.
and, last but not least, you, the participants of this seminar, for its success and its
influence on the continued success and growth of all future meetings at our University of young
people and university teachers working in the field of all kinds of energetic materials.

Pardubice, March 25th, 2006

Pardubice, March 25, 2006

Svatopluk
eman
SvatoplukZZeman

Invited lectures

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

EUROPEAN GOVERNMENTAL COLLABORATION IN


ENERGETICS RESEARCH THE FUTURE
A S Cumming
Dstl Fort Halstead
Sevenoaks
Kent
UK
ascumming@dstl.gov.uk

Abstract
The level of joint working in Energetics has increased significantly in the last few
decades. It has developed from informal meetings and discussions to formal
intergovernmental agreements on joint programmes. The need to and benefits of
working together have been obvious and this has produced greater interactions. This is
changing again with the setting up of the European Defence Agency, and while there is
some confusion over what this means, it will provide new routes to collaboration. This
paper will describe some of the achievements as well as discussing the possibilities for
the future.

INTRODUCTION
In the mid-eighties formal collaboration was limited. There were links, usually for
information exchange, and scientists met at conferences where they presented on work in
progress and discussed common problems. There were formal links within NATO and also
some other agreements in place. One reason for this was that the most nations were able to
sustain their own needs through their own investment. The needs differed as did the
investments, but there was no awareness of major limitations.
At the same time equipment was becoming increasingly complex and expensive so that
while is was possible to have national main battle tanks for example in Germany, France the
UK and the US, continuing that development was becoming more difficult to do. This
increasingly led to joint projects and the same time policy looked towards the privatisation of
defence manufacturing. The UK privatised the Ordnance Factories in the mid eighties and
also turned the research establishments into an agency aimed at improving efficiency and
providing value for money. One aim was to assist in the transfer of technology for civil use.
It was realised that better value for many could be gained by working with partners on
joint research programmes, and so existing links were developed with Memoranda of
1

Adam Cumming

11

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Understanding to do much more than exchange information but to include staff exchanges
and joint programmes with financial and delivery commitment.
However the existing links did not meet all the perceived needs, and a new Group the
Independent European Programme Group, IEPG, was set up to develop joint research
programmes as well as trying to harmonise what would now be called capability
requirements.
Energetics played a major part in this with the formation of Technical Area (TA) 12,
looking at gun propellants, TA25 with three programmes covering extrusion of energetic
compositions, pressed shaped charges and synthesis and characterisation of new materials.
Each programme required a separate MOU leading to delays and complications so that
attempts were made to provide a more flexible route to collaboration in Europe. This led to
the CEPAs and to WEAG.
WEAG and the CEPAs
The first group of CEPAs did not include Energetics. The made use of the EUCLID
MOU, which was based round competition and the assumption of industrial participation. It
had the draw back of not allowing closed groups and making the participation of government
agencies more difficult.
However when CEPA 14 was started it was possible to devise and begin programmes on
new Rocket Propellant Technology and Insensitive Munition Technology. While these were
started two further developments made collaboration easier. First the creation of the
Research Cell in Brussels to support the activity of each CEPA and also to place central
contracts for the agreed programmes, and second the signature of the THALES MOU aimed
at Joint Programmes amongst government laboratories. This was rapidly followed by the
SOCRATE MOU allowing other nations, non-NATO members such as Sweden to take part.
A third factor was the introduction of the EUROFINDER process where industries were
encouraged to propose programmes for inclusion. There were partially industrially funded
and closed, being limited to the proposing nations only.
Finally the EUROPA MOU was developed and signed, streamlining the processes and
providing for closed programmes amongst groups of members. This provided the final and
most effective set of tools for use in collaboration.
Within CEPA 14 this provided a spur to new work and the development of joint
understanding and information exchange for the development of joint programmes. For
nearly ten years the CEPA acted as a focus a clearing house for most of European Energetics
research. One effect was the reduction of activity within many of the older agreements,
though specific items might still be developed by those methods.
12

European Govermental Collaboration in Energetics Research The Future

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The main problem was always the limited amount of collaboration possible when there
was no harmonisation in requirements. That harmonisation might include capability need,
finance, timescales and other factors. The more nations involved the harder is remains to
agree programmes beyond a limited level. This is illustrated by the Venn diagram below
where the degree of total overlap is very small. (Fig 1)
The limitations in the WEAG process made it harder to improve on this, and it meant a
practical limit of about 5% of the available research funding. The only effective way to
improve on this is by improving the coordination of the capability requirements so that the
factors listed above are dealt with at an early stage. If this is done then the level of mutual
involvement will increase significantly,

R&T Priorities Multilateral

Fig. 1.

ACHIEVEMENTS OF CEPA 14
Despite all the problems outlined above the CEPA achieved a significant amount and
even though it has now officially ended, that achievement continues through its legacy.
Two programmes were mentioned above. The IM programme added significantly to the
European capability to manage Munition vulnerability. Within the UK it provided the
technology to ensure that a new and major system met our IM Policy requirements. In the
same the way the programme on LOVA ammunition assisted in the decision to reinvest in
such technology with the UK.
The CEPA has provided the means of holding Workshops including the European
Survivability Workshop; Workshops on Environmental Effects on Munitions of Out of Area
Operations, High Energy Density Materials, The Effect of Shock on Energetics; Hybrid
Missile Propulsion and others. Many of these have led to proposals for new joint research
programmes.

Adam Cumming

13

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In addition there are running programmes on Flares, the development of a European


Virtual Underwater Testing Centre for munitions, Fragment and Bullet attack on Munitions,
again supporting IM, Particle Size optimisation and of course the Synthesis of Nitro
compounds, involving the Czech Republic. All these will continue and older ones have
already proposed new programmes to continue the relevant research. This is the legacy of
the CEPA.
The Future
The European Defence Agency is in its infancy, and has yet to develop it processes. It is
an intergovernmental organisation, not a supra-governmental one, and exists to support the
member states. There remain issues over the dealing with nations who are signatories of
EUROPA but not in the EU, and hence EDA (Norway, Turkey and Romania) or have
refused to join the EDA (Denmark). The EDA has inherited most of the existing WEAG
programme and is attempting to produce mechanisms to develop this. The major mechanism
is by means of Capability/Technology Areas or CapTechs, with national coordinators
(CNCs) supported by an EDA Moderator and by national and industrial experts.
The EDA will have very limited central funding for the foreseeable future and it
therefore dependent on the requirements and policies of the member states. One of its main
roles is to begin the harmonisation of national requirements so that more extensive and
effective collaboration can be undertaken. This is not a simple task, and is made more
complex by the variation in money available for R&T. The relative levels are shown in
Figure two below.

5000

MUSD

97
%

77
%
4000

3000

2000

1000

0
UK

FR

GE

IT

SP

SW

PL

NL

TU

GR

NO

CZ

HU

AU

BE

FI

PO

DK

LU

Fig. 2. R& Figure 2 Expenditure Across WEAG

14

European Govermental Collaboration in Energetics Research The Future

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The EUROPA MOU still exists and will continue to exist having been chosen for certain
other activities outside the EDA. It requires some modification but that is being done. It
may be extended to cover other non-EDA countries.
There are, therefore mechanisms for collaboration in Europe, at least as good as those
that formerly existed. Within the Energetics community, those involved with CEPA 14 have
proposed an Expert Group covering the same area and supporting both EDA CapTech
activities as well as those outside it. How this develops depends on the way the various
needs. However, for most nations it is agreed that the EDA will be the prime method of
operation, but other methods may be employed if they are more appropriate.
In short it will be possible to work as closely together and opportunities should be
sought and developed. The mechanism employed to deliver the result may vary but does
exist and the processes will continue to develop to meet the need to provide results for the
national MODs. It may look complex and uncertain, but the same was true when WEAG
was created and that grew and developed to meet real requirements. For that reason it is
worth looking back to see how we arrived where we now are, and to reassure ourselves that
there has been real progress. No mechanism is perfect or remains useful in all
circumstances. It must evolve to meet new needs or be replaced by methods that do meet
requirements, learning from the experience.
In this changing Europe there is no reason to suppose that these changes will not assist in
providing the defence capabilities we require.

Adam Cumming

15

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

NEW DIAGNOSTIC TECHNIQUES IN BLAST WAVES

Prof. Dr. Manfred Held


TDW/EADS Schrobenhausen
E-Mail: manfred.held@eads.com
Tel.: 49-8252-996-345;

Fax: 49-8252-996-126

Abstract
The author has developed and used methods to measure the blast contour of not
spherical charges and the momentum distribution of antitank mines under different
ground conditions lying on the ground, levelled to the ground or buried in the ground.
Further he has defined a diagnostic method to measure the combined transferred
impulse of fragments and blast loads on steel-plates.

1.

IMPULSE LOAD IS DOMINANT

The pressure durations of the positive pressure history phase (Fig. 1) of detonating high
explosive charges in the near field at less than two (Z < R/W1/3) are shorter than 2 ms (Fig.
2). In such short times the maximum pressure is not responsible for any damage, because the
movement of the target is minimal in this short time interval. But the transferred momentum
to the target structure can be compared to a ballistic pendulum. Therefore the transferred
momentum has to be measured. The maximum pressure has no relevance to the damage
caused by small charges history in the fireball or if the charge is lying in sand or is confined.
High frequency pressure gauges have also some prices. They can not be used in large
quantities for high angle resolutions, what is easily possible for the momentum gauges which
the author has used and is using as a diagnostic tool.

2.

DIAGNOSTIC WITH THE MOMENTUM METHOD

Momentum gauges are arranged on a steel band around the charge (Fig. 3). The heights
of the momentum gauges and of the high explosive charge has to be exactly levelled, so that
only horizontal forces are acting on the rectangular shaped gauges. At closer distances of 0,5
m and 0,75 m the gauges of 2,5 cm5 cm are oriented with the smaller surface of 12,5 cm to
the charge. At 0,5 m distance 60 steel momentum gauges are used with 490 g weight in the
semi-circle, which gives an angle resolution of 3, and at 0,75 m distance 90 aluminium

16

New Diagnostic Techniques in Blast Waves

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

momentum gauges with 172 g weight, which results in 2 angle resolution (Fig. 3). They
were supported by a 3 mm thick and 30 mm wide steelband, where the cross section to the
charge was sharpened with an angle of about 45.
At the 1,0 m and 1,5 m radii the same gauges are oriented with their quadratic cross
sections of 25 cm to the high explosive charge in the centre, which gives also the angle
resolutions of 3, respectively 2 at 1,0 m and 1,5 m radius. A typical picture of the test setup
with the cylindrical charge in the centre of the two half circles is shown in Fig. 4. By the
horizontally acting blast wave the momentum blocks are thrown radially away, where the
first impacts in the sand are measured on the levelled ground.
To get the transferred momentum the velocity of the momentum gauges has to be found
out. The velocity can be easily measured from their displacement D. The falling time t from
the height H is constant and is given by the equation t = 2 H / g , independent from the
horizontally transferred velocities.

For the height H of 1 m the falling time t is 0,45 sec (t =

2 1(m) / 9,81(m / sec ) ).

The trajectories of exact horizontally loaded momentum gauges with different initial
velocities are shown in Fig. 5. Therefore the horizontal displacement D is direct proportional
to the velocity v (Fig. 6).
This method allows to measure easily the transferred momentum M, respectively the
impulse of non-spherical charges with very high angle resolution and with very low
investment costs. Also the momentum distribution of confined charges can be very well
measured with this method.
It should be shortly summarised the by the author used blast wave terminologys with
the corresponding dimensions.
Pressure p = force/area = F/A

(1 kp/cm = 1 bar ~ 10 kgm/s/cm = 105 kgm/s/m = 105 Pa)


Impulse I, respectively momentum M

Impulse density

= mass x velocity = mv

(kgm/s)

= force x time

(kps)

= Ft

ID = I/A = mv/A = Ft/A

(1 kp s/cm ~ 10 kgm/s/cm = 105 kg m/s/m = 105 Pa s)


Scaled impulse density IDS = ID/W1/3

Manfred Held

kgm/s/(cm W1/3) = 105 Pa s/ kg1/3

17

New Trends in Research of Energetic Materials

3.

Czech Republic

2006

[Content]

TYPICALLY NEAR FIELD BLAST MOMENTUM RESULTS OF


NON-SPHERICAL CHARGES

The thrown distances of the momentum blocks on the ground are shown, as an example
in Fig. 7, where the comparison is made between plastic bonded charges of HMX charge
(left side) to RDX (right side). The impression is correct, that the differences for the 1 kg
high explosive charges are only marginal.
The results can be much better compared in the diagrams of the impulse density as a
function of the polar angle in the semi-circles, where 0 is the axis of the charge in
detonation direction, 90 exactly the radial load and 180 the axis of the charge opposite to
the detonation direction. In these diagrams the characteristics of the blast momentum of the
cylindrical charges with the length to diameter ratio of 2/1 are nearly identical as shown at
1,0 m (Fig. 8) and 1,5 m distance (Fig. 9). In the axial direction (1) surprisingly high values
are achieved, but only in a very narrow angle of less than 10. The minima (2) from 15 to
30 are lying beneath 0,01 IDS. Values beneath 0,01 IDS are not taken into account, because
no damage will be more caused. By the interaction of the axial and radial wave a small
maximum (3) in the range between 40 and 50 is generated. It follows again a minimum (4)
in the angle around 65. Then the impulse density rises exponentially (linear in the log
diagram) to the radial maximum (5) at the angle of 85 90. The rearwards first minima (6)
are typically around 120, followed by a small maximum (7) at 130 and the next minima (8)
at around 155 and finally the rearward impulses (9) against the detonation directions.

4.

BLAST MOMENTA OF DIFFERENT CHARGE COMPOSITIONS

The direct comparison of the charges KS 33 (90% HMX and 10% Binder) and KS 13
(88% RDX and 12% Binder) shows, that the HMX charge just presents a little higher
impulse densities in some regions. Generally the advantage of the more powerful HMX
charge in the blast impulse output seems to be only very small. It has to be mentioned, that
the logarithmic scale, which is necessary for the large differences of the impulse densities in
the different directions is strongly levelling out small differences.
As standard charge was used the pasty flexible or formable charge SK DM 12 B 1,
which consists of 87% PETN and 13% lubricant (Fig. 10). At 1 m distance the cast
TNT/RDX 35/65 charge seems very similar, if one takes out some random noises of
individual momentum gauges. The comparison at 1,5 m gives a slight advantage for the
TNT/RDX 35/65 charge.
A comparison was also made with aluminised high explosive charges with different
binder systems. Both charges had 67% RDX; 28% aluminium and 15% binder. The KS 22a
charge had an HTPP and the KS 22S a silicon binder type. The blast contours at 1,0 m and
1,5 m seems to show that the silicon binder system gives slightly higher impulse density
values (Fig. 11).
The composition B type TNT/RDX 35/65 charge is also compared to a plastic bonded
aluminised high explosive charge from type KS 22 A, which contains a RDX/Al/B 67/28/15

18

New Diagnostic Techniques in Blast Waves

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

(HTPP) composition. At 1 m distance (Fig. 12) the impulse densities look very similar, but
at 1,5 m distance the aluminised charge has definitely less impulse density values.
A very interesting aspect was found on 4 tests with not aluminised and aluminised high
explosive charges, once containing 90% RDX and 10% binder and the other type 70% RDX,
20% aluminium and also 10% binder. If the charges are initiated by a booster with 37 mm
diameter, the impulses were generally lower compared to a stronger initiation. In 1 m
distance the aluminised charge looks a little beneath the not aluminised charge (Fig. 13) but
stronger or more pronounced difference exists at 1,5 m distance. If the charges are strongly
initiated by a 10 mm thick disc of PBX N9 over the full end surface of 75 mm, the
aluminised RH 29 charge looks a bit better at 1 m distance, but definitely better at 1,5 m
distance (Fig. 14).
For cylindrical charges the impulse densities are not regularly decreasing with increasing
distance (Fig. 15 and Fig. 16), as the simple impulse equation for spheres would predict. The
narrow axial blast fan in detonation direction has to be specially treated. The bridge wave
sectors seem to be more filled up on costs of the axial and radial blast waves at larger
distances. The radial blast wave seems to be stronger reduced than linear. This are few
general remarks on the distance dependence of the blast impulse and finally for the damage
behaviour.

5.

CUBICAL CHARGE AND CHARGES WITH DIFFERENT L/D


RATIOS
Cubical charges

1 kg heavy cubical charges made from the pasty high explosive type Seismoplast or SK
DM 12 B1 are also fired under two test conditions in half circles of 1,0 m and 1,5 m
distance. Once the cube was initiated on the flat side in the centre and one at the corner point
oriented in the semi-circle in such a way, that in the first case the flat areas were oriented to
0, 90 and 180 (reverse direction), respectively to the diagonal directions.
Corresponding to the two different orientations, the momentum contours are remarkably
different (Fig. 17 and Fig. 18). In the case 1 the maxima are oriented in the flat areas. By the
interactions of these strong blast waves smaller maxima are just build-in the diagonal
directions at 45 and 135. Extreme low impulse density values are registered between these
maxima (see Fig. 17). If the cube is oriented in the semicircles in a diagonal direction, then
no maxima are measured and generally a lower level of impulse density values can be
analysed (see Fig. 18).

Different L/D ratios


Here also the pasty, high explosive charge of 86% PETN and 14% lubricant were taken.
The 1 kg heavy charges are formed in cylinders of the L/D ratios of 0,5, 1, 2 and 4. All these
charge Types are one end centrally initiated by a number 8 electric detonator (Fig. 19).

Manfred Held

19

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In all cases the specific impulse density has a sharp maximum around 0 or in the
forward axial direction at 0,5 m distance (Fig. 20). The contours are described in detail for
the L/D = 1 charge. The value falls steeply down around to 0,03 at around 30. The
interaction of the axial and radial blast waves give a small maximum with 0,07 in the bridge
wave around 40. These then follows a second minimum, lower than 0,02. From this very
low value the impulse rises to a shoulder at 78 with a value of 0,27. Then it increases
further to 0,4 at around 90. After this the pressure decreases in the rearward direction,
where the total contour is relatively smooth around 0,05 to 0,06 in the range of 135 to 165.
Then it rises to the rearward direction near the axis.
The contours look surprisingly very similar, still some differences exist. Some main
features are presented in the following 2 tables and in the Figures 21 and 22 at the distances
of 0,50 m and 0,75 m.

Table 0,50 m distance


Geometry
L/D

M radial
-1

-2

Cone angles for IDS 0,1

kg_m_s cm _kg

-1/3

Forward F

Radial Ra

Rearward R

0,5

0,16

2 x 37

30

2 x 43

0,40

2 x 20

50

2 x 10

0,40

2 x 20

38

2 x 15

0,20

2 x 25

52

2 x 40

These values are influenced by the scatter of the data, but they show the clear general
trends. The radial Mradial impulse is the largest for the L/D = 1 and L/D = 2 ratios with 0,4 at
0,5 m distance, as expected. The larger peak value in the case of 0,75 m distance for the L/D
= 1 is a little surprising.

Table 0,75 distance


Geometry
L/D

20

M radial
-1

-2

kg_m_s cm _kg

Cone angles for IDS 0,1


-1/3

Forward F

Radial Ra

Rearward R

0,5

0,16

2 x 29

12

2 x 33

0,40

2 x 10

27

0,25

2 x 7

27

2 x 12

0,18

2 x 25

13

2 x 20

New Diagnostic Techniques in Blast Waves

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

As expected, the forward and rearward cone angle has the largest values for disc shaped
charge with the L/D = 0,5. Surprising is, that the second largest angles in these directions are
achieved by the charge with the L/D = 4.
The radial blast angle around the 90 direction is wider for the L/D = 1 and L/D = 2. The
large opening angle of 52 is not very well understood for the L/D = 4 at 0,5 m distance up
to now.
For high radial blast loads the shapes L/D = 1 and L/D = 2 are the favourites, which have
also large radial cone angles.

6.

CHARGES WITH DIFFERENT CONFINEMENTS

Also the pasty compound, called Seismoplast for commercial applications or SK DM


12 B1 in military use, with L/D ratios of 2:1, respectively 1:1, with 75 mm diameter and
150 mm length, respectively 95 mm diameter and 95 mm length were used for the high
explosive charge. The charge weight was constantly 1 kg. The charges were fired unconfined
and confined with steel casings of 1 mm, 3 mm and 10 mm (Fig. 23). The charges were
initiated in the axis at one end-surface (Fig. 24).
The fragments of casings naturally hit these frames, which were distorted after each test
with confined charges (Fig. 25). A number of momentum gauges in forward and rearward
axial directions, and in the radial direction were also hit by fragments (Fig. 26).
In figure 27 the specific impulse densities of the confined cylindrical charges are
compared to bare cylindrical high explosive charges in the distance of 0,50 m and the
geometry L/D 2:1.
These detailed structures of the blast wave in the near field from the forward direction
with the narrow axial blast fan, the large minimum in the forward bridge-wave direction, the
maximum of the radially expanding blast wave to the second minimum in rearward bridgewave direction are already discussed.
If casings of steel are attached, the blast contour is changed by the additional fragment
impacts on the momentum gauges, which was expected.
This is caused in the axial forward and rearward directions by the endplates fragments of
the charge, and by the fragments of the cylindrical confinement, expanding radially in
perpendicular direction to the charge axis. By the Taylor angle they are pushed a little in the

Manfred Held

21

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

detonation direction which means in the range of 80. The transferred momenta in these
areas covered by fragment covered areas are much higher by the combined momenta of the
fragment impacts and the blast wave, although the blast wave impulses are partially reduced
in these areas (Held 2003b). But the blast waves are continuously increasing, in the bridge
wave fields, as shown by the impulse density diagrams from 1 mm to 3 mm and 10 mm
confinement.
The reason for this is, that the products of explosives cannot expand fast axially and
radially due to the confinements, as on unconfined charges, because then the bridge wave
gaps are much more filled up with the expanding reaction products. Therefore the blast loads
are strongly increasing with confinements in these directions (Fig. 28).
In summary: confined cylindrical charges are increasing the momenta in all directions by
the fragments in axial and radial, and by the enhanced blast wave in the diagonal directions.
This investigation shows that the damage potential of confined charges in free air is
remarkably increased by confinements, because in the fragment zones the momenta are
increased by the larger loads, caused by the fragments, and the bridgewave areas are filled
up by factors of 5 to 10 to higher blast momenta, compared to unconfined charges.

7.

COMBINED FRAGMENT AND BLAST LOAD

Momentum sleds were used to measure the combined blast and fragment loads. For all
described tests cylindrical C4 charges of 60 mm diameter and 290 mm length were used,
giving an explosive weight of about 1,2 kg. These were surrounded by steel casings with
three different thicknesses 2,5 mm, 5 mm and 10 mm. For reference of the airblast load
alone, a bare charge in a very thin plastic tube was also used. The radially transferred
momenta of the blast and fragments were measured with sleds. The front of each sled
consisted of a 15 mm thick mild steel plate, 500 mm in height and 600 mm in width. The
sleds were fixed on roller bearings. The friction of the sleds was found to be less than 10 N.
The velocities of the 51,5 kg sleds were measured with electronic displacement transducers
(wire position sensors) and with video cameras with 25 frames/sec, but 0,1 ms exposure
times on an extended rod with distance marks. The vertical positions of the target plates with
respect to the charge positions were arranged according to the Taylor angles of the fragment
trajectories ( = arc sin Fr/2d) to ensure maximum fragment impacts. For each casing
thickness, two tests were conducted, one with two sleds at 1,0 m and 1,5 m and the second at
2,0 m and at 3,0 m stand-off distances (Fig. 29). The sled rails were mounted on a heavy
concrete block, as shown in Fig. 30. The front side of the concrete block and the edge of the
tracks were protected by a massive steel plate (Fig. 31).
The measured velocities of the bare and the 3 different cased charges at the 4 distances
are shown in Table1:

22

New Diagnostic Techniques in Blast Waves

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 1. Measured initial sled velocities by the blast and fragment loads
Casing

Velocities in m/s at the four investigated distances

thickness

1,0 m

1,5 m

2,0 m

3,0 m

0 mm

5,5

2,65

1,2

0,36

2,5 mm

6,4

3,6

2,2

0,93

5,0 mm

7,1

4,6

3,2

1,8

10,0 mm

10,4

6,8

5,0

2,5

These values have to be multiplied with the sled weights of 51,5 kg to get the achieved
momenta in kgm/s or Ns of the combined blast and fragment loads (Table 2).

Table 2. Measured sled momenta by the combined blast and fragment loads
Casing

Momenta in kg m/s or N s at the four distances

thickness

1,0 m

1,5 m

2,0 m

3,0 m

0 mm

283

137

62

19

2,5 mm

330

185

113

48

5,0 mm

366

237

145

93

10,0 mm

536

350

258

129

Individual blast and fragment loads


The fragment impulses can be calculated by using the fragment velocities, which are
predicted by the Gurney equation with the Gurney constant 2 E of 2165 m/s for the used
C4 charges (5), and the fragment masses in the azimuth hit sector which is covered by the
600 mm wide plates. It is assumed that at 1 m distance the 500 mm vertical height of the
target plate captures 100% of the vertical spread of the fragments. Then linear reduction of
the fragment hits is assumed with increasing distance 0% at 1 m, 2,5% at 1,5 m, 5% at 2 m
and 10% at 3 m distance. The initial fragment velocities are taken as constant along the four
short distances between 1 m and 3 m.
Table 3 gives the maximum mass of the fragments that can impact the plates assuming
100% hit the 600 mm wide target plates. This is calculated with the following equation:
mfrag.sector = (De - Di) /4L Steel(arc tan 0,3/d)2/360
mfrag.sector = (De - 6) /4297,85(arc tan 0,3/d)2/360
8

Manfred Held

23

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

where De and Di are the external and internal diameters of the steel casings, L their
length of 29 cm, Steel the steel density and d the distance of the plates to the charge.

Table 3. Fragment masses in grams impacting the 600 mm wide target plates
Casing

De

(mm)

(cm)

= c/m

1,5

2,5

6,5

0,93

104

70

53

35

7,0

1,94

2156

146

110

74

10

8,0

4,17

464

316

237

159

Distance d (m)

Table 4. Calculated fragment momenta in the given azimuth angles on the 600 mm
wide target plates with 100% fragment hits at 1 m, 97,5% at 1,5 m, 95% at
2 m and 90% at 3 m distance.
Casing

= m/c

(mm)

VFragment

Momenta (kgm/s or Ns)

Distance

1,0 m

1,5 m

2,0 m

3,0 m

Hit %

100%

97,5%

95%

90%

2,5

0,93

1805 m/s

188

123

91

57

1,94

1386 m/s

299

196

145

92

10

4,17

1002 m/s

465

309

226

143

Approximate values of the blast loads alone can be calculated by subtracting the
calculated fragment loads in Table 4 from the measured momenta in Table 2. These values
are summarised in Table 5.

Table 5. Estimated contribution of blast load alone in kgm/s or Ns


Casing

24

Differences between the measured and calculated fragment momenta

Thickness

1,0 m

1,5 m

2,0 m

3,0 m

0 mm

283

137

62

19

2,5 mm

330188 = 142

185123 = 62

117-91 = 26

48-57 = -9

5,0 mm

366-299 = 67

237-196 = 41

145-145 = 0

95-92 = +3

10 mm

536-465 = 71

350-309 = 41

258-226 = 32

129-143 = -14

New Diagnostic Techniques in Blast Waves

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

These are the blast contributions of the momenta in kgm/s or Ns. These values can be
also expressed in percentages to the blast loads of the corresponding (table 6) bare charges.

Table 6. Fraction of the blast loads of the confined charges to the bare charge
Casing

Ratio of the blast loads of confined to bare charges

Thickness

1,0 m

1,5 m

2,0 m

3,0 m

2,5 mm

142/283 = 0,50

62/137 = 0,45

26/62 = 0,42

(-9/19 = -0,47)

5 mm

67/283 = 0,24

41/137 = 0,30

0/62 = 0

(3/19 = +0,16)

10 mm

71/283 = 0,25

41/137 = 0,30

32/62 = 0,52

(-14/19 = -0,74)

From Table 6 the conclusion can be drawn that a greater portion of momentum is
transferred by the fragment impacts of encased charges. Furthermore these data demonstrate
that, if a combined blast and fragment load impacts a target, the blast part in radial direction
is more reduced in the case of more heavily confined charges.

Radial blast reduction by confinement


A number of equations exists for predicting the reduction of the blast effect by radial
confinements of the detonating high explosive charge. One of these is the well known FANO
equation which is described for the positive impulse as an extension of the Gurney formula.
W/W =

0,2 + 0,8/(1+2 m/c)

From the former Naval Ordnance Labortaory NOL the following equation was given for
the positive impulse of the positive phase for confined charges:
W/W =

(1+m/c(1- m ))/(1+m/c)

where m is equal to m/c, if m/c<1 and 1, if m/c>1.


The reflected specific impulse density can be expressed with the following equation:
IDS,refl/W1/3 = CW0,84/R0,84
where C is a constant, W the charge weight and R the distance:
IDS,refl = CW0,84W0,333/R0,84 = CW1,173R-0,84
A reduction factor F, generally called casing factor, has to be used for the charge weight
W if the charges are confined. Therefore the above equation can be written in the following
form:
IDS,refl = C(FW)1,173/R1,25

10

Manfred Held

25

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

This means the impulse values have to be corrected by the casing factors F1,173 for the
different confinements. In the following Table 7 the calculated casing factors, using F alone
and their F1,173 values, are summarised for IDS,refl with the confinements for the above
equations.

Table 7. Reduction factors of the radial momenta without and with their exponential
values
Equation/Thickness

2,5 mm / = 0,93

5 mm / = 1,94

10 mm / = 4,17

FANO

0,480 / 0,422

0,364 / 0,302

0,286 / 0,320

NOL Impulse

0,552 / 0,498

0,344 / 0,286

0,193 / 0,146

These values are independent of the distance between the charge and the target. Reasonable
experimental values were found only in the distances of 1,0 m and 1,5 m. Therefore these
values are only used for the prediction of the blast load alone in radial direction of
cylindrical charges with the FANO and NOL impulse load reductions with different
confinements in Fig. 32.
The experimentally achieved blast reduction factors by confinements are only rough
indications. But they give the same trend. Shortcomings in these considerations are the
assumptions of the accuracy of the calculated fragment velocities and also the percentages of
fragment hit in the elevation directions spalling fragments on the end faces of the charge
may have angles which are not captured by the vertical heights of the target plates.

Combined fragment and blast load


In an EMI report [6] the specific impulse densities are summarised for blast and
fragment loads of cylindrical confined charges with different m/c ratios of 1, 2 and 4 against
perpendicular targets in the dimensions kgm/skg-1/3m-2 (Fig. 6). These lines solid can
be expressed with the following exponential equations:
bare charge:

IDS,bare[6] = 370(R/W1/3)-1

confined charge:

IDS,conf. [6] = 1500(R/W1/3)-1 0,6

In this diagram the results achieved in the momentum sled tests are added as dashed
lines. the values, also expressed in the dimensions kgm/skg-1/3m-2 are summarised in Table
8. They give similar trends in the log log diagram although the values are generally lower
than the published data in [6].

26

New Diagnostic Techniques in Blast Waves

11

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 8. Specific impulse densities of the combined blast and fragment loads in
kgm/skg1/3m-2 of the described test series
Casing

= m/c

Distances
1,0 m

1,5 m

2,0 m

3,0 m

0 mm

888

430

194

60

2,5 mm

0,93

1035

580

354

151

5 mm

1,94

1148

783

455

292

10 mm

4,17

1681

1098

809

405

These data can be described with the following equations, which have generally the
same trends.
370(R/W1/3)-1,25

bare charge:

IDS,bare[6] =

confined charge:

IDS,conf. [6] = 1010(R/W1/3)-1,25 0,33

The impulse density needs a selection of target area. This can lead to some problems. If
a larger area with less dense fragment hits is selected, then lower values will be calculated
for the specific impulse density.
With the sled diagnostic tests the combined effects of blast and fragment loads were
measured for m/c ratios of about 1, 2 and 4 and compared to bare cylindrical charges in
distances form 1 m to 3 m or in scaled distances Z of about 1 to 3. The momenta were
measured with a new diagnostic technique, namely with the involved velocity of momentum
sleds, on which the target plates were arranged perpendicularly. The measured results
generally fit the data of the reflected impulse of bare charges referenced in the literature and
also approximately the casing factors of confined charges, using the FANO equation.

8.

BLAST CONTOURS AT MEDIUM DISTANCES

To get the blast contour at scaled distances Z of 3 and 6, tests are performed for the 1 kg
charges at 3 m and 6 m radius. From the impulse density over the scaled distance diagram
was predicted impulse density values around 0,02 ID and 0,008 ID (Fig. 34).
To get displacements of a few meters wooden momentum gauges were selected (Fig.
35). In the first test it was fired a 1 kg sphere, made from Seismoplast. The momentum
gauges, arranged at 3 m radius, are laying in less than 0,5 m distance, instead of more than 3
m (Fig. 36). In 6 m radius they were still partially lying on the holding fixtures. The
achieved results of impulse densities in 0,5, 3 and 6 m compared to the expected values are
shown in Fig. 37 for 1 kg charge.

12

Manfred Held

27

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The momentum gauges are loaded by the positive and also the negative phase of the
pressure history (see Fig. 1). People are normally not taken into account the negative
pressure phase. The author has measured the pressure and Impulse history at 3 m and 4 m
distance with pressure transducer and has found out that the total impulse over the integrated
time of 30 ms has given only a very small value (Fig. 38 and Fig. 39). It is rarely mentioned
in the literature ([..] and [..]) that the resulting impulse at larger distances is remarkably
reduced by the succeeding wave (Fig. 40).
Beside the 1 kg charges also 2,83 kg and 8 kg heavy spheres of Seismoplast were fired.
The analyzed impulse density values over the polar angle are shown in Fig. 41. The impulse
density values over the scaled distances - which means the distances have to be divided by
the cubic root of the charge weights and the impulse density by the 2/3 root of the charge
weight are presented in Fig. 42.
The impulse density contours for a 1 kg Seismoplast charge of a L/D ratio 2/1 from
0,5 m until 4 m distance are presented in Fig. 43. It seems that the axial load in detonation
direction is constant form 0,5 to 1,5 m distance and is remarkably reduced from 3 m on. The
blast loads in the bridge wave direction around 45 have a similar behaviour and looks
constant between 0,5 m and 1,5 m distances. From 4 m distance on the blast contour is more
or less levelled out.
All the tests are done in open air with sometimes small wind conditions where such
influences cannot be fully neglected. For better controlled conditions, such tests should be
done in large enough buildings where all such details can be better controlled.

9.

GHOST EVENT

Video pictures of the detonation event of cast TNT-charge with 1 kg weight and L/D
ratio of 1 has given a surprising phenomena. At first the author has thought, that this is an
optical reflection in the camera from the bright detonation light (Fig. 44). After some
discussion of this surprising effect with experts of the proofing ground Meppen [..] he has
got the picture of a 100 kg cast TNT charge with also a L/D ratio of 1 (Fig. 45).
Very narrow blast effects with very high impulse densities happens generally for
cylindrical charges in the detonation directions up to scaled distances of 3. But the ghost
effect is in around 3 m altitude above the fireball of the cylindrical charge.
This surprising event happens in distances of 20 to 30 times charge diameter above the
charge.
Reasonable explanation was not found up to now and is an open point for a research and
some investigations
28

New Diagnostic Techniques in Blast Waves

13

New Trends in Research of Energetic Materials

10.

Czech Republic

2006

[Content]

DIAGNOSTIC OF THE MOMENTUM DISTRIBUTION OF ANTITANK MINES

In the very near field of detonating high explosive charges scaled distances beneath
one Z < 1 - , again the damage is only caused by the momentum transferred to the target.
Therefore it is sufficient to measure directly the momentum and not the pressure time history
which is at least very difficult, if not really impossible in a mostly very nasty environment.
The pressure duration of spherical charges of a few kilograms weight in a distance of around
0,5 m is in the range of 0,1 ms [6]. Therefore, the blast wave only induces the movement to
the target which reacts with frequencies of fractures of seconds, similar to a ballistic
pendulum.
The author could use flash X-ray units which were installed in a Leopard 1 chassis at the
Proving Ground Meppen, Germany [10]. To level up the observation height to the flash Xray tube altitude concrete blocks were arranged to a quadrat, which was then filled up with
sand. The high explosive charge was layed up in the middle (Fig. 46). 500 mm above the
levelled sand the momentum gauges were arranged. They were lying on two steel bands of 5
mm width and 50 mm height. The charge axis was oriented exactly on the symmetry planes.
To provide momentum gauges, plates with 25 mm width and 100 mm length and the
following heights of thicknesses were arranged: 150, 100, 70, 50, 25, 20, 10, 5 mm, flush
with the symmetry plane in blocks of always 4 with equal thickness (Fig. 47). The distance
between the flash X-ray tubes and the charge was about 4,0 m and between the charge and
the film 1,5 m (Fig. 48). This provides a magnification factor of 1.6. With two flash X-ray
units first a still picture of the arrangement was made on a separate film and then two
exposures of the dynamic event, typically 2 ms and 10 ms after the detonation of the charge
(Fig. 49).The delay generators were triggered by an ionization pin, a twisted isolated copper
wire in contact to the charge.
The flash X-ray pictures show the displacements of the momentum gauges at the two
time differences. The velocities are obtained by dividing the measured distances by the
magnification factor and then divided by the time difference.
A horizontal plate arrangement perpendicularly to the blast wave would not undergo
horizontal stretching forces by the blast load and therefore this vector component is not taken
into account.
The achieved impulse densities ID on the line of the momentum gauges are presented in
Fig. 50 as a function of the horizontal distance for the high explosive disks lying on the
ground, levelled to the ground and buried 100 mm deep.
These results are a cut to the circular load on a plate. These values have to be multiplied
by ring areas for the individual 25 mm wide momentum gauges (Fig. 51). The values of the
100 mm deep buried charge are a little jumping, caused by the sand fountains. The load on
circles as a function of their radii, are given by the sum of the individual ring zones (Fig. 52).
Over the analyzed distance of 700 mm, the charge levelled to the ground gives an about 50%
14

Manfred Held

29

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

and the charge 100 mm deep buried a 90% higher value than the charge laying on the
ground.
This test method gives reproducible results of the blast load by the diagnostic of the
achieved/measured momenta. With the high space resolution in this differential method
they are ideal for the calibration and verification of blast models with the charges arranged
on top of sand, levelled to the sand and buried 100 mm deep.
This method allows to measure the real loads of existing mines. Any change of the high
explosive charge type and the shape render different results, which was expected.

11.

DIAGNOSTIC OF THE PLATE ACCELERATION BY ANTI-TANK


MINES

The bulge of different thick steel plates under anti-tank mine blast loads should be
measured with high time and space resolution. For this purpose a special device was
designed to get the displacement, the velocity and the acceleration. The diagnostic tool had
to be fitted in the dome device of a special German mine-protection test rig (MinenschutzSprengtisch) (Fig. 53). A height of 380 mm was available. This diagnostic tool could not be
fixed in the upper part, because it had to be brought in by lifting up the test plate above the
ground.
The drawing of the displacement device is shown in Fig. 54. A 20 kg heavy cylinder
acting as a drag mass and was held via 3 shear pins of 2 mm diameter on the tripod. The
static cutting force of one shear pin is about 3 KN. In the dynamic case a force of 20 KN can
be expected for all three pins. This leads to an acceleration of the 20 Kg heavy mass to 1.000
m/s (a = F/m = 20.000 Kg m/s2/20 Kg = 1.000 m/s).
This force causes a movement of the drag mass after the equation s = 0,5 g t to 510-3
mm after 100 s time difference, what is a negligible value.
The drag mass had a hole of 20,2 mm diameter in the axis, in which a tube of 20 mm
outside and 16 mm inside diameter with a sharp knife edge on top was guided. This moving
up tube was cutting and shortening the wires of a flat band cable. This is tightened from the
side in the drag mass. The signals were recorded with transient recorders. A tube was
selected so that the weight is small and the cut flat band pieces can fall down inside and are
out of the way. Fig. 55 shows the displacement device assembled with the flat band cables,
standing on the testing plate, just before the plate is brought up into the test rig (Fig. 56).
An 8 kg heavy cylindrical cast TNT-charge, with 267 mm diameter and 89 mm length,
given in a so-called steel pit arrangement, was exactly positioned to the centre of the plate
and displacement device. It had 500 mm distance to the 25 mm thick high graded steel plate
(Fig. 57). At this test a 64 fold flat band cable was used, where only the first 16 contacts
where measured with transient recorders at 0,05 s time resolution. The copper bundles have
1,25 mm distances from one another. The shortening times as function of distance are shown
30

New Diagnostic Techniques in Blast Waves

15

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

in the diagram of Fig. 58. The spacing of 1,25 mm divided by the time differences between
two contacts gives their velocities, which are presented as a function of displacement
distance in Fig. 59. These velocities as a function of the shortening times are presented in
Fig. 60. But the velocities are not values at the individual end times. They represent always
the mean velocities between two contacts. Therefore they are drawn up over the mean time
of two contacts respectively (Fig. 60 left). The velocities rise linear with the time, which
means constant acceleration. Only the tangent to the first point cannot be very well
predicted, because it has, with 40 s in the starting phase, a long interval time. Nevertheless
the acceleration is constant with 920 000 m/s between 20 s and 140 s or 1,25 mm and
125 mm. From this time and distance on there cannot be made any further predictions with
the gained data.
The knife tube moves on with its maximum velocity because it is not mechanically
connected to the plate and has nearly no friction forces by cutting the wires and is therefore
not decelerated. The maximum velocity has achieved around 140 m/s after 12,5 mm bulge
and 140 s time delay.
This is an autonomous displacement device or acceleration gauge which allows to
measure accurately the increasing velocity and the magnitude of accelerations of the bulge
on a steel plate caused by a mine blast. Such devices can be also fitted in real tanks or
vehicles. It is independent, how the total system reacts on the mine blast, because the heavy
drag mass is virtually stable during the measuring time.

SUMMARY
The author tried to summarize some new, partially unusual diagnostic methods to get
answers to the damage levels by blast waves. Partially they are low costs in investment, but
partially a little labour intensive. The methods with momentum gauges around the energy
source has given new results on the momentum distribution of not spherical charges in the
near, respectively in the damage distances. Surprisingly he has found out, that the negative
phase, which has a great influence on the transferred momenta at medium distances is mostly
neglected.
The ghost events in the axial direction of a detonating cylindrical high explosive charges
is also a new task.
The diagnostic of anti-tank mine blast distribution, using flash X-ray equipments, allows
to calibrate the numerical codes.
The diagnostic of the acceleration of a bulging plate, caused by a detonating blast mine,
was also a special challenge.

16

Manfred Held

31

New Trends in Research of Energetic Materials

32

Czech Republic

2006

[Content]

New Diagnostic Techniques in Blast Waves

17

New Trends in Research of Energetic Materials

18

Manfred Held

Czech Republic

2006

[Content]

33

New Trends in Research of Energetic Materials

34

Czech Republic

2006

[Content]

New Diagnostic Techniques in Blast Waves

19

New Trends in Research of Energetic Materials

20

Manfred Held

Czech Republic

2006

[Content]

35

New Trends in Research of Energetic Materials

36

Czech Republic

2006

[Content]

New Diagnostic Techniques in Blast Waves

21

New Trends in Research of Energetic Materials

22

Manfred Held

Czech Republic

2006

[Content]

37

New Trends in Research of Energetic Materials

38

Czech Republic

2006

[Content]

New Diagnostic Techniques in Blast Waves

23

New Trends in Research of Energetic Materials

24

Manfred Held

Czech Republic

2006

[Content]

39

New Trends in Research of Energetic Materials

40

Czech Republic

2006

[Content]

New Diagnostic Techniques in Blast Waves

25

New Trends in Research of Energetic Materials

26

Manfred Held

Czech Republic

2006

[Content]

41

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THE THERMAL BEHAVIOUR OF BTAW, A HIGH NITROGEN FUEL


D.E.G. Jones*, K. Armstrong, T. Parekunnel, Q.S.M. Kwok
Canadian Explosives Research Laboratory, Natural Resources Canada
555 Booth Street, Ottawa, ON, Canada K1A 0G1
djones@nrcan.gc.ca
Abstract:
BTAw (bis-(1(2)H-tetrazole-5yl)-amine monohydrate) has recently been considered for use
as a low-smoke pyrotechnic fuel. There is relatively little information available in the
literature concerning the thermal properties of BTAw or its precursors. In the present
work, various thermoanalytical experiments were performed on BTAw and BTA (bis(1(2)H-tetrazole-5yl)-amine) in an effort to better characterize the thermal stability and
decomposition of these compounds.
Variable heating rate studies were carried out on BTAw samples in a helium atmosphere
using DSC and TG. Two steps were seen in the results: dehydration followed by
decomposition. Kinetic parameters were determined for both of these steps using a number
of methods. Experiments using simultaneous TG-DTA coupled with FTIR and MS were
performed on BTAw in both helium and dry and CO2 free air atmospheres, and evolved gas
analysis was used to determine the gaseous decomposition products. The thermal stability
of BTAw and BTA was examined using accelerating rate calorimetry (ARC).

INTRODUCTION
Typically, pyrotechnic formulations consist of carbonaceous fuels with metal salts added to
oxidize the fuel and colour the flame. These combinations tend to produce an abundance of
noxious smoke and ash, the result of partial combustion, and the generation of non-gaseous
products, particularly metal oxides. Chlorinated polymers were introduced for colour
enhancement but they still produce large amounts of smoke. High-nitrogen compounds were
also introduced with more success. Chavez and Hiskey have reported work on 3, 6-dihydrazinos-tetrazine (DHT) in which it was mixed with non-metallic oxidizers and 5 mass % coloring
agents, resulting in the production of little smoke and ash [1].
Other high-nitrogen compounds studied for use as low smoke pyrotechnic fuels are BTAw,
shown in Figure 1, and 5,5-bis-1H-tetrazole (BT). These compounds are diprotic acids that
react with basic amines to form mono- and di-aminated salts, and react with metal carbonates or
hydroxides to form metal salts. Chavez et al have recently reported that, by preparing BTAw
and BT with different ratios and compositions, it is possible to achieve greater control and
variability in burn rate and ignitability of high-nitrogen pyrotechnic formulations than with DHT
alone. Various characteristics of these two molecules, their hydrates and salts, were examined
by Chavez, including impact sensitivity, thermal behaviour by thermogravimetry (TG) as well
as the dehydration of BTAw to BTA [2]. It was noted that BTA exhibited greater impact
sensitivity than the hydrated form, BTAw. Nedelko et al. studied the kinetics and

42

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

decomposition products of BTA using thermogravimetry, volumetry, calorimetry, IRspectroscopy, and mass spectroscopy [3].
Earlier results from our laboratory showed that BTAw was relatively insensitive to impact
and friction, thermally stable at 75 C and not reactive when exposed to flames.
The present paper includes a detailed study of the current work on BTAw. Two variable
heating rate studies on BTAw, one using DSC and the other using TG, a study of the thermal
decomposition using simultaneous TG-DTA coupled with FTIR and MS, and an investigation of
the thermal stability of BTAw and BTA in air and Ar atmospheres using accelerating rate
calorimetry (ARC) are described in this paper.

EXPERIMENTAL
Samples of BTAw containing 20 % water were used as starting materials. These asreceived samples were used in the DSC heating rate study. Two methods were used to pre-dry
the BTAw. One method involved drying the BTAw in situ, i.e. in the ARC studies, for 3 hours
at 120 C in an open system. The other method consisted of drying BTAw in an oven at
temperatures of either 50 C, in order to remove excess water, or at 120 C to remove water of
hydration, for several hours. With pre-drying at 120 C, the end result was BTA. The pre-dried
sample at 50 C was used in the TG and TG-DTA-FTIR-MS studies. Samples obtained from
both methods were used in the ARC studies.
A TA Instrument (TAI) 2910 DSC was used for a variable heating rate study of BTAw
following ASTM standard test method E 698-01 [4]. Hermetic pin-hole Al pans (75 m hole)
containing about 0.25 mg of BTAw were heated from 30 C to 350 C at the rates of 0.5, 1, 2, 5,
8 and 10 C min-1. The DSC was calibrated for heat flow [5] and temperature [6]. An empty
pan was used as the reference.
The TG studies on BTAw were conducted using two different modules: the TAI 2950 and
the TAI Q500. The consistency of results between both modules was verified by performing
repeat measurements at the same heating rate. Sample sizes of 1 mg placed in aluminum pans
were run in helium at a flow rate of 100 mL min-1, split 60:40 between the furnace and the
balance. The samples were heated from 30 C to 350 C at rates of 1, 2, 5 and 10 C min-1. The
TG instruments were calibrated for mass using the procedures recommended by the
manufacturer and for temperature using the Curie point method as described in ASTM method
E1582 [7].
Details of the instruments used to obtain simultaneous TG-DTA-FTIR-MS data as well as
calibration information can be found in an earlier publication [8]. The TG-DTA-FTIR-MS data
were acquired simultaneously to study the thermal behaviour of BTAw and to identify the gases
that evolved while the BTAw sample was heated. Samples and reference (Pt foil) of 5 mg were
heated in alumina pans at a rate of 5 C min-1. Samples were run in helium and in dry and CO2free air atmospheres, purged at rates of 100 and 50 mL min-1, respectively.
The ARC measurements were performed on two instruments: one adiabatic calorimeter was
distributed by TIAX LLC (formerly known as Arthur D. Little Inc.) while the other was a
Thermal Hazard Technology (THT) instrument. Samples of about 0.5 g were placed in
lightweight spherical titanium vessels. The ARC experiments were started at an ambient

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

43

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

pressure of either air or argon. The standard ARC procedure of heat-wait-search (HWS) was
used [9,10] with a 5 C heat step.

RESULTS AND DISCUSSION


Dehydration
DSC Results
Typical thermal curves of BTAw obtained at various heating rates are shown in Figure 2.
The inset shows the entire range of the thermal curve obtained at 5 C min-1, including an
endotherm in the 125 to 150 C range, which is attributed to the dehydration of BTAw. The
average enthalpy of dehydration is 268 15 J g-1 or 43 2 kJ per mol of BTAw. Hydrogen
bonding energies are generally found to be between 8 and 42 kJ mol-1. This result suggests that
there is strong hydrogen bonding present between the water and BTA molecules.
Kinetic parameters for the dehydration step were determined from a series of variable
heating rate experiments following the ASTM method E698 [4]. The results were determined
from Figure 3, which is a plot of ln(T-2/K-1min-1) vs. 103 K/T for the peak temperatures (T) of
the dehydration endotherm at different heating rates (). The activation energy, E, was found to
be 197 5 kJ mol-1, indicating that the water molecule is strongly bound to BTA. Using this
value of E in conjunction with the calculated pre-exponential factor, ln(Z/min-1) = 46 1, the
rate constant was determined to be ln[k(144 C)/min-1] = -11 2.
TG Results
A two-step mass loss was observed in all experiments, the first step represents dehydration
and the second step is a result of decomposition. Figure 4 shows a typical TG curve obtained
from the variable heating rate study. An initial 9 10 % mass loss was observed in the 130
160 C range. This corresponds to the loss of 1 mol of H2O from 1 mol of BTAw, for which the
calculated mol % H2O is 10.5.
In a study [2] by Chavez et al, it was reported that hydrated water in a BTAw sample was
lost from 74 C to 116 C. In this particular study the heating rate used during the TG
experiment was 0.1 C min-1, which is substantially lower than the heating rates used in the
present work. Thus, the loss of hydrated water at lower temperatures was expected because of
the much slower heating rate.
Kinetic parameters, E/kJ mol-1 and ln(Z/min-1), were obtained for the dehydration step using
ASTM method E 1641 [11], which is based on the assumption that the dehydration obeys firstorder kinetics. As recommended by the ASTM method, calculations were carried out at several
different levels of conversion: 5, 10, 15 and 20 %. The results can be seen in Figure 5, which is
a plot of ln vs. 103 K/T, where T is the temperature of these different conversion levels, ,
obtained at various heating rates (). Table 1 shows that the values of E do not vary between
the range of conversion level examined and an average value of 193 5 kJ mol-1 was
determined. This result is in good agreement with the DSC results in this study, as seen from

44

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 3. An average value of the pre-exponential factor was also calculated, giving ln(Z/min-1)
= 57 2.
With the peak temperatures from the DTG data, ASTM method E 698 [4] was employed as
another technique to determine kinetic parameters. The results are shown in Figure 3, which is a
plot of ln(T-2/K-1min-1) vs. 103 K/T, where T is the peak temperature from the DTG data.
Calculations of the kinetic parameters yielded E = 182 4 kJ mol-1 and ln(Z/min-1) = 41 1.
Using the calculated kinetic parameters, E and lnZ, previously mentioned, the values for lnk
were calculated for the dehydration step at a common temperature of 144 C in order to
compare the results obtained by ASTM E698 and ASTM E1641. The value for the DTG data
was ln(k/ min-1) = 11 1, which is in agreement with the value calculated from the DSC data.
From the TG data, an average was taken of the rate constants and found to be ln(k/min-1) = 1
2, which is significantly higher than those found using the DSC and DTG data. This was
expected given the difference in pre-exponential factors. This difference could be due to the
fact that the values calculated using the TG data were obtained in the early stages of the
dehydration in comparison with the values from the DSC and DTG data, which were obtained at
much later stages.
The software IsoKin [12] was then used to perform a model free kinetic (MFK) analysis on
the first mass loss from the TG results. The model-free approach makes use of an
isoconversional method that allows for evaluating a dependence of the effective activation
energy on the extent of conversion [13]. These results can be found in Figure 6, which shows
the dependence of the activation energy on the extent of reaction. The MFK results show that
the activation energy remains consistent throughout the extent of the dehydration step,
indicating a single-stage process, with a mean value of 178 3 kJ mol-1 for 0.1 < < 0.8. From
the MFK analysis, the mean value of ln[f().(Z/min-1)] is 49.7 0.8.

Decomposition
DSC Results
The results of the heating rate study, using ASTM method E698 [4], on the decomposition
of BTAw are shown in Figure 7, which is a plot of ln(T-2/K-1min-1) vs. 103 K/T. This plot was
used to determine the values of the kinetic parameters, E and ln(Z/min-1), for BTAw. Only
results from the 0.5, 1, 2 and 5 C min-1 heating rates were included as the heat generated for the
8 and 10 C min-1 runs exceeded the maximum heat generation limit of 8 mW recommended in
the ASTM method E968 [5]. The kinetic parameters determined are E = 222 16 kJ mol-1 and
ln(Z/min-1) = 38 4. A ln(k/min-1) value of 12 6 was calculated at 265 C.
To confirm the kinetic results, a sample of BTAw was aged at 234 C for 64 minutes, the
calculated half-life of BTAw at this temperature from analysis of the data, and then immediately
quenched to a temperature at least 50 C lower. This aged sample was then heated at 5 C min-1
and its thermal curve recorded. The peak area for this curve was 737 5 J g-1. When compared
to 1.48 0.01 kJ g-1, the peak area for the unaged sample of BTAw run at the same heating rate,
it is seen that the peak area of the aged sample shows good agreement with the predicted value.

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

45

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Thus, the values determined for the kinetic parameters are good estimates for BTAw. The DSC
thermal curves for the aged and unaged samples of BTAw are shown in Figure 8.
Using the data collected from the DSC, a MFK analysis was performed with the IsoKin
software package [12]. The results for E and ln[f().(Z/min-1)] are shown in Figure 9. E has
increasing values as the extent of conversion, , increases to 0.6. This observation suggests that
parallel reactions are occurring from 0.1 < < 0.6.

TG Results
The second observed mass loss corresponds to thermal decomposition. This step occurred
with a mass loss of approximately 60 % in the 250 290 C range, as shown in Figure 4. When
using ASTM method E1641 [11] to evaluate kinetic parameters for the decomposition step, the
calculated values for E and ln Z were observed to increase as the extent of conversion increased
(Table 1 and Figure 10).
Decomposition of BTAw was reported [2] to commence at approximately 175 C using a
heating rate of 0.1 C min-1. With a heating rate significantly lower than those used in this
study, it was expected that the onset temperature in the literature work would also be lower due
to the dependence of the onset temperature on the heating rate.
Nedelko et al reported [3] kinetic parameters corresponding to both first order
decomposition and a first order catalytic reaction. For the first order decomposition of BTAw,
kinetic parameters of E = 197 5 kJ mol-1 and ln(Z/min-1) = 40 5 were given. A value for lnk
= -6 5 was calculated from this data at 242 C. In order to compare this value with the rate
constants from this study, values for lnk were also calculated at 242 C. From the TG data at 5
% conversion, it was found that ln(k/min-1) = 1 2, and at 10 % conversion, ln(k/min-1) = 0 3.
These values agree with the results from Nedelko within the margin of error. From the DSC
data, a value of ln(k/min-1) = -14 5 was found. This value was also within the margin of error
in comparison with Nedelkos results; however, using data from the DTG results to calculate the
rate constant gave ln(k/min-1) = -14 2, which did not fall within the range of error.
Using the DTG peak temperatures and ASTM method E 698 [4], kinetic parameters of E =
236 5 kJ mol-1 and ln(Z/min-1) = 41 1 were calculated. Both of these values are consistent
with those obtained from the DSC heating rate study and provide a good estimate of the kinetic
parameters for BTAw. Figure 7 is a plot of ln(T-2/K-1min-1) vs. 103 K/T for both DTG and
DSC data. Values for ln(k/min-1) were calculated using a common temperature of 265 C. The
DTG lnk value was found to be 12 1 min-1 while the TG value at 10 % conversion was 2 3
min-1. This difference was not unexpected as the results from ASTM method E 1641 (Figure
10) showed that the kinetic parameters vary with extent of conversion. The DTG value for lnk
was consistent, however, with that obtained from the DSC results.
The MFK results for this second step, seen in Figure 11, revealed that E slowly increases for
< 0.6. In this range, MFK analysis also showed that the values of ln[f().(Z/min-1)] slowly
increased from about 30 to 50. The variation of E with is consistent with the results obtained
using ASTM method E 1641 (Figure 10). As in the DSC decomposition results, this
dependence of E on and its increasing nature suggest that parallel reactions are occurring in
the range 0.1 < < 0.6.

46

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

TG-DTA-FTIR-MS Results
The thermal decomposition of BTAw was studied using TG-DTA-FTIR-MS. A TG-DTAFTIR-MS plot for the thermal decomposition in helium is shown in Figure 12. A 10 15 %
mass loss is observed in the 125 to 150 C range. This was attributed to dehydration, verified
by the evolved gas analysis: the absorbance peak in the FTIR spectrum at 3854 cm-1 and the ion
current peak for m/e = 18 in the MS. The mass of the water lost indicates that there is
approximately 1 mol of water per mol of BTAw.
Hydrogen azide (2153 and 1165 cm-1), ammonia (967 cm-1) and hydrogen cyanide (714 cm) appear as the major products in the FTIR trace in helium beginning at approximately 240 C.
This is supported by the MS data with peaks appearing for m/e = 17, 29, and 43. A strong peak
in the MS for m/e = 28 could be due to N2 formed either by electron impact fragmentation of
HN3, or by the thermal decomposition of BTAw. The total mass loss for this decomposition
slowly approaches 80 % by 350 C. These results for mass loss are in good agreement with
those obtained by TG.
1

A TG-DTA-FTIR-MS plot for the thermal decomposition of BTAw in air is shown in


Figure 13. The same dehydration is seen in the 125 to 150 C range. However, at
approximately 262 C, a sharp mass loss totalling almost 100 % was observed. This was
accompanied by the evolution of mainly hydrogen cyanide and ammonia, as well as small
amounts of carbon monoxide, carbon dioxide and various other gases for which the
identification is inconclusive. FTIR absorbance in the 2280 to 2285 cm-1 region suggests the
evolution of a nitrile compound in both He and air. The presence of m/e = 42 suggests this
nitrile may be cyanamide but further work is needed for a definitive identification. Nitrogen gas
is thought to be evolved based on previous studies of the thermal decomposition of 5aminotetrazole done by Levchik et al [14]. However, there is no supporting evidence at this
time due to the presence of nitrogen in the purge gas.
Tetrazole and 5-aminotetrazole both exhibit tautomerism, which is known to affect the
mechanism of their thermal decomposition. When the thermal decomposition of tetrazole occurs
in a melt, nitrogen is evolved as a result of breakdown of the azido-form of the molecule and
approximately 5 % of the sample decomposes with the elimination of hydrogen azide from the
ring form of the molecule [15,16]. With 5-aminotetrazole, fragmentation of the imino form, the
main isomer of the molecule in the solid state, most likely gives HN3 [14, 17]. However, upon
heating the imino form converts to the amino form and decomposition results in the elimination
of nitrogen gas and ammonia. Using these two molecules for comparison, BTAw may undergo
a tautomeric isomerization in which one form of the molecule has one or both tetrazole rings in
an imino form and results in the elimination of hydrogen azide. If this is correct, the lack of
hydrogen azide production in a reactive atmosphere would suggest that the imino form is not
present in air. The decomposition of the other form of the molecule results in the formation of
nitrogen, hydrogen cyanide and ammonia.

ARC Results
The thermal behaviour of BTAw and BTA using ARC was studied in atmospheres of air
and argon at ambient pressure. Figure 14 is an example of the results for a typical ARC

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

47

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

experiment, showing temperature and pressure vs. time. A comparison of ln(R/C min-1 g-1)
versus reciprocal temperature for the various methods used is shown in Figure 15. Table 2
presents a summary of the ARC results using the various drying methods.
BTAw samples were dried in-situ at 120 C followed by a HWS performed in an ambient
air atmosphere using an initial temperature of 150 C. In this set of experiments, each reaction
concluded with an exothermic reaction with a self-heating rate greater than the set maximum
rate of 1 C min-1. The experiments displayed consistent results, having onset temperatures of
213 and 215 5 C and each ending in a runaway reaction.
In addition, samples of BTAw were dried externally at 50 C. HWS experiments were
performed on these samples in both ambient argon and air. The results in argon exhibited
similar behaviour with onset temperatures of 205 and 203 5 C. These runs also showed some
unusual behaviour in which the heating rate attained the pre-determined self-heating rate limit, 1
C min-1, which triggered cooling of the ARC, but after cooling had begun, the heating rates
were in excess of 10 C min-1 for a short time before cooling resumed. Similar experiments in
ambient air did not exhibit this unusual behaviour and the onset temperatures for these
experiments were 211 and 201 5 C. In spite of the potential oxidizing potential in air, there is
no apparent difference between the onset temperatures for the ARC experiments conducted in
inert gas or air.
Finally, HWS runs in ambient argon were carried out on samples of BTAw that were predried at 120 C. This method of pre-drying removed all water content and resulted in BTA
(verified by conducting a TG study on the sample). Onset temperatures of 198 and 203 5 C
were observed, which are similar to, but slightly lower than, the previous experiments conducted
in argon on a sample pre-dried at 50 oC, suggesting that BTA is less thermally stable than
BTAw. In addition, when comparing this sample to the sample dried in-situ at 120 C, a
significant decrease was seen in the onset temperature. This indicated that an increase in drying
time approximately 16 hours for the external drying compared to 3 hours for in-situ drying
results in a decrease in the thermal stability and thus, the onset temperature.
From Figure 15, it is apparent that all the experiments demonstrate similar variation of R
with reciprocal temperature, with the exception of the sample dried in-situ. For this sample, the
onset temperature is slightly higher and the values of R are significantly larger than the other
experiments throughout the entire temperature range. It should be noted that the initial
temperature was also higher for this sample and this may have had an effect on these
observations.

CONCLUSIONS
A DSC heating rate study was performed on the dehydration and decomposition of BTAw
using the rates of 0.5, 1, 2 and 5 C min-1. This yielded, for the dehydration step, an activation
energy of 197 5 kJ mol-1 and a ln(Z/min-1) value of 46 1. For the decomposition step, values
of 222 16 kJ mol-1 and 38 4 min-1 were calculated for the activation energy and the preexponential factor, respectively. The results from the half-life test indicated that the ASTM
method E698 [4] provided a good estimate of the kinetic parameters for BTAw.
A second heating rate study, this time using TG, was performed on BTAw using 1, 2, 5, and
10 C min-1. This yielded an activation energy of 236 5 kJ mol-1 and a ln(Z/min-1) value of 41

48

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

1. These results, in conjunction with the DSC results, provide a good estimate of the kinetic
parameters for decomposition of BTA.
Results from DSC, TG, TG-DTA-FTIR-MS and ARC all suggest that BTAw first loses
water before undergoing any exothermic reaction and that BTAw is more thermally stable than
BTA. The enthalpy for the dehydration step is consistent with the energy of strong hydrogen
bonding. Results from TG-DTA-FTIR-MS experiments indicate that BTAw decomposes with
similar gaseous products as those from tetrazole and 5-aminotetrazole, which may suggest a
similar mechanism of decomposition is involved. However, further studies are required,
including an analysis of the solid residue left over after decomposition.

REFERENCES
[1]
[2]
[3]

[4]
[5]
[6]

[7]
[8]
[9]
[10]

[11]
[12]
[13]
[14]
[15]
[16]
[17]

1 D.E. Chavez and M. A. Hiskey, J. Pyrotechnics, 7 (1998) 11.


2 D.E. Chavez, M. A. Hiskey and D. L. Naud, J. Pyrotechnics, 10 (1999) 1.
3 V.V. Nedelko, B.L. Korsounskii, A.V. Shastin, N.V. Chukanov, T.S. Larikova, in: Proceedings
of the 29th Annual Conference of ICT on Energetic Materials and Technology, Karlsruhe,
Germany, 30 June 3 July 1998.
4 ASTM E 698 01, Standard Test Method for Arrhenius Kinetic Constants for Thermally
Unstable Materials, American Society for Testing and Materials, West Conshohocken, PA, U.S.A.
5 ASTM E 968, Standard Practice for Heat Flow Calibration of Differential Scanning
Calorimeters, American Society for Testing and Materials, West Conshohocken, PA, U.S.A.
6 ASTM E 967, Standard Practice for Temperature Calibration of Differential Scanning
Calorimeters and Differential Thermal Analyzers, American Society for Testing and Materials,
West Conshohocken, PA, U.S.A.
7 ASTM E 1582 00, Standard Practice for Calibration of Temperature Scale for
Thermogravimetry, American Society for Testing and Materials, West Conshohocken, PA, U.S.A.
8 R. Turcotte, R.C. Fouchard, A.-M. Turcotte and D.E.G. Jones, J. Therm. Anal. Cal., 73 (2003)
105.
9 D.I. Townsend and J.C. Tou, Thermochim. Acta, 37 (1980) 1.
10
ASTM E 1981-98, Standard Guide for Assessing the Thermal Stability of Materials by
Methods of Accelerating Rate Calorimetry, American Society for Testing and Materials, West
Conshohocken, PA, U.S.A.
11
ASTM E 1641 04, Standard Test Method for Decomposition Kinetics by
Thermogravimetry, American Society for Testing and Materials, West Conshohocken, PA, U.S.A.
12
C.A. Wight, Isoconversional Data Analysis Program, Center of Thermal Analysis,
University of Utah, Version 1.42, 2000.
13 S. Vyazovkin and D. Dollimore, J. Chem. Inf. Comp. Sci., 39 (1996) 42.
14
S.V. Levchik, O. A. Ivashkevich, A. I. Balabanovich, A. I. Lesnikovich, P. N. Gaponik
and L. Costa, Thermochim. Acta, 207 (1992) 115.
15
A.I. Lesnikovich, O.A. Ivashkevich, G.V. Printsev, P.N. Gaponik and S.V. Levchik,
Thermochim. Acta, 171 (1990) 207.
16
A.I. Lesnikovich, S.V. Levchik, A.I. Balabanovich, O.A. Ivashkevich and P.N. Gaponik,
Thermochim. Acta, 200 (1992) 427.
17 A.I. Lesnikovich, O.A. Ivashkevich, S.V. Levchik, A.I. Balabanovich, P.N. Gaponik and A.A.
Kulak, Thermochim. Acta, 388 (2002) 233.

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

49

New Trends in Research of Energetic Materials

Czech Republic

[Content]

N
H

H
N

N
N

2006

N
C

N
H

H2O

Figure 1 Structure for BTAw.

Figure 2 DSC curves for BTAw at 5 C min-1 (____); 2 C min-1 (__ __ __); 1 C min-1 ();
..
0.5 min-1 (
).

50

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

-2
-1
-1
-ln(
T /K min )

9.0

10.0

11.0

12.0

DSC
DTG
13.0
2.30

2.35

2.40

2.45

2.50

10 K/T

100

80

60

40

-1

120

20
50

100

150

200

250

300

DTG/% C

TG/mass %

Figure 3 Plot of ln( T-2/K-1 min-1) vs. 103 K/T for the dehydration step using data from DTG
and DSC.

-1
350

Temperature/C

Figure 4 TG and DTG mass loss curves for BTAw for = 5 C min-1.

10

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

51

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

2.5

-1
ln(
/K min )

2.0

1.5

1.0

0.5

0.0

Conversion/%
5
10
15
20

-0.5
2.35

2.40

2.45

2.50

10 K/T

Figure 5 Plot of ln(/K min-1) vs. 103 K/T for the dehydration step with various heating rates.

Figure 6 BTAw (pre-dried at 50 C) IsoKin TG dehydration results ( = 1 to 10 C min-1).

52

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

11

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

-2 -1
-1
-ln(
T /K min )

10

11

12

13

DSC
DTG
14
1.75

1.80

1.85

1.90

1.95

103 K/T

Figure 7 Plot of ln(T-2/K-1 min-1) vs. 103K/T for the thermal decomposition step using data
from DTG and DSC.

endo

Heat Flow/W g

-1

BTA
Aged at T = 234 C
Unaged

200

225

250

275

300

Temperature/C

Figure 8 Comparison of the DSC thermal curve of an aged BTA sample with that of an unaged
sample.

12

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

53

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 9 BTAw IsoKin DSC results for = 0.5 to 5 C min-1.

2.5

-1
ln(
/K min )

2.0

1.5

1.0

0.5

0.0

Conversion/%
5
10
15
20

-0.5
1.80

1.85

1.90

1.95

2.00

103 K/T

Figure 10. Plot of ln(/K min-1) vs. 103 K/T for the thermal decomposition step with various
heating rates.

54

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

13

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 11. BTAw (pre-dried at 50 C) IsoKin TG decomposition results for = 1 to 10 C min-1.

14

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

55

Czech Republic

100

1.5
1.0

= 5 C min -1
5 mg

60

0.5
0.0

20

7
6

CO + HN3

CO2
H2O
HCN
NH3

4
3

HN3

[N C]

1
0

10000

MS Ion Current/pA

c)

[Content]

-0.5

FTIR 10 Absorbance

b)

2006

DTA
T/C

a)

TG Mass %

New Trends in Research of Energetic Materials

1000

m/e
17
18
28
29
30
42
43
44

100

10

1
50

100

150

200

250

300

350

Temperature/C

Figure 12 5 mg BTAw heated in helium at 5 C min-1: a) TG-DTA; b) FTIR; c) MS.

56

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

15

a)

Czech Republic

100

80

= 5 C min
5 mg

60

-1

1
0

40

-1

20
0
b)

FTIR Absorbance

0.6
0.5

2006

[Content]

DTA
T/C

TG Mass %

New Trends in Research of Energetic Materials

-2
0.25

0.20

CO + HN3
CO2

0.15

0.4

H2O
0.10

HCN
NH3

0.3
0.05

N2O

0.2

[N C]

0.00

0.1

50

100

150

200

250

300

350

0.0

c)

MS Ion Current/nA

100

10

m/e
1

17
18
42
43
44
52
61

0.1

0.01

0.001
50

100

150

200

250

300

350

Temperature/C

Figure 13 5 mg BTAw heated in air at 5 C min-1: a) TG-DTA; b) FTIR; c) MS.

Figure 14 Example of ARC results for BTAw: pre-dried at 50 C, and run in a closed system of
ambient Ar.

16

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

57

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

ln(R/C min -1 g-1)

-1
4.3

4.4

4.5

4.6

4.7

4.8

4.9

5.0

10 K/T
Figure 15 Comparison of rate vs. temperature for various methods. Dried in-situ at 120 C and
then run in ambient air ( ); Pre-dried externally at 50 C and then run in ambient Ar (- - -);
Pre-dried externally at 50 C and then run in ambient air ( ); Pre-dried externally at 120 C
and then run in ambient Ar ( ).

58

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

17

18

Method
Conversion % E/kJ mol
DSC ASTM E698
197 5
DTG
182 4
MFK
178 3
TG ASTM E1641
5
192 6
10
194 5
15
194 5
20
193 5
a)
Calculated at 144 C
b)
Calculated at 265 C

-1

ln(Z/min )
46 1
41 1
57 2
57 2
57 1
57 1

-1

Dehydration
ln(k/min )
-11 2
-11 1
-1 1
23
13
12
12

-1 a)

Table 1 Comparison of kinetic parameters from the variable heating rate studies.

E/kJ mol
222 16
236 5
180 6
197 9
203 10
209 8

-1

ln(Z/min-1)
38 4
41 1
43 1
46 2
47 2
48 2

Decomposition
ln(k/min-1) b)
-12 6
-12 1
23
23
13

New Trends in Research of Energetic Materials


Czech Republic

David Jones, K. Armstrong, T. Parekunnel, Q. Kwok

2006
[Content]

59

60

air

Atmosphere Sample Mass/g

Ti/C

0.48
150
0.49
150
Externally at 50 C
Ar
0.53
100
0.53
100
Externally at 50 C
air
0.51
100
0.49
100
Externally at 120 C
Ar
0.50
100
0.50
100
a)
Uncertainty from the extrapolation of the onset temperature.
b)
Pressure recorded at ambient temperature following the run.
c)
lnk calculated at 220 C.

In-situ at 120 C

Drying Method
213
215
205
203
211
201
203
198

(To 5)/C a)

Table 2 Comparison of ARC results from the various drying methods.

100 2
101 2
77.2 0.7
76.2 0.7
87.2 0.7
75 1
84 5
78.1 0.6

E/kJ mol-1 ln(Z/min-1)

0.37 @ 20 C 418 7
0.30 @ 22 C 425 10
0.27 @ 38 C 322 3
0.68 @ 21 C 319 3
0.52 @ 20 C 365 3
0.51 @ 26 C 315 4
0.48 @ 25 C 351 2
0.53 @ 39 C 326 2

Pf/MPa b)

-2 3
-3 3
-1 1
-2 1
-2 1
-2 1
-2 5
-1.4 0.8

ln(k/min-1) c)

New Trends in Research of Energetic Materials


Czech Republic
2006
[Content]

The Thermal Behaviour of BTAW, a High Nitrogen Fuel

19

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

NEW TRENDS IN DETECTION OF EXPLOSIVES


P. Mostak, M. Stancl
Research Institute of Industrial Chemistry, Explosia
53217 Pardubice-Semtin, Czech Republic

Abstract:
Detection methods used in screening procedures looking for hidden explosive charges
and improvised explosive devices (IED) are presented and factors influencing the
detection efficiency discussed. The simultaneous use of x-ray screening and electronic
detectors adds the substantial synergy to the efficiency of explosive detection.
Exploitation of dogs and colour reactions are another important detection methods. The
improvement of imaging ability of X-ray systems and the increasing sensitivity of
electronic detectors should be expected in near future. The progress in neutron
technologies should be also achieved. LIBS, MW and radar systems have a chance to
fulfil the gap in the stand- off detection. The detection on distance is still the serious
problem in the defence against bomb attacks performed by suicide bombers and big
explosive charges situated in cars. The effort in search for principally new detection
methods is required.
Keywords:
detection

1.

detection of explosives, improvised explosive devices, marking for

INTRODUCTION

The experience obtained in the fight with terrorist attacks by explosives have proved,
that the tools used by terrorists in bomb attacks are more sophisticated, more dangerous,
effective countermeasures complicated and in some cases not adequate.
The broad use of remote control of IED initiation by mobile phones gives to terrorists
the choice of exact moment of the bomb attack and also the attacks performed by suicide
bombers opened new challenge to efficient countermeasures including quick detection of
explosive charge or/and IED on distance.
Therefore, more effective detection systems and following countermeasures are
required and the research and development leading to substantial improvement of detection
methods and a development of new detection technologies is a permanent important task
The aim of this presentation is to analyse the efficiency of detection techniques and
discuss main factors, which influence the explosive detection. The important tendencies,
leading to the improvement of detection efficiency, will be discussed.

Petr Mostak, Miroslav Stancl

61

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

MAIN DETECTION METHODS

The important task in screening operations is covered by the visualisation of the


checked object content by X-rays or other imaging techniques. In fact X-ray equipment is
the most frequently used tool for detection of hidden explosives at stabile checking points,
the visualisation enables also the search of further contraband as weapons and narcotics.
Electronic detectors of explosives are further effective method for detection of
explosive charges by analysing vapours and/or particles of explosives.
The very important tools of explosive detection are special trained dogs. The
sensitivity of the dog nose is unique and electronic detectors have not yet overcome the high
detection ability of dog. Further technologies as nuclear methods, NMR, NQR are in
development stage or in field tests.
The simultaneous use of different detection systems usually gives the substantial
synergy to the detection efficiency, typical example is integration of X-ray and electronic
detectors at stabile checking points in airports, other transport terminals and important
buildings.

3.

FACTORS INFLUENCING THE DETECTION EFFICIENCY

IED design
There are many types of explosives used in IEDs and many possible designs, differing
in initiation chain, starting mechanism and packing arrangement. The range of explosive
mass used in IEDs should be very different, starting with several grams in letter bombs and
ending with the mass of several tons of explosive in a lorry. There are also many different
scenarios in which IEDs are used.
Very dangerous are mobile or radio controlled IEDs situated in transport means, on
roads, streets and places with high concentration of people.
Extremely dangerous are:
-

IEDs fixed on bodies or carried in personal luggage of suicide bombers

IEDs in cars containing big amount of explosives or ammunition

Electronic detectors
Electronic detectors based on IMS and GC are able to detect, by analysing vapours
and/or particles, most of military and civil explosives and also organic peroxides. The
sensitivity of this equipment is fully sufficient in particle detection, collection of 1 particle
usually enables the positive result.
The vapour detection is dependent on conditions at detection, from which most
important are the vapour pressure of explosive, barrier properties of packing, the
temperature, air movement and the sampled air quantity.
In unfavourable conditions, the vapour detection is not fully reliable method.
Therefore, the systems using simultaneously vapour and particle detection are effective and

62

New Trends in Detection of Explosives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

often used. The detection efficiency is increased by pre-concentration of explosive traces


from sampled air, some equipment is able to work with pre-concentration factor as high as
100 000. Such high pre-concentration factor is typical for stabile screening equipment.
Dogs
The dog remains the unique tool for trace detection. The dog is able to detect
explosives even at low temperatures at which most detectors are not fully effective. No
exact knowledge about the mechanism of dog sniffing is at disposal, but it seems, that dog is
able to sniff not only vapours but also particles of explosives and even explosives fixed on
the surface of dust.
Dogs have their place in search for hidden IEDs in various scenarios and the efficiency
of dogs has been confirmed by steady rise of dog numbers used in security missions.
Colour reactions
Explosives produce colour reactions with specific reactants, and this knowledge has
been exploited in trace detection. The sensitivity of this detection method is lower in
comparison with electronic detectors, but when we are able to collect several particles of
explosive, the detection is reliable. Usually, collection of 1 average particle is sufficient for
positive detection. Detection sets for colour reaction can detect also AN, chlorates and
perchlorates and organic peroxides, it means important components of some civil and
improvised explosives.

Imaging and radiation technologies


X-ray systems are broadly used as the basic screening tool for control of luggage,
postal sending and further objects. The progress of X-ray technology is steady and enables
the better imaging of the object content, scanning and 3 dimensional screening. New
developed X-ray method can identify the chemical compounds present in suspicious part of
object by angle diffraction spectrum. Diffraction spectrum is typical for each compound and
substantially increases the possibility of detection and identification of explosive charge.
Further radiation and wave technologies as neutron in-gamma out detection, LIBS
spectroscopy, radar, NMR, QMR and other wave systems are in development phase and/or
in
field tests.
Detection of marking agents in plastic explosives
Practical importance among marking agents, specified in the Technical Annex of
Montreal Convention on the Marking of Plastic Explosives for the Purpose of Detection,
have only following compounds:
-

2,3-Dimethyl-2,3-dinitrobutane (DMNB)

para-Mononitritoluene (p-MNT).

Vapour detection of DMNB, the most popular marking agent for plastic explosives, by
IMS is not very effective. This is caused by the fact, that cluster ions formed from molecules
of DMNB by irradiation have the low thermal stability and therefore, the yield obtained after

Petr Mostak, Miroslav Stancl

63

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

acceleration in drifting tube is low. The detection of DMNB by GC is excellent.


Unfortunately, most detectors used in detection practice at this time are based on IMS. The
reason are some advantages of IMS in comparison with GC such as lower price, lower
weight, absence of carrying gas, and a very short time of analysis. This discrepancy is to be
solved by some acceptable way, but the solution will be not easy.

4.

STAND- OFF DETECTION

Extremely difficult problem, which emerged in last years, is the massive use of IEDs
controlled on distance by radio or mobile phone and attacks by bombs fixed oh the body of
suicide bombers. Also attacks by cars filled by big amount of explosives present high danger
and consequences of such attacks are very heavy.
The experience, obtained in Iraq, Chechnya and other countries, confirms, that the
detection of these terrorist tools on distance and effective countermeasures is a very uneasy
task. No effective, service ready technology for stand-off detection is at disposal now.
Therefore, research and development of effective detection technology for stand-off
detection should have the highest priority.

5.

EXPECTED PROGRESS IN DETECTION TECHNOLOGIES


The progress in electronic detectors should be expected in following directions:
-

increasing sensitivity of vapour and particle detectors

higher efficiency of pre-concentrators

broader use of micro and nano technologies to decrease mass, size and cost
of instruments

Further increase of performance of x-ray screening systems should be attained and


broader application of angle diffraction spectra will be achieved in the near future.
Some progress is to be expected in the detection using neutron technologies. The
research and development in imaging and wave technologies should lead to efficient
detection methods.
Strong effort is concentrated on techniques enabling the stand-off detection of
explosives charges. It is probable, that the results obtained in the field of LIBS, MW and
possibly radar technology will be able to bring the feasible detection method to detect
explosive charges on distance.
We can also expect, that an intensive investigation of new principles and methods
suitable for detection of explosives and IEDs will bring new knowledge and experience. This
should lead to improvement of the efficiency in the fight with the bomb terrorism and better
safety to people.

64

New Trends in Detection of Explosives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

MOLECULAR SURFACE ELECTROSTATIC POTENTIALS IN THE


COMPUTATIONAL CHARACTERIZATION OF ENERGETIC
COMPOUNDS
Peter Politzer and Jane S. Murray
Department of Chemistry
University of New Orleans
New Orleans, LA 70148
USA

Abstract:
The electrostatic potential of a molecule is a fundamental determinant of its properties
and behavior. When computed on the molecular surface of an energetic compound, it is
characterized by strong and highly variable regions of positive potential, which can be
related to the compounds impact sensitivity. The features of molecular surface
electrostatic potentials can also be used to establish quantitative expressions for heats
of sublimation and vaporization, thus permitting gas phase heats of formation of
energetic (and other) compounds to be converted to the more useful solid and liquid
values. Since these procedures are carried out computationally, they can be applied to
the assessment of proposed target compounds as well as those that have already been
prepared.
Keywords:

1.

molecular electrostatic potentials, impact sensitivities, solid and liquid


heats of formation, heats of sublimation, heats of vaporization

THE ELECTROSTATIC POTENTIAL

The remarkable Hohenberg-Kohn theorem[1] established rigorously that the


properties of a system of nuclei and electrons are determined by its electronic density (r).
Thus its energy, E, is a functional of (r):
E = [(r ) ]

(1)

The electronic density is in turn related by Poissons equation to the electrostatic potential
V(r) that is created by the systems nuclei and electrons:

2 V(r ) = 4(r ) 4

Z (r R )
A

(2)

V(r) is simply the Coulomb potential,

V(r ) =

R
A

Peter Politzer, Jane Murray

ZA
(r)dr

r r
A r

(3)

65

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

in which ZA is the charge on nucleus A, located at RA. V(r) is a physical observable; it can
be obtained experimentally, by diffraction techniques,[2-4] as well as computationally. The
sign of V(r) in any region depends upon whether the positive contribution of the nuclei or
the negative one of the electrons is dominant there.
On the basis of the Hohenberg-Kohn theorem in conjunction with eq. (2), it can be
argued that the electrostatic potential is as fundamental a determinant of the properties of a
system as is the electronic density.[5,6] There have been derived exact expressions for atomic
and molecular energies[6-9] and the chemical potential (electronegativity)[5,6] in terms of the
electrostatic potentials at the nuclei. Functional relationships between the energies and V(r)
have been developed,[9] as well as effective formulations for ionic and covalent radii.[10,11]
In this paper, we will review work that has shown how electrostatic potentials can
help to quantify and predict two important properties of energetic compounds: impact
sensitivity and condensed phase (solid or liquid) heat of formation. For these purposes, we
compute V(r) on the surfaces of the respective molecules, which we define, following the
suggestion of Bader et al,[12] as the 0.001 electrons/bohr3 three-dimensional contour of the
electronic density (r). The resulting surface potential is labeled VS(r).
In seeking to develop quantitative relationships involving VS(r), we have found it
useful to characterize it in terms of certain statistically-defined features. The ones that will
be relevant to the present discussion include:
(1) the most positive and negative values of VS(r), the VS,max and the VS,min;
(2) the positive and negative averages, VS+ and VS- , given by

VS+ =

1 m +
VS (ri )
m i =1

VS- =

1 n VS (rj )
n j=1

(4)

(3) the total variance of VS(r), 2tot , and its positive and negative components, 2+ and

2 ,
2tot = 2+ + 2 =

2
2
1 m
1 n
VS+ (ri ) VS+ + VS- (rj ) VS-

m i =1
n j=1

(5)

(4) an electrostatic balance parameter,


=

2+ 2

2tot

(6)

The first summations in eqs. (4) and (5) are over the points where VS(r) is positive, the
second over those where it is negative, on a grid encompassing the molecular surface.
The quantities 2tot , 2+ and 2 are measures of the variabilities of the total, positive
and negative surface potentials, reflecting how extensive are the ranges that they cover. The
parameter is an indicator of the degree of balance between the positive and negative
potentials; when 2+ = 2 , then reaches its maximum value of 0.25.
We have shown, in a series of studies, that the quantities in terms of which we
characterize VS(r) (which include also its average deviation) provide a basis for developing
analytical relationships for a variety of condensed phase physical properties that depend
66

Molecular Surface Electrostatic Potentials in the Computational . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

upon noncovalent interactions: boiling points and critical constants, heats of phase
transitions, solubilities and solvation energies, partition coefficients, diffusion constants,
viscosities, surface tensions, etc. For reviews of this work, see Murray et al.[13-15]

SENSITIVITY

2.
2.1

Background

In the present context, sensitivity denotes the vulnerability of an energetic compound


to accidental detonation, i.e. caused by an unintended external stimulus. Such stimuli can
include impact, shock, heat, friction and electrostatic charge.[16,17] These various types of
sensitivity do show some approximate similarities in trends. For example, Storm et al have
found correlations between impact and shock sensitivity, and betweeen the latter and the
critical temperature at which thermal decomposition becomes self-sustaining.[18] Zeman et
al obtained relationships involving impact and electric spark sensitivities.[19]
Minimizing sensitivity while maximizing detonation performance is a major
challenge, to which a great deal of attention and effort is being devoted. Energetic
compounds are inherently metastable. The introduction of external energy, e.g. via impact,
can lead to decomposition, first endothermic and then highly exothermic, releasing gaseous
products and a large amount of energy and producing a shock front that moves through the
compound at supersonic speed causing continuing and self-sustaining exothermic
decomposition (detonation).[16,17,20,21]
The likelihood that such a series of events will take place for a particular compound,
i.e. its sensitivity, depends upon several factors. These include its molecular composition
and structure, crystal properties, and physical state, e.g. sizes of particles. Thus, polymorphs
of a given energetic solid may differ in sensitivity, as may crystals of various sizes and
shapes.
Since crystal and physical features can have such importance, it is somewhat
remarkable that quite good correlations have also been obtained between sensitivity
(especially toward impact) and diverse properties of individual molecules (usually of a
specific chemical type, e.g. nitramines). Among these properties are different indicators of
the stabilities of CNO2 and NNO2 bonds (trigger linkages), electronic excitation
processes, rates of energy transfer into certain vibrational modes, and stoichiometry. There
have been several reviews of these studies.[22-24] A cautionary note was sounded by Brill and
James,[25] who pointed out that the existence of these correlations does not necessarily imply,
in each instance, any fundamental significance.
2.2

VS(r) and Sensitivity

As was discussed above, sensitivity has been found to correlate quantitatively with
some rather disparate molecular features, and it has been linked qualitatively to yet others,
such as the ability to form strong intermolecular hydrogen bonds.[22-24] One possible
interpretation of these observations is that they reflect different aspects of the overall global
nature of the molecule, and that a useful approach might be to try to characterize it as a
whole, globally, instead of focusing upon certain bonds, vibrational modes, etc. We have
explored, with encouraging success, the possibility of doing this in terms of the electrostatic
potential, the fundamental nature of which has already been demonstrated.
There is a great deal of empirical evidence indicating a relationship between the
electrostatic potential and molecular stability. Paulings recognition of this is embodied in

Peter Politzer, Jane Murray

67

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

his electronegativity postulate,[26] which states essentially that the charge distributions in
molecules tend to be such as to make each atom as close to electrically neutral as possible.
This translates into the electrostatic potential on the molecular surface, VS(r), being
relatively uniform and fairly weak; deviations from this promote instability.
This reasoning has long been widely applied in chemistry; delocalization of charge is
commonly associated with enhanced stability. For example, relative pKa values of
carboxylic acids are often interpreted on the basis of how well the charge of the resulting
anion can be delocalized, thereby increasing its stability and lowering the acids pKa. We
have recently shown that the magnitudes of the most positive potentials on the surfaces of
carbocations, the VS,max, are good measures of their stabilities;[27] the less positive is VS,max,
the more has the charge delocalized and the more stable is the carbocation.
Before examining the VS(r) of the molecules of energetic compounds, it is useful to
gain some perspective by looking at a group of more representative ones. In Table 1 are
listed, for a diverse series of organic molecules, some of the computed quantities that we use
to characterize VS(r): (a) its most positive and negative values, VS,max and VS,min, (b) the
positive and negative averages, VS+ and VS- , and (c) the variances, 2tot , 2+ and 2 . Also
included are the postive and negative surface areas, AS+ and AS .
An overall trend is immediately apparent. For most organic molecules, the regions of
positive surface potential are larger in area but weaker than are the negative ones. There are
some exceptions, such as anthracene and 1,4-dichlorobenzene (see Table 1), but these are
unusual. Thus, it is normally the case that AS+ > AS , but VS,min > VS,max , VS- > VS+ and 2 >

2+ . The differences are often quite large, as can be seen in Table 1. A more extensive data
set confirms that these are general patterns.[29]

Table 1. Computed surface quantities (HF/STO-5G*//STO-3G*) for some representative


molecules.a,b
Molecule
AS+
A-S VS,max VS,min
VS+
VS 2+
-2
benzene
50.6
58.0
9.5
10.2
4.6
4.9
7.3
8.5
butylbenzene
123.6
78.3
8.5
12.4
3.7
4.5 4.9
13.0
anthracene
100.1 112.6 10.6
9.4
5.3
5.2
8.8
6.8
dimethylsulfide
68.3 32.4 11.2
17.0
4.1
5.7
7.2
25.7
1-hexanol
131.2
51.7 12.6
36.7
5.0
9.2 9.9
132.5
cyclohexanol
118.7
26.7 26.7
43.5
5.6
15.9 18.6
207.5
methanol
44.8 21.5 29.9
41.6 10.3
18.4 48.9
181.9
ammonia
27.7 19.7 18.1
29.0
9.2
12.8 27.6
73.7
methylamine
52.0 20.9 21.4
53.2
9.3
20.2 34.6
263.0
piperazine
92.2
38.5 22.9
47.8
9.3
17.4 26.3
207.2
phenol
65.0
63.0 34.9
36.6
8.6
8.5 64.0
73.1
acetamide
62.7 31.8 32.0
40.6 12.7
20.6 68.0
150.8
1,4-dichlorobenzene 95.8 56.7 15.2
11.1
6.2
7.0 18.0
10.2
3-nitrochlorobenzene 101.9 54.2 22.3
37.5 12.3
16.7 21.7
124.4
________________________________________________________________________
a
Units: AS+ and A-S are in A2; VS,max, VS,min , VS+ and VS- are in kcal/mole; 2+ and -2 are in
(kcal/mole) 2.
b
Data taken from reference 28.
68

Molecular Surface Electrostatic Potentials in the Computational . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The situation is quite different for molecules of energetic compounds, as is shown in


Table 2. The available data are more limited, but it is clear that the positive regions are now
stronger than the negative. In every instance, VS+ > VS- and 2+ > 2 . The striking contrast
between energetic and typical organic molecules is due to the electron-attracting powers of
the nitro groups[30] and the ring nitrogens[31] of the former. These withdraw electronic
charge from the central portions of the molecules, thereby creating strong and extensive
positive regions, sometimes having several local VS,max. Thus, the negative VS(r) above the
center of the ring in benzene, associated with its electrons, is not present in TNB, TNT and
picric acid. The only negative surface potentials of energetic molecules are likely to be
above parts of their peripheries, e.g. NO2 oxygens, ring nitrogens, the carbonyl oxygen in
NTO, etc. These various features of the VS(r) of energetic molecules can be seen in color
figures in several earlier reports, by us[24,32] and by Rice and Hare.[33] It should be pointed
out, however, that a single NO2 or ring nitrogen is not sufficient to produce these effects, as
is evident from the examples of nitrobenzene, which has VS+ < VS- and 2+ < 2 [34], and
piperazine (Table 1). It is necessary that these components constitute some critical portion
of the entire molecule, just as is required for energetic performance.

Table 2. Computed surface quantities (HF/STO-5G*//STO-3G*) for some energetic


molecules.a,b
Molecule
AS+
A-S VS,max VS,min
VS+
VS 2+
-2
1,3,5-trinitrobenzene,
TNB

---

---

39.2

---

23.9

15.3

109.0

55.3

2,4,6-trinitrotoluene,
TNT

125.0

77.8

37.5

28.8

20.6

14.7

104.3

53.2

2,4,6-trinitrophenol,
picric acid

---

---

---

---

24.9

17.0

70.7

53.1

---

---

21.0

96.3

73.8

---

---

21.8

3-nitro-1,2,4-triazol-5-one,

NTO
1,3,3-trinitroazetidine,
TNAZ

18.0
12.2

174.3
111.0

142.0

87.5
81.0

42.6

octanitrocubane,
ONC
203.3 110.3
----23.8
6.9
407.3 16.5
________________________________________________________________________
a
Units: AS+ and A-S are in A2; VS,max, VS,min , VS+ and VS- are in kcal/mole; 2+ and -2 are in
(kcal/mole) 2.
b
Data are from references 28 and 34 and unpublished results.

In the present context, it is also relevant to mention a particular unusual buildup of


positive potential, with a local VS,max, that is found above the CNO2 bonds in
nitroaromatics, nitroheterocycles and nitroalkanes.[32,35,36] This is not true of chemical bonds
5

Peter Politzer, Jane Murray

69

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

in general. We have shown that there is a limited inverse correlation between this VS,max and
the CNO2 bond energy,[35,36] which is consistent with early observations of a relationship
between nitroaromatic impact sensitivities and the roughly estimated electrostatic potentials
of the CNO2 bonds.[36] (Another interesting aspect of the positive buildup above CNO2
bonds is that it has been demonstrated to provide a channel for the initial approach of
nucleophiles.[37,38])
Can it be inferred from the preceding discussion that the metastabilities of energetic
compounds are reflected in their characteristic molecular surface electrostatic potentials,
specifically the unusually strong and highly variable positive components? If so, can this
provide a basis for ranking their crystalline sensitivities? We have shown that this is indeed
possible.[34] We focused upon the imbalance between the positive and negative surface
potentials. There are of course a number of ways to express this imbalance, e.g. in terms of
VS,max VS,min , VS+ VS- , VS+ / VS- , 2+ -2 , 2+ / -2 , etc., or combinations of these. We
were most successful when we used a different combination for each class of compounds
considered nitroaromatics, nitramines and nitroheterocycles. This produced very good
correlations with impact sensitivity, as obtained by the drop-weight method; [16-18] they all
had R2 0.980, with standard deviations of 8 to 14 cm, for experimental databases that
covered, respectively, 284, 307 and 256 cm. The detailed relationships are presented in the
original paper.[34] We certainly do not view these as final. As larger databases become
available, or with more judicious choices of variables, they are likely to be modified, perhaps
simplified. Nevertheless, it is gratifying that such good correlations are already attainable.
2.3

Discussion

In would clearly be preferable to be able to include all chemical classes in a single


correlation. This would be desirable both conceptually, since it might provide insight into
key factors determing impact sensitivity, and also from a practical standpoint, because there
can occasionally be some ambiguity concerning the chemical class to which a compound
should be assigned. Developing one general relationship remains an important objective.
What is very significant, however, is the unifying concept that the impact sensitivities
of energetic compounds are related to specific features of their molecular surface
electrostatic potentials: their characteristic anomalously strong and variable positive
components. The extensive study by Rice and Hare supports this concept, although they feel
that the key positive regions are those above covalent bonds within the inner molecular
framework.[33]
In section 1 of this paper, we noted that a number of condensed phase physical
properties have been expressed analytically as functions of quantities characterizing their
molecular VS(r). It is therefore at least plausible that by linking sensitivity to VS(r), we are
taking some account of the effects of crystal properties upon sensitivity, which was also
mentioned earlier (section 2.1).

3.

SOLID OR LIQUID PHASE HEAT OF FORMATION

The heat of formation Hf is a key property of an energetic compound. It allows the


determination of the energy release upon combustion or decomposition, the latter in turn
being used to calculate detonation pressure and velocity,[17,21,39] two primary measures of
explosive performance.

70

Molecular Surface Electrostatic Potentials in the Computational . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Since experimental Hf are often not available, or may be unreliable,[40,41] a great


deal of effort has gone into developing procedures for producing them computationally. One
option is to utilize an ab initio or density functional technique to obtain H for either the
formation of the molecule from its elements or for its separation into the constituent atoms.
The first yields Hf directly; the second may require appropriate inclusion of some
dissociation energies. Due to the remarkable advances in methodology, software and
processor technology in recent years, these approaches can give quite accurate results for
molecules of practical interest, e.g. propellant combustion products,[42] although size is
certainly a limitation. For larger molecules, it is common to introduce empirical correction
terms. Another widely-used route to Hf is by calculating H for isogyric, isodesmic or
homodesmic reactions, which are designed to achieve some cancellation of error. For
comparisons and assessments of these and other procedures, see Stewart,[40] Irikura and
Frurip,[43] Thiel[44] and Politzer et al.[45]
The various schemes that have been mentioned typically lead to predictions of gas
phase heats of formation. What is normally relevant for energetic compounds, however, is
Hf for the corresponding solid, or less often the liquid. One can arrive at these from the gas
phase value if the heats of sublimation and vaporization, Hsub and Hvap, are known:
H f (solid) = H f (gas) H sub

(7)

H f (liquid) = H f (gas) H vap

(8)

Our present interest is not in the different ways of obtaining H f (gas) computationally, but
rather in H sub and H vap . If these are not known, as is often the case, then some method
for estimating them is needed.
It was already brought out in section 1 that heats of phase transitions are among the
physical properties that can be expressed in terms of the quantities that characterize VS(r),
defined by eqs. (4) (6). The relationships for H sub [46] and H vap [47] are:
2
H sub = 1AS,tot
+ 2 ( 2tot ) + 3
0.5

H vap = 1 ( AS,tot ) + 2 ( 2tot ) + 3


0.5

0.5

(9)

(10)

In eqs. (9) and (10), AS,tot is the total surface area of the molecule, AS,tot = AS+ + AS- . The
parameters i and i , i = 1-3, can be evaluated by fitting to experimental H sub and H vap .
Our most recent parametrizations of eqs. (9) and (10), at the density functional B3PW91/631G** level, was with general databases of 66 H sub and 30 H vap .[48] We were able to
reproduce these properties for 105 and 44 compounds, respectively, including many that

Peter Politzer, Jane Murray

71

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

were not involved in the parameter fittings, with average absolute deviations from
experiment of 2.8 kcal/mole for H sub and 2.0 kcal/mole for H vap .
We have used eqs. (9) and (10) in computing heats of formation for numerous
energetic solids (and one liquid),[45,46,49-53] many of them being proposed target compounds.
Rice et al have carried out very extensive and detailed calculations of the gas, liquid and
solid phase heats of formation of energetic compounds, utilizing eqs. (9) and (10) to find
H sub and H vap .[54,55] They parametrized their entire procedure with databases containing
exclusively C/H/N/O energetic compounds. Their root-mean-square deviations in the solid
and liquid phase heats of formation of compounds not included in the fitting were 5.6 and
3.2 kcal/mole, respectively.[55]
The focus in this section has thus far been entirely upon molecular solids and liquids.
For an ionic solid, the heat of formation can be obtained by converting its lattice energy to
H for the process[56]
MmXn(s) mM+a(g) + nXb(g)

(11)

and combining it properly with the experimental or computed Hf of the gas phase ions. We
have shown, for three different cations (NH4+, Na+ and K+), that the lattice energies of their
salts can be represented in terms of the quantities characterizing the VS(r) of the anions.[57]
In this manner, we have predicted Hf for NH4O2 and NH4O3.[58]

4.

CONCLUDING REMARKS

We have shown how the electrostatic potential, a fundamental determinant of


molecular properties and behavior, can be applied to the characterization of energetic
compounds, with respect to their impact sensitivities and their solid or liquid phase heats of
formation. A point to be emphasized is that the procedures that have been described are
entirely computational in nature. They can accordingly be used to predict the impact
senstivities and heats of formation of proposed target compounds as well as others that may
have been prepared but are not readily amenable, for whatever reason, to experimental
analysis. These procedures can therefore play key roles in the assessment of potential
energetic performance.
REFERENCES
[1] P. Hohenberg and W. Kohn, Phys. Rev. B 136, 864 (1964).
[2] R. F. Stewart, J. Chem. Phys. 57, 1664 (1972).
[3] P. Politzer and D. G. Truhlar, eds., Chemical Applications of Atomic and Molecular
Electrostatic Potentials, Plenum Press, New York, 1981.
[4] G. Naray-Szabo and G. G. Ferenczy, Chem. Rev. 95, 829 (1995).
[5] P. Politzer and J. S. Murray, in Molecular Electrostatic Potentials: Concepts and
Applications, J. S. Murray and K. Sen, eds., Elsevier, Amsterdam, 1996, ch. 16.
[6] P. Politzer and J. S. Murray, Theor. Chem. Acc. 108, 134 (2002).
[7] P. Politzer and R. G. Parr, J. Chem. Phys. 61, 4258 (1974).
[8] P. Politzer, P. Lane and M. C. Concha, Int. J. Quantum Chem. 90, 459 (2002).

72

Molecular Surface Electrostatic Potentials in the Computational . . .

New Trends in Research of Energetic Materials

[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]

[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
[43]

[44]

Czech Republic

2006

[Content]

P. Politzer, Theor. Chem. Acc. 111, 395 (2004).


K. D. Sen and P. Politzer, J. Chem. Phys. 90, 4370 (1989); 91, 5123 (1989).
P. Politzer, J. S. Murray and P. Lane, J. Comp. Chem. 24, 505 (2003).
R. F. W. Bader, M. T. Carroll, J. R. Cheeseman and C. Chang, J. Am. Chem. Soc. 109,
7968 (1987).
J. S. Murray and P. Politzer, J. Mol. Struct. (Theochem) 425, 107 (1998).
P. Politzer and J. S. Murray, Trends Chem. Phys. 7, 157 (1999).
P. Politzer and J. S. Murray, Fluid Phase Equilib. 185, 129 (2001).
S. Iyer and N. Slagg, in Structure and Reactivity, J. F. Liebman and A. Greenberg, eds.,
VCH Publishers, New York, 1988, ch. 7.
J. Khler and R. Meyer, Explosives, 4th ed., VCH Publishers, New York, 1993.
C. B. Storm, J. R. Stine and J. F. Kramer, in Chemistry and Physics of Energetic
Materials, S. N. Bulusu, ed., Kluwer, Dordrecht (The Netherlands), 1990, ch. 27.
S. Zeman, Z. Kamensky, P. Valenta and J. Jakubko, Proceedings of the Study Days on
the Sensitivity of Energetic Components and Substances to Electrostatics, Aussois,
France, 1996, p. 197.
P. A. Urtiew and B. Hayes, J. Energet. Mater. 9, 297 (1991).
C. L. Mader, Numerical Modeling of Explosives and Propellants, 2nd ed., CRC Press,
New York, 1998.
P. Politzer and H. E. Alper, in Computational Chemistry: Reviews of Current Trends,
vol. 4, J. Leszczynski, ed., World Scientific, Singapore, 1999, ch. 6.
P. Politzer and S. Boyd, Struct. Chem. 13, 105 (2002).
P. Politzer and J. S. Murray, in Energetic Materials. Part 2. Detonation, Combustion,
P. Politzer and J. S. Murray, eds., Elsevier, Amsterdam. 2003, ch. 1.
T. B. Brill and K. J. James, Chem. Rev. 93, 2667 (1993).
L. Pauling, J. Chem. Soc. 1461 (1948).
A. M. Robbins, P. Jin, T. Brinck, J. S. Murray and P. Politzer, Int. J. Quantum Chem.,
in press.
P. Politzer, J. S. Murray and Z. Peralta-Inga, Int. J. Quantum Chem. 85, 676 (2001).
J. S. Murray, T. Brinck, P. Lane, K. Paulsen and P. Politzer, J. Mol. Struct. (Theochem)
307, 55 (1994).
O. Exner, Correlation Analysis of Chemical Data, Plenum Press, New York, 1988.
P. Lane, J. S. Murray and P. Politzer, J. Mol. Struct. (Theochem) 236, 283 (1991).
J. S. Murray, P. Lane and P. Politzer, Mol. Phys. 85, 1 (1995).
B. M. Rice and J. J. Hare, J. Phys. Chem. A 106, 1770 (2002).
J. S. Murray, P. Lane and P. Politzer, Mol. Phys. 93, 187 (1998).
P. Politzer and J. S. Murray, Mol. Phys. 86, 251 (1995).
P. Politzer and J. S. Murray, J. Mol. Struct. 376, 419 (1996), and references cited.
P. Politzer, P. R. Laurence, L. Abrahmsen, B. A. Zilles and P. Sjoberg, Chem. Phys.
Lett. 111, 75 (1984).
J. S. Murray, P. Lane and P. Politzer, J. Mol. Struct. (Theochem) 209, 163 (1990).
M. J. Kamlet and S. J. Jacobs, J. Chem. Phys. 48, 23 (1968).
J. J. P. Stewart, J. Mol. Model. 10, 6 (2004).
P. Politzer, P. Lane and M. C. Concha, Struct. Chem. 15, 473 (2004).
P. Politzer, P. Lane and M. E. Grice, J. Phys. Chem. A 105, 7473 (2001).
K. K. Irikura and D. J. Frurip, in Computational Thermochemistry, K. K. Irikura and D.
J. Frurip, eds., ACS Symposium Series 677, American Chemical Society, Washington,
1998.
W. Thiel, in Computational Thermochemistry, K. K. Irikura and D. J. Frurip, eds., ACS
Symposium Series 677, American Chemical Society, Washington, 1998.

Peter Politzer, Jane Murray

73

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

[45] P. Politzer, P. Lane and M. C. Concha, in Energetic Materials. Part 1. Decomposition,


Crystal and Molecular Properties, P. Politzer and J. S. Murray, eds., Elsevier,
Amsterdam, 2003, ch. 9.
[46] P. Politzer, J. S. Murray, M. E. Grice, M. DeSalvo and E. Miller, Mol. Phys. 91, 923
(1997).
[47] J. S. Murray and P. Politzer, in Quantitative Treatments of Solute/Solvent Interactions,
J. S. Murray and P. Politzer, eds., Elsevier, Amsterdam, 1994, ch. 8.
[48] P. Politzer, Y. Ma, P. Lane and M. C. Concha, Int. J. Quantum Chem. 105, 341 (2005).
[49] P. Politzer, J. S. Murray and M. E. Grice, Mater. Res. Soc. Symp. Proc. 418, 55 (1996).
[50] P. Politzer and P. Lane, Adv. Mol. Struct. Res. 3, 269 (1997).
[51] P. Politzer, M. C. Concha, M. E. Grice, J. S. Murray, P. Lane and D. Habibollahzadeh,
J. Mol. Struct. (Theochem) 452, 75 (1998).
[52] P. Politzer, P. Lane and J. J. M. Wiener, in Carbocyclic and Heterocyclic Cage
Compounds and Their Building Blocks, K. K. Laali, ed., JAI Press, Stamford, CT, 1999,
p. 73.
[53] P. Politzer, M. E. Grice and J. S. Murray, Recent. Res. Devel. Phys. Chem. 3, 95
(1999).
[54] B. M. Rice, S. V. Pai and J. Hare, Combust. Flame 118, 445 (1999).
[55] E. F. C. Byrd and B. M. Rice, J. Phys. Chem. A 110, 1005 (2006).
[56] D. R. Lide, ed., Handbook of Chemistry and Physics, 78th ed., CRC Press, New York,
1997.
[57] P. Politzer and J. S. Murray, J. Phys. Chem. A 102, 1018 (1998).
[58] P. Politzer, J. S. Murray, M. E. Grice and T. Brinck, J. Energ. Mater. 18, 89 (2000).

74

Molecular Surface Electrostatic Potentials in the Computational . . .

Lectures

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THE INFLUENCE OF GEOMETRY AND MATERIAL PROPERTIES


ON AN EXPLOSIVES GURNEY VELOCITY AND ENERGY
Joseph E. Backofen
BRIGS Co.
2668 Petersborough Street, Oak Hill, VA, USA 20171-2443

Abstract:
Experimental data from many scientific publications reveal that the copper cylinder test
used internationally to measure an explosives Gurney Velocity and Gurney Energy falls
within a unique combination of geometry and material properties factors affecting an
explosives measured performance. These data also support the need to use two
separate propulsion events to model detonation-driven propulsion: a brisant first stage
and a gas-dynamic second stage.
Keywords:

1.

detonation performance, detonation, explosive, cylinder test,


Gurney models, Gurney Velocity, Gurney Energy

INTRODUCTION

This paper presents and employs a practical means by which to model an explosives
detonation-driven propulsive effects by separating these into two parts: 1) initial motion
imparted by a brisant shock-dominated process that depends upon intimate contact of an
explosive with the propelled material, and 2) subsequent acceleration by a gas-push (gasdynamic) process. Initial motion is envisioned as being caused by the higher-pressure region
of a detonation front (i.e. envision the von Neumann spike or reaction zone region as being a
finite thickness of solid material squeezed at high pressure). The gas-push process is
envisioned similar to that assumed by Gurney modeling. The two-stage modeling employs
an engineering methodology consistent with the Zeldovich-von Neumann-Doering (ZND)
detonation model.
The need for a new model was experienced more than thirty years ago when studying
the effects of explosives during explosive welding, explosive compaction of powders, and
the development of commercial and military hardware. However, this need remained
unsatisfied with proprietary rules-of-thumb and practical experience being used as
substitutes for a rigorous well-defined model. For example, in the shaped charge
community, a liners initial velocity was presumed equal to zero in acceleration models
derived both analytically and through the use of hydrocode analyses even though the initial
free-surface velocity imparted by an explosives detonation against a plate formed the basis
of many equation-of-state measurements.[1-5] Similarly, the motion of material driven by a
grazing detonation (a detonation traveling along the surface rather than impacting
perpendicular to the surface) was taken to occur along the direction of half of the angle
through which a plate was bent a result of practical experience rather than analysis.[4,6]
During development of the BRIGS methodology for explosive device design and
analysis, old rules-of-thumb were used in its acceleration and Gurney models. An
eyeballed initial free-surface velocity (sometimes called the jump-off velocity) started
76

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

an exponential curve rising to that calculated using 86% of an explosives Gurney Velocity
with the asymmetric-sandwich formula when the travel distance was 5-mm and linearly from
there to a published 100% Gurney Velocity at 19-mm travel. These parameters represented
the two travel distances associated with the one-inch cylinder test at approximately 2 and 6.5
gas-volume expansions.[4,7] Although this model was only roughly approximate and even
discontinuous in nature, it generally worked well enough for more than twenty years.[8]
Some questions and problems still affect the use of Gurney modeling and development
of an analytical acceleration model. What is an explosives Gurney Velocity? Why is the
Gurney Velocity inconsistent with different materials? What should the initial free-surface
velocity be for grazing in contrast to head-on (perpendicular-impact) detonation waves?
How might these affect the Gurney Velocity? Should the factor be 86% at two gas-volume
expansions? This paper presents answers to some of these questions.

2.

THEORY, DATA, AND DISCUSSION


2.1

Gurney Models and the Gurney Velocity Constant

Gurney Energy modeling is commonly used for predicting the velocity of boundary
material debris from explosive-filled devices of various geometry assuming that a fraction of
an explosives chemical detonation energy is converted into the kinetic energy of both
boundary material debris and the explosives detonation products.[9] A negligible amount of
energy is generally assumed lost to deforming and fracturing the boundary material.[10,11]
The methodology assumes that the explosive was instantaneously converted to static
homogeneous all-burned high-pressure gaseous products which can expand at uniform
density with a linear velocity gradient within a system that expands from a momentum massbalanced virtual point or virtual plane at which some detonation products have zero velocity.
The explosives chemical detonation energy continues to be converted into kinetic energy
until the boundary debris attains a steady-state velocity a velocity from which one then can
calculate a Gurney Velocity and a Gurney Energy. These Gurney constants are now
typically obtained during experiments using a standard-size copper cylinder and then used to
approximate velocities which could be imparted to other boundary materials when using the
same explosive in different geometrical arrangements.[12] However, it has long been
recognized, as shown in Table 1, that Gurney constants are not single-value constants.[13]
Table 1. Comparison of Gurney Velocities derived from experiments using steel
cylinders versus those using copper cylinders [14,15]

Steel (m/s)

Copper (m/s)

Comp. A-3 (RDX)

2416

2630

Cyclotol (75/25 cast)

2320

2790

Comp. B

2310

2700

TNT (cast)

2040

2370

Tetryl

2209

2500

Joseph Backofen

77

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The Gurney formulas for simple, symmetric geometry and a relationship to approximate
the Gurney Velocity (Vg or (2Eg)1/2) using the detonation rate (D) in km/sec and adiabatic
expansion constant () for an explosives gaseous detonation products are:
Vf cyl = (2Eg) 1/2 [M/C + 1/2] -1/2 (cylinder)

(1)

Vf plate = (2Eg) 1/2 [M/C + 1/3] -1/2 (symmetric sandwich)

(2)

Vf sph = (2Eg) 1/2 [M/C + 3/5] -1/2

(sphere)

(3)

(Roths formula)

(4)

(2Eg)1/2 0.605 D / [ - 1]

Where M and C represent the masses of the inert boundary material and explosive,
respectively.[9,16,17] (Cooper provides an alternative approximation that the Gurney Velocity
is about 0.337 times an explosives detonation rate [3].)
One of the problems with so-called standard tests, such as the copper cylinder test, is
that once a procedure is accepted as a standard, very little other data is produced
examining variations of materials or geometry. Figure 1 contains Gurney Velocity data for
Composition B explosive which were derived using final steady-state cylinder expansion
velocities in equation (1) showing that the data also depend upon the cylinders material and
wall thickness.[18] These data clearly show that experiments using steel cylinders with a
cylinder wall areal density (tcyl cyl, in units of g/cm2) equal to that of copper cylinders yield
a lower Gurney Velocity. The data also show that aluminum cylinders behave similar to
steel cylinders.
7.0
Aluminum in Air

6.5

Gurney Velocity (km/s)

6.0
Aluminum in Helium

5.5
5.0

Steel in Air

4.5
Copper in Air

4.0
3.5
3.0
2.5
2.0
0

Cylinder Wall Areal Density (g/cm2)

Fig 1. Gurney Velocity versus Cylinder Wall Areal Density for Comp. B Explosive
The data in Fig.1 imply that explosive energy is somehow lost when steel and aluminum
are used instead of copper or that at least more energy is lost to a process not included in
Gurney modeling.

78

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

1.2

Czech Republic

2006

[Content]

Gurney Velocity Trends affected by Thickness and Material

To date, fifteen published papers have been found to contain additional, useful cylinder
test data.[19] Mostly, the data were from so-called standard, scaled copper-cylinder
expansion experiments taken at the time at which the gaseous detonation products had
expanded to about seven times their initial volume. However, some data were available for
copper cylinders of greater or lesser thickness as well as for other materials. Where the data
were good enough, the final steady-state velocity was used to calculate the Gurney
Velocity (Vg) if it was not explicitly provided in the paper.
The useful Gurney Velocity data were examined to see if a pattern might emerge
revealing how either cylinder geometry or materials may have affected the experimental
results. Figure 2 presents a portion of this data where the cylinder wall areal density is
normalized by an expression for explosive areal density (Rex ex) using the cylinders inner
radius in cm and explosive density in g/cm3 along a radius through the cylinder wall. (Only
some representative data were chosen from the many standard cylinder tests available for
modern explosives and their mixtures with binders since most copper-cylinder tests have
been performed at (tcyl cyl / Rex ex) 1). Figure 2 reveals that the data suggest a trend.

0.001

0.010

Gurney Velocity / Detonation Rate (Vg / D)

Comp.B in Aluminum

0.100

1.000
1.000

Vg / D = 0.302 ( tcyl cyl / R ex ex ) (--

10.000

5 / 30 )

Comp.B in Steel
Comp.B in Copper
RDX in Steel
RDX in Copper
Octol 75/25 in Copper
LX-14 in Copper
PBX9404 in Copper
LX-04 in Copper
TATB in Copper
LX-17 in Copper
PETN in Copper
Trendline

0.100

Ratio of Cylinder Wall Mass to Explosive Mass in the Radial Direction ( tcyl cyl / Rex ex )

Fig 2. Ratio of an Explosives Gurney Velocity to Its Detonation Rate Compared to the
Normalized Cylinder Wall Areal Density
The ideal situation assumed for using Gurney modeling is rarely attained in engineering
practice. For example, premature bursting of a cylindrical or spherical shell can permit the
detonation products to escape before they can perform the same level of gas-dynamic work
that is usually found during standard copper cylinder tests. However, data exhumed from
scientific publications show in Fig.3 that similar material property and geometry effects
appear in fragmentation test data even though the cylinder wall velocities were measured at
4

Joseph Backofen

79

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

time of cylinder fracture rather than after a fully-contained gas-dynamic volumetric


expansion to a steady-state velocity.[20-23] These data from tests involving five explosives
(TNT, Comp B, Octol, Baratol, and an unknown explosive) and five materials (aluminum,
brass, copper, steel, and lead) trend parallel to the previous trend equation. (For clarity,
Fig.3 is a semi-log plot containing only a subset of the data.)
0.40

Gurney Velocity / Detonation Rate (Vg / D)

Vg / D = 0.302 ( tcyl cyl / Rex ex ) ( -- 5 / 30 )

0.35

0.30

Comp B in Aluminum 2S
Comp B in Unk. Aluminum
Unknown in Aluminum D16

0.25
Comp B in Copper
Unknown in Copper

0.20

Unknown in Brass
Unknown in Lead
Calc. Comp B in Aluminum

0.15
0.10

1.00

10.00

Normalized Cylinder Wall Areal Density ( tcyl cyl / Rex ex )

Fig 3. Gurney Velocity Data from Measurments of Cylinder Wall Velocity at Fracture
Similar effects from material properties and geometry were found in experiments using
spherical charges. Figure 4 presents data extracted from four papers in comparison to data
from cylinder tests using comparable explosives and materials.[24-27] (Filled symbols
represent data from cylinder tests while open symbols present data from experiments
involving spheres.) If one were to disregard effects caused by the dimensions and properties
of the explosive and confining material, one might estimate that the ratios of the Gurney
Velocities to the detonation rates fall within +/- 17% of 0.350 with the spread representing
experimental error. (This is not much different from 0.337 found by Cooper from cylinder
test data representing sixteen military explosives.[3]) However, it is rather clear in Fig.4 that
the Gurney Velocities derived from the experiments using spherical charges demonstrate a
distribution and a trend similar to those trends found for cylindrical charges. (The trend for
spherical charges has been marked using 0.250 vice 0.302 in the trend formula.) Thus, these
data appear also to confirm that the relative dimensions and properties of the confining
material and explosive play a role beyond that expressed only by the mass-to-charge ratio
available in Gurney formulas.

80

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

0.45
Comp B in Al 2S Cyl
Comp B in Al Sphere

Gurney Velocity / Detonation Rate (Vg/D)

TG 50/50 in Al Sphere
PBX9404 in Al Sphere
LX-04 in Al Sphere

0.40

RDX/TNT in Al Sphere
Comp B in Cu Cyl
PBX9404 in Cu Cyl
LX-04 in Cu Cyl
TG 50/50 in Cu Cyl

0.35

RDX/TNT in Cu Sphere
Comp B in Steel Cyl
LX-04 in Mild Steel Cyl
LX-04 in Stainless Steel Cyl
Unk Expl, Stainless St. Sphere
TG 50/50 in Steel Sphere

0.30

RDX/TNT in Zn Sphere
RDX/TNT in Pb Sphere
Unk Expl in U Sphere
Calc. Comp B in Al Cyl
Calc. Comp B in Al Sphere

0.01

0.10

0.25
1.00

Normalized Boundary Areal Density (tdevice matl / Rex ex)

Fig 4. Normalized Gurney Velocity Data for Some Explosives Detonated in Cylinders and
Spheres of Different Materials
The following equation can be used to include the trends identified in Figs.2-4 into
Gurney modeling by expanding Roths formula so that the currently known effects of
geometry and material properties can affect a final steady-state velocity calculation:
Vg / D = A (tcyl cyl / Rex ex ) B

(5)

Where:
A = 0.302 for Comp B explosive in aluminum cylinders (or for = 3 explosives),
A 0.605 / [ - 1] for explosives in copper cylinders (Roths approximation), and
B = (-5 / 30).
For spherical and plate geometry, sph, tsph and pl, tpl are substituted for cyl and tcyl with
Tex also replacing Rex for plate geometry.
It must be noted, however, that extensive experimental data must yet be collected in
order to fully identify and affirm the relationships between geometry (planar, cylindrical,
spherical, wall thickness and explosive thickness), material properties, and explosives that
affect the total propulsion imparted by an explosives detonation.

Joseph Backofen

81

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Brisant Propulsion the 1st Stage Propulsion Effect Producing the


Initial Free-Surface Velocity

1.3

The fact that both material properties and the normalized areal density appear to affect
detonation propulsion performance brought into consideration the initial shock effects
measured long ago during equation-of-state experiments. As previously mentioned, the
initial free-surface velocity (Vi) which is affected by shock waves during such experiments
appears to be important to the acceleration of a boundary material.

ETRi (head-on) Vi/D * (RhoP/RhoE) ^ 0.5

In ref.[5], the initial velocity imparted by the shock action of perpendicular impact of a
detonation wave with plates was found to be predictable by two simple equations. These
equations resulted from analysis of over 600 summarized data-sets representing data from
thousands of experiments involving ten materials and nineteen explosives as plotted in Fig.5.
0.7

Uranium

Plexiglass

Reasonable Charges &


Very Thin Plates

Zinc

0.6

Magnesium
Brass

0.5

Dural

Thick Plates &


Thin Charges

Copper
Dural

0.4

Reasonable Plates &


Very Long Charges
0.3

0.417 * (t/L) ^ -3/40


Aluminum

Reasonable Plates &


Reasonable Charges

0.380 * (t/L) ^ -0.30

0.2
0.001

0.01

0.1

Plate Thickness / Charge Length

Brass

10

(t/L)

Fig 5. Free-Surface Velocity Ratio (ETRi) versus the Ratio of Plate Thickness to Charge
Length for Detonation Waves Impacting Perpendicular to a Plate
As described in ref.[5], the initial effort attempted to use a conventional momentum
balance concept in order to create a two-stage propulsion model suitable for describing
coupling of detonation waves to inert boundaries as a function of the angle between the
detonation front and the boundary. However, the following Energy Transference Ratio
(ETR) was found to best represent the wide range of experimental data shown in Fig.5 in the
region within which explosives might be employed in practical devices.
ETRi = (Vi / D) (cyl / ex) 1/2 = 0.417 (tcyl / Rex) -3/ 40

(6)

(For plate geometry, pl, tpl, and Tex are used for plate density, plate thickness, and explosive
thickness, respectively.)
In contrast to the perpendicular impact of a detonation wave, there is a paucity of freesurface initial velocity data driven by a grazing detonation simply because of two factors: 1)
no practical applications seemed to require such free-surface velocity data, and 2) such data
82

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

were very difficult to acquire and to interpret before reliable Fabry-Perot interferometry was
applied to detonation wave coupling experiments. Very simply, it was difficult to determine
the instantaneous free-surface velocity while material rotated through the initial bend
angle. The limited amount of data available in scientific literature, however, has been
summarized in Fig.6 [28]. As noted in ref.[28] and Fig.6, the data can be approximately
represented by equation (6) using 0.2085 (half of 0.417). This also provides a simple means
by which to estimate the initial free-surface velocity as a function of the angle between the
detonation wave and the driven boundary as it changes from a gazing to perpendicular
impact wave.

ETRi (grazing)

Vi/D (RhoP/RhoE)^0.5

0.40
0.35
Magnesium

Tantalum
0.30

0.189 (tp/Te) ^-3/8


0.25

0.2085 (tp/Te) ^-3/40


0.20
0.15

Aluminum

Uranium, Lead
Plexiglas, Antimony
Copper

0.10
Iron
0.05

0.00
0.001

0.01

0.1

10

Plate Thickness / Explosive Thickness (tp/Te)

Fig 6. Free-Surface Velocity Ratio (ETRi) versus the Ratio of Thickness to Charge
Thickness for Grazing Detonation Waves Running Along a Plate or in a Cylinder
The initial free-surface velocity equation can also be used to analyze the effect that the
initial acceleration might play in overall detonation-driven propulsion. Substituting
equations (5) and (6) into (1) and (2) provides the following ratios between the initial
velocity (Vi) and the final steady-state velocity (Vf) for a = 2.75 explosive:
(Vi/Vf) cyl = (0.2085 / 0.3457) (tcyl /Rex) -3/ 40 [(tcyl /Rex)2 + 2 (tcyl /Rex) + 0.5 (ex / cyl)] 1/2
(Vi/Vf) plate = (0.2085 / 0.3457) (tpl /Tex)

-3/ 40

[2 (tpl /Tex) + 0.333 (ex / pl)]

1/2

(7)
(8)

Equation (6) is also used in Fig.7 using 0.2085 in order to estimate the initial velocity for
some cylinder tests for which only final velocity data were available. Since aluminum was
used in most of the experiments from which equation (6) was derived, it was used to
calculate Vinitial for the aluminum cylinders. Values calculated for steel cylinders were
simply divided by 2 to approximate the decreased initial velocities observed in currently
available experiment data. As can be seen in Fig.7, this simple approximation for the energy
lost to the iron to phase transition an effect well known since the 1950s produces
data which nearly overlays some experiment data. (See Table 2 for some other materials
8

Joseph Backofen

83

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

commonly used in explosive-driven devices and susceptible to phase transitions at pressures


less than an explosives Chapman-Jouguet detonation pressure.)
Table 2. Phase Transition Pressure for Some Materials Commonly Used with
Explosive Charges
Transition Pressure (GPa)
Aluminum

20.5

Carbon (pressed graphite)

23

Iron (0.2 wt% Carbon)

14.7

Iron (0.5 wt% Carbon)

13

Titanium

9.4

Zirconium

23

0.9

Copper

0.8

Iron
0.7

Magnesium
Tantalum

V initial / V final

0.6

Fe (est.Vi/2)
Al (est.Vi)

0.5
Cu Cyl Calc.
Fe Cyl Calc.1.9
0.4
Al Cyl Calc.
Cu (BN03-5)

0.3

Steel (BN03-5)
Ta (BN03-5)

0.2

Al (BN03-5)
0.1
0

0.1

0.2

0.3

0.4

0.5

Ratio of Cylinder Wall to Explosive Radius ( tcyl / Rex ) or


Plate Thickness to Sandwich Explosive Thickness ( tp / Tex )

0.6

0.7

= 2.75 for Calc.

Fig 7. Vinitial / Vfinal Data for Cylinders and Plates of Various Inert Materials Subjected to
Grazing Detonation by Various Explosives
Fig.7 reveals that the initial free-surface velocity is not only a major portion of the total
final steady-state velocity but also their ratio (Vi/Vf ) is affected by the properties and
thickness of the boundary materials. As shown in Fig.7, approximately 50% of the final
velocity (representing approximately 25% of the final kinetic energy) is imparted to the
cylinder wall by the initial brisant process during experiments having the geometry ratios
employed during typical standardized cylinder tests the regions of 0.1 and 0.2 thickness
ratio. Other authors have also recorded similar observations for plates accelerated by
84

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

grazing detonations. [29-31] It remains for future research, however, to show whether this is
the reason why a shaped charge liners motion follows a trajectory at approximately one half
of the total bend angle (Taylor angle).

Gas-Dynamic Propulsion the 2nd Propulsion Process Responsible


for Most of the Final Kinetic Energy

1.4

The findings for the ratio Vi/Vf suggest that propulsion during the gas-dynamic (gaspush) stage should also be modeled using Vf to normalize the instantaneous velocity (Vx,
Vt, or Ver) respectively coordinated to distance, time or the explosives gaseous product
expansion ratio.[32] Figure 8 presents data for sixteen (ideal and non-ideal) explosives
demonstrating the similarity by which copper cylinders are propelled when Ver is
normalized by Vf as an explosives gaseous products expand from 1.0 < Expansion Ratio
(ExR) < 13. However, the similarity might actually be much better than shown in Fig.8
since its data are normalized using published Vf values values mostly measured at arbitrary
final expansion ratios (ExRf), such as 6.5 or 7. Such published Vf do not necessarily
represent the maximum velocity that could have been reached during experiments at larger
expansion ratios.
1.05

Velocity / Final Velocity ("Steady-State")

0.95

0.85

0.75

TNT

Comp B

Cyclotol

PETN

HMX

LX-04

LX-07

LX-09

LX-10

LX-11

LX-14

H-6

H-6 fine

9404

9011

Nitromethane

0.5_ExpR10

0.45_ExpR10

0.65

0.55

0.45
0.39_ExpR10

0.35
1

11

13

Expansion Ratio (Volume / Initial Volume)

Fig 8. Normalized Velocity Data for Sixteen Explosives Driving Copper Cylinders
The data shown in Fig.8 demonstrate a rapid velocity rise from the initial free-surface
velocity during 1< ExR < about 1.5. After this, the velocity rises at a less steep slope until
eventually reaching Vf at ExRf. This implies that two terms should be chosen to represent
such velocity data:
Ver / Vf = V1 / Vf + V2 / Vf
V1 / Vf = (Vi / Vf) {(e -- ExR e ExR ^ 3 ) + ([ExR / ExRf ] -- 0.5 / [1 / ExRf ] -- 0.5 )}
10

Joseph Backofen

(9)
(10)
85

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

V2 / Vf = [1 {(Vi / Vf) / (1 / ExRf) -- 0.5 }](ExR / ExRf) -- 0.333 [log (ExR) / log (ExRf)] (11)

The first term (V1/Vf) appears to represent a lingering effect from the brisant 1st
propulsion stage imparting the initial free-surface velocity; and the second term (V2/Vf)
appears to represent acceleration by detonation product gas pressure during their gaseous
expansion. Some experimental data exhumed from published literature appear to support
that typical ideal explosives continue pushing out to expansions beyond 10 and nonideal explosives towards 14 as in the case of some aluminized explosives. (See also Fig.9.)
This means that most data in textbooks and technical papers probably do not provide a sound
basis for accurately modeling an explosives gas-dynamic propulsion since data are almost
always only measured up to or normalized to ExRf = 6.5. (In other words, cylinder wall
movement was normalized to travel of 19 mm during a standard 1-inch copper cylinder
test.) Nevertheless, equation (9) offers a means to estimate velocity versus time, expansion
ratio or boundary movement by means of direct measurement of Vi and Vf at ExRf. (The
former velocity can be measured using a Fabry-Perot interferometer and the latter velocity
using a streak camera.)

Velocity / Final Velocity ("Steady-State")

1.05

0.95

0.85
Nitromethane

0.75

Pourex
Aquanal

0.65

Unigel
ANFO

0.55
0.39_ExpR10
0.45_ExpR 8

0.45

0.35
1

11

13

Expansion Ratio (Volume / Initial Volume)

Fig 9. Normalized Velocity Data for Five Commercial Explosives Driving Copper
Cylinders
Figure 10 shows that the time derivative of equation (9) creates a smoothly varying
equation comparable to a JWL equation-of-state (EoS).[33] As shown in Fig.10, both are
found to be comparable for military explosives such as LX-07. However, there are distinct
differences. Firstly, the new equation only represents the gas-push 2nd stage propulsion
cycle whereas the JWL model was formed to describe the entire acceleration provided by the
explosive. Thus, the JWL EoS must initially employ a high gas pressure in order to play
catch-up since the JWL method assumes an initial velocity of zero rather than recognizing
86

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

an initial free-surface velocity. The JWL EoS pressure also drops more quickly to a
negligible pressure well before the time ExR reaches 6.5 whereas the time derivative of
equation (9) continues a decaying pressure until the expansion ratio at which the final
steady-state velocity is achieved in experiments.

Pressure (JWL Along Isentrope) Compared To Acceleration


Equation

0.14

0.12

Time Derivative of Eqn 4; Vi / Vf = 0.52; ExRf = 6.5


0.10

JWL EoS for LX-07 Explosive


JWL EoS for Comp B Explosive

0.08

Time Derivative of Eqn 4; Vi / Vf = 0.6; ExRf = 6.5


0.06

0.04

0.02

0.00
1.0

2.0

3.0

4.0

5.0

6.0

7.0

Expansion Ratio

Fig 10. Comparison of JWL Calculated Pressures for Two Explosives to the Time
Derivative of Equation (9) for These Explosives Driving Copper Cylinders

3.

SUMMARY AND CONCLUSIONS

This paper has shown that experimental data reveal that geometry and material
properties affect an explosives performance beyond the mass-to-charge ratio used in Gurney
modeling. As a result, these factors affect the measurement of an explosives Gurney
Velocity and its corresponding Gurney Energy parameters used not only in analytical
formula Gurney modeling but also in the formulation of gas-dynamic equations-of-state.
This paper has also provided a new model a two-stage detonation propulsion model
capable of describing a materials velocity beginning from its initial free-surface velocity,
during its gas-dynamic-driven acceleration, and at its final steady-state velocity. However,
the experiments and their data upon which this model and previous Gurney modeling in
scientific literature are based are not well described, are not statistically significant in and by
themselves, and contain measurement limitations and errors that affect the accuracy of work
based on them. Nevertheless, the new model and its equations can be used to guide
experiments to create a consistent, robust database from which future Gurney Velocity and

12

Joseph Backofen

87

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Gurney Energy data can be extracted. Such work to clarify, to fully understand, and to
accurately describe how a detonating explosive actually provides propulsion needs to:

Measure the brisant 1st stage propulsions initial free-surface velocity,

Thoroughly measure the gas-push 2nd stage propulsion out to large expansion ratios,

Conduct the experiments over a wide range of explosives, materials, geometry


(planar, cylindrical, spherical), and normalized areal densities.

REFERENCES
[1]

[2]

[3]
[4]
[5]

[6]
[7]
[8]
[9]

[10]
[11]
[12]

[13]
[14]
[15]

[16]

[17]

88

G. RANDERS-PEHRSON: An Improved Equation for Calculating Fragment Projection Angles,


Proc. of 2nd Int. Symp. Ballistics, 9-11 March, Daytona Beach, FL, American Defense
Preparedness Assoc., Washington, D.C., 1976
P.C. CHOU, J. CARLEONE, E. HIRSCH, W.J. FLIS, R.D. CICCARELLI: Improved Formulas for
Velocity, Acceleration, and Projection Angles, Proc. Sixth Int. Symp. Ballistics, J.E. Backofen,
Jr., Ed., 27-29 Oct., Orlando, FL., American Defense Preparedness Assoc., pp. 286-296, 1981
P.W. COOPER: Explosives Engineering, VCH Publishers, Inc., New York, 1996
J.A. ZUKAS, W.P. WALTERS, Eds: Explosive Effects and Applications, Springer-Verlag, New
York, 1998
J.E. BACKOFEN, C.A. WEICKERT: The Effects of Plate Thickness and Explosive Properties on
Projection from the End of a Charge, Proc. 16th Int. Symp. Ballistics, 23-28 Sept., San
Francisco, CA, American Defense Preparedness Assoc., Rosslyn, VA, pp. 641-650, 1996
W.P. WALTERS, J.A. ZUKAS: Fundamentals of Shaped Charges, John Wiley & Sons, New
York, 1989
M. SUCESKA: Test Methods for Explosives, Springer-Verlag, New York, 1995
J.E. BACKOFEN: The Use of Analytical Computer Models in Shaped Charge Design,
Propellants, Explosives, Pyrotechnics, 18, pp. 247-254, 1993
J.E. KENNEDY: Explosive Output for Driving Metal, (Behavior and Utilization of Explosives in
Engineering Design, Proc. 12th Annual Symp. of the ASME (New Mexico Section), L.
Davison, J.E. Kennedy, F. Coffey, Eds.), Albuquerque, NM, pp. 109-124, 1972
J.E. REAUGH, P.C. SOUERS: A Constant-Density Gurney Approach to the Cylinder Test,
Propellants, Explosives, Pyrotechnics, 29(2), pp. 124-128, 2004
W.J. STRONGE, MA XIAOQING, ZHAO LANTING: Fragmentation of Explosively Expanded Steel
Cylinders, Int. J. Mech. Sci., 31(11/12), pp. 811-823, 1989
J.E. KENNEDY: The Gurney Model of Explosive Output for Driving Metal, (Explosives Effects
and Applications, J.A. Zukas, W.P. Walters, Eds.), Springer-Verlag, New York, pp.221-257,
1998
D.R. KENNEDY: The Elusive (2E) , 21st Annual Bomb & Warhead Section Meeting,
American Ordnance Assoc., Picatinny Arsenal, NJ, 22 Oct. 1969
D.R. PRICE: Dependence of Damage Effects Upon Detonation Parameters of Organic High
Explosives, Chem. Rev., pp. 801-825, 1959
B.M. DOBRATZ, P.C. CRAWFORD: LLNL Explosives Handbook: Properties of Chemical
Explosives and Explosive Simulants, UCRL-52997, Lawrence Livermore National Lab., 31 Jan.
1985
J. ROTH: Correlation of the Empirical Gurney Constant with Detonation Parameters of the
Driver Explosive, Proc. Int. Symp. Pyrotechnics and Explosives, 12-15 Oct., Beijing, China,
China Academic Publishers, pp. 629-634, 1987
J. ROTH: The Adiabatic Exponent a Key to the Numerical Evaluation of the CJ State, Proc.
24th Int. Pyrotechnics Seminar, Chicago, pp. 469-481, 1998

The Influence of Geometry and Material Properties on an Explosives Gurney . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

[18] J.E. BACKOFEN, C.A. WEICKERT: Effect of an Inert Materials Thickness and Properties on the
Ratio of Energies Imparted by a Detonations 1st and 2nd Propulsion Stages, (Shock
Compression of Condensed Matter-2001, M.D. Furnish, N.N. Thadhani, Y. Horie, Eds.), Amer.
Inst. Physics, pp. 954-957, 2002
[19] J.E. BACKOFEN: The Effects of Cylinder Geometry and Material on Gurney Velocities and GasPush Gurney Velocities Measured During Cylinder Test Experiments, BRIGS Report 02-1,
BRIGS Co., Oak Hill, VA, 24 Feb. 2002
[20] J.E. BACKOFEN: Confirmation of the Effects of Cylinder Wall Thickness and Material
Properties on Measurement of an Explosives Gurney Velocity, BRIGS Note 03-1, BRIGS Co.,
Oak Hill, VA, 9 Feb. 2003
[21] J.E. BACKOFEN: Additional Information on How a Cylinders Wall Thickness and Material
Properties Can Affect the Measurement of an Explosives Gurney Velocity, BRIGS Note 03-2,
BRIGS Co., Oak Hill, VA, 23 March 2003
[22] M. SINGH, H.R. SUNEJA, M.S. BOLA, S. PRAKASH: Dynamic Tensile Deformation and
Fracture of Metal Cylinders at High Rates of Strain, Int. J. Impact Engng., 27(9), Oct., pp. 939954, 2002
[23] A.I. VOROBEV, M.S. GAINULLIN, G.V. ZLYGOSTEV, A.P. RYBAKOV: Experimental
Investigation of the Motion of Cylindrical Shells Under the Action of the Products of an
Explosion in a Cavity, J. Appl. Mech. Tech. Phys., pp. 872-877, 1976 (trans. Zh. Prik. Mekh.
Tekh. Fiz., No.6, Nov.-Dec., pp. 165-169, 1974)
[24] P.M.B. SLATE, M.J.W. BILLINGS, P.J.A. FULLER: The Rupture Behaviour of Metals at High
Strain Rates, J. Inst. Metals, 95, pp. 244-251, 1967
[25] E.L. LEE, H.C. HORNIG, J.W. KURY: Adiabatic Expansion of High Explosive Detonation
Products, UCRL-50422, TID-4500, UC-4, Chem., Lawrence Radiation Lab., Univ. California,
Livermore, CA, 2 May 1968
[26] F. OLIVE, A. NICAUD, J. MARILLEAU, R. LOICHOT: Rupture Behaviour of Metals in Explosive
Expansion of Shells, Inst. Phys. Conf. Series No.47: Ch.2, pp. 243-251; and M. STELLY:
Metallurgical Aspects of the Dynamic Expansion of Shells, ibid., pp. 252-253, (Mechanical
Properties at High Rates of Strain, J. Harding, Ed.), Inst. Phys., Bristol and London, 1979
[27] A.I. IVANOV, M.A. SYRAININ, A.G. FEDORENKO, A.P. TSOI: Fragmentation of Spherical Shells
Under Blast Loading, Strength of Materials, 33(2), pp. 150-156, 2001 (trans. Prob. Proch.,
No.2, Mar.-Apr., pp. 78-87, 2001).
[28] J.E. BACKOFEN, C. WEICKERT: Initial Free-Surface Velocities Driven by Grazing Detonation
Waves, (Shock Compression of Condensed Matter-1999, M.D. Furnish, L.C. Chhabildas, R.S.
Hixon, Eds.), Amer. Inst. Physics, pp. 919-922, 2000
[29] O.B. DRENNOV, A.I. MIKHAILOV: Initial Stage in the Acceleration of Thin Plates in the
Grazing Detonation Mode of a High Explosive, Fiz. Gor. Vzry. 15(4), July-Aug., pp. 143-146,
1979, (trans. Comb. Expl. Shock Waves, 15(4), pp. 539-542, 1980)
[30] V.V. KISELEV: Estimation of the Properties of Metal Plates Propelled by the Sliding
Detonation of Charges of Condensed Explosives in the Initial Phase of the Process, Fiz. Gor.
Vzry. 31(1), Jan.-Feb., pp. 138-142, 1995, (trans. Comb. Expl. Shock Waves, 31(1), pp. 134137, 1995)
[31] A.A. DERIBAS: Acceleration of Metal Plates by a Tangential Detonation Wave, Prik. Mekh.
Tekh. Fiz., 41(5), Sept.-Oct., pp. 68-74, 2000, (trans. J. Appl. Mech. Tech. Phys., 41(5), pp.
824-830, 2000)
[32] J.E. BACKOFEN: Modeling a Materials Instantaneous Velocity during Acceleration Driven by
a Detonations Gas-Push, in publication (Shock Compression of Condensed Matter-2005),2006
[33] H. HORNBERG: Determination of Fume State Parameters from Expansion Measurements of
Metal Tubes, Propellants, Explosives, Pyrotechnics, 11, pp. 23-31, 1986

14

Joseph Backofen

89

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

COMPUTER SIMULATION OF THERMAL DECOMPOSITION


MECHANISM FOR COMPOUNDS WITH NITROGUANIDINE
FRAGMENT
Eugeniya A. Bakhmatova*, Vyacheslav L. Korolev*, Aleksey A. Porollo**,
and Tatyana S. Pivina*
*

Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences,


Leninsky Prospect 47, Moscow 119991, RUSSIA
Fax: +7 (495) 135 5328, E-mail: tsp@ioc.ac.ru
**

Biomedical informatics, Childrens Hospital Research Foundation,


3333 Burnet Av, Cincinnati, OH 45229, USA

Abstract:
The application efficiency of energetic compounds depends on the set of their
characteristics. The most important of them is a thermal stability, which is connected
with the thermal decomposition mechanism of compounds.
Nitroguanidine and his analogs have been of interest as an example of energetic
compounds. However, currently there is no general view for the thermolysis of
nitroguanidine analogs. Having labile hydrogen atoms, theoretically, nitroguanidine
and his analogs may exist in different tautomeric forms. Meanwhile, there are no
experimental evidences if this or other tautomeric form domination during a
decomposition process.
In order to fill this gap, the simulation of mechanism of all nitroguanidine tautomers
and 3-nitramine-1,2,4-triazole thermal decomposition was carried out. Subsequent
evaluation of different tautomeric forms in terms of thermodynamic stability and
activation energy for initial steps of their decomposition reactions has been conducted
using DFT approach (B3LYP/6-31G*). Thermo-chemical preferences of some
decomposition pathways have been determined.

Keywords: nitroguanidine, 1,2,4-triazole, simulation of mechanism, decomposition


reaction.

1.

INTRODUCTION

As is known, experimental investigations of multi-stepped thermolysis processes for


energetic materials are complicated. That is why in a set of works a theoretical approach to
simulate mechanisms of organic compounds thermolysis processes has been proposed /1-5/.
In the present work, this approach has been used for nitroguanidine (1) and 3-nitramino1,2,4-triazole (2) thermolysis channels modeling (the latter compound was considered
because it contains nitroguanidine in the nitrimine form as a fragment of the structure).
90

Computer Simulation of Thermal Decomposition Mechanism for Compounds . . .

New Trends in Research of Energetic Materials

H2 N
H2 N

Czech Republic

N NO2

N
NH

2.

[Content]

NH NO2

2006

METHODS OF INVESTIGATIONS

Based on analysis of experimental data on mechanisms of thermal decomposition


reactions for nitroguanidine and 1,2,4-triazoles, a generator of hypothesis for the
mechanisms of the compounds thermolysis has been elaborated. At the heart of the
methodology was the representation of decomposition processes as Recombination Reaction
Nets (RRN). The approach was realized in the CASB computer program /1-2/ and allows
one to get the whole spectrum of reactions occurring in the course of compounds thermal
decomposition. Having probable mechanism, to evaluate the thermochemical preference of
these or other pathways of decay, quantum-chemical calculations of the activation energy for
reactions at the first step of decomposition were carried out. All calculations were
accomplished by the method of density functional (DFT) using hybrid potential B3LYP and
standard valence-split basic 6-31G* /6/. Calculations were executed using GAUSSIAN 98
/7/ in the Computational Centre of Zelinsky Institute of Organic Chemistry, Russian
Academy of Sciences.

3.

RESULTS ANF DISCUSSION

For nitroguanidine (1) it is possible ten tautomeric forms and rotational isomers.
Experimental evidences of one or other tautomer prepotency under thermolysis conditions
are lacking. Apparently, the opportunity of one or another tautomer or their mixture
existence depends on particular conditions. That is why the simulation of nitroguanidine
thermolysis mechanism was carried out for all (1-1) tautomeric forms.
As the measure of isomer stability the calculated tautomerisation energy /9/ of
different forms we used. The results (Table 1) show that in gas phase the tautomer (1b) is the
most thermodynamic stable: its total energy is the least relatively to all others forms.
Table 1.

Energy characteristics of nitroguanidine tautomers (1a-c).

Total
energy
(E), ..

Compounds

Tautomerization
energy,
kcal/mol

Total
energy
(E), ..

Tautomerization
energy,
kcal/mol

23.4

1c4 O

409.739156

Compounds
NH2H
HN

HN
C
H2 N

H
N

1a

NO2

409.75874

11.09

O
N

Eugeniya Bakhmatova, Vyacheslav Korolev, Aleksey Porollo, Tatyana Pivina

91

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

NH
H2 N

1b

H2 N

NO2

H2N

0.00

O
N
1c5

409.77642

N
O

20.7

409.743446

0
NH

NH2
HN

O
N
1c1

N
O

H2N

20.8

O
N

409.74323

1c6

NH2
O

1c2

H2N

20.7

O
N
1c7

409.74336

23.9

409.738276

NH H
O

1c3

NH2
N

409.743129

NH

HN

HN

20.9

H2N

24.0

1c8

409.73813

O
N

25.3

409.736059

The energy of tautomerization was calculated by formula [(E + ZPE) (E +


ZPE)]627.5 kcal/mol /9/, where ZPE is a correction for the energy of zero-point
oscillations, x the data for the tautomer (1b) and y the data for any other isomers.

Computer generation of possible pathways for (1b) decomposition gave three


hypotheses of the initial stage (Fig. 1): homolytic cleavage of nitro-group (channel 1), NH2
group breakage (channel 2), and nitro-nitrite rearrangement (channel 3).

H 2N

H 2N

NO2
3

H2N
H2N

Fig. 1.

92

NO2

H2 N

NO2

NH2

H2 N

NO

H2 N

The initial stages of nitroguanidine (1b) thermal decomposition.

Computer Simulation of Thermal Decomposition Mechanism for Compounds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The results of activation energy (Ea) calculations of the presented decomposition


reactions shown the thermochemical preference of (1b) destruction along the channel 1 with
= 46.4 kcal/mol. The radical NH2 cleavage energetic input (channel 2) is higher (=56.1
kcal/mol). Unfortunately, we failed to localize the nitro-nitrite rearrangement transition state
(channel 3). However, according to /10/ for N-nitro derivatives this rearrangement is
thermochemicaly less favorable than homolytic cleavage of nitro-group.
At the Fig. 2 the fragment of consequent stages for nitroguanidine (1b)
decomposition along the channel 1 is presented as an illustration of generated net
intermediates and of the order of their transformations.

H2N

NO2

H2N
H2N

H2N

H2 N

+HONO

C
NH2
O

H2N

+M

+M

ONO

HN

HNO O

+M

H
N

H2O

NH2
HN

C
H

N
N2

NH

NH2

O O

N2

H2 O

CO

NH3
+HONO

H2O
H2O

NH2

NH2 N

+M

NH3
NH2

CO2

CO2
HNO

-NO

H2O

+ONO

CO
+NO2

N2 O

NH2

+NO

CO2
N2O

+NO2-NO

O
+M

N2

H2 N

OH

+M

H2O

H2O

+M

+NO2

N2O

CHO
+R

CH2O
+N O

CO2
H2N

N
OH
H2 N

+M

NH2

NH3

HCN

+M

+M

+2NO

H2O HNO NO2 N2O

NH2

H2N

+HONO

N2O OH

NO
HNO O

+ONO

O
HN

H2O

NH2

CN

NO
N

NO

+M

NH2
+H2O

N2O

H2 O N 2

C
O

+NO2

+M

NH2
HN

NH2
NH3

NH

HONO

OH

NH

H2 N

+M

ONO

H2N

+R

HN

Condensed residuum

N2

Fig. 2. The fragment of consequent stages for nitroguanidine (1b, channel 1)


thermal decomposition.
The generated mechanism of (1b) thermolysis allows one to describe the whole
spectrum of reactions occurring in the course of thermal decomposition and of the
experimentally confirmed final products /11/ of thermolysis: N2O, NH3, CO2, HNCO,
cyanamide, dicyanamide, urea, cyanuric acid, melamine.
For the second compound that we present in the paper (C-nitramino-1,2,4-triazole (2))
30 tautomeric forms and rotational isomers are possible. The same as in the case of
nitroguanidine, for the compound (2) there is no evidence of thermochemical preferences of
this or other form. The results of our calculations (DFT/B3LYP/6-31G*) done for all
isomers (Table 2) show that in gas phase the tautomer (2a) is the most thermodynamic stable
one.
4

Eugeniya Bakhmatova, Vyacheslav Korolev, Aleksey Porollo, Tatyana Pivina

93

N
H2

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 2. Energy characteristics of C-nitramino-1,2,4-triazole tautomers (2a-h).


Compounds

Total
Energy
(E), ..

Tautomerization
energy, kcal/mol

Compounds
N NH

N NH

-502.0856

0.0

N N

NH

-502.0854

N
H

0.3

NO 2

2b

N
O

NH

N
H

NO2

-502.0748

9.5

-502,0597

15.6

-502,0639

13.1

-502,0637

13.3

-502,0662

11.6

-502,0593

15.9

-502,0637

13.0

-502,0672

10.6

6.6

O
H

2c

HN N

-502.0835

1.4

N
O

2d1

N
O

N
O

-502.0842

N O
H

0.9

2h4

2d2
O
N

HN NH
N

2h3

N
O

HN N

N NH
N
H

2h2
O

NH

HN N
O

-502.0581

16.8

2e1

H O
N
N O

2h5

HN NH

HN N
N
N
O

-502.0472

23.8

O H
N
N O

2h6

2e2

HN N
N NO2

NH

-502.0699
O

HN N

HC

+
NH N

9.0

2h1

N N

-502,0711

HN N

NH N

N
H

2g7

2a

Tautomerization
energy, kcal/mol

NO 2

NH

Total
Energy
(E), ..

-502.0460

24.5

O
N NO H

2h7
2f1

NH N
NH
+

2f2

94

N N

N
NO2

-502.0602

15.5

N
H

H O
N
N O

2i1

Computer Simulation of Thermal Decomposition Mechanism for Compounds . . .

New Trends in Research of Energetic Materials


N NH

Czech Republic
O H
N
N O

N N

-502,0771

N
H O

5.0

N
H

2006

[Content]

-502,0531

19.7

-502,0540

18.9

-502,0559

17.8

-502,0604

15.1

-502,0611

14.7

-502,0523

19.8

2i2

2g1
N NH
N N
N
O O
H

-502,0637

13.3

NN O
N O
N N
H
H

2i3

2g2
N NH

O
N H
O

NN
N

N
O

-502,0657

12.1

N
H

2i4

2g3
N NH

N N
N

-502,0672

N O
O

11.3

2i5
O H

N N

N N

-502,0635

13.3

N
H

2g5
N NH
N

N
O

N O

2i6
N N

O
N N
H

2g6

N O
H
O

2g4
N NH

N
H

-502,0716

8.7

N
H

N
N O
O
H

2i7

Therefore in the paper we present the result of simulation of thermal decomposition


mechanism for tautomer (2). Generation of possible pathways of (2) decomposition gave
three probable channels of the initial stage (Fig. 3): homolytic cleavage of nitro-group
(channel 1), N-N bond breakage in heterocycle (channel 2), and C-N out-of-cycle bond
breakage (channel 3).
In nitro-1,2,4-triazoles /12/ it takes 23.7-34.3 kcal/mol more energy for heteroaromatic
C-N bond breakage than for transannular N-N bond breakage (channel 2). According to our
calculations for (2) tautomer at the initial step the most benefit is the C-NO2 homolysis
(channel 1). This mechanism is estimated to have the activation energy of 27.9 kcal/mol. The
initial steps of nitro-1,2,4-triazole decomposition processes of other channels (2 and 3) are
almost twice as many (= 68.1 and 87.1 kcal/mol, correspondingly) compared with the
channel 1 process.

Eugeniya Bakhmatova, Vyacheslav Korolev, Aleksey Porollo, Tatyana Pivina

95

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

N NH

. .

N NH

N HN
N

N
H

N
H

II

NO2

NO2
III

NH

N NH

ONO

+M

+M

OH

.
N

N
H2N

+M

H2N

H2N

NH2

.C

N2O H2O

+M

NH3

CO2

N
N

+R

HN

H2 N

NH3

O
HN

NH3
NH2

CN

+ONO

N N

OH N

NO

N
N NH

NO2

+NH2

N NH

NH

+CN

N N
O
N

+M

NH2

ONO HCN
+HONO

NO

CH N N OH

+M

N NH

+OH

N NH

OH

N2O

+M

H2O

Condensed residuum

CHO

+NO

O
NH

NH2

HNO
H
N

NH2

NH2

+M

+NO2

H2O

+M

+M

H2O

+NO

NH2

NH2

+M

H2NNO2

.
NH2

N2

N NH

H2N

NO2

N NH

+NO2

H2N

HNO

.NH

+R

H2N

N2O

+M

H2O

HN

+2 NO

NO

+M

NH2

OH

HONO

N NH

N2O

+NO2

CO2

+R

CO

N2
+NO2

HNO
+NO

-NO

N2O

CO2
N2

OH
O

+M

+M

H2O

Fig. 3. Consequent stages for 3-nitramino-1,2,4-triazole (2a) thermal decomposition.


The computer generation of the possible pathways of (2) thermal decomposition (Fig.
3) provided a wide number of reactions and the final decomposition products. It is
significant that during HS-thermolysis of 3-nitramino-1,2,4-triazole /13/ only NO, NO2,
HCN, CO, N2O have been detected experimentally. However, in this study we shown (Fig.
3) the possibility of formation for a set of some other products being experimentally not
found: H2O, N2, O2, CO2, HNCO, ammonia, cyanamide, dicyanamide, dicyane, urea,

96

Computer Simulation of Thermal Decomposition Mechanism for Compounds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

cyanuric acid, melamine, 3(5)-amino-1,2,4-triazole, 4-amino-1,2,4-triazole, 1,2,4-triazole,


1,2,4-triazolon, 3(5)-nitro-1,2,4-triazole, 3(5)-nitrozo-1,2,4-triazole.

4.

CONCLUSION

As the result of nitrimine form for nitroguanidine and of 3-nitramino-1,2,4triazole thermal decomposition mechanisms simulation the comprehensive spectrum of
possible pathways for their thermal decay has been suggested. Based on the suggested
hypotheses on the mechanisms of thermal decomposition the estimation of thermo-chemical
preferences of decomposition pathways have been carried out. The data obtained can be used
to propose the thermal stability of compounds and for understanding of experimental thermal
decomposition kinetic investigations.

REFERECES
[1]
[2]
[3]

[4]

[5]

[6]
[7]
[8]

[9]
[10]
[11]
[12]

[13]

POROLLO A.A., LUSHNIKOV D.E., PIVINA T.S., IVSHIN V.P. J. Mol. Struct.
(TEOCHEM), Vol. 391, 1997, 117.
POROLLO A.A., LUSHNIKOV D.E., PIVINA T.S., IVSHIN V.P., ZEFIROV N.S., Izvestiya
Akademii nauk, 48, 1845, 1999 (in Russian).
POROLLO A., PETUKHOVA T., IVSHIN V., PIVINA T., and LUSHNIKOV D. Azetidine
and its mono-, di- and tri-nitro substituted derivatives: computer modeling of decomposition
reactions, in Proceedings of the 30th International Annual Conference of ICT, Karlsruhe,
Germany, p. 15 /1-3, 1999.
PIVINA T. S., POROLLO A. A., PETUKHOVA T. V., and IVSHIN V. P. Basic scheme for
computer simulation of decomposition reactions for energetic compounds, in Proceedings of
the 5th International Symposium on Special Topics in Chemical Propulsion (5-ISICP), Italy, p.
343-344, 2000.
PETUKHOVA T. V., POROLLO A. A., KOROLEV V. L., IVSHIN V. P. V. P., SURIKOVA
Y. N., and PIVINA T. S. Computer Modeling of Decomposition Reactions for Varies
Representatives of Nitroamines, in Proceedings of the 33rd International Annual Conference of
ICT, Karlsruhe, Germany, p. 60/1-10, 2002.
KOCH W. and HOLTHAUSEN M. C. A Chemistss Guide to Density Functional Theory,
Wiley-VCH, Weinheim, p.300, 2001.
CLARK ., Computational Chemistry (In Russian), Mir, Moscow, 383 p., 1990.
Programme Complex Gaussian-98 of the Computer Centre for Chemical Investigations,
Russian Academy of Sciences.
POLITZER P., GRICE M. E., SEMINARIO J. M., J. Quantum Chem., Vol. 61, p. 389, 1997.
MANELIS G. B., NAZIN G. M., RUBTSOV Yu. I., and STRUNIN V. A., Thermal
Decomposition and Combustion of Explosives and Powders, Nauka, Moscow, p. 223, 1996.
DAVIS T. L., ABRAMS A.-J.J., Proc. Am. Acad. Sci., Vol. 61.p. 437-457, 1926.
BAKHMATOVA E.A., PETUKHOVA T.V., KOROLEV V.L., PIVINA T.S., IVSHIN V.P., in
Proceedings of the 8th Seminar "New Trends in Research of Energetic Materials", Pardubice,
the Czech Republic, 2005.
GAO A., REINGOLD A.L., BRILL T.B. //Propellants, Explosives, Pyrotechnics, 1991, Vol.
16, 3, 97.

Eugeniya Bakhmatova, Vyacheslav Korolev, Aleksey Porollo, Tatyana Pivina

97

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THERMAL DECOMPOSITION OF SOLID ENERGETIC MATERIALS

Yurii M. Burov
Institute of Problems of Chemical Physics, Russian of Sciences, 142432
Chernogolovka, Moscow Region, Russian Federation.
E-mail: yuburov@icp.ac.ru

Abstract:
Decomposition of the organic compounds in the solid phase occurs much more slowly
than in the liquid and gas phases. The retardation effect of the crystalline lattice (REL),
which can be defined as the ratio of rate constants of decomposition in the liquid phase
to that in the solid phase (kliq/ks), is often the only factor that causes the stability of the
substance and its suitability for use after prolong storage. This situation is
characteristic of medicines, explosives, and initiators of chain processes. Experimental
determination of REL and development of methods suitable to predict the properties of
new substances are an important aspect of the general theory of the stability of organic
compounds. Development in this field is restricted because the date on ks values are
limited and the theoretical models of monomolecular reactions in the solid phase cannot
be verified. Solid phase decomposition reactions are usually accompanied by several
side processes, such as evaporation and fast decomposition of compounds in the gas
phase, melting on admixture and products autocatalysis at the early stages of the
process, effects of premelting near points of phase transitions, and others. Due to side
reactions, the observed ks values often exceed the true values by one and even two
orders of magnitude. Therefore, the purpose of this work is to obtain, firstly, a
sufficiently representative array
of correct ks and REL = kliq/ks values and, secondly, the dependence of REL on the
physical properties of the crystal and other parameters used in theoretical models. The
REL values for 22 compounds calculated using the correct ks values are presented in
this work. The REL values vary within very wide limits from 4 to 104. The reaction on
defects limits the kliq/ks ratio to 104, and the reaction in the bulk can be observed only if
REL does not exceed 100. Reaction in the bulk of the crystal requires the formation of a
cavity with a volume exceeding the activation volume s V , so that the leaving group
does not experience forces of interatomic attraction. Cavities can be formed by two
mechanisms energetically or entropically. Accordingly, two mechanisms were
observed. Both of them are in agreement with experiment.
Keywords: crystalline lattice, thermal decomposition, kinetics

1.

INTRODUCTION

Decomposition of the organic compounds in the solid phase occurs much more slowly
than in the liquid and gas phases [1-3]. The retardation effect of the crystalline lattice
(REL), which can be defined as the ratio of rate constants of decomposition in the liquid
phase to that in the solid phase (kliq/ks), is often the only factor that causes the stability of the
98

Thermal Decomposition of Solid Energetic Materials

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

substance and its suitability for use after prolong storage. This situation is characteristic of
medicines, explosives, and initiators of chain processes. Experimental determination of REL
and development of methods suitable to predict the properties of new substances are an
important aspect of the general theory of the stability of organic compounds. Development in
this field is restricted because the date on ks values are limited and the theoretical models of
monomolecular reactions in the solid phase cannot be verified.
Solid phase decomposition reactions are usually accompanied by several side processes,
such as evaporation and fast decomposition of compounds in the gas phase, melting on
admixture and products autocatalysis at the early stages of the process, effects of premelting
near points of phase transitions, and others. Due to side reactions, the observed ks values
often exceed the true values by one and even two orders of magnitude. The methods for
taking into account secondary factors [2] are used to a full extent only for several
compounds.
Therefore, the purpose of this work is to obtain, firstly, a sufficiently representative array
of correct ks and REL = kliq/ks values and, secondly, the dependence of REL on the physical
properties of the crystal and other parameters used in theoretical models.

2.

EXPERIMENTAL

Decomposition rates were measured by the manometric method. The reaction course
was monitored at conversions of 0.01 1.00%, which allowed one to avoid the influence of
autocatalytic processes and topochemical regime of the reaction. To eliminate unstable or
catalytic admixtures, the substances were purified by sublimation onto a heated support or y
recrystallization from different solvents with thorough drying. The purification was carried
until a constant decomposition rate was achieved.
To increase the sensitivity of the method, we used such loadings of the substance that the
ratio of the sample weight to the volume of the reaction vessel was 0.5 g cm-3. Thus, the
amount of the substance in the vapor phase was insignificant as compared to the weight of
the solid sample, and the reaction in vapor could be neglected even at the highest possible
difference in the rates in the gas and solid phases (four orders of magnitude) [4 6].
The ks values were calculated from the time of achievement of the degree of
decomposition of 0.1%. For calculation of the conversion, the stoichiometric coefficient of
gas release determined by the decomposition of the substance in melt was used.

3.

RESULTS AND DISCUSSION

The REL values for 22 compounds calculated using the correct ks values are presented in
Table 1. The kinetic data using for composing Table 1 are presented in Table 2. The rate
constants obtained for the decomposition in melt (noncatalytic stages) or in inert solvents
ware used as kliq. When several data were available, the minimum ks and kliq values taken for
the calculation of REL
The REL values ware calculated at temperatures 200C lower then the melting point of
the substance when pre-melting effects do not affect the ks value. The melting points of the
substances and the volume compressibility ( ) (the parameter characterizing the force of
interatomic interaction in the crystal) are also presented in Table 1. The values were
calculated by the Rao method [19], and the group increments were borrowed from the
literature date [20].
2

Yurii Burov

99

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The mechanisms of initial monomolecular stages of decomposition are known for all
compounds presented in Table 1: bond cleavage of C-NO2 (2, 3, 5 - 7, 10, 13), N-NO2 (1, 14,
16, 17, 19 22), O-NO2 (18), and N-N (4) and elimination of N2 (11) via formation of the
linear transition state or elimination of N2 (9) and CO2 (12, 15) molecules through the cyclic
transition state.
Table 1.

Retardation effect of the crystalline lattice and physical properties of the


substances

Compound

kliq/ks

m.p., 0

.1010 V0,
, P-1

cm3/mol

[FC(NO2)2CH2]2NNO2

86

3,530

173,9

[F2NC(NO2)2CH2]2NNO2

102

198,0

C(NO2)3C(NO2)3

140

5,500

169,4

PhN=N-NHPh

101

[C(NO2)3CH2]2NNO2

10

95,5

1,050

199,0

F2NC(NO2)2CH2NNO2(CH2)2C(NO2)2NF2

10

120

[NF2(NO2)2CCH2N(NO2)CH2]2

20

158

25

158

1,3,5-Trinitro-2,4,6-triazidobenzene

28

131

10

1,3,5-Tris(trinitromethyl)benzene

38

113

11

N3(CH2NNO2)4CH2N3

87

177

12

HOOCCH2COOH

90

135

13

[(NO2)3CCH2N(NO2)CH2]2

92

180

1,060

254,5

14

2,4,6-(NO2)3C6H2N(NO2)Me

100

130

1,200

164,5

15

HOOC-COOH

122

189

16

1,4-Dinitro-1,4-diazacyclohexane

210

213

1,310

107,3

17

[MeC(NO2)2CH2]2NNO2

230

177

0,968

189,5

18

C(CH2ONO2)4

360

142

178,5

19

1,1,3,5,5,7-Hexanitro-1,4-diazacyclohexane

500

250

0,925

203,2

20

1,4,6,9-Tetranitro-1,4,6,9-tetraazadecalin

1000

236

21

1,3,5-Trinirro-1,3,5-triazacyclohexane

1400

202

0,809

122,0

22

1,3,5,7-Tetranitro-1,3,5,7-tetraazacyclooctane ( -modification)

8380

277

0547

155,8

.
100

Thermal Decomposition of Solid Energetic Materials

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The majority of reactions are characterized by a positive activation volume, i. e., they
occur with a volume increase in the transition state ( V ). For compounds 9 and 11, one
can expect V values close to zero, whereas for 12 and 15, they can even be negative.
It is seen from the data in Table 1 that the REL values vary within very wide limits from
4 to 104. To explain such a broad interval of REL variation, the theory of solid phase
reactions should be considered.

Table 2. Kinetic parameters of decomposition of organic compounds in liquid and


solid states
Com-

Mediuma

pound
Solution in TNT
1
Solid phase

T / 0C

Eb

170-210
66

40.0

lgA /s-1
14.92

k c /s-1

References

1.310-

4
29

11

3.01012

2
3
4
5

6
7
8
9
10
11
12
13
14

15
4

Yurii Burov

Melt
Solid phase
Gas
Solid phase
Melt
Solid phase
Melt
Solution
Solid phase
Solution in DNB
Solid phase
Solution in TNT
Solid phase
Solution in NB
Solid phase
Solution in xylene
Solid phase
Solution in TNT
Solid phase
Solution in DNB
Solid phase
Melt
Solid phase
Solution in DNB
Solid phase
Melt
Melt
Melt
Solid phase
Gas

105-120
82
90-135
120
100-160
80
110-150
110-165
75
100
100
130-170
100-140
114
114
70-115
50-100
95
95
120-180
120-175
136-160
115
130-180
158
140-160
131-155
150-175
80-120
150-190

38.5

15.70

35.8

16.52

39.9

17.50

36.8
36.1

15.59
15.06

40.3
42.6
26.0
28.8
36.3
36.7
32.2
40.7

16.9
16.4
12.10
12.20
14.59
12.86
13.50
16.80

35.2
36.0
35.9
38.9
35.2

13.50
13.80
13.50
13.20
13.80

310-8
8.810-9
4.010-4
6.610-5
7.210-8
1.210-8
2.910-8
2.410-8
2.410-9
1.310-7
1.310-8
310-4
1.510-6
3.310-4
1.310-4
1.610-3
5.710-5
1.210-3
3.510-5
1.410-4
1.610-4
2.310-5
2.610-7
2.610-4
2.810-4
2.610-7
1.810-7
1.010-7
1.010-9
2.510-4

29
29
2
29
5
29
4
4
29
29
29
29
29
29
29
6
29
29
29
7
29
29
29
4
29
8
9
10
2
29
101

New Trends in Research of Energetic Materials

16

17
18
19
20

21
22

Solid phase
Melt
Solution in TNB
Solution in NB
Solid phase
Solution in DNB
Solid phase
Melt
Solid phase
Solution in TNT
Solid phase
Gas
Solid phase
Solid phase
Solution in DNB
Solid phase
Solution in DNB
Solid phase

120-180
135-200
120-260
225-245
185
145-170
150
120-130
170-210
230
216
216
204-234
160-200
140-190
171-215
130-180

Czech Republic

2006

38.6
37.1
37.9
37.1

13.40
13.90
14.00
12.00

39.7

15.7

40.0
39.0
38.0

15.8
12.7
14.5

50.1
39.7
39.8
44.9
37.9

18.9
14.3
11.2
16.0
9.2

2.110-6
1.510-8
8.110-8
2.010-6
9.510-9
3.310-5
2.310-7
4.710-7
1.310-9
110-2
210-5
210-3
2.110-5
3.210-4
1.410-5
110-8
3.110-3
3.710-7

[Content]

29
11
12
12
29
13
29
14
2
29
29
29
29
15
16
17
16
17

Based on the published date [1], a generalized physical model of monomolecular


reactions in the solid phase can be envisaged. A reaction can proceed in the bulk of the
crystal or on its surface and on defects of the crystalline lattice. The reaction can proceed in
the bulk of the crystal or on its surface and on defects of the crystalline lattice. The reaction
occurring on defects has the same activation energy (E) as in the liquid phase (or close to it),
but the pre-exponential factor includes the coefficient that takes into account the fraction of
molecules in disordered sites of the lattice. According to estimates [1, 24, 25] based on the
calculation of the number of molecules arranged on the network of dislocations that separate
microblocks of the crystal (the linear sizes of the microblocks are 10-3 10-5 cm), the
fraction of such reactive molecules is 0.01 1.00%. Thus, the reaction on defects limits the
kliq/ks ratio to 104, and the reaction in the bulk can be observed only if REL does not exceed
100.
Reaction in the bulk of the crystal requires the formation of a cavity with a volume
exceeding the activation volume Vs , so that the leaving group does not experience forces
of interatomic attraction. Cavities can be formed by two mechanisms energetically or
entropically.
The macroscopic approach [22, 31 - 33], considering the crystal as an elastic continuous
medium, leads to the equation

E = E s Eliq = Vs

) /(2V ) = RT ln k
2

liq

,
k s

(1)

where V0 is the molar volume of the substance, and Vs is the activation volume in the
0

solid phase. This volume is not equal to the value V . According to the published data
[21], it can be estimated as an increase in the volume of the cell occupied by a molecule due
to elongation (by 10 15%) of the cleaving bond in the transition state [27], i. e.,
approximately by 0.2 . Extension of the cell by 0.2 results in weakening of
intermolecular interactions and allows atoms of molecule to converge freely, forming cyclic

102

Thermal Decomposition of Solid Energetic Materials

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

transition states. Thus, the Vs value depends slightly on the reaction type and on the value
and sign of the true activation volume.
The pre-exponential factor of the reaction occurring in the bulk should be the same as
that for the reaction in the liquid phase [22]. In [23] added influence of solid state on the preexponential factor.
The calculation of REL by Eq. (1), assuming that the cell volume is V0 /(6.02 10 23 ) and
its expansion is 0.2 , gives values that often coincide by an order of magnitude with the
experimental values. For compounds 1, 3, 5, 13, 14, 16, 17, 19, 21, and 22, i. e., when
kinetic data for calculations were available, the calculated REL value was 8, 3, 100, 386, 10,
453, 31, 250, 400, 400, and 3040, respectively. The highest deviations from the experimental
values (see Table 1) are observed for molecules with a long chain. In this case, local motions
of, not the whole molecule, but only of its fragments containing the reaction center have a
substantial effect on the REL values.

1
10 13 / Pa mol cm 3
V0
Fig. 1. Dependence REL on volume compressibility.

As it's shown in the works [31 - 33], the main assumptions of this theory are contrary to
the principles of the thermodynamics. Nevertheless, this theory rather well describes the
experiment (see Fig. 1 and 2). Therefore, the new approach - "free volume model" was
offered in the works [31 - 33].

At temperatures is higher 2 , where - characteristic temperature, the crystal


properties are in the best way described in frameworks cell model. Within the framework of
v
the Lennard Jones cell model [29], the "free volume" f , which one can be computed by a
method atom atom potential [30]
6

Yurii Burov

103

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

U (q ) U (q )
(2)
dq
kT
q - generalized coordinates, U (q ) - energy of crystal lattice, U (q) - interact energy of a
molecula with the neighbours in the supposition, that the adjacent moleculas take
midpositions, the integrating is carried out on volume basis cells.
v f = exp(

Probability of originating of a cavity by the size Vs

P (Vs ) = exp(

Vs

vf

Supposing, that [31 33]


k s = k liq P(Vs )

(3)
the problem of an estimation REL is reducible to finding of a free volume, which one
can be computed by Eq. (2), or can be estimated from the degree of thermal expansion
T

v f 3 (T )V0 dT ,
0

(T ) - is coefficient of thermal expansion.


The absence of a method of exact estimation of Vs restricts the possibility of using
Eq. (1) and (3) for practical calculations, although there is some doubt that considered
models adequately reflect the real pattern. Equation (1) is important because it predicts, first,
low REL, i.e., the possibility of the reaction occurring purely in the bulk of the sample, and,
second, the dependence of REL on the elastic properties of the crystal. Both predictions
agree with the experimental data. Only the nonconstant character of the Vs values disturbs
the linearity of the dependence of REL on 1
(V0 ) , which is demonstrated by the data in
Fig. 1.
The dependence of REL on the melting point of the substance, which is also a good
measure of forces of intermolecular interaction, especially for short-chain molecules [28], is
presented in Fig. 2, and points for substances with a high molecular weight lie closer to the
lower boundary in Fig. 1. For compounds with m.p. < 150 0C, the REL values are mainly
lower than 100 and, hence, we may consider that the reaction in the volume predominates in
this case as well.
For the substances with higher melting points, the reaction in the volume is changed
by the decomposition on crystal defects, which, as shown above, can limit REL by the
interval of 102 104, which is experimentally observed. Compound 3 (it is not shown in Fig.
1), which represents plastic crystals, falls out of the general dependence. For long-chain
compounds 6 and 7, according to the absolute REL value, decomposition in the bulk is also
possible, despite sufficiently high melting points.
The experimentally determined E value is low and does not usually exceed the
measurement error, which is 3 4 kcal/mol. Therefore, it is difficult to use this value to
estimate REL at different temperatures and to separate volume and local reactions. The latter
can be performed using absolute values of REL or melting points of the substances. The
experimental data in Table 1 agree with considered models and suggest that for organic
compounds with m. p. 100 0C the irreversible monomolecular reaction occurs in the ideal
104

Thermal Decomposition of Solid Energetic Materials

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

part of the crystal lattice with REL < 10, and at m. p. > 200 0C the reaction proceeds on
defects only. Within the interval m. p. = 150 200 0C, REL is equal to 102 103, and the
indicated reactions compete with each other. These conclusions give a clear semiquantitative pattern of changing REL; however, they need further verification.

Fig. 2. Dependence REL on melting point.

v
It is necessary to obtain new data on f , Vs and REL, including those for reaction
with a negative activation volume, as well as to identify directly localized reactions and to
compare the decomposition rate with the defectiveness of crystal.

4.

REFERENCES

[1]
[2]

Chemistry of Solid State, Ed. W. E. Garner, Butterworths, London, 1955.


Yu. M. Burov, G. B. Manelis, and G. M. Nasin, Dokl. Akad. Nauk SSSRR, 1984, 279, 1142
[Dokl. Chem., 1984 (Engl. Transl.)]
[3] O. N. Karpukhin, T. V. Pokholok, and V. Ya. Shlyapintokh, Vysokomol. Sedin. A, 1971, 13,
(Engl. Transl.)
[4] B. L. Korsunski, V. G. Matveev, L. D. Nazina, and G. M. Nazin, 29th Intern. Annual
Conference of ICT (June 3- - July 3, 1998), Karlsruhe, 1998, 60.
[5] N. Okawara, H. Jamazaki, and E. Imato, J Chem. Soc. Jpn., Ind. Chem. Sect., 1955, 58, 991.
[6] B. L. Korsunski, T. A. Apina, and F. I. Dubovitskii, Izv. Akad. Nauk SSSR, Ser. Khim., 1970,
2080 [Bull. Acad. Sci. USSR, Div. Chem. Sci., 1971, 20 (Engl. Transl.)].
[7] Yu. M. Burov and G. M. Nazin, 29th Intern. Annual Conference of ICT (June 3- - July 3, 1998),
Karlsruhe, 1998, 129.
[8] F. I. Dubovitskii, G. B. Manelis, and L. P. Smirnov, Zh. Fiz. Khim., 1961, 35, 521 [J. Phys.
Chem. USSR, 1961, 35 (Engl. Transl.)].
[9] F. I. Dubovitskii, Yu. I Rubtsov, and G. B. Manelis, Izv. Akad. Nauk SSSR, Otd. Khim. Nauk
[Bull. Acad. Sci. USSR, Div. Chem. Sci.], 1960, 1763 (in Russian).
[10] R. S. Stepanov, V. N. Shan'ko, I. P. Medvetskaya, and V. m. Gorodetskaya, in Khimicheskaya
fisika protsessov goreniya i vsryva. Kinetika khimicheskikh reaktsii [Chemical Physics of
8

Yurii Burov

105

New Trends in Research of Energetic Materials

[11]

[12]

[13]

[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]

[25]
[26]
[27]

[28]
[29]
[30]
[31]
[32]
[33]

106

Czech Republic

2006

[Content]

Combustion and Explosion. Kinetics of Chemical Reaction], OIKhF AN SSSR,


Chernogolovka, 1977, 56, (in Russian).
G. V. Sitonina, B. l. Korsunskii, N.F. Pyatakov, V. G. Shvaiko, I. Sh. Abdrakhmanov, and F. I.
Dubovtskii , Izv. Akad. Nauk SSSR, Ser. Khim., 1979, 311 [Bull. Acad. Sci. USSR, Div.
Chem. Sci., 1979, 28 (Engl. Transl.)].
B. A. Lur'e and V. N. Ivakhov, in Khimicheskaya fizika kondensirovannykh vzryvchatykh
veshcestv [Chemical Physics of Condensed Explosives], Moscow, 1979, 12 (Tr. MkhTI, 1979,
No 104) (in Russian).
B. L. Korsunskii, L.Ya Kiseleva, V. I. Ramushev, and F. I. Dubovitskii, Izv. Akad. Nauk
SSSR, Ser. Khim., 1974, 1778 [Bull. Acad. Sci. USSR, Div. Chem. Sci., 1974, 23 (Engl.
Transl.)].
K. K. Andreev and Kaidymov, , Zh. Fiz. Khim., 1961, 35, 2676 [J. Phys. Chem. USSR, 1961,
35 (Engl. Transl.)].
H. Ronqzu, Y. Zhengounan, and L. Yanjun, Thermochim. Acta, 1988, 123, 135.
Yu. Ya. Maksimov, in Teoriya vzryvchatykh veshcestv [Theory of Explosives], Vysshaya
Skola, Moscow, 1967, 73, (Tr. MkhTI, 1967, No.53) (In Russian).
Yu. M. Burov and G. M. Nazin, Khim. Fiz., 1984, 3, 1126 [Chem. Phys., 1984, 3. No. 7 (Engl.
Transl.)].
Yu. M. Burov, G. M. Manelis, and G. M. Nazin, Khim. Fiz., 1985, 4, 956 [Chem. Phys., 1985,
4. No. 7 (Engl. Transl.)].
M. R. Rao, J. Chem. Phys., 1941, 9, 682.
I. M. Voskoboinikov, A. N. Afanasenkov, and V. M. Bogomolov, Fizika Goreniya I Vzryva,
1967, 3, 585 [Phys. Comb. Explos., 1967, 3,4 (Engl. Transl.)].
G. M. Shutov, Zh. Fiz. Khim., 1967, 39, 2817 [J. Phys. Chem. USSR, 1961, 39 (Engl. Transl.)].
G. B. Manelis, in Problemy kinetiki elementarnykh khimicheskikh reaktsii [Problems of
Kinitics of Elementary Chemical Reactions], Nauka, Moscow, 1973, 93 (in Russian).
L. D. Zusman and A. B. Gel'mam, Zh. Strukt. Khim., 1980, 21, 72 [J. Struct. Chem. (USSR),
1980, 21 (Engl. Transl.)].
M. S. Belyaeva, G. K. Klimenko, L. T. Babaitseva, and N. P. Stolyarov, in Khimicheskaya
fisika protsessov goreniya i vsryva. Kinetika khimicheskikh reaktsii [Chemical Physics of
Combustion and Explosion. Kinetics of Chemical Reaction], OIKhF AN SSSR,
Chernogolovka, 1977, 47, (in Russian).
A. I. Kitaygorodskii, Moleculyarnye kristally [Molecular Crystals], Nauka< 424 pp. (in
Russian).
D. A. Charman, S. Rozak, P. B. Keegstra, R. C. Hariharan, J. J. Kaufman, and R. S. Buenker,
Int. J. Quant. Chem., 1991, 38, 541.
M. G. Gonikberg, Khimicheskoe ravnovesie i skorost' reaktsii pri vysokikh davleniyakh
[Chemical Equilibrium and Reaction Rate at High Pressures], Khimiya, Moscow, 1969, 427 pp.
(in Russian).
M. S. Westwell, M. S. Searle, D. J. Walles, and D. H. Williams, J. Am. Chem. Soc., 1995, 117,
5013.
J. Lennard-Jones, Proc. Phys. Soc., 52, 729, 1940.
A. I. Kitaigorodskii, Usp. Fiz. Nauk, 127, 391, 1979, (In Russian).
Yu. M. Burov, XIV Symposium Sovremennaya Khimicheskaya Fizika [ Modern Chem.
Physics], 18 29 Sept. 2002, Tuapse, 2002, 59, (in Russian)
Yu. M. Burov, Zh. Fiz. Khim., 1961, 35, 2676 [Russian J. Phys. Chem., 78, No. 4, 2004, 579
(Engl. Transl.)].
Yu. M. Burov, XVII Symposium Sovremennaya Khimicheskaya Fizika [ Modern Chem.
Physics], 18 29 Sept. 2005, Tuapse, 2005, 32, (in Russian).

Thermal Decomposition of Solid Energetic Materials

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

USE OF COMPRESSED GAS GUN TO INVESTIGATE RESPONSE OF


EXPLOSIVES TO A RANGE OF INITIATION STIMULI
Michael Cartwright and Andrew Lewis
Department of Environmental and Ordnance Systems,
CRANFIELD UNIVERSITY
At DEFENCE ACADEMY of the U.K.
Shrivenham
Oxfordshire, U.K.
SN6 8LA
Abstract
Low velocity Impacts are a source of explosive accidents. Several different
mechanisms may be concurrently responsible for the initiation. Studying the effects of
low velocity, comparatively long duration, impacts on energetic materials can assist
in the development of explosive safety procedures and Insensitive Munitions. Careful
design of the target configuration can eliminate some of the problems associated with
the original Steven tests and enable the various mechanisms such as friction, shear
and pinch to be isolated from other factors. In the new configuration the impact is
delivered to the explosive by a spigot driven into the target by collision from a gun
launched sabot. The spigot geometry i.e. flat or hemi-spherically nosed was varied as
was the surface roughness and diameter of the spigot. The rate of energy delivery was
varied, by varying the mass of the sabot and its velocity. The gun, 50 mm diameter,
operated in the normal atmosphere conditions, NAG. Results show that pinch may be
the most critical stimulus but also that friction is more significant than shear.
Examination of the target debris following an event indicated that for some materials
the limited confinement arising during the firing may result in a burn to detonation
condition whereas for other materials the confinement was insufficient for the DDT to
occur. Further work on the sample perfection and homogeneity are planned.
Keywords accidents, impact , initiation, explosives, energy, output
Introduction
In the past a number of costly accidents with munitions, both in peacetime and
war zones has resulted in the loss of five USN warships, four ammunition storage
depots and two Soviet submarines all with significant loss of human lives
More recent accidents during EOD work have demonstrated that there is a gap in test
procedures. Operators sent out to sample failed munitions prior to disposal have
variously taken chisels, screwdrivers and broom handles in conjunction with large (3
kg) hammers. In one accident a rocket motor ignited and left the chisel in the
operators skull. In Afghanistan a multinational team were trying to remove an
explosive warhead from a Soviet SAM missile when an ignition occurred with a
resulting deflagration or low order detonation event killing five people1. In these two
incidents the maxim elicited by F.P. Bowden 2 hitting a solid with a hammer or
1

Michael Cartwright, Andrew Lewis

107

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

rubbing it with sandpaper is a proper experiment for a Carpenter than for a physicist
appears to have been forgotten. Nevertheless impact either by dropping munitions
from modest heights or striking munitions with implements is a major source of
accidental initiation in explosives and thus requires further study.
As a result of accident initiation the following series of safety tests for
explosive fillings using different forms of impact have been developed in the NATO
countries
Bullet Attack (STANAG 4241)
Fragment Attack (STANAG 4296)
Shaped Charge Attack(STANAG 4526)
Sympathetic reaction (STANAG 4625)
Some of these tests are particularly applied to completed weapons in their
packaging. Impact test methodologies are almost as numerous as the test houses using
them and quantitative agreement between them is rare3. JSP 333 4quotes 38 separate
sensitivity tests of which a number are impact in one form or another. Impact test
sample size ranges from the mg level of the Rotter and Drop or Fall Hammer tests to
the 200 g of the Susan and Steven tests. In the Susan test5, the explosive, contained in
a specified projectile, is fired at an armoured target using a propellant gun and in the
Steven6 test the explosive, contained in a target, is struck by a gun launched projectile.
Basic principles of the Steven test and modified assembly trialled are shown in figure
1. Computational fluid dynamic calculations on the Steven test7 predict that ignition
occurs from a variety of mechanisms and the response can be sensitive to preparation
procedures. Modelling and prediction are thus difficult.
Ignition occurs in accidents as a result of hot spot generation when the rate of
heat liberation is greater than the rate of heat loss. The major stimuli responsible for
hot spot creation are :Compressive effects - adiabatic heating of trapped gases and pore collapse
Frictional effects - both inter-crystalline and at surface boundaries
Mechanical effects - shear, tribochemical and viscous flow
Most of the accidental ignitions occur at low velocities so that the shock wave
impact mechanism is not operative. Driving a spigot into explosive samples could
produce systems in which the various ignition mechanisms could each be given
prominence and thus an estimate of the response of materials to different stimuli
could be obtained. A target vehicle, which would confine an explosive and deliver a

108

Use of Compressed Gas Gun to Investigate Response of Explosives to a Range of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

measured quantity of energy through an impact by a projectile, launched from a gas


gun, has been designed manufactured and used.

Experimental
Equipment
The gun used was 50 mm calibre firing in air against a mounted target NAG.
Initially mechanical rupture of a diaphragm was used to fire the gun, but later gas
pressurisation behind the sabot, through a venting valve was used as shown below

Figure 1 Schematic diagram of Gas gun arrangement with Sabot loaded.


This latter had the advantage of providing more sensitive pressure
control..Three sabots were used see figure ;

Figure 2 Typical sabots used in Tests.


The simple near spherical sabot was an attempt to reduce the effect of yawing
of the projectile on impact energy and later the extended projectiles with the widely
3

Michael Cartwright, Andrew Lewis

109

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

spaced O rings to obscure the venting orifices between the gas storage chamber and
the barrel were used. Both flat and round nosed projectiles had masses of 1.5 Kg.
Lighter projectiles, 0.8 Kg, consisted of front and rear steel sections bolted together
through a PVC tube which carried the obturation O rings as shown in the gun
schematic. Sabot velocity was determined from the time taken break of a series of
equally spaced wires stretched across a plastic tube through which the sabot passed
after it had exited the barrel and before striking the target
Target Design
A range of targets assemblies was used. Initially simple and modified Steven
test assemblies were used see figure 3

Figure 3 Modified Steven test Assembly showing increased confinement


Notice that the confining ring in the modified test was extended considerably over the
cover plate leaving an uncovered area slightly larger than the sabot impact area and
also down the sides of the body thus ensuring significant increase in confinement of
the explosive over the original target design. More recent tests used an assembly
shown below figure 4 .

Figure 4 Schematic Target Assembly using moving piston mounted spigot


in which a piston, with attached spigot, was driven into the explosive by impact of a
gas gun launched sabot with the top of the piston.
110

Use of Compressed Gas Gun to Investigate Response of Explosives to a Range of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Mounted within these assemblies were pellets of the explosive under test. Two pellet
sizes were used, 70 mm and 25mm diameter, by 15 mm thick Columns up to three
pellets thick could be mounted in the target assembly and held in place by PTFE or
PVC rings. Spigot penetration into the explosive was controlled by shims in front or
behind the pellets, figure .

Rear closure
Sample ring
washer

Spigot

Piston

Sample

shim

Sample holder

Mounting
plate

Figure 5 Target assembly (schematic)


This enabled the ignition mechanisms i.e. friction shear and nip sensitivities to be
investigated. Some pellets had dimples machined into one face allowing air to be
trapped between two pellets to investigating aidiabatic heating effects.
In house manufactured PVDF pressure gauges were enclosed within the
targets between pellets and also between pellets and housing surfaces. The intention
was to monitor the pressure as the projectile struck and after reaction commenced.
Experiments investigated the effect of spigot characteristics, diameter, shape and
surface texture as well as penetration and nip on sensitivity. The range layout for the
normal air gun NAG are shown in the figure

Michael Cartwright, Andrew Lewis

111

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 6 Schematic range layout for 50 mm NAG


Events were monitored by high speed video camera and data recorded on fast
recording oscilloscopes. All firings were performed remotely on the COTEC ranges
on the edge of the Salisbury Plain Training Area at West Lavington
Explosive compositions investigated were a PBX formulation based on HMX
and PETN/WAX (92:8), both supplied by B.Ae. systems Glascoed, U.K. The PBX
was hot, 363 K, pressed into 70 mm diameter 13 mm thick pellets at 100 MPa. The
PETN/wax was pressed, at ambient temperatures without additional binder, into 25
mm diameter pellets 13 mm thick at 50 Mpa. The lower pressure was used for the
PETN to reduce the risk of events even though the pressing was done remotely on the
range facilities at the college. Some PETN pellets were pressed after removing the
wax by solvent washing but these showed poor mechanical strength and so were not
used beyond the preliminary stages.

112

Use of Compressed Gas Gun to Investigate Response of Explosives to a Range of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Results and Discussion


(a) PBX
Some of the observed events occurring from the modified Steven tests were
recorded at velocities >100 m s-1 by direct impact of the projectile. The induced
reactions appeared to be deflagrations and metallographic examination of the broken
edges confirmed ductile failures. In the spigot impact tests the velocities of the spigot
were less than the velocity of the impacting projectile but nevertheless events
occurred at lower sabot speeds. This is an indication of the extent of the energy lost
during the deformation of the steel cover of the Steven test. The results from a series
of tests on the PBX composition with the spigot intrusion are shown in the table 1.
Table 1 Impact trial results PBX showing spigot configuration sabot
velocity nature of event and mass of explosive recovered
Trial

Velocity
ms

event

-1

mass %
Recovered

13 mm flat spigot

100

Yes def

42

13 mm flat spigot

65

No

92

13 mm diam. Flat Spigot air bubble

53

Yes def.

48

As above

45

Yes def

45

As above

36

No smell

96

13 mm diam round nosed spigot

86

No sl. evi

95

103

yes

52

4 mm round nose

115

No smell

97

Flat nosed

115

yes

30

When events occurred they were deflagrations confirmed my metalographic


examination of the fragments. In these PBX tests not all the explosive was consumed
in the deflagration. As soon as the pressure generated exceeded the tensile strength of
the bolts holding the target together the vehicle separated, the pressure released and
the burning process was extinguished. About 50% of the contents were recovered
from the floor of the test facility and the fragments of the target assembly. Even
addition of a reinforcing plate over the piston housing, although it resulted in
apparently more violent reaction, which distorted the covering plate, it was still a
deflagration. The interesting point is that the smaller diameter piston seemed to
7

Michael Cartwright, Andrew Lewis

113

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

require more energy to be deposited by the spigot since events only occurred at higher
impact velocities. This is opposite to the effect predicted by Lee8 based on energy
density. The same energy over a smaller area should have an increased ignition effect.
The reason for this may come from the motion of the spigot. Not all the energy
imparted to the piston is delivered to the explosive. The piston is arrested by the
carrier rather than the spigot in the explosive causing the stop. The explosive sample
offers less resistance to the smaller spigot and hence the same energy may not be
delivered for the same impact velocity. Methods for measuring the energy transferred
to the explosive by the spigot rather than just the piston K.E. are under evaluation.
Simple conservation of momentum principles need modification.
Tests to confirm the effects of detonation were performed by mounting two
explosive pellets which had been impacted but failed to initiate into a target assembly
and inserting an electric detonator L2A1 in the cavity with the firing wires passing out
through the gauge slots. The piston was inserted into the holder but was free to move.
The whole assembly was mounted on the target plate in the normal mode and the
detonator fired. Typical fragments are shown in the figure below.

Figure 7 Showing Rear Plate fragments recovered from full detonation in target
The back plate was destroyed and the confining ring fragmented with the
bottom plate perforated. The surprising effect was the piston was retained in the
carrier housing because the shock wave had mushroomed the end protruding beyond
114

Use of Compressed Gas Gun to Investigate Response of Explosives to a Range of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

the carrier before the piston had chance to move backwards in the carrier. Notice how
similar the fragments are to those obtained from the PETN samples mentioned below
where the impact induced a run to detonation
(b) PETN/wax
The results from these firings are summarised in the table below.
Table 2 PETN/wax Results Summary

Shot
No
(a)
1
2
3
4
5
6
7
8
9
10

Trial Type

(b)
10mm intrusion
10mm intrusion
10mm intrusion
10mm intrusion
10mm intrusion
10mm intrusion
15mm Shear
15mm Shear
Pinch
Pinch

Spigot
Surface
(c)
Plain
Plain
220 grit
plain
80 grit
80 grit
plain
plain
plain
plain

Impact
Velocity
Ms-1
(d)
28.2
47.77
41.8
42.857
42.25
37.1
43.86
47.78
47.32
84.74

Response

(e)
Event ( old o rings)
Spigot damaged
No reaction
Event
Confinement loss
No reaction
No reaction
Event
Event

The points of interest are that the spigots surface roughness appeared to
influence the sensitivity. Shear was not found to be a particularly exciting stimulus in
that none of our shear experiments produced an event even though we were firing at
velocities at which both friction and pinch had given violent deflagrations. Pinch was
deemed to be the most sensitive since reaction occurred at similar velocities to simple
intrusion experiments but the event at the higher velocity of 87 m s-1 was a detonation
which did not occur with any of the other stimuli at that velocity. Examination of the
target fragments recovered showed brittle fracture

Michael Cartwright, Andrew Lewis

115

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 8 Firing 10 against PETN/wax showing target destruction


The added wax is a desensitiser for the PETN but the sensitivity has not been reduced
to the level of the PBX tested. Examination of the traces from the PVDF gauges show
typical pressure increase leading to DDT following mechanical impact with increased
velocity but below the shock threshold

Figure 9 Showing Oscilloscope traces recorded from PVDF gauges mounted at


front and rear of the target pellet.

116

Use of Compressed Gas Gun to Investigate Response of Explosives to a Range of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Conclusions
The most significant stimuli for accidental initiation are :pinch or nip of the explosive between end of spigot and rear steel plate
gas compression within the explosive
Friction between explosive and spigot surface
Increasing surface roughness of spigot increases response to impact
Shape of the spigot is important. Flat fronted spigot impact results in events at lower
velocities than hemispherical fronted spigots.
Shear does not appear to be a significant source of initiation but may be a contributing
factor
PETN when suitably confined can run to detonation before confinement is lost if the
energy input is above the simple deflagration threshold

Acknowledgements
Staff at COTEC ranges for assembly of the protected firing facilities and providing a
TCO and pickets during the firing.
Andrew Jones, Peter Williams and other members of staff from AWE, Aldermaston,
who provided both the instruments and the technical expertise to operate the fast
video cameras and recording oscilloscopes

Lt. Col. T B Rapple HQ Land Command, Wilton, private communication


F.P. Bowden and A.D. Yoffe, Initiation and Growth of Explosions in Liquids and Solids,
Cambridge University Press, 1952
3
Encyclopaedia of Explosives and related Items 8, 102, 1978, R and D Command, US Army
4
JSP333, Services Text book of Explosives, chapter 15, 2nd edition, HMSO(U.K.) 1989
5
P.W. Cooper & S.R.Kurowski, ch 4, Introduction to the technology of explosives Wiley VCH 1996
6
S.K. Chidester, L.G. Green & C.G. Lee, A frictional work predictive model for the Initiationof Solid
High explosives from Low Pressure Impacts, Proc. 10th Int. Det. Symposium
7
Alec Milne private communication
8
P.R. Lee, Critical Power Density: A Universal Quantitative Initiation Criteria, Proc 10 Int. Symp on
Ballistics, Peking, 1987
2

11

Michael Cartwright, Andrew Lewis

117

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THE RELATIONSHIP BETWEEN SHOCK SENSITIVITY AND


MORPHOLOGY IN GRANULAR RDX
Helen Czerski, William G. Proud, John E. Field
PCS Group, Cavendish Laboratory, J.J. Thomson Avenue
Cambridge, CB3 0HE, UK

Abstract: It is known that batches of the secondary explosive RDX from different
manufacturers show significant variation in their shock sensitivity.
No obvious
correlation between shock sensitivity and either chemical composition or morphology
has previous been identified which explains this. We use a range of techniques to study
the microstructure of RDX crystals and the bulk morphology of granular beds in order
to assess which hotspot mechanisms tend tot be dominant. Crystals were characterized
using mercury porosimetry, environmental scanning electron microscopy (ESEM) and
optical microscopy. This range of methods yields quantitative and qualitative data on
internal void size and number and surface structure. Shock sensitivity is quantified
using small-scale gap tests, and this demonstrates the clear differences in sensitivity
between batches from different manufacturers.
The samples used are from three
manufacturers, produced by both the Woolwich and Bachmann processes, and all have
an average particle size of approximately 1200 m.
Keywords:

1.

RDX, gap tests, sensitivity, morphology

INTRODUCTION

Explosives can be found in many different forms such as liquids, pressed powder compacts,
polymers, and two-phase mixtures such as polymer-bonded explosives. These can be
divided into homogenous and heterogeneous materials and this division provides a basis for
describing the details of their response to mechanical stimuli. A homogenous explosive is
simpler to understand since it is a continuous medium; each region has the same material
properties as those adjacent to it and parameters such as temperature, pressure and density
vary continuously across the bulk. In general the mechanisms of ignition and initiation in
these materials are well-understood. For example, ignition in a liquid containing a number
of bubbles is likely to start as the bubbles collapse to form hotspots(1).
Most commonly used explosives fall into the second category. Solid explosives such as
RDX and HMX perform very well, releasing a considerable amount of energy per unit
volume and having high detonation pressures and velocities. However, they are produced in
a crystalline form and are usually either pressed into high-density compacts or are mixed
with a polymer binder to produce a polymer-bonded explosive.
In either case, the
heterogeneities in the system make the details of ignition or initiation much harder to
understand than is the case for homogenous explosives. These heterogeneities lead to a
highly non-uniform temperature and stress field in the shocked material, and it is the highesttemperature regions within this distribution which will form critical hotspots(2) and lead to
the reaction of the bulk.
118

The Relationship between Shock Sensitivity and Morphology in Granular RDX

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The commonly used secondary explosive RDX is produced by many different manufacturers
and it has been found that there are considerable variations in its sensitivity(3). The batches
are generally found to be chemically identical and so the explanation for these differences
seems to lie in the crystal and bed morphology. In this research, four batches of RDX of a
single size from three different manufacturers were examined. Their sensitivity to shock was
assessed using a small-scale gap test, and their morphology was examined using a wide
range of techniques.
The exact processes causing ignition and initiation in granular
explosives are not yet completely understood and so this study should shed light on these
details. We have the opportunity to compare a wide range of large crystals with varying
morphologies.
The processes which occur when a shock wave passes through a granular material do not
depend only on the features of individual crystals. A granular bed consists of individual
particles, supported by each other, with a small number of contact points per particle. Low
level stresses are not supported by all the particles equally, but are transmitted by force
chains which depend on how the particles fit together. Considerable work has been done on
the structure of granular materials and how they transmit stress, but almost all of it relates to
the quasi-static regime. During a shock, there will be a qualitatively different response(4)
which will significantly modify the material behaviour, with the result that quasistatic
studies may be of limited use.
In the past, many studies have been done to try to link particular RDX particle features, for
example closed internal pores, to sensitivity. There has been some success(5,6,7,8), but a
complete explanation has not been found. It is possible that part of the explanation relates
to how the particles interact with each other through the contact points and not only how
particles respond individually. In general, the mechanisms which occur will contribute to
the inhomogeneity of the temperature field and critical hotspot formation are friction, viscoplastic work, jetting, and adiabatic compression of gas-filled pores(9).
The approach that will be taken here is to examine the morphology in a general sense,
including how particles are arranged and how they interact as a bulk material, as well as
studying the features of individual particles.
It should be noted that the sensitivity differences between different batches of RDX have
mostly been observed for material cast in a polymer binder. We have chosen to test the
sensitivity of free-poured material in order to study the simplest possible system, and to
investigate how much the reduced sensitivity effects are due to the behaviour of the asreceived material and not due to interaction with the binder or damage caused by pressing.

Helen Czerski, William Proud, John Field

119

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

EXPERIMENTAL METHODS
2.1

Samples

Granular RDX from three different manufacturers in a single size class was acquired. The
particles had an average size of 1200 m and all fell within the range 600-1500m. Laser
particle size analysis confirmed the size distributions.
To make the samples, as-received crystalline material was poured into the confinements in
small increments and tapped. This resulted in a very reproducible porosity for each sample
batch.
For the purposes of this paper, samples are labelled in order of sensitivity within
each size class, with (1) being the most sensitive material.

2.2

Sensitivity tests

Shock sensitivity was measured using a small scale gap test(10) (see figure 1). The detonator
generates a reproducible shock wave which is then attenuated by a PMMA gap. By
conducting tests with various gaps, the critical gap can be found the largest gap
(corresponding to the lowest pressure) at which ignition will occur.

Figure 1. Diagram of the small-scale gap test. The charge was 9mm in diameter and
25mm long, contained within a 25mm diameter PMMA confinement. The shock pressure
from the detonator alone was approximately 15GPa.

The gap test results are shown in figure 2. It can be seen that there are significant variations
in shock sensitivity and that some samples require twice the input pressure of others in order
for reaction to start.
The criterion that must be satisfied for a go result in a gap test is that part of the charge
detonates. In these materials, almost all the no go samples ignited, but did not transition to
a detonation. From the streak photography used to follow how the reaction progressed with
time, it can be seen that there was a steady increase in the burn speed up to the point where
the material actually detonated. All detonations were prompt detonations and the input
pressure necessary to cause a go result is therefore the detonation pressure. This is
shock-to-detonation or SDT. It is likely that the sensitivity hierarchy would be the same

120

The Relationship between Shock Sensitivity and Morphology in Granular RDX

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

if longer charge lengths had been used to allow deflagration-to-detonation processes to


occur, but this has not been demonstrated for these samples.

5
4.5
4

Gap /mm

3.5
3
2.5
2
1.5
1
0.5
0
0

Figure 2: Gap test results for the four batches. Filled symbols represent go results and
open symbols represent no go results. An additional 2mm of PMMA gap attenuates the
shock by approximately a factor of 2. These results show significant differences in shock
sensitivity, since nearly twice the critical pressure for sample (1) is required to initiate
sample (4).

2.3

Investigation of morphology

A range of techniques are available to study individual particle morphology. Environmental


scanning electron microscopy yields information on surface features of 1m and above in
size. Optical microscopy of particles which are surrounded by a refractive index-matched
fluid can be used to observe closed internal voids down to 1m in size, and also shows the
general particle morphology. Mercury porosimetry provides surface roughness data in the
form of specific surface area (with a resolution down to 0.01m).
Figures 3 and 4 show typical crystals viewed using optical microscopy and ESEM.
discernable surface or internal features correlate with sensitivity in the material.

No

The contents of the closed internal voids seen is unknown, but it has been shown(5) that it is
likely to be a mixture of water, solvent and air. Gas-filled voids are expected to affect shock
sensitivity the most as the gas is compressible and allows the void to collapse and the
material around it to plastically deform.
Previous studies(11) have shown that shock
sensitivity can decrease with void content, i.e a higher pressure is required for the system to
detonate. This can be understood if a variety of mechanisms can cause critical hotspots and
there is no single ignition process for all particle morphologies.

Helen Czerski, William Proud, John Field

121

New Trends in Research of Energetic Materials

Czech Republic

2006

(1)

(2)

(3)

(4)

[Content]

Figure 3 Optical microscopy showing typical crystals from the four batches. The field of
view in each case is 1.3mm wide.
Quantifying bulk morphology is far more difficult, but an indication of intergranular friction
(which may affect the packing and co-ordination number) is provided by measuring the
angle of repose. The angle of repose is the angle between the horizontal and the slope of a
poured pile of the granular material (see figure 5). Particles which interlock more and have
greater intergranular friction will be able to support a steeper slope.
For all of these samples, the measured bulk properties (porosity and angle of repose) were
very similar in spite of the large differences in overall shape. However, they are very
different to the measured quantities for batches of RDX with smaller particle sizes(5), so this
may be relevant in a wider context.

122

The Relationship between Shock Sensitivity and Morphology in Granular RDX

New Trends in Research of Energetic Materials

Czech Republic

2006

(1)

(2)

(3)

(4)

[Content]

Figure 4. ESEM images of the surface of each of the four crystal batches. Since any
critical hotspot is expected to be 0.1-10m in size, it might be expected that features of
approximately that size would be potential sites for critical hotspots.
However, no
correlation is seen here between any surface feature and sensitivity. (1) is the most
sensitive and (4) is the least sensitive material.
Table 1. All quantifiable data for each batch. Samples are shown with the most sensitive
at the top of the list and the least sensitive at the bottom.
Samples from different
manufacturers were supplied with up to 4% of HMX as an impurity. No average void
content is shown because this data does not capture the highly irregular void shape and size
in some samples, and such a comparison would be misleading. The percentage of the
theoretical maximum density (% TMD) data relates to the free-poured and tapped samples
which were used for the sensitivity tests. No trends can be seen in any single parameter as
the shock sensitivity decreases.

Sample

Critical Gap
/mm

%
HMX

(1)
(2)
(3)
(4)

3.6 0.2
3.2 0.4
2.9 0.3
1.9 0.4

0
4.1
0
0.3

Helen Czerski, William Proud, John Field

Specific
Surface
Area m2/g
0.0044
0.0087
0.0018

Angle of
repose

% TMD

35 2
35 1
35 2
32 2

62 2
62 1
59 1
63 1

123

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 5. Angle of repose of a powder.

3.

DISCUSSION

Am empirical study of the morphology and shock sensitivity of a material cannot prove in
absolute terms whether any one particular mechanism alone is responsible for sensitivity.
Direct observation of these mechanisms causing critical hotspots in a shocked material is not
possible in these systems. However, any correlation seen would be a strong indication of
the importance of a particular feature and a corresponding mechanism.
These results show that there are significant differences in sensitivity between different RDX
batches. Some manufacturers claim to produce a "reduced sensitivity" product, but no
causal explanation of these effects has been found. The batches tested here are a mixture of
"reduced sensitivity" products and standard products from manufacturers. Much of the
current effort to quantify and explain the difference has made the assumption that there is
one dominant property or mechanism which extends across all the products. No clear
correlation is seen here between any measured property and the shock sensitivity.
Large particles are more likely to fracture than smaller particles, which tend to plastically
deform, and it seems that some of the stress in the shock could be relieved by particle
fracture before ignition. The samples studied here contained large grains of millimetre
dimensions, and so are also likely to contain large internal cracks and defects initially. This
means that fracture is more likely than plastic deformation to relieve the initial stress.
Fracture surface energy in these materials is not sufficient to cause critical hotspots, but the
fractured morphology could respond differently to the continued loading.
Jetting from surfaces concave to the incident shock can be a source of critical hotspots. As
the radius of curvature of the concave region increases, the resulting jet increases in speed.
Jets causing critical hotspots are most likely from cavities with a radius of a few microns or
less. Significant numbers of external particle features which meet this criterion were not
observed in any one sample more than the others and only the least sensitive material had
significant numbers of closed internal voids of this size. It seems that jetting alone is not a
likely critical mechanism in this case.

124

The Relationship between Shock Sensitivity and Morphology in Granular RDX

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

There are large closed internal voids in the most sensitive material examined, but their
contents is not known. Hotspots caused by adiabatic heating of trapped gas are more likely
when the gas spaces are larger, but in this case the closed internal voids are extensively
interconnected. No other sample contained significant numbers of larger internal voids and
there is no correlation between internal void content and sensitivity here.
The most
important factor here could be the number of gas-filled spaces only, since closed internal
voids could also contain water or solvent.
No clear correlations were seen between the bulk material properties and the sensitivity.
Quantities such as co-ordination number are difficult to measure directly, but could be very
relevant, since the lower the number of interparticle contacts, the higher the stress
concentration at each contact. The importance of such parameters has not been investigated,
but it seems very likely that the nature of the packing and the interparticle contacts must play
a large role in localising stress. Contacts such as these will experience considerable stress
and plastic deformation at such regions could cause very high temperatures locally, making
these potential sites for critical hotspots.
The fact that the go results were all prompt detonations implies that the input pressures
necessary to cause a go were above or equal to the detonation pressure. However, it
seems likely that the resulting sensitivity hierarchy will be the same as one measured with
longer charges.

4.

CONCLUSIONS

In the case of free-poured granular samples, there is no correlation between any individual
morphological feature and shock sensitivity for these large particle sizes. Previous work(7)
has shown that there is no such correlation for smaller particle sizes either. This suggests
that there is no single quality leading to the differences in sensitivity between different RDX
products.
The critical pressure for these samples varies by a factor of two between the most and least
sensitive, so there are significant sensitivity differences between samples of the raw material
from different manufacturers. The observed variation in shock sensitivity may be explained
by qualitative differences in how a shock interacts with the bulk material as well as with
individual particle features. It may be the case that the combination of material properties in
RDX mean that several critical hotspot mechanisms become significant at a similar shock
pressure. Separating out specific mechanisms and quantifying their contribution would be
extremely difficult at this stage. For the samples examined here, we cannot identify which
mechanisms are dominant in causing critical hotspots, but we can make some suggestions.

Bulk morphology is difficult to quantify since properties such as average co-ordination


number cannot be directly measured. Only angle of repose and the particle size analysis
provide quantitative data on bulk morphology and neither of these quantities correlates
directly with sensitivity. However, it seems likely that the interaction of particles in the
granular bed must make some contribution to the inhomogeneity of the stress field. A nonuniform stress field will cause a non-uniform temperature field, because of the
micromechanical response of the material to shock. There are many processes that could
8

Helen Czerski, William Proud, John Field

125

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

cause critical hotspots (viscoplastic deformation, jetting, friction, adiabatic gas collapse) and
all are directly affected by both the microstructure and how the particles are arranged.
Further investigation of how the interaction of bulk and individual particle morphology
might affect shock sensitivity in free-poured granular beds is required. Future papers will
deal with samples which has smaller particle sizes than those reported here.

5.

ACKNOWLEDGEMENTS

We would like to acknowledge Adam Cumming, Dave Tucker and Richard Biers of [dstl]
for funding this research and Ron Hollands or BAe Land Systems for providing samples and
background knowledge. We are grateful for technical assistance from Dave Powell of the
Cavendish Laboratory.

6.

REFERENCES

[1]

N. K.BOURNE and J. E. FIELD : Explosive ignition by the collapse of cavities, Proceedings of


the Royal Society series A, 455(1987): 2411-2426 Jul 8 (1999)
[2] F. P.BOWDEN and A. D. YOFFE : Initation and Growth of Explosion in Liquids and Solids,
Cambridge University Press, London, 1952
[3] F. PEUGEOT and D. WATT: RS-RDX, Literature Review and Discussions, NIMIC report, 2004
[4] V. F.NESTERENKO: Dynamics of Heterogeneous Materials, Springer Series: High Pressure
Shock Compression of Condensed Matter, 2001
[5] L. BORNE: Explosive Crystal Microstructure and Shock Sensitivity of Cast Formulations, 11st
Detonation Symposium, 1998, p657-663
[6] L. BORNE and A. BEAUCAMP: Effects of Explosive Crystal Internal Defects on Projectile
Impact Initiation, 12th Detonation Symposium, 2002, p35
[7] H. MOULARD, J.W.KURY, and A. DELCLOS : The effect of RDX particle size on the shock
sensitivity of cast PBX formuations, 8th Detonation Symposium, 1986
[8] L. BORNE and J. PATEDOYE: Quantitative Characterization of Internal Defects in RDX
Crystals, Propellants, Explosives, Pyrotechnics 24, 255-259, 1999
[9] J. E. FIELD, N.K. BOURNE, S.J.P.PALMER, S.M. WALLEY, J.M. SMALLWOOD: Hotspot ignition
mechanisms for explosive and propellants, Proceedings of the Royal Society series A,
339(1654): 269-283 May 15 1992
[10] A. CHAKRAVATY, M. J. GIFFORD, M. W. GREENAWAY, W.G. PROUD and J. E. FIELD : Factors
affecting shock sensitivity of energetic materials, Shock Compression of Condensed Matter,
1007-1010, 2001
[11] H. CZERSKI, M.W.GREENAWAY, W.G. PROUD and J. E. FIELD : Links between the morphology
of RDX crystals and their shock sensitivity, Shock Compression of Condensed matter, 2006
[12] H. CZERSKI, W. G. PROUD and J. E. FIELD: Aspects of Material Insensitivity, Submitted to
Insensitive Munitions & Energetic Materials Technology Symposium, Bristol 2006

126

The Relationship between Shock Sensitivity and Morphology in Granular RDX

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

GUANIDINIUM NITROFORMATE SALTS:


POSSIBLE NEW OXIDIZERS FOR HIGH PERFORMANCE,
HALOGEN FREE SOLID PROPELLANTS
Michael Gbel, Thomas M. Klaptke and Peter C. Thumbs
Department of Chemistry and Biochemistry
Ludwig-Maximilians University of Munich
Butenandtstr. 5 - 13 (Haus D)
D-81377 Munich, Germany
Abstract:
Energetic salts of the nitroformate anion with several guanidinium cations were
investigated, including guanidinium nitroformate monohydrate (GNFH),
aminoguanidinium nitroformate (AGNF), diaminoguanidinium nitroformate
(DAGNF) as well as triaminoguanidinium nitroformate (TAGNF). Guanidinium
nitroformate was obtained according to the literature procedure as the
monohydrate (GNFH), whereas AGNF, DAGNF and TAGNF are reported for the
first time. All salts were characterized using vibrational spectroscopy (IR,
Raman), mass spectrometry, multinuclear NMR spectroscopy and elemental
analysis. In addition, the impact, friction and electrostatic sensitivity data were
determined for all compounds. The crystal structures of GNFH, AGNF, DAGNF
and TAGNF were determined using single crystal X-ray diffraction. The thermal
decomposition of the salts was monitored using differential scanning calorimetry.
Keywords:

1.

nitroform, X-ray crystallography, GNFH, AGNF, DAGNF, TAGNF

INTRODUCTION

Although several energetic salts containing the nitroformate anion have been reported in the
literature, so far only one compound has gained practical application. Since its discovery in
1951, more than eighty publications [1] have since appeared, showing that there is continued
interest in investigating the properties of hydrazinium nitroformate (HNF). Furthermore, it
has been shown that a formulation containing HNF, aluminum (Al) and glycidyl azide
polymer (GAP) has improved properties in comparison with other composite propellants,
and a theoretically predicted performance that is superior to even advanced propellants
based on the oxidizer ADN. Investigations into the use of HNF and HNF containing
propellants continues to be undertaken by various companies including Nobel's Explosives
Company, ICI (Great-Britain), Fiat Avio (Italy) and SNPE (France) [2].
Recently,
two
derivatives
of
hydrazinium
nitroformate
(HNF),
namely
monomethylhydrazinium nitroformate (MMHNF) and dimethylhydrazinium nitroformate
(DMHNF) were synthesized and suggested to be not only promising new high performance
energetic materials, but also potential ingredients for melt castable high explosive
formulations [3].
In the course of our investigations into high energy density materials (HEDM) [4-13], we
recently focussed our attention to various guanidinium nitroformate salts, in order to explore
their potential as possible new oxidizers for high performance, halogen-free solid
propellants, which would avoid the problematic formation of hydrogen chloride when used
as oxidizers in rocket propellant formulations.
1

Michael Gbel, Thomas Klaptke

127

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

SYNTHESIS

NHR3

NO2

K/Ag

C
O2N

NO2

R1HN

NHR2

NO2

Cl

O2N

NO2

NHR3
R1HN

NHR2

KCl/AgCl

GNFH: R1 = R2 = R3 = H
AGNF: R1 = NH2, R2 = R3 = H
DAGNF: R1 = R2 = NH2, R3 = H
TAGNF: R1 = R2 = R3 = NH2

All of the compounds were obtained as bright yellow solids according to a slightly modified
procedure [14] in high yields by the reaction of either the potassium or silver salt of nitroform
with the corresponding guanidinium chloride in acetonitrile.

3.

CRYSTAL STRUCTURES

Fig 1. Asymmetric unit of Guanidinium nitroformate hydrate (GNFH) (1) (left) and
Aminoguanidinium nitroformate (AGNF) (2) (right).

Fig 2. Asymmetric unit of Diaminoguanidinium nitroformate (DAGNF) (3) (left) and


Triaminoguanidinium nitroformate (TAGNF) (4) (right).
128

Guanidinium Nitroformate Salts: Possible New Oxidizers for High Performance, . . .

New Trends in Research of Energetic Materials

TABLE 1.

2006

[Content]

Composition of Crystallographic data of GNFH, AGNF, DAGNF and


TAGNF.

compound
formula
formula weight
temperature / K
crystal system
space group
a/
b/
c/
/
/
/
volume / 3
Z
absorption coefficient /mm-1
density exptl. / g / cm3
F(000)
2 theta /
index ranges
reflections collected
reflections unique
parameters
GOOF
R1 / wR2 [I > 2(I)]
R1 / wR2 (all data)

Czech Republic

GNFH (1)
C2 H8 N6 O7
228.14
200(2)
monoclinic
C 2/c (no. 15)
8.1859(4)
14.3541(7)
7.7556(5)
90.0
101.132(2)
90.0
894.15(8)
4
0.167
1.695
472
54.88
- 10 h 9,
- 18 k 17,
- 9 l 10
1021
811
87
1.072
0.0513 / 0.1293
0.0645 / 0.1395

Michael Gbel, Thomas Klaptke

AGNF (2)
DAGNF (3)
TAGNF (4)
C2 H8 N8 O6
C2 H9 N9 O6
C2 H7 N7 O6
225.15
240.16
255.18
200(2)
200(2)
200(2)
triclinic
monoclinic
triclinic
P 21/n (no. 14)
P -1
P-1
5.0690(7)
10.980(1)
8.021(2)
7.559(1)
7.7524(8)
8.347(2)
11.320(1)
11.416(1)
8.515(2)
84.567(9)
90.00
105.49(2)
84.425(9)
105.336(8)
95.03(2)
79.61(1)
90.00
111.10(2)
423.29(9)
937.1(2)
501.7(2)
2
4
2
0.170
0.162
0.159
1.766
1.702
1.689
232
496
264
52
52
52
- 6 h 6,
-13 h 13,
-9 h 9,
- 9 k 9,
-9 k 9,
-10 k 9,
- 13 l 13
- 14 l 14
- 6 l 10
1657
1832
1962
1515
1573
1617
164
178
190
1.086
1.168
1.099
0.0312 / 0.0748 0.0467 / 0.1064 0.0391 / 0.1014
0.0346 / 0.0773 0.0604 / 0.1155 0.0491 / 0.1113

129

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Considering the relationship of the orientation of the nitro groups relative to the central C-N3
plane it appears that the relation of bond lengths vs. dihedral angles of the nitro groups works
well [15]. A correlation between the length of the C-N and N-O bonds and the corresponding
dihedral angles can be observed as expected, whereby the C-N bonds of out-of-plane nitro
groups are elongated and C-N bonds of in-plane nitro groups are shortened. There is also a
reverse tendency for N-O distances. The structural characteristics such as bond distances and
dihedral angles of the guanidinium salts of nitroform are presented in Table 2, and fall within
the range of expected values.

TABLE 2.

Selected structural parameters of the nitroformate anion in the guanidinium,


aminoguanidinium, diaminoguanidinium and triaminoguanidinium salts.

nitro groups of:


GNFH
1
2
3
AGNF
1
2
3
DAGNF
1
2
3
TAGNF
1
2
3

130

dihedral angle []

C-N distance
[]

N-O distance []

80.9(3)
4.8(6)
4.8(6)

1.454(2)
1.363(5)
1.363(5)

1.217(4)
1.246(9)
1.246(9)

16.1(2)
35.7(2)
24.4(2)

1.385(2)
1.418(5)
1.409(6)

1.243(2)
1.228(1)
1.235(1)

24.5(5)
31.7(8)
9.7(8)

1.397(2)
1.411(2)
1.396(2)

1.235(5)
1.229(5)
1.238(2)

2.6(7)
95.2(7)
2.7(2)

1.373(2)
1.461(2)
1.367(2)

1.251(6)
1.217(6)
1.254(2)

Guanidinium Nitroformate Salts: Possible New Oxidizers for High Performance, . . .

New Trends in Research of Energetic Materials

4.

Czech Republic

2006

[Content]

NMR SPECTROSCOPY

An investigation of the NF anion and the corresponding guanidinium cation in


solution was performed using 1H, 13C and 15N NMR spectroscopy. The 13C and 15N
spectra of the anion showed one signal at room temperature. A comparison of the
obtained signals can be found in Table 3. The signals of the NF anion fit well with
those reported by A.A. Gakh et al. [15] who reported the 13C chemical shifts of the
cesium and the tetrabutylammonium salts of the nitroformate anion at 150.3 ppm and
151.6 ppm (both recorded in acetone-d6), respectively.
Table 3.

NMR Spectroscopic data of GNFH, AGNF, DAGNF and TAGNF.


13

GNFH
AGNF

5.

C (DMSO-d6) / :
158.0 (C(NH2)3+)
150.3 (C(NO2)3-)
158.9(H2N-NH=C(NH2)2+)
150.3 (C(NO2)3-)

DAGNF

159.9 (H2N-C(NH-NH2)2+)
150.4 (C(NO2)3-)

TAGNF

161.8 (C(NH-NH2)3+)
150.6 (C(NO2)3-)

15

N (DMSO-d6) / :
-306 (NH2)
-31 (NO2)
-326 (NH-NH2)
-311 (NH-NH2)
-284 (C-NH2)
-30 (NO2)
-328 (NH-NH2)
-313 (NH-NH2)
-288 (C-NH2)
-31 (NO2)
-249 (NH-NH2)
-95 (NH-NH2)
-31 (NO2)

H (DMSO-d6) / :
6.9 (NH2)
3.39 (H2O)
8.52
6.94
4.28

8.50 (NH-NH2)
4.48(C-NH2,NH-NH2))

9.79 (C(NH-NH2)3+)
3.53 (C(NH-NH2)3+)

DECOMPOSITION EXPERIMENTS

Differential Scanning Calorimetry (DSC) was used to determine the relative thermal
stabilities of compounds 1-4. Figure 3 shows the DSC curves that were obtained for 2-4.
Whereas 2 shows only one decomposition signal (80C (onset)), 3 and 4 show melting
points ((78C (onset)) and (83C (onset))) as well as subsequent decomposition ((110C
(onset)) and (138C (onset)) using a heating rate of 20C/min. The thermal stability
increases from compounds 2-4. Since 1 is a hydrate, it was not included in Figure 3. Table 3
summarizes the thermal behaviour of 1-4 using heating rates of 2/min. IR spectroscopy was
used to show that slow decomposition of 2 and 3 occurs at room temperature. AGNF and
DAGNF were converted to the corresponding nitrates on standing at room temperature and
are believed to decompose in much the same way as potassium nitroformate which is
converted to potassium nitrate [16].

Michael Gbel, Thomas Klaptke

131

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 3. DSC curves of AGNF, DAGNF and TAGNF ( = 20/min).

The sensitivity data of GNFH, AGNF, DAGNF and TAGNF were measured in order to
establish safe handling procedures for these compounds. Two different instruments were
used to determine the friction and impact sensitivity: the Bundesanstalt fr
Materialforschung und Prfung (BAM) drop hammer (BAM fh (Fallhammer)) and friction
tester (BAM ft) [17]. GNFH was found to meet the United Nations (UN) recommendations
for the transport of dangerous goods, with a friction sensitivity of greater than 360 N and an
impact sensitivity of greater than 29.6 J. In contrast, great care should be taken when
handling the other salts (Table 2) which are considerably more sensitive.
Table 4.

Comparison of the sensitivity data of GNFH, AGNF, DAGNF and TAGNF.


The values of the melting and the decomposition points correspond to a
heating rate of = 2/min.
Parameter
GNFH
AGNF
DAGNF
TAGNF
C2H8N6O7
C2H7N7O6
C2H8N8O6 C2H9N9O6
formula
228.12
225.15
240.16
255.18
weight [g/mol]
1.70
1.77
1.70
1.6
density [g/cm3]
110-117
80
84
melting point [C]
71
82
104
decomp. point [C]
deflagration deflagration deflagration deflagration
flame
negative
negative
negative
negative
electrostatic discharge
> 360
144
32-40
20-24
friction sensitivity [N]
30
10
5-6
2-2.5
impact sensitivity [Nm]

132

Guanidinium Nitroformate Salts: Possible New Oxidizers for High Performance, . . .

New Trends in Research of Energetic Materials

6.

Czech Republic

2006

[Content]

CONCLUSION

Although guanidinium nitroformate monohydrate (GNFH) has already been mentioned in


the literature,[14] the characerization reported in the original paper was limited to the melting
point as well as IR spectroscopy and mass spectrometry. This compound has now been
extensively characterized, as well as the related anhydrous species aminoguanidinium
nitroformate
(AGNF),
diaminoguanidinium
nitroformate
(DAGNF)
and
triaminoguanidinium nitroformate (TAGNF), which have been synthesized for the first time.
The sensitivity data and the thermal behaviour of the compounds appears to follow a general
trend. Correlations of analytical, sensitivity and performance properties of these compounds
within the series of homologous compounds, as well as between these compounds and HNF
have been established. Triaminoguanidinium nitroformate (TAGNF) seems to be a more
promising candidate for practical application than the GNFH, AGNF and DAGNF salts.

Acknowledgements
We thank Priv.-Doz. Dr. K. Karaghiosoff for carrying out the NMR experiments. We are
grateful to the Cusanuswerk for a scholarship to MG. Ludwig-Maximilan University and the
Fonds der Chemischen Industrie are thanked for generous financial support.

Michael Gbel, Thomas Klaptke

133

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REFERENCES
[1]
[2]
[3]

[4]

[5]

[6]
[7]

[8]

[9]

[10]

[11]
[12]

[13]

[14]
[15]
[16]
[17]

134

SciFinder enquiry 11/2005


www.appbv.nl/hnf.html
Synthesis, characterization and thermal behaviour of hydrazinium nitroformate (HNF)
and its new N-alkyl substituted derivatives. Jadhav, H. S.; Talawar, M. B.; Dhavale,
D. D.; Asthana, S. N.; Krishnamurthy, V. N., Indian Journal of Chemical Technology
2005, 12(2), 187.
Mononitrobiuret (MNB) and 1,5-Dinitrobiuret (DNB): Pyrolysis Experiments and
Thermochemistry, J. Geith, G. Holl, T. M. Klaptke, J. J. Weigand, Combustion and
Flame 2004, 139, 358.
1,4-Bis-[1-methyltetrazol-5-yl)-1,4-dimethyl-2-tetrazene: A Stable, Highly Energetic
Hexamer of Diazomethane (CH2N2)6 ,T. M. Klaptke, P. Mayer, A. Schulz, J. J.
Weigand, Propellants Explosives and Pyrotechnics 2004, 29, 325.
Improved Synthesis and X-Ray Structure of 5-Aminotetrazolium Nitrate, M. v.
Denffer, T. M. Klaptke, G. Kramer, G. Spie, J. M. Welch, G. Heeb, Propellants
Explosives and Pyrotechnics, 2005, 30(3), 191.
1,5-Diamino-4-methyltetrazolium dinitramide, T. M. Klaptke, P. Mayer, A. Schulz,
J. J. Weigand, J. Am. Chem. Soc., 2005, 127, 2032.
Derivatives of 1,5-Diamino-1H-tetrazole A New Family of Energetic Heterocyclicbased Salts, J. C. Glvez-Ruiz, G. Holl, K. Karaghiosoff, T. M. Klaptke, K.
Lhnwitz, P. Mayer, H. Nth, K. Polborn, Ch. J. Rohbogner, M. Suter, J. J. Weigand,
Inorg. Chem. 2005, 44(12), 4237.
A study on the thermal decomposition behavior on derivatives of 1,5-diamino-1Htetrazole (DAT): a new family of energetic heterocyclic-based salts G. Fischer, G.
Holl, T, M. Klaptke, J. J. Weigand, Thermochim Acta 2005, 437, 168.
The Elimination of NO2 from Mixtures of the Nitramines HMX, RDX and CL20 with
the Energetic Binder Glycidyl Azide Polymer (GAP) - A Computational Study I, M.
A. Bohn, A. Hammerl, K. Harris, T. M. Klaptke, Central European Journal of
Energetic Materials, 2005, 2(2), 29.
Azidoformamidinium and Guanidinium 5,5'-Azotetrazolate Salts, M. A. Hiskey, A.
Hammerl, G. Holl, T. M. Klaptke, K. Polborn, J. Stierstorfer, J. J. Weigand,
Chemistry of Materials, 2005, 17, 3784.
The Dianion of 5-Cyanoiminotetrazoline: C2N62-, T. M. Klaptke, C. Kuffer, P.
Mayer, K. Polborn, A. Schulz, J. J. Weigand, Inorg. Chem., 2005, 44(16), 5949.
Further Decomposition Pathways of Mixtures of the Nitramines HMX, RDX and
CL20 with the Energetic Binder Glycidyl Azide Polymer (GAP) - A Computational
study II, M. A. Bohn, A. Hammerl, K. Harris, T. M. Klaptke, Central European
Journal of Energetic Materials, 2005, 2(3), 3.
Ashwin M. Krishnan, Per Sjoberg, Peter Politzer and Joseph H. Boyer, J. Chem. Soc.
Perkin Trans. II, 1989, 1237.
A. A. Gakh, J. C. Bryan, M. N. Burnett, P. V. Bonnesen, J. Mol. Struct. 2000, 520,
221.
M. Gbel, T. M. Klaptke, P. Mayer, ZAAC, 2006, in press.
Test methods according to the UN Recommendations on the Transport of Dangerous
Goods, Manual of Tests and Criteria, fourth revised edition, United Nations
Publications, New York and Geneva, 2003, ISBN 92-1-139087-7, Sales No.
E.03.VIII.2.; 13.4.2 Test 3(a)(ii) BAM Fallhammer.

Guanidinium Nitroformate Salts: Possible New Oxidizers for High Performance, . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

PRELIMINARY STUDY OF SUPERFINE METAL POWDERS EFFECT


ON THE DECOMPOSITION OF ENERGETIC MATERIALS
Alexander Gromov *, Elena Popenko **, Alexey Sergienko**
Young-Soon Kwon***, Ulrich Teipel****
* Tomsk Polytechnic University, 30, Lenin Ave., 634050, Tomsk, Russia
** Biysk Technologic Institute, 29, Trofimov Str., 659315, Biysk, Russia
*** Research Center for Machine Parts and Materials Processing, School of Materials and
Metallurgical Engineering, University of Ulsan, Mugeo-2Dong, Ulsan 680-749, South Korea
**** Fraunhofer Institute for Chemical Technology, P.O. Box 1240, D-76318 Pfinztal, Germany

Abstract:
The preliminary study describes the decomposition of energetic materials (AP, HMX) in
the presence of superfine metal powders (SFP) of Al, Ni, Cu-Ni, Fe, Cu and W. The
decomposition temperature for the energetic substances could be lowered because of
catalytic effect of superfine metal powders. The effect of several types of metal particles
produces by electro-explosive method to the rate of decomposition, temperature of
decomposition of modern energetic substances have been discussed. The of DTA-TG
investigation were conducted under non-linear heating of mixes (SFP-HMX) and (SFPAP) in air.
Keywords: decomposition, HMX, AP, catalysis, superfine metal powders, DTA-TG.

1.

INTRODUCTION

It is known that the usage of superfine Al powders (SFAP, as ~ 100 nm) instead of
industrial Al powders (as ~ 10 m) results in achievement of the double increase in the
burning rate for aluminized solid propellant, which contain 55 % of nitrocellulose (13,5 %
N), 30 % of dioxyethylnitramine and 15 % of Al [1]. The complete substitution of industrial
Al powders to superfine Al powders does lead to an increase in the burning rate of
combustion for aluminized solid propellant on the basis of active binder (20 %), AP (27 %),
HMX (35 %) and Al (18 %) [2,3]. The value of pressure exponent is reduced in the burning
law u=Ap, at the same time the combustion efficiency of metal increases and the degree of
agglomeration is reduced [4]. It is known that decomposition temperature for the energetic
substances could be lowered because of catalytic effect of superfine metal powders [5], but
the experimental data needs to be confirmed as well as the mechanism of such catalytic
process has to be developed. In this work we tested the effect of several types of metal
nanopowders, produces by electro-explosive method [6]. The purpose of this work is a
preliminary study of catalytic activity of superfine metal powders in their mixtures with AP
and HMX. The three types of superfine Al powders, SFP Fe, SFP W, SFP Ni, SFP Cu and
SFP (Cu-Ni) composition as well as micron sized Al powder (ASD-4) was tested.

2.

EXPERIMENTAL RESULTS AND DISCUSSION

Production of the samples of superfine metal powders was carried out by using an electrical
explosion of wires (EEW) machine UDP-4G which was constructed for the production of
1

Alexander Gromov, Elena Popenko, Alexey Sergienko, Young-Soon Kwon, . . .

135

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

metal powders at the High Voltage Research Institute, Tomsk Polytechnic University,
Russia.
Non-passivated superfine metal powders self-ignite if exposed to air, making
passivation for them essential. Hence, after the EEW machine stopped, the collector with
prepared powders was placed into a separate hermetic passivation chamber immediately
after powder production. Ten samples, studied within this work and their specific surface
area, determined by BET method, as well as metal content, measured by DTA analysis, are
shown in table 1.
Table 1. Properties of nano-metals
BET, as,
m2/g m

Cmetal,
wt. %

1. SFAP-Al2O3

Initial wire Gas media in Passivation


condition
composition explosive
chamber
Al
Ar
Air

7.0

0.3

76.1

2. SFAP -AlB2

Al (B)

Ar

Air

17.0

0.1

78.2

3. SFAP -NC

Al

Ar

0.4

67.5

4. ASD-4

Industrial powder

Nitrocellulose 5.0
in ethanol
0.4

5.6

98.5

5. Fe

Fe

Ar

Air

18.2

0.1

88.9

6. Ni

Ni

Ar

Air

20.3

0.1

91.0

7. Cu

Cu

Ar

Air

16.8

0.1

93.2

8. W

Ar

Air

5.2

0.4

98.0

9. Cu-6 % Ni

Cu-6 % Ni

Ar

Air

5.9

0.4

96.5

10. Cu-45 % Ni

Cu-45 % Ni

Ar

Air

6.4

0.3

96.7

Sample code

The powders, passivated with AlB2 (SFAP-AlB2), nitrocellulose (SFAP-NC) and


SFAP, which particles had the oxide passivation layer (SFAP-Al2O3) as well as other metals
were studied by SEM (Fig.1). The application of the AlB2 passivation coating was
performed for the increase in the stability of particles to the oxidation and for the decrease in
the agglomeration of Al particles during production [7]. The specific surface area of powder
with the AlB2 coating is twice as much as for the powder, which has the oxide passivating
layer (table 1). The application of NC coating was proceed with the purpose of the study of
aluminum oxidation activation due to the high exothermic effect of NC decomposition at
T=180oC. Coating of powders by nitrocellulose, which contains 12.4 wt. % of nitrogen, was
used by the method of SFAP treatment in the 10 wt. % NC solution in acetone with the
subsequent evaporation of acetone at room temperature. As a result SFAP had the content of
NC - 33 wt. %. The other metals were passivated by air in standard procedure [8] and had the
oxide layers.

136

Preliminary Study of Superfine Metal Powders Effect on the Decomposition of . . .

New Trends in Research of Energetic Materials

Al-Al2O3

Al-AlB2

Ni

Fe

W
Fig 1.

Czech Republic

Cu-6 % Ni

2006

[Content]

Al-NC

Cu

Cu-45 % Ni

SEM images of superfine metal powders (JSM-6500 F JEOL, Japan).

The superfine metal powders were tested by the method of differential-thermal


analysis (DTA). DTA-TG (Universal 2.4 F TA Instruments) was used for testing of
components non-isothermal oxidation in air with heating rate 10 K/min. DTA results are
given in fig. 2. The temperature of the oxidation onset (Ton, oC) and the degree of the
oxidation (, %), calculated under the (eq. 1) of the studied Al powders were different
depending on the particle size.
+m
(1),
(AlAl2O3) = ---------------- 100%
Al 0.89
+m, % - weight increase on TG curve;
Al, % - Al metal content in the powder.
Industrial powder ASD-4 had Ton = 580oC, while for SFAP-Al23 this temperature
was 510 (fig. 2, a). For the SFAP-Al23 the considerably higher degree of oxidation
(=28.3 % up to T=660oC) is observed in comparison with the industrial powder ASD-4
(=1.8 %) (fig. 2, b). SFAP with the coating of AlB2 had higher Ton = 560oC in comparison
with the SFAP-Al23 (510). Probably, this is caused by the protective film of refractory
AlB2. The degrees of the oxidation of the samples SFAP-Al23 and SFAP-AlB2 are higher at
3

Alexander Gromov, Elena Popenko, Alexey Sergienko, Young-Soon Kwon, . . .

137

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

660oC and approximately identical. In the DTA curve for sample Al-NC two exothermic
effects are characteristic, whose are caused by NC decomposition in the temperature range of
180-280 with the 25.0 % weight decrease and Al metal oxidation (510-620). The
gasification of the carbonaceous species, which were being formed during the NC
decomposition, occurs in the range of temperatures 280-510. The total weight loss at the
stage of NC decomposition (on TG) it achieved 31.2 %. The intensive Al oxidation for the
sample SFAP-NC occurs with T=525-620 and the degree of oxidation in this case was
=19.4 %. Thus, the processes of NC and Al decomposition under the heating in air proceed
independently of each other (with DTA curves). The parameters of thermal decomposition of
NC and the parameters of SFAP oxidation remain at the same level as for individual
substances.
The DTA results of mixtures (SFAP-Al23 + 50 % HMX) showed that the
temperature of the beginning of the decomposition (Td) of HMX in the studied mixture is
Td=260oC (fig. 2, c), which is 28oC lower than for the pure HMX (Td=288oC). Hence, SFAPAl23 initiates the decomposition of HMX at the lower temperature. The parameters of the
oxidation of SFAP-Al23 sample in presence of HMX do not change. SFAP, covered by
AlB2 and NC, have an analogous effect on the Td of HMX, i.e., the Td is decreased down to
240 and 250, respectively. The small decrease of Td for the HMX (to 270) was
observed also on the DTA curve of mixture (ASD-4 + 50 % HMX). The degree of ASD-4
oxidation (=1.1 % to 660) in the mixture with the HMX was considerably lower in
comparison with the oxidation of the sample SFAP -Al23, for which =13.3 % at 660
(fig. 2, d). The parameters of AP decomposition are practically not change in the mixture
(SFAP-Al23+50 % AP). Oxidation of SFAP-Al23 in the mixture with AP begins on 30
lower (d=480) in comparison with the oxidation of pure SFAP -Al23 (fig. 1, e,f).
The results of DTA-TG investigation of the other metal SFPs are presented in table 2.
Compared to ASD-4 and SFAP, the samples 5-10 have the very low temperatures of the
oxidation onset (174-352oC), while oxid for all metals were relatively high, except for SFP
W.
The effect of superfine metal particles on the HMX decomposition was studied with
DTA-TG (fig. 3). The mixtures (SFP Me + 50 % HMX) was suspended in the ethanol
(C2H6O), dried in air and than analysed. Pure HMX had a Td=288oC. The Cu, Fe and (Cu-6
% Ni) powder (samples 7, 5 and 10, table 3) strongly effect on HMX decomposition. The
decomposition temperature of HMX was 96oC lower as a result of the Fe powder interaction
with HMX. Hence, Fe and Cu powder has been selected for the future study.

3.

CONCLUSION

The results of experimental studies of superfine metal powder effect on the HMX and
AP decomposition shows the strong effect of Fe, Cu and Cu-Ni superfine powders on the
HMX decomposition temperature it is lowered on more than on 90oC. Superfine aluminum
powders coated by AlB2 react with HMX at the same temperatures, but with twice higher
heat release in comparison with SFAP, coated by Al2O3 and NC.

4.

ACKNOWLEDGEMENT

This work has been financially supported by INTAS grant YSF 03-55-671 and Russian
President grant 1812. 2005.8.

138

Preliminary Study of Superfine Metal Powders Effect on the Decomposition of . . .

70

Alexander Gromov, Elena Popenko, Alexey Sergienko, Young-Soon Kwon, . . .

m, %

150

150

a)

450

300

SFAP-AlB2

c)

450

ASD-4 (Micron-sized)

SFAP-NC

SFAP-Al2O3

300

NC decomposition

SFAP-AlB2

ASD-4 (Micron-sized)

SFAP-NC

SFAP-Al2O3

600

600

, o

, o

750

750

-45

-35

-25

-15

-5

m, %

-20

20

40

60

80

DTA, mV

300

300

SFAP-Al2O3+HMX
SFAP-NC+HMX
ASD-4+HMX
SFAP-AlB2+HMX

150

150

d)

b)

450

450

600

600

, 0

SFAP-AlB2+HMX

ASD-4+HMX

SFAP-NC+HMX

SFAP-Al2O3+HMX

750

750

, 0

2006

-40

-30

-20

-10

DTA, mV

Czech Republic

10

20

30

40

-20

-10

10

20

30

40

50

60

80

New Trends in Research of Energetic Materials


[Content]

139

140
450

600
0

750

-100

-80

-60

AP

SFAP-Al2O3 + AP

750

8.

284

Fig 2.

1766*

2408*

611

4040

4870

4180

220

5790

6232

4730

oxid Me, J/g

DTA and TG curves of the studied SFAP.

*two peaks of oxidation (total).

174

181

10. Cu-45 % Ni

600

Cu

7.

337

-40

450

Ni

6.

352

239

300

Fe

5.

Cu-6 % Ni

150

300

9.

150

580

525

560

-20

m, %

ASD-4

SFAP-NC

3.
4.

SFAP-AlB2

SFAP-Al2O3 510

1.

AP

2.

Sample code on, o

SFAP-Al2O3 + AP

Table 2. DTA-TG study of metal SFPs

16

11

14

23

19

35

19

22

28

+m (to 600oC), %

Czech Republic
2006

f)

e)

-5

15

25

DTA, mV

New Trends in Research of Energetic Materials


[Content]

Preliminary Study of Superfine Metal Powders Effect on the Decomposition of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 3. DTA-TG study of mixes (SFP Me + 50 % HMX)


Sample code

on, o

T, oC

oxid Me, J/g

-m (to 300oC), % Remark

HMX

288

3240

70

5.

Fe

192

96

1500

48

6.

Ni

283

747

45

7.

Cu

228

60

2270

34

8.

284

611

14

9.

Cu-6 % Ni

193

95

1800

39

10.

Cu-45 % Ni

196

92

4831

48

No effect

No effect

REFERENCES
[1]

[2]

[3]

[4]
[5]

[6]
[7]
[8]

B. BASCHUNG, D. GRUNE, H.H. LICHT, SAMIRANT. Combustion Phenomena of a Solid


Propellant
Based
on
Aluminum
Powders.
Internet
Site:
www.argonide.com/Gun_Propellant_Report.pdf
V.N. SIMONENKO, V.E. ZARKO. Comparative Studying of the Combustion Behavior of
Composite Propellants Containing Ultra Fine Aluminum. In: Energetic materials. 30th Int.
Annual Conf. of ICT, 1999. Karlsruhe, Germany. Report 21, 14 P.
O.G. GLOTOV, V.E. ZARKO, M.W. BECKSTEAD. Agglomerate and Oxide Particles Generated
in Combustion of Alex Containing Solid Propellants. In: Energetic materials. Analysis,
diagnostics and testing. 31th Int. Annual Conf. of ICT, 2000. Karlsruhe, Germany, Report 130,
14 P.
U. TEIPEL, I. MIKONSAARI. Size Reduction of Particulate Energetic Material, Propellants,
Explosives, Pyrotechnics. 2002, Vol. 27, P. 168-174.
LEILI LIU, FENGSHENG LI, LINGHUA TAN, LI MING, YANG YI. Effects of Nanometer Ni, Cu, Al
and NiCu Powders on the Thermal Decomposition of Ammonium Perchlorate, Propellants,
Explosives, Pyrotechnics. 2004, Vol. 29, P. 34-38.
Y.S. KWON, Y.H. JUNG, N.A. YAVOROVSKY, A.P. ILYIN AND J.S. KIM: Ultrafine Metal
Powders by Wires Electric Explosion Method, Scripta mater. 2001, Vol. 44, P. 2247-2251.
Y.S. KWON, A.A. GROMOV, A.P. ILYIN. Reactivity of Superfine Aluminum Powder Stabilized
by Aluminum Diboride, Combustion and Flame. 2002, Vol. 131. P. 349352.
Y. S. KWON, A.A. GROMOV, A.P. ILYIN AND G.H. RIM: Passivation Process for Superfine
Aluminum Powders Obtained by Electrical Explosion of Wires, Applied Surface Science, 2003,
Vol. 211, P.57-67.

Alexander Gromov, Elena Popenko, Alexey Sergienko, Young-Soon Kwon, . . .

141

New Trends in Research of Energetic Materials

Fig 3.

142

Czech Republic

2006

[Content]

DTA and TGA curves for superfine metal powders mixtures with HMX under nonisothermal heating in air (m = 2.6 mg, vheat= 10 K/min, etalon -Al2O3).
Preliminary Study of Superfine Metal Powders Effect on the Decomposition of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

PREPARATION, PROPERTIES AND APPLICATION OF ENERGETIC

COMPLEX PERCHLORATES OF D-METALS


Mikhail A. Ilyushin, Igor V. Tselinsky, Irina V. Bachurina, Yurii A. Gruzdev,
Valery P. Sinditskii#, Valery V. Serushkin# and Vyacheslav Yu. Egorshin#
Saint-Petersburg State Institute of Technology (Technical University), 26 Moskovsky pr., SaintPetersburg, Russia
#

D.I.Mendeleev Russian University of Chemical Technology, 9 Miusskaya sq., Moscow, Russia.


Abstract:
Methods for the synthesis of novel energetic complex perchlorates of d-metals have been
developed. According to derivatographic analysis data, the complexes are thermally
stable compounds. One of the promising ways of regulating ballistic characteristics of
highly filled energetic formulations is the introduction of burn modifiers. Complex
compounds of d-metals are known to be effective modifiers of the process of burning.
Ballistic tests were made in a constant pressure bomb and demonstrated that the
prepared compounds are promising modifiers of the burn rate and can be used in the
developed formulations of rocket propellants.
Keywords:
complex perchlorates, 1,5-pentamethylenetetrazole, modifiers of
burning, derivatographic analysis, drop hammer test, ballistic test, bomb of constant
pressure.

INTRODUCTION

1.

The report is devoted to the synthesis and characterization of coordination compounds in


the series of perchlorates of d-metals containing substituted tetrazole as ligand. Considerable
attention to this class of energetic materials is caused by their application in safe detonators
and rocket propellants.

2.

DISCUSSION
2.1

Synthesis and properties of the perchlorate complexes

We have carried out the synthesis of new energetic d-metal complex perchlorates with
1,5-pentamethylenetetrazole (PMT) as ligand.

Mikhail Ilyushin, Igor Tselinsky, Irina Bachurina, Yurii Gruzdev, . . .

143

New Trends in Research of Energetic Materials

Czech Republic

CH2

CH2

CH2

CH2

2006

[Content]

N
N

H2C

1,5-Pentamethylenetetrazole (PMT)
Preparation of metal complexes was performed according to the following reactions:
For Cu(II) complexes
Cu(H3COO)2 + n(CH2)5CN4 + 2HClO4 [Cu((CH2)5CN4)n](ClO4)2 + 2H3COOH
n = 2, 4
For Co(III) complex
[Co(NH3)5H2O]3+(ClO4-)3 + P {Co(NH3)5[P]}3+(ClO4-)3 + H2O
The following compounds were prepared and investigated: Cu(PMT)4(C1O4)2 (1),
Co(NH3)5(PMT)(ClO4)3 (2). The composition and structure of complexes 1 and 2 were
supported by the results of elemental analyses; IR-, UV-, and NMR(H1)-spectroscopy. The
results of thermochemical studies of complex perchlorates by means of derivatographic
method are presented in Tables 1 and 2.
Table 1. Derivatographic analysis of complex 1
Products of
thermal decomposition
[Cu((CH2)5CN4)4] (ClO4)2

T 0C

Effect

30-175

Endo

Residue, %
Found Calculated
100
-

Cu ((CH2)5CN4))3 (ClO4)

175-240

Exo

29,3

29,6

Cu((CH2)5CN4))2

240-305

Exo

29,3

29,6

CuCxNy

350

40,9

Table 2. Derivatographic analysis of complex 2


Products of thermal decomposition T 0C

144

Residue, %
Found Calcul-ated
100

[Co(NH3)5(C6H10N4)](ClO4)3

60-245
(max 243)

Co(C6H10N4)(ClO4)3

245-360
85.0
(max 275, 295, 320)

85.3

1/2Co2O3

360-500

14.3

14.0

Preparation, Properties and Application of Energetic Complex Perchlorates of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The data listed in Tables 1 and 2 shows that complexes 1 and 2 are decomposed in
several steps. Decomposition of the complexes begins when temperature rises above 200 0C.
Removal of PMT molecule is the first step of decomposition of perchlorate complex 1. For
complex 2 during the first stage the process of inactivation and removal of NH3 molecules
from inner sphere is observed. Subsequently the thermal decomposition of the tetrazole
ligand occurs with the involvement of the perchlorate anion into oxidation reactions at the
late stages of the complex destruction. According to the data of derivatographic analysis, the
complexes are thermally stable compounds.
The drop hammer test of impact sensitivity according to GOST 4545-88 demonstrated
that the sensitivity to impact of perchlorate complex 2 is similar to that of PETN. Sensitivity
to impact of copper complex 1 was at the level of tetryl. Complex 2 has a crystal density
near 1,82 g/cm3. Sensitivity to impact of complex 2 (drop hammer -44-II, mass of hammer
2 kg, H=25 cm, % of explosion) is 16% (the impact sensitivity of PETN is 12% under
similar conditions). Complex 2 has a short distance of deflagration-to-detonation transition
and can initiate charges of RDX in blasting cap No 8. Minimal charge of complex 2 for
blasting RDX is near 0.4 g under this condition [1]. Detonation velocity of complex 2 at
maximum density is near 6.98 mm*s-1. (Detonation velocity calculated by method of A.A.
Kotomin [2]). Consequently, complex 2 may serve as an initiating material with the
detonation velocity close to that of TNT and sensitivity to mechanical stimuli like other
modern high explosives. Perchlorate complex 1 does not initiate HE.

200

100
80
60

Burn rate, m m /s

40

20

10
8
6
4

1
0.1

0.2

0.4

0.6 0.8 1

8 10

20

Pressure, MPa

Fig 1.

The rate of deflagration of complex 2 as function of pressure

The burn rates of perchlorate complexes 1 and 2 in the bomb of constant pressure (BP
400) were investigated. Charges of the perchlorate complexes were pressed into
polymethylmethacrylate tubes with the internal and external diameters equal to 4 and 7 mm,
respectively. Photoregister FR-14 was used for measuring the burn rates of complex 2. Fig. 1
demonstrates the rate of deflagration of complex 2.

Mikhail Ilyushin, Igor Tselinsky, Irina Bachurina, Yurii Gruzdev, . . .

145

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The pressure exponent in the equation for the burn rate of perchlorate complex 2 ()
equals to 0,874. This value is a little lower than the pressure exponent in equation for the
burn rate of hexaammine cobalt (III) perchlorate ([Co(NH3)6](ClO4)3) (=1) and
aquapentaammine cobalt (III) perchlorate ([Co(NH3)5(H2O)](ClO4)3) (=1) which have the
structures analogous to that of complex 2 [3]. It is noteworthy that perchlorate complex 2
burns partly in the condensed phase. This effect may be the result of presence of CH2groups as a fuel in the ligand.
Complex 1 has a lower burn rate. Digital camera JVC GR-DVL 9800 was used to
measure burn rates of complex.1. Fig. 2 shows the rate of deflagration of complex 1.
100
80
60

40

Burn rate, mm/s

20

10
8
6

1
0.1

0.2

0.4

0.6 0.8 1

8 10

20

Pressure, MPa

Fig 2.

The rate of deflagration of complex 1 as function of pressure.

The lower pressure limit of self-sustaining burning of perchlorate complex 1 is 0.5 MPa.
In the interval from 0.5 MPa to 2 MPa the pressure exponent in the equation of burning ()
equals to 1.089 and shows that the process of burning takes place entirely in the gas phase.
The line illustrating the dependence of the burn rate on pressure for values greater than 2
MPa changes the slope ( =0,383) because of the change in the rate limiting stage. As the
pressure rises, the processes in the condensed phase become dominant.
Experimental values of burn rate were used to calculate the burn temperatures and
composition of burning products for complexes 1 and 2 at the studied pressures . The
operation was carried out by means of REAL computer program. The program was
developed by G.V. Belov for computer simulating of complex chemical reactions taking
place at high temperatures and pressures.
The experemental formulae of perchlorate complexes (Cu1C24H40N16Cl2O8 (1) and
Co1N9H25C6Cl3O12 (2)) and the values of enthalpies of formation (f0 = 439 kJ/kg for
complex 1, f0= - 883 kJ/kg for complex 2) were used to perform calculations. The

146

Preparation, Properties and Application of Energetic Complex Perchlorates of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

calculated burn temperatures and composition of burning products of perchlorate complexes


at pressure 10 MPa are shown in Tables 3-5.
Table 3. The calculated burn temperatures for perchlorate complexes 1 and 2
Complex
[Cu((CH2)5CN4)4] (ClO4)2
[Co(NH3)5(C6H10N4)](ClO4)3

Burn temperature, K
1451
2329

Table 4. Table 4.The calculated composition of principal burning products of


perchlorate complex 1
Burning product
C(condensed)
Cu(condensed)
HCl
N2
H2
CH4
CO

Concentration, mol/kg
18.270
1.218
2.444
9.788
17.459
2.436
8.534

Trace amounts of the following burning products: CO2 -0.184 mol/kg, H2O-0.910 mol/kg, HCN0.013 mol/kg, NH3 -0.040 mol/kg are shown to be present

Table 5. The calculated composition of principal burning products of perchlorate


complex 2
Burning product
Co(condensed)
HCl
N2
H2O
H2
CO
CO2

Concentration, mol/kg
1.468
4.684
7.749
8.957
10.206
8.964
1.369

Trace amounts of the following burning products: NH3 0.003 mol/kg, CoCl2 0.235 mol/kg are
shown to be present

The above regularities in the calculated composition of main burning products of


perchlorate complexes 1 and 2 may be the result of strongly negative oxygen balance of the
salts (the oxygen balance of perchlorate complex 1 equals to -117.77 %, the oxygen balance
of perchlorate complex 2 equals to - 33.07 %)

2.2

Perchlorate complexes as the modifiers of burning processes in


solid rocket propellants.

One of the promising ways of regulating ballistic characteristics of highly filled


energetic formulations is the addition of burn modifiers. Complex compounds of d-metals
are effective modifiers of the burning [4]. An energetic copolymer of 2-methyl-5-

Mikhail Ilyushin, Igor Tselinsky, Irina Bachurina, Yurii Gruzdev, . . .

147

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

vinyltetrazole with methacrylic acid (PVMT) (Technical Specification 38-403-208-88) was


used as a high-molecular binder for a model solid rocket propellant formulation.

-CH-CH 2| N
N

CH 3

CH 3
|
-CH 2-C|
COOH

PVMT
A liquid aliphatic azide, 1,5-diazido-3-nitro-3-azapentane (3) was used as active
plastisizer of the polymer base (Technical Specification 130176)[4,5]..
N3-(CH2)2-N-(CH2)2-N3
|
NO2
1,5-Diazido-3-nitro-3-azapentane
Plastisizer 3 has the density of to 1.344 g/cm3, and value of the enthalpy of formation
f, = 3025 kJ/kg
Model formulations had the following composition: ammonium perchlorate of bimodal
granulometric composition:grade (<50 m) - 32%; :grade A according to Technical
Specification -84-942-82 (Sspecific = 12000 cm2/g) - 32%, fuel - aluminum powder of grade
ACD - 6 (Sspecific = 6000 cm2/g) - 6%, binder and plastisizer - 28%, additions of the
complexes were introduced into the formulations in amount of 2%. Ballistic tests were made
in a constant pressure bomb. Results of ballistic tests are presented in Table 6.
Table 6. Effect of additions of perchlorate complexes 1 and 2 on the burning of a
model propellant

Index of formulation
Basic formulation
[Cu((CH2)5CN4)4))](ClO4)2
[Co(NH3)5((CH2)5CN4)](ClO4)3

Equation of deflagration
U=AP

11,1
0,71
16,1
0,56
10,5
0,60

U10,
mm/s
47,2
58,4
41,8

U4,
mm/s
26,6
35,0
24,1

U10- rate of burning at 10 MPa pressure;


U4- rate of burning at 4 MPa pressure;

The results obtained permit to make the following conclusions:


-The pressure exponent of burn rate law in the presence of additions decreases from 0,71
to 0,60-0,56 and this is an indirect confirmation of the fact that these additions affect the
burning processes in the condensed phase.
-omplex perchlorate 1 increases the burn rate of a model propellant by 35.5%.

148

Preparation, Properties and Application of Energetic Complex Perchlorates of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Therefore, the prepared energetic complexes are promising modifiers of the burn rate
and can be used in the development of solid rocket propellant formulations.

ACKNOWLEDGMENTS
This work was performed under financial support of RF Ministry of Education and
Science via scientific and technical program Development of scientific potential of higher
school, Subprogram II Applied research and developments of priority directions of
science and engineering, Project 0120.0504746 and State Contract
02.444.11.7051(RI-111.0/001/028)

REFERENCES
[1] M.A.Ilyushin, A.Yu. Zhilin, I.V.Tselinsky, Yu.A. Nikitina, A.S.Kozlov and I.V Shugalei.
Energetic ammine cobalt (III) perchlorates with cyclopentamethilentetrazole as ligand. //Russian
Journal of Applied Chemistry. 78. No.2. 195-199 (2005) (in Russian).
[2] M.A.Ilyushin, A.V.Smirnov, A.A.Kotomin and I.V Tselinsky. Coordination complexes as
inorganic explosives for initiation systems //Hanneng Cailiao=Energetic materials. 2. No 1. 1620.(1994)
[3] A.E. Fogelzang, V.Yu. Egorshev, V.P.Sinditsky,.V.V. Serushkin, A.R.Saclanty. A study of
burning behavior of chemical analogs of explosive CP// Proceedings of the 18-th International
Pyrotechnics Seminar. Chicago. Illinois. 13-15 July 1992. 1992. P. 303- 320.
[4] A.S. Dudyrev, I.V. Tselinsky and M.A. Ilyushin Chapter 6.Technology of energetic materials and
products on their base in Chemical technologies, edited by P.D. Sarkisov. Ministry of Education of
Russia. D.I.Mendeleev Russian University of Chemical Technology. Moscow. 2003. pp. 403-488. (in
Russian).
[5] M.A.Ilyushin, I.V.Tselinsky, I.V.Bachrina, L.O.Novocelova, E.N.Konyushenko, A.S.Kozlov ,
Yu.A.Gruzdev. Application of energy-saturated complex perchlorate.//International Conference
Zababakhin scientific talks. September 5-10.2005. Snezhinsk. Chelyabinsk Region. Russia.2005.
p.83-84.

Mikhail Ilyushin, Igor Tselinsky, Irina Bachurina, Yurii Gruzdev, . . .

149

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

BALISTIC STABILITY OF A SPHERICAL PROPELLANT:


COMPARISON WITH A FLATTENED SPHERICAL PROPELLANT.
L. Jeunieau, M. H. Lefebvre,
Ecole Royale Militaire, Avenue de la Renaissance 30, 1000 Bruxelles (BEL)
e-mail: laurence.jeunieau@rma.ac.be

P. Guillaume
PB Clermont SA, Rue de Clermont 176, 4480 Engis (BEL)

_____________________________________________________________________
Abstract
The physical ageing of a spherical deterred propellant has been monitored by infra-red
microscopy, closed vessel tests and ballistic firing. This result has been compared with the
ageing of a flattened spherical propellant; it has been observed that the ageing of the
spherical propellant is far less important.
Keywords: spherical propellant, flattened propellant, deterrent, ageing
_____________________________________________________________________

INTRODUCTION

Rolled ball propellants are used widely in small calibre weapons. They are easy to
manufacture and their properties can be specially adjusted for different ammunition types by
chemically modifying their surface composition. This surface modification usually consists
of an impregnation by a deterrent, which is only present along the surface up to a certain
depth and with a certain gradient. This deterrent gradient gives a nearly constant pressure
generation during the combustion process, compensating the reduction of surface by an
increase of the burning rate with time.
The disadvantage of this special design is the possible migration of the deterrent
throughout the propellant. In this case, the burning rate increases at the beginning of the
burning process, which leads to higher pressures in the weapon. This can damage the
weapon or give rise to malfunctions of the ammunition.
The service life of propellants consists of two components: the chemical shelf life and
the functional or ballistic shelf life. The chemical shelf life covers the period of time during
which the propellant can be safely stored without representing a hazard to its environment.
This one is limited by the extent of chemical ageing reactions, such as decomposition of
nitric esters and reactions of the decomposition products with the stabilizer. The functional
life or ballistic shelf life is the period of time during which the propellant or ammunition can
be used safely or during which the internal requirements remain fulfilled. In this work only
the ballistic shelf life has been investigated.
Previously, we have extensively investigated the ageing of a flattened spherical
propellant [1-4]. In this work, the ageing of a spherical propellant has been studied. To ease
the comparison between the results, the same conditions of ageing have been used at 80C,
70C and 65C.

150

Balistic Stability of a Spherical Propellant: Comparison with a Flattened . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

EXPERIMENTAL

2.1 Samples
The propellant, which is used for the investigation is the double base ball powder
propellant SPH27. It is a spherical propellant with an average diameter of 756 m. It
contains around 10% of nitroglycerin. It is stabilized with DPA. The deterrent applied on
the surface is dibutylphthalate (DBP). The average concentration of DBP is 7.4%.
The flattened propellant K5810 which has been previously studied can be
geometrically modelled by an ellipsoid with a great axis of 660 m and two small axis of
360m. It contains 10% of nitroglycerin and 4.6% of DBP.
The two types of propellant differ by their geometry and their DBP concentration.
2.2 Experiments and work program
Three different areas of investigation are covered: migration analysis, closed vessel
tests and ballistic firing.
Infrared microscopy
Migration of the deterrent was made visible by IR microscopy: the propellant grain is
placed into an adhesive, cut by a microtome into small slices (thickness 7m) and analyzed
by IR microscopy. A Bruker Hyperion infrared microscope mounted on a Vector 33 Fourier
transform spectrometer is used in this study. A medium-band MCT (HgCdTe) detector in
the microscope gives high sensitivity in the 4000-600cm-1 range. A 15X cassegrain mirror
objective is used to obtain the infrared spectra. The IR spectrometer is operated at a
resolution of 4cm-1 and 32 scans are acquired for each measurement posision. The
measuring window has a aperture of 10m x 50m, the great side of the aperture is placed
perpendicularly to the measured diameter. The step for each data point is 3m. The DBP
concentration is obtained using the ratio of two IR bands, one typical of the DBP (1720cm-1),
one typical of the nitrocellulose (1160cm-1).
Closed vessel tests
Closed vessel tests experiments were carried out in a vessel of 132cm3 using a
piezoelectric transducer (Kistler 6215) coupled to a charge amplifier (Kistler 5011A) for
recording the pressure. The output voltage of the pressure gauge was transferred to a data
recording acquisition system (Nicolet Multipro, resolution 12 bit, sampling frequency 250
KHz). The ignition system consists of two electrodes, which are connected with a nickeline
hot wire. The vessel is equipped with a valve to introduce the gaseous ignition mixture. The
partial pressures of this mixture are measured with a piezoelectric transducer (Kistler 4070).
The composition of the used ignition mixture is 1bar CH4 1.4 bar O2. For the actual
composition the initial atmospheric pressure must be added. The used loading density is
0.15g/cm3.

Laurence Jeunieau, Michel Lefebvre, Pierre Guillaume

151

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Ballistic firing
For the spherical propellant, the experimental work is carried out in a .50inch weapon
instrumented with a piezoelectric transducer (Kistler 6215) coupled to a charge amplifier
(Kistler 5011A) in order to record the pressure. The pressure measurement is located in the
cartridge case. The output charge of the pressure is amplified and filtered to get an output
voltage which is transferred to a data acquisition system (resolution 12bit, sampling rate
1MHz). 14g of the propellant are poured in a standard 12.7 x 99 mm cartridge. The
corresponding loading density is 0.79g/cm3.
The ballistic firings for the flattened propellant have been performed in a 5.56
calibre, for more details see [3].
Propellant ageing
The ageing times at 80C are nearly identical for the flattened and the spherical
propellant.
The ageing times of the spherical propellant are presented in Table 1. Throughout
this paper S stands for the spherical propellant ageing. The two numbers behind the first
letter describe the ageing temperature whereas the last letter describes the ageing state (see
table 1). By example, S80B corresponds to a propellant aged at 80C during 1.81 days. For
the ballistic firing, the aged propellant is poured into the cartridge.
The ageing times for the flattened propellant are presented in table 2. For these
propellant K stands for the flattened propellant ageing.

T
[C]
80
70
65

Table 1: Ageing plan of the spherical propellant.


C
D
E
F
G
H

1.20

1.81

2.28
9.82

5.53 7.22
21.74

9.02

10.91 20

J
15.73

29.8

T
[C]
80
70
65

Table 2: Ageing plan of the flattened propellant.


B
C
D
E
F
G

1.20
4.83

1.81
7.25

2.28
9.66

3.61 5.41 7.22 9.02 10.83


14.50 21.74 29.00 36.23 43.48
30.00

3. RESULTS
3.1 Deterrent migration
Figure 1 shows some deterrent concentration profiles of the spherical propellant aged
at 80C. It can be observed that the deterrent concentration decreases in surface with ageing
and increases in the centre part of the propellant with ageing.

152

Balistic Stability of a Spherical Propellant: Comparison with a Flattened . . .

New Trends in Research of Energetic Materials

Czech Republic

[Content]

w ithout ageing
S80B
S80C
S80E
S80H
S80J

14
12
10

%DBP

2006

8
6
4
2
0
0

50

100

150

Figure 1: Deterrent concentration profiles of DBP in the spherical propellant aged at 80C.
Figure 2 shows the deterrent concentrations profiles that have been previously
obtained for the flattened spherical propellant for the same ageing time. It can be observed
that the variation of the deterrent concentration profiles is far more important for the
flattened spherical propellant than for the spherical propellant.

14
w ithout ageing
K80B
K80C
K80E
K80H

12

%DBP

10
8
6
4
2
0
0

50

100

150

Figure 2 : Deterrent concentration profiles of DBP in the flattened spherical deterred


propellant.
3.2 Closed vessel tests
The two combustion phases resulting for the chemical composition gradient of a
deterred propellant can be observed in closed vessel tests if an appropriate gaseous ignition
system is used. The influence of the ignition system on the obtained combustion rate has
been extensively studied elsewhere [5-6]. Figure 3 shows the obtained combustion rate of
the spherical propellant before ageing and the corresponding derivative of the pressure
versus time. The pressure derivative permits to distinguish clearly two phases in the
combustion process, as two different slopes can be observed during the combustion process.
4

Laurence Jeunieau, Michel Lefebvre, Pierre Guillaume

153

Czech Republic

20

2006

[Content]

100
dP/dt (Mpa/ms)

combustion rate (cm/s)

New Trends in Research of Energetic Materials

15
10
5

80
60
40
20
0

0
0

50
100
Pressure (MPa)

150

-5

0
time (ms)

Figure 3: Combustion rate of the spherical propellant without ageing and the corresponding
derivative of the pressure versus time
Figure 4 shows the obtained combustion rate of some of the samples aged at 80C. It
can be observed a decrease of the combustion rate for the high pressure value. In fact, an
increase of the combustion rate pressure for the low pressure value is also present but is not
visible in this figure.
For this reason, figure 5 shows the variation of the combustion rate at 20MPa and at
100MPa as a function of ageing time.
These variations can be related to the propellant migration: the deterrent concentration
decreases near the surface which is correlated with the increase of the combustion rate at the
low pressure values and the concentration increases in the centre part of the propellant;
which is correlated with the decrease of the combustion rate at the high pressure values.
20

Combustion rate (cm/s)

18
16
14
12
10

without ageing

8
6

S80J

S80B, S80C,
S80E, S80H

2
0
0

20

40

60
80
Pressure (MPa)

100

120

140

Figure 4: Combustion rate of the spherical propellant aged at 80C for different ageing time.

154

Balistic Stability of a Spherical Propellant: Comparison with a Flattened . . .

2.2

flattened
propellant

spherical
propellant

1.8
1.6
1.4
1.2
1
0

10

20

Czech Republic
Combustion rate @ 100MPa (cm/s)

Combustion rate @ 20 MPa (cm/s).

New Trends in Research of Energetic Materials

2006

[Content]
flattened
propellant

12.5
12
11.5
11
10.5
10
9.5
9
8.5
8

spherical
propellant

ageing time @ 80C (day)

10

20

ageing time @ 80C (days)

Figure 5: Variation of the combustion rate at 20MPa and at 100MPas a function of ageing
time at 80C for the spherical and the flattened propellant.
Figure 5 shows the variation of the combustion rate at 20MPa and at 100MPa as a
function of ageing time for the spherical and the flattened propellant. It can be observed that
the variations are greater for the flattened propellant than for the spherical one. This can be
correlated to the more important deterrent migration in the flattened propellant. As the
variation of the deterrent concentration is greater with ageing, the variation of the
combustion rate is also more important.
3.3 Ballistic firing
Ballistic experiments were performed with aged and unaged cartridge. The pressure in
the cartridge and the velocity of the bullet have been determined. No variation of the bullet
velocity at the muzzle position have been observed. This shows that the ageing time has not
modified significantly the energy content of the propellant.
Figure 6 shows the pressure recorded in the cartridge for a double base propellant and a
deterred propellant at a same loading density. When the propellant begins to burn, the
pressure increases. When a threshold pressure is reach, the force applied on the projectile is
enough to engrave the rotating band. The movement of the projectile down the gun bore
causes the volume of the reaction chamber to increase. As the propellant continue to burn,
the pressure still increases, but this increase is slower as the volume increases. After a
certain time the increase of the volume has a more important effect than the increase of the
gas mol number and the pressure decreases. It can be observed that for a deterred propellant
the pressure increase is more progressive and the pressure stay at its maximal value for a
greater time than for a not deterred propellant. This is due to the presence of the deterrent in
surface. At the beginning of the combustion, the deterred part of the propellant burnt, the
burning rate is low but the propellant surface is high, when the deterred part of the propellant
is burnt the combustion rate is higher but the propellant surface is smaller and the gas
production stay more or less constant. For a not deterred propellant the combustion rate is
high at the beginning of the combustion and as the propellant surface is important, the
pressure increase is rapid and by consequence the maximal pressure in the cartridge is high.

Laurence Jeunieau, Michel Lefebvre, Pierre Guillaume

155

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

350

double base propellant

Pressure (MPa)

300

deterred
propellant

250
200
150
100
50
0
0

time (ms)

Figure 6: Variation of the pressure in the cartridge as a function of time for a double base
propellant and a deterred propellant in a .50inch weapon for a loading density of 0.6g/cm3.
Figure 7 shows the pressure recorded in the cartridge for different aged propellants. It
can be observed that the maximal pressure increases with ageing and that the pressure peak
for an aged propellant has a shape which resembles more to the one of a not deterred
propellant. Due to deterrent migration, for an aged propellant the transition between the
deterred and the not deterred part of the propellant is smoother. By consequence, the
pressure increases is more rapid at the beginning of the combustion and the maximal
pressure is higher.
300

Pressure (MPa)

250

without ageing

200
150

S80E and
S80H

100

S80B
S80C

50
0
2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

tim e (m s)

Figure 7: Variation of the pressure in the cartridge for the spherical propellant at different
ageing time.
The same type of variation has been obtained for the flattened propellant (fig.8) but
in this case the pressure increase is more important for the same ageing time. This greater
variation is emphasized by fig.9 which shows the variation of the maximal pressure as a
function of ageing time for the two types of propellant.

156

Balistic Stability of a Spherical Propellant: Comparison with a Flattened . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

300

Pressure (MPa)

250

without ageing

200
150

S80E and
S80H

100

S80B
S80C

50
0
2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

tim e (m s)

Maximal pressure in the cartridge


(MPa)

Figure 8: Variation of the pressure in the cartridge for the flattened spherical propellant at
different ageing time.
flattened propellant

500

spherical propellant

450
400
350
300
250
200
150
0

10
15
20
ageint time @ 80C (day)

25

Figure 9: Variation of the maximal pressure in the cartridge as a function of ageing time at
80C for a flattened and a spherical propellant.

3.4 Correlation between the different experiments


3.4.1 Correlation between infrared measurement and closed vessel tests
The combustion rate at 20MPa corresponds to the beginning of the combustion and
can be related to the deterrent concentration in surface. In fig.10, the combustion rate at
20MPa has been plotted as a function of the deterrent concentration in surface for the
flattened and the spherical propellant. It can be observed that the combustion rate decreases
with the deterrent concentration and that a single linear regression can be drawn for the
flattened and the spherical propellant. This shows that the combustion rate at 20MPa
depends only of the deterrent concentration in surface and that the geometrical difference
between the two propellants is appropriately treated in the combustion rate determination.

Laurence Jeunieau, Michel Lefebvre, Pierre Guillaume

157

combustion rate @ 20MPa

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

2.5
2
1.5
1

flattened propellant

0.5

spherical propellant

0
0

10
%DBP @ 0m

15

Figure 10: Variation of the combustion rate at 20MPa as a function of the DBP
concentration in surface for the flattened and the spherical propellant.
3.4.2 Correlation between infrared microscopy and ballistic firing
Figure 11 shows the variation of the maximal pressure measured in the cartridge as a
function of the DBP concentration at the outer surface of the propellant grain for the
flattened and spherical propellant. It can be observed that the maximal pressure in the
cartridge decreases with the DBP concentration in surface. This confirms that the variation
of the maximal pressure in the cartridge depends on the deterrent migration.

Maximal pressure in the


cartridge (MPa)

The apparent difference between the two types of propellant is due to the different
uses: the flattened propellant is used in a 5.56 mm cartridge and the spherical propellant is
for a .50inch cartridge.
500
450
400
350
300

flattened propellant
spherical propellant

250
200
0

10
%DBP @ 0m

15

Figure 11: Variation of the maximal pressure in the cartridge as a function of the DBP
concentration in surface for the flattened and spherical propellant.
4

CONCLUSIONS

The ageing of a spherical deterred propellant has been monitored by infrared


microscopy, closed vessel tests and ballistic firing. These results have been compared to the
one obtained for a flattened spherical propellant. It has been observed that the effect of
ageing is more important for the flattened propellant than for the spherical. This can be due
to the effect of stress during the flattening operation. Some crack may appear in the
propellant during this operation. A microscopic observation should confirm or infirm this
hypothesis.
A good correlation has been observed between the combustion rate at 20MPa and the
deterrent concentration in surface and between the maximal pressure in the cartridge and the
deterrent concentration in surface.
158

Balistic Stability of a Spherical Propellant: Comparison with a Flattened . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

BIBLIOGRAPHY
[1] S.Wilker, P. Guillaume, M.H. Lefebvre, S. Chevalier, L. Jeunieau, Ballistical and
chemical stability of rolled ball propellants, 7 Seminar New Trends in Research of
Energetic Materials , Pardubice, April 2004.
[2] S. Wilker, P. Guillaume, M.H. Lefebvre, S. Chevalier, L. Jeunieau, G. Pantel, U.
Ticmanis, T. Stottmeister, Stability analyses of rolled ball propellants Part I :
Microcalorimetric studies and stabilizer deplation, 34th International Annual Conference of
ICT, Karlsruhe, June, 2003.
[3] L. Jeunieau, M.H. Lefebvre, P. Guillaume, S. Wilker, S. Chevalier, S. Eibl, Stability
analyses of rolled ball propellants. Part II- Ballistic stability, 35rd International Annual
Conference of ICT, Karlsruhe, June, 2004.
[4] L. Jeunieau, M.H. Lefebvre, P. Guillaume, S. Wilker, S. Chevalier, Stabillity analyses of
rolled ball propellants Part III- Correlation between closed vessel tests, ballistic firing and
deterrent migration, 36th International Annual Conference of ICT & 32rd International
Pyrotechnics Seminar, June, 2005
[5] L. Jeunieau, M.H. Lefebvre, A. Papy, M.C. Pirlot, C. Reynaud, Closed vessel test:
influence of the ignition method on the combustion rate, 33rd International Annual
Conference of ICT, Karlsruhe, June 2002.
[6] L. Jeunieau, M.H. Lefebvre, M.C. Pirlot, P. Guillaume, Characterization of deterred
propellants by closed vessel tests: importance of the ignition method, 8th Seminar New
Trend in Research of Energetic Materials, Pardubice, April 2005.

10

Laurence Jeunieau, Michel Lefebvre, Pierre Guillaume

159

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

SYNTHESIS AND CHARACTERIZATION OF


1,2-DINITROGUANIDINE (DNG) AND ITS DERIVATIVES
Nikolaj V. Latypov, Martin Johansson, Larisa N. Yudina-Wahlstrm, Stefan Ek,
Carina Eldster and Patrick Goede
FOI, Swedish Defence Research Agency, Department of Energetic Materials,
S-147 25 Tumba, Sweden

Abstract:
1,2-Dinitroguanidine (DNG), (1), is a novel explosive with high performance and
moderate sensitivity1,2. Synthesis of DNG and its salts has recently been reported3. X-ray
studies of the crystal structure of DNG can also be found in the literature4. In this paper
we present complementary studies on the synthesis, reactivity and thermal
characteristics of this compound and some of its derivatives.
Keywords:

1.

1,2-Dinitroguanidine, DNG, synthesis, characterization

INTRODUCTION

N,N-Dinitroguanidines have been known compounds for a rather long time5. Despite the
extensive studies6,7, which resulted in the synthesis of several interesting substances in this
series, the un-substituted 1,2-dinitroguanidine itself had never been reported until recently
when different groups of Russian researchers presented simultaneously both the synthetic
approach3 and basic properties of the compound3,4.
Judging by the molecular composition of DNG, which is an oxidizer (+5,4%) one can
treat the substance and its derivatives as attractive ingredients in both explosive and
propellant formulations, - surprisingly, very little information about their properties as
energetic materials has so far been published1,2.
For this reason we have performed a preliminary estimate of the most important
properties of these compounds.

160

Synthesis and Characterization of 1,2-Dinitroguanidine (DNG) and its . . .

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

RESULTS AND DISCUSSION


2.1 Synthesis
1,2-dinitroguanidine (DNG) was easily synthesised by the reported method3 with
some minor modifications: higher content of nitric acid in the nitration mixture, in
comparison with the published procedure, was used to ensure a controllable reaction
course. In addition to the observations made by Astratev et al.3, showing that the
nitration of mono-nitro guanidine (MNG) to DNG is a slow process taking place in a
very reactive nitrating media (figure1), we

Fig 1.
UVspectra of the reaction mixtures for the nitration of MNG to DNG;
nitrating mixture HNO3: oleum (65% SO3) = 0, 55: 0, 45.

7
6

Absorbency

5
4
3
2
1
0
-1
150

200
1__after 2,5 hrs
1__after 6 hrs
1__after 20 hrs

250

300

350

400

450

Wavelength (nm)

have also noticed the instability of DNG in such mixtures so that its concentration is
strongly dependent on the reaction time (figure 2) and the composition of the nitrating
media. The yield of DNG has never exceeded 80%.

Nikolaj Latypov, Martin Johansson, Larisa Yudina-Wahlstrm, Stefan Ek, . . .

161

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 2.
Formation of DNG by nitration of MNG in different nitration mixtures;
dependence of the yield on the reaction time.
%
80

1
2

60

1__ HNO3 : oleum (65% SO3) = 1,0:1,5


2__ HNO3 : oleum (65% SO3) = 1,0:0,9
3__ HNO3 : oleum (30% SO3) = 1,0:1,0

40

20

12

16

20

24

72

Hrs

For this reason, the reaction course was in all cases monitored by UV spectroscopy to
detect the full conversion of MNG to DNG and the maximum concentration of the
latter. Despite these efforts, the product separated by dilution of the nitration
mixtures and extraction by appropriate solvent always contained small amounts (25%) of MNG.
All these findings can be easily understood in view of the known fact that N-nitration
is a reversible process and nitramines are prone to irreversible decomposition in
strongly acidic media8 (Scheme1).
Scheme 1.

Reaction sequence in the nitration of MNG to DNG.

NH
H2N

NO2+, -H+

NH2

-NO2+, H+

H2N

NO2

NO2+, -H+

NH2

-NO2+, H+

H+

O2N

N
H

NO2
NH2

H+

N2O + (NH2)2CO + H2O

2.2

Physicochemical properties.

Samples of DNG produced by this method needed further purification which was
achieved by dissolution of the raw product in diisopropylether, followed by removal of the
insoluble MNG by filtration. The purified product was identified by elemental analysis and
mass-spectroscopy; its spectral characteristics (UV, NMR) were identical with those
published in the literature3.

162

Synthesis and Characterization of 1,2-Dinitroguanidine (DNG) and its . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

DNG is indeed a very acidic product: our measurements resulted in the pKa value of
0.95, which agrees reasonably good with the figure (1.11), reported by Astratev et al.3 The
density of crystalline DNG (1.895), estimated by pycnometry in the present work, is also
rather close to that (1.884) obtained from X-ray measurements4. Salts of DNG were easily
obtained by careful neutralisation of its solutions in an alcohol with an appropriate base. The
properties of DNG and some of its salts are summarized in Table1.

Table 1. Properties of DNG and its salts.


Density
3

g/cm
DNG

H f,

Sensitivity

Sensitivity

kcal/mol

Friction, N

Impact, cm/J

90

22 / 4,5

1.906*/

0,0 ***

1.884**

ADNG

1,86*

- 36,60 *

250

78 / 14,2

K salt

n.m.

n.m.

130

28 / 7,3

ADN

1,805

-34,5

153

12 / 3,0

RDX

1,802

14,7

195

32 / 6,4

- present work

** - reference [4]
*** - reference [2]
Unlike alkylation and substitution reactions3, the reactivity of DNG, or , strictly
speaking, its anion in reactions of the Henry and Michael type is quite low (cf. Table 2); this
is in line with the behaviour of some anions of primary nitramines, which undergo such
reactions only with difficulty9,10.
Table 2. Attempted reactions of DNG.
Michael donor Michael acceptor
DNG, NaDNG
Methyl acrylate
DNG, NaDNG
Acrylic acid
DNG
Methyl vinyl ketone
NaDNG
Methyl vinyl ketone
DNG
Acrylonitrile
NaDNG
Acrylonitrile

Temperature
RT, reflux
RT, Reflux
RT
RT
RT
RT

Result
Denitration
Denitration
Denitration
Addition
Denitration
Product mixture

Several N-alkyl-N, N-dinitroguanidines has been also synthesized by nitration6,7of


corresponding mono-nitro-compounds, which in turn were obtained by the known method 11.
Of these, iso -butyl-DNG showed interesting properties as a potential plasticizer. (fig.3)

Nikolaj Latypov, Martin Johansson, Larisa Yudina-Wahlstrm, Stefan Ek, . . .

163

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 3. DSC thermogram showing the low temperature properties of i-butyl-DNG.


^exo
!&SEK_iBuDN G
SEK_iBuDNG, 20,0000 mg

Step
Onset
Inflect. Pt.
Ends et
Midpoint

-140

-120

-0,21 Wg ^- 1
-4,24 mW
-130,29 C
-125,71 C
-123,56 C
-127,03 C

-100

FOI: Vapen och Skydd.

2.3

0,5
Wg^-1

-80

-60

-40

-20

20 C

METTLER TOLEDO STARe SW 7. 01

Thermal stability

According to DSC measurements 1, 2-DNG has a rather good thermal stability and
decomposes without melting. Similar behaviour was observed for salts of DNG, though their
thermal stability varies significantly (cf. Figures 4-6).
Fig 4.

164

DSC thermogram of DNG and ADNG.

Synthesis and Characterization of 1,2-Dinitroguanidine (DNG) and its . . .

New Trends in Research of Energetic Materials

Czech Republic

Fig 5.

DSC thermogram of HDNG and LiDNG.

Fig 6.

DSC thermogram of NaDNG and KDNG.

2006

[Content]

These promising results prompted us to further investigate the long term stability of the
products. This was performed according to standard method (STANAG4582). As it can be
seen from fig. 7-8 DNG itself has excellent thermal stability. On the contrary, salts of DNG
(Fig.8-9) and iso -butyl-DNG (Fig.10) did not pass the test: in all cases violent, and, most
probably, autocatalytic decomposition was observed.

Nikolaj Latypov, Martin Johansson, Larisa Yudina-Wahlstrm, Stefan Ek, . . .

165

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 7.
Heat flow produced by DNG in isothermal heat flow calorimetry
measurements at 75 C for 19 days. The signal is jagged due to large
amplification.

Fig 8. Heat flow evolved during isothermal heat flow calorimetry measurements on
different samples of ADNG at 75 C for 19 days.

166

Synthesis and Characterization of 1,2-Dinitroguanidine (DNG) and its . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 9.
Heat flow evolved during isothermal heat flow calorimetry measurements on
DNG and salts at 75 C for 19 days.

Fig 10. Heat flow evolved during isothermal heat flow calorimetry measurements on
iso-butyl-DNG at 75 C for 19 days.

These dissapointing properties , such as enormously high acidity of DNG itself and
unacceptable thermal stability of its derivatives made us to suspend further work on this
interesting group of compounds.

Nikolaj Latypov, Martin Johansson, Larisa Yudina-Wahlstrm, Stefan Ek, . . .

167

New Trends in Research of Energetic Materials

3.

Czech Republic

2006

[Content]

CONCLUSIONS

1,2-dinitroguanidine is an interesting energetic material; unfortunately, any


formulation work with this highly acidic compounds does not seem feasible. On the
contrary, this namely property favours synthesis of its derivatives with variable
characteristics.
Several compounds of this series, - salts of DNG and also iso-butyl-DNG- synthesized and
characterized in the present work, are not stable enough to find some application.
Further work in this area in search of more stable derivatives of DNG might be interesting.
Acknowledgements
The authors would like to thank the Swedish Defence Forces for financial support of this
research.

4.
[1]
[2]
[3]

[4]

[5]
[6]
[7]
[8]
[9]
[10]
[11]

168

REFERENCES
V.I. Pepekin, L.M. Kostikova, G.T.Afanasev, Explosive properties of 1,2dinitroguanidine, 35th International Annual Conference of ICT, June29-July2, 2004.
V.I. Pepekin, G.T. Afanasev, Explosive properties of 1,2-dinitroguanidine and its
compositions with graphite Chimitcheskaya fyzicka, v.24, N4, p.p. 66-68, 2005(in Russian).
A. A. Astratev, D.V. Dashko, L.L. Kuznetsov, Synthesis and some properties of 1, 2dinitroguanidine, Russian Journal of Organic Chemistry, 2003, v.39, N4, p.p.537-548. (in
Russian ).
A. D. Vasilev, A.M. Astachov, M.S. Molokeev, L.A. Kruglyakova and R.S. Stepanov, 1, 2dinitroguanidine, Acta Crystallographica Section C, Crystal Structure Communications,
2003, C59, o550-o552
A.F.McKay Nitroguanidines, Chemical Reviews, 1952, 51, 301-346.
A.F.McKay and G.F.Wright, The nitration products of 2-nitramino-2-1,3diazacycloalkenes, JACS, 1948, v.70, p.p. 3990-3994.
R.H.Meen and G.F.Wright, The Scission of New 1-alkyl-1,3-dinitroguanidines and an
analog, JACS, 1952, v.74, p.p.2077-2079.
K.Schofield, Aromatic nitration, Cambridge University Press, 1980, part 15,N-Nitration
and Nitramines rearrangement and references cited therein.
L.W.Kissinger and M.Schwartz, Some Michael-Like Additions of Primary Nitramines,
JOC , 1958, v.23, p.p. 1342-1344.
David Woodcock, The reactions of Nitramines with Formaldehyde, J.Chem.Soc., 1949,
p.1635-1637.
Lawrence Fishbein and John A.Gallaghan The preparation and reactions of 2-Alkyl-1(or3)nitro-2-thiopseudourea, JACS, 1954, v.76, p.p.1877-1879.

Synthesis and Characterization of 1,2-Dinitroguanidine (DNG) and its . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

ENERGETIC CHARACTERISTICS OF SOLID COMPOSITE


PROPELLANTS AND WAYS FOR ENERGY INCREASING
Lempert D.B., Nechiporenko G.N., Manelis G.B.
Institute of Problems of Chemical Physics of Russian Academy of Sciences, (IPCP RAS),
Chernogolovka, Semenov av.1. Moscow reg. 142432. Russia
lempert@icp.ac.ru
Abstract: The creation of solid composite propellants is one of the most important
applications of high energetic materials. If in the very beginning of the development of
this branch (the middle of the 20th century) achieved energetic level increased
considerably every decade, up to 70-80 years it seamed that energy had already come to
maximal possible value, that is the power contained in chemical compounds (in
individual substances as well as in their mixtures) was already limited. Detailed
investigations, that have been carrying out in IPCP RAS, showed that there is a definite
potential for further energy increase though this potential is rather limited. The growth
of energy capacity of chemical substances is attended with the degradation of many
operating characteristics thermal stability decreases, impact sensitivity and friction
sensitivity increase, compability of compounds degrades, probability to transfer the
combustion process into explosion decreases, the cost increases drastically etc.
However, the increasing of propellants power is necessary therefore we have to agree to
a compromise energy growth against definite degradation of other performances.
Hereat investigations of thermal stability began to develop strongly, ways to increase
thermal stability of many substances (which could not be used in propellants earlier)
have been found, many other branches (polymer chemistry for new binders creation,
chemistry of inorganic and organic oxidizers, studying of combustion low etc) began to
develop.
In the presentation energetic characteristics of solid composite propellants of different
type and application are described metal-free compositions containing an oxidizer
and a fuel-binder only; compositions containing additionally an energetic compound
(Al, Be, Mg, B), metal hydrides (AlH3, BeH2, BxNyHz). A considerable attention is
attracted to analysis of oxidizers (ammonium perchlorate, hydroxylammonium
perchlorate, ammonium dinitramide etc) because oxidizer occupies the greatest mass
fraction in propellant and mostly that is the oxidizer that determines energetic
characteristics of the propellant. Solid composite propellant contains always a fuelbinder (for achievement of necessary physico-mechanical characteristics), therefore a
considerable attention is attracted to principles of the using of different kinds of binder.
It was shown that the competence to choose the binder (even among the set of accessible
ones) to given oxidizer can increase considerably the energetic characteristics even
without creation of new compounds. The ways to optimize solid composite propellants
destined for missile complexes of different mass characteristics are described in the
presentations. Different ways to further development of propellants energetic
characteristics are presented optimization of formulations basing on accessible
components; creation of new more effective binders, that have additional functions of
oxidizer or gasifying component etc. Ecological problems of solid composite propellants
are also a topic of the presentation, i.g. perchlorates outthrow a great amount of HCl,
that consumes greenery at large radius from the launch location. Berillium-containg
propellants are extremely dangerous for humanity, fluorine-containing propellants are
dangerous too. Ways to decrease ecological danger while solid composite propellants
using are discussed in the presentation.

David Lempert, G. Nechiporenko, G. Manelis

169

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The creation of solid composite propellants (SCP) is one of the most important
applications of high energetic materials. In the very beginning (the middle of the 20th
century) of the SCP development the achieved energetic level has been increasing
considerably every decade up to 70-80 years; however later it seamed that energy had
already come to the maximal possible value, that is the power contained in chemical
compounds (in individual substances as well as in their mixtures) was already limited. The
fact is that the growth of energy capacity of chemical material is attended with the
degradation of many operating characteristics thermal stability degrades, impact sensitivity
and friction sensitivity increase, compability of compounds degrades, probability of
combustion becoming a detonation decreases, the cost increases drastically etc. So, when we
try to increase energetic characteristics we must be ready to a compromise - energy growth
against definite degradation of other performances.
Rocket flight range and mass of the useful load depend mainly on specific impulse (Isp)
value (though many other parameters are influence too), and this dependence is very
considerable. For example, when the specific Isp is increased from 255 to 260, or 2%, useful
load is increased by 10% at the same flight range. Which is a significant achievement both
for military and non-military rockets.
Specific impulse, together with the content of combustion products is easily calculated
nowadays with computers if the brut-formula and enthalpy value of propellant, and the
pressure in the combustion chamber and nozzle section are known. When we compare
different types of propellant, we do not compare experimental parameters, but the theoretical
(i.e. calculated) ones, because experimental determination is very expensive (the sample
mass must be at least several kg) and even impossible when there is a lack of any compound.
The calculation of Isp is a regular procedure using current computers and necessary
software. It was shown that the difference between the calculated and real Isp values is 4-6
% and this difference depends much more on the type of rocket engine than on the type of
propellant.

BASIC CHEMICAL CONTENT AND BASIC COMPONENTS OF SCP.


The major chemical elements from which one can create a real SCP are H, O, F, C, N,
and Cl. The most important for thrust creation are H, O and F atoms. The value of specific
impulse definitely increases as the enthalpy of the composition or pressure in the combustion
chamber increases. The specific impulse usually also increases if the percentage of H, O or F
increases. However in real SCP one cant exclude C and N, because without them its
impossible to create molecules in solid state. It is very useful to take notice that a simple
presence of high energetic components in composition does not guarantee a high value of
Isp, the optimal ratio among components is always necessary.
Oxidizer is one of the most important components of SCP because it takes the greatest
mass fraction in propellant and mostly that is the oxidizer that determines energetic
characteristics of the propellant. Solid composite propellant contains always a fuel-binder
(for achievement of necessary physico-mechanical characteristics). This binder often plays
function of combustible when the heat of its oxidation by the oxidizer is the main source of
energy. The introduction of metal (Be, Al, B, sometimes Mg) into composition is one of
ways to increase the Isp value because of high combustion heat of these metals. There is
another kind of component, which properly speaking can be named neither oxidizer nor fuel
because these components play the both functions actually they may be named gas
170

Energetic Characteristics of Solid Composite Propellants and Ways for Energy . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

generating components. Cyclotetramethylene tetranitroamine (octogen or HMX) is the most


well-known and most used substance of this type.

REQUIREMENTS TO SOLID COMPOSITE PROPELLANTS, A


COMPROMISE AMONG BASIC PROPERTIES.
While in liquid-propellant rocket engines all components are kept in tanks and are
injected with pumps into the combustion chamber, solid-propellant rocket engines look
simpler - the propellant fills up almost entirely the whole interior of the rocket body and
burns from the surface. But this simplicity requires many specific properties for the
propellant. Since the propellant is subjected to much tension during acceleration - it has to be
both strong and elastic, it has to burn uniformly with a given rate, the combustion rate has
not to depend considerably on the pressure (in the formula U=A*P, where U is combustion
rate, P pressure, the exponent usually must be lower than 0.6), otherwise during the
combustion a sharp increase in pressure is possible and consequently the destruction of the
rocket body. It is necessary to prevent the possibility of combustion becoming a detonation.
Other restrictions include: low hygroscopicity, enough thermal stability, rather low impact
and friction sensitivity, not too high temperature in the combustion chamber (<3800-3900K)
and a lot of other restrictions. All these restrictions hamper developers of new more powerful
compositions. For example, a substitution of any hydrazonium salt (as oxidizer) for its
ammonium analogue would definitely increase the Isp value, however it would significantly
degrade thermal stability, impact sensitivity and compatibility of components. Introducing of
the most effective elements (beryllium, active fluorine) into the propellant formulation
damages the environment. Increasing of the enthalpy of formation increases the Isp
(Isp/Hf 0.06 s/(kcal/kg)), but the thermal stability usually degrades etc.

DENSITY AND BALLISTIC EFFECTIVENESS.


The density is one of the important properties determining ballistic effectiveness of
propellant. It happens often that density is in a compromise with specific impulse. As draft
estimation one uses some special functions for comparing ballistic characteristics of
propellants with different Isp and density values. One of these functions is a so-called
ballistic effectiveness (Efn) of the propellant on the nth stage of the multistage rocket
complex. Efn can be written as Efn =Isp*da, where d is the density, the value of the exponent
a depends on the mass characteristics of the rocket complex and the definite stage number
n. Usually for three-staged rocket complexes the values a are the following: ~0.6 for
the first stage, ~0.4 for the second, and ~0.2 for the third one. Thus, on the upper stages the
density has less influence on the ballistic effectiveness. For example if we have five different
compositions, where the first has Isp=260 with d=1.95; the second Isp=270 with d=1.8; the
third Isp=275 with d=1.7; the forth Isp=295 with d=1.0, and the fifth Isp=230 with
d=2.3, the best composition for the first stage is #1; for the second stage - #2; for the third
stage -#3. Though the composition #4 is of the highest Isp value its low density makes it less
effective than ## 1-3 on all stages. Thus, developing ballistic effectiveness of SCP, we have
to find on what stage of the rocket complex the SCP under investigation can be used with the
most effectiveness.

David Lempert, G. Nechiporenko, G. Manelis

171

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THE SIMPLEST SCP BINDER+OXIDIZER


The set of oxidizers is not too large, we can divide them conditionally on three basic
categories: a) oxidizers free of halogen; b) perchlorates; and c)oxidizers containing active
fluorine.
Oxidizers of type a are ecological safe, of type c are very dangerous ecologically,
oxidizers of the type b occupy the intermediate position they are wide used though their
combustion product (hydrogen chlorine) damages environment.
a) SCP basing on halogen-free oxidizers.
The cheapest oxidizer of this type is ammonium nitrate (AN). It has the best element
content, however its too low enthalpy of formation (-1080 kcal/kg) makes it low effective as
oxidizers for SCP. In table 1 several oxidizers are represented for give an impression about
Isp values and difference among Isp values of different compositions. All Isp values
represented in this paper are calculated in condition when the combustion chamber pressure
is 40 atm and nozzle section pressure is 1 atm. Dimension of a Isp value is (kg-force)(kgmass)-1s (that is Isp is the resultant thrust by propellant burning with the rate 1 kg/s), or
simply s. All compositions represented in this paper contain at least oxidizer and a binder
with the optimal content of the binder, but in condition that binder percentage is not lower
than 18-20 volume % (that is about 8-10 mass% for compositions with total density ~1.8 and
binder density ~0.9) otherwise its impossible to produce a composite with satisfactory
physico-mechanical properties. Actually in binary compositions (oxidizer+binder) a binder
is necessary not only as a binder itself, but as a combustible too. In binary compositions
simple hydrocarbon binders may be rather optimal.
Table 1.

Energetic characteristics of SCP basing on the most important NOHoxidizers.

NH4NO3

Hf,
kcal/kg
-1080

d,
Additional info
g/cm3
1,725 M.p.
160C,
stable.

N2H5NO3

-661

1,685

M.p 71C,
hygroscopic

NH4N(NO2)2

-270

1,82

M.p.~89C,
enough stable

Oxidizer

NO2

NH4 N
NO2

ADNA

172

Composition, mass
%%
12%CH*
12%PMVT**
10%CH +20%Al
11%CH
12%PMVT
17%Act ***
15%Act +20%Al
12%PMVT+20%Al
10%CH +20%Al
12%CH
12%PMVT
10%CH +20%Al
15%Act +20%Al
12%CH +13%Be
15%CH +13%Be
34%Act +13%Be
34%Act +14%Be

Isp, s
211
220
246
219
230
240.5
256.5
255.5
255
247
250
262
264
281
293
291
294.5

d,
g/cm3
1.57
1.60
1.70
1.54
1.62
1.64
1.78
1.7
1.67
1.63
1.72
1.76
1.86
1.63
1.34
1.42
1.40

Energetic Characteristics of Solid Composite Propellants and Ways for Energy . . .

New Trends in Research of Energetic Materials

N2H5N(NO2)2

-52

1,80

Czech Republic

2006

[Content]

M.p.76-80C,
hygroscopic

12%CH
251
1.62
16%PMVT
255
1.70
19%Act
241.3 1.64
10% CH +20%Al
267
1.75
C(NH2)3N(NO2)2 -229
1,71
M.p
135- 10%CH
208
1.58
139C, enough 18%Act
233
1.67
stable
17%Act +20%Al
255
1.80
*-CH hydrocarbon binder (#1, table 3); ** PMVT - polymethylvinyltetrazole
(#2, table 3); *** Act- active binder (#3, table3).
The most interesting among NHO-oxidizers is ammonium dinitramide (ADNA). It was
first synthesized in the former USSR in 1972 (however first published in 1992 [1]). ADNA
is enough stable and compatible with many compounds and nowadays it is used in real
compositions. ADNA is almost the best oxidizer for SCP for upper stages while because of
its rather low density it is not enough powerful for first stage SCP, where ammonium
perchlorate (see below) is more appropriate.
b) SCP basing on perchlorates
Among all perchlorates (we consider only oxidizers free of metal, because mostly metal
salts are considerably less effective) only ammonium perchlorate (AP) is really used.
Nitronium perchlorate (NP) is the best energetic oxidizer, but it is so much chemical active,
that it can not be used in mixture with any other compound. Hydroxylammonium perchlorate
(HAP), hydrazine monoperchlorate (HMP), and hydrazine diperchlorate (HDP) are much
more effective than AP, especially for lower stages (because of high density). However these
high-density oxidizers have some serious drawbacks they are less thermostable than AP,
less compatible with other compounds, and, finally, they are too hygroscopic. In IPCP RAS
methods of thermal stabilization of HAP, HMP, HDP have been found [2], methods of
stabilizing mixtures of HAP with some compounds have been developed [3], but the
hygroscopicy can not be decreased yet.

Table 2. Energetic characteristics of SCP basing some perchlorates.

Oxidizer

Hf,
kcal/k
g

d,
g/cm3

Additional info

NH4ClO4

-597,5

1,95

Stable

NH3OHClO4

- 494

2,07

N2H5ClO4

- 320

1,94

Formulation, mass
%%

10%CH*
20%Act***
18%PMVT**
9%CH+20%Al
m.p.
89C, 13%CH
hygroscopic
20%Act
21%PMVT
9%CH+20%Al
m.p.
137C, 9%CH
hygroscopic,
10%CH
high
impact 26%Act
sensitivity .
9%CH+20%Al

David Lempert, G. Nechiporenko, G. Manelis

Isp, s

d,
g/cm3

240
204
238
250
250
204
247
253.5
251
246.5
240
259

1.75
1.83
1.78
1.86
1.77
1.92
1.82
1.94
1.76
1.74
1.70
1.86
173

New Trends in Research of Energetic Materials

N2H6(ClO4)2

- 297

2,21

C(NH2)3ClO4

-465

1,75

C(NHNH2)3ClO

-116,5

1,56

65

2,22

NO2ClO4

Czech Republic

hygroscopic,
high
impact
sensitivity
m.p.240C,
stable,
hygroscopic
high imp. sensit.,
mp.132C.
Extremally
reactive

2006

[Content]

12%CH
9%CH+20%Al

251
254

1.88
2.02

18%Act
15%Act+20%Al

226
246.4

1.70
1.82

18%Act
15%Act+20%Al
20%CH
14%CH+20%Al

233.5
255
261
258

1.54
1.69
1.72
1.91

c) Oxidizers containing active fluorine


The fluorine is a more active oxidizer than oxygen is, the heat of H2O formation is much
lower than the heat of HF formation. There are some inorganic oxidizers with active fluorine
(e.g. FClO4, some complex salts (NF4)aMeFx) but their physical and chemical properties do
not allow us to use them in SCP. The only real way to introduce active fluorine into
molecule is the using of >N-F or -NF2 groups because if fluorine is introduced with a
component containing C-F bond no power increase is gained because of high energy of the
C-F bond. Nowadays many NF2-derivatives of organic substances are known, some of them
are used in energetic compositions. If such compounds are used as oxidizer its possible to
create SCP with Isp 280 and even higher. However these substances have some serious
drawbacks they are not ecologically soft, and their impact and friction sensitivity are rather
high.

SCP, CONTAINING METALS.


Only four metals (Be, Al, B and Mg) can increase the Isp of the pair binder+oxidizer.
The effectiveness of metal is due to the great heat of their oxidation into oxides. In case of
other metals the heat of their combustion does not compensate the loss of gas fraction in
combustion products. The effectiveness of metal introduction falls with the range
BeAlBMg, and it depends much on the oxidizer enthalpy. In fig.1 the dependence of
the effectiveness of metal introduction (dIsp) on oxidizer enthalpy is demonstrated. dIsp =
IspMe - Isp0, where IspMe is the Isp value of the optimal composition
metal+oxidizer+hydrocarbon binder; Isp0 is the Isp value of the optimal binary formulation
oxidizer+hydrocarbon binder. Circlets on the figure represent very different oxidizers, with
different oxygen balance, organic and inorganic, real and designed. Thats why all circlets
for each metal do not represent an absolute straight lines. However, the tendency is
evidential effectiveness of metal introduction falls with the growth of the oxidizer
enthalpy. Magnesium introduction may be useful only if Hf of the oxidizer is rather low
(see fig.1), while the introduction of beryllium is useful until Hf of oxidizer is about +600800 kcal/kg. However Be-containing SCP are extraordinarily toxic because of BeO toxicity,
thats why their development ended about 30 years ago. Regardless the introduction of other
metals always decreases the Isp values, some metals with high density (Zr, Ti) can be useful
in definite cases when the propellant density is of higher importance (see above), that is
when the exponent a in the formula Efn=Isp*da is 0.7-0.8 and higher.

174

Energetic Characteristics of Solid Composite Propellants and Ways for Energy . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

dIsp
60

NH4NO3

ADNA

50
Be

40
30

N2H5N3O4

Al

20
10
0

Mg

-1000

-500

500

1000

oxidizer's specific enthalpy, kcal/kg

Fig. 1. The dependence of the dIsp on CNHO-oxidizer enthalpy [4].

OPTIMAL BINDERS.
We found that regardless binder does not occupy a big fraction in the composition
(usually 9-12 mass % if binder density is 0.9-1.0 and about 20-25 mass % if the density is
high, up to 1.5-1.55) the choice of binder type is very important for the maximal energy
achievement. Element content of the binder has to compensate excess or lack of oxygen in
the leftover formulation (the summary of compounds except binder). Thus, if binary
formulation contains a binder and an oxidizer with a high oxygen balance ( =
O/(2C+0.5H)), e.g. ADNA or HAP, the best binder is usually a simple hydrocarbon binder,
because it contains a lot of hydrogen. As soon we begin to add so called gasifying
component, e.g. HMX, the value of the mixture oxidizer + HMX begins to decrease and
the hydrocarbon binder can not any longer secure maximal energetic potential, that is kept in
these compounds we have to use more active binders with higher values. Because of
the same reason active binders are necessary if the formulation contains metal hydrides (see
below). Thus, when we study a new compound (mainly oxidizer, gasifying or energetic
compound) we have to test formulations with different kinds of binder. For example, in our
investigations we use a listing like that represented in table 3. Binder #2 is a representative
of high-enthalpy binders, binder #4 is a representative of binders with a middle value, but
higher enthalpy and hydrogen content than #3. In some cases such binders as #4 may
become the optimal binders among all other types.
As for gasifying compounds the best of them is widely used well-known HMX. Actually
we may consider a gasifying compound as a part of a binary oxidizer. Regardless gasifying
compound is not usually a good oxidizer itself, its mixture with the main oxidizer can be
more effective than if the main oxidizer is used alone. For example in the system 15.5
mass% (20 volume %) active binder + 20 mass %Al +AP + HMX the maximum value of Isp
is 260.5 at 57 % HMX, while in the system 10 mass % (20 volume %) hydrocarbon binder +
20 %Al +AP + HMX the maximum value of Isp is 254 at 25 % HMX.

David Lempert, G. Nechiporenko, G. Manelis

175

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 3. Energetic properties of some binders.


#

Name, formula

Hydrocarbon, C73,17H120,9

Polymethylvinyltetrazole, C37,34H56,007N32,69O2,32
Active, containing 20% polyvinylmethyltetrazole,
nitroglycerine with some additives. C18,96H34,64N19,16O29,32

3
4

Polyvinylmethoxydiazen-N-oxid, C3H6N2O2

Hf,
kcal/kg

d,

g/cm3

-94

0,91

+300

1,28

~0

-181

1,49

0.53

-14

1,31

0.22

SCP WITH METAL HYDRIDES.


The using of metal hydrides (BeH2, AlH3) in SCP offers great opportunity to increase
much Isp values. In table 3 its evident that Isp values can be increased on 10-15s using the
same oxidizer, because hydrides serve not only as metal fuel, but as a rich source of
hydrogen too. However new SCP with hydrides required new binders so called active
binders (see above). If in propellants oxidizer+binder and oxidizer+binder+metal a binder
was necessary not only as properly binder, but as a fuel (in system oxidizer+binder) and a
gas source (both quoted systems) too, in the presence of hydride SCP has already a powerful
heat source as well as a gas source. So, binder is not energetically necessary any more.
However so long as a polymer binder is still necessary, it has to possess new qualities,
namely to serve as oxidizer too. So, new active binders, containing NO2, ONO2, NNO2
groups etc have been created. Simultaneously active binders may serve as energetic
component if they contain high energetic functional fragments, e.g. N-heterocycles. The
creation of new SCP with metal hydrides was a great step in development of high energy
SCP. Isp value of the composition containing 25% AlH3, 25% active binder and ADNA is
276-277 (d=1.64), of the composition containing 14% BeH2, 34% active binder and ADNA
is 293.5 (d1.40). However AlH3 and especially BeH2 have low density (1.48 and ~0.7 g/cm3
respectively), therefore the SCP density is lower than in similar SPC with free metals. Thus,
SCP with BeH2 is energetically useful only for upper stages (really for the third stage only)
when the exponent a in the formula Efn=Isp*da is 0.2 and lower; with AlH3 - if the
exponent a is ~0.4 and lower, i.e. for the second and third stages. Moreover, BeH2 has the
same obstacle than free Be has, this is an extraordinary high toxicity of the combustion
product (BeO).

SCP BASING ON HIGH-ENTHALPY OXIDIZERS.


This kind of SCP is not developed yet. Regardless many organic oxidizers with Hf
+500 kcal/kg and even higher are known, their application in SCP is limited by many factors
(high impact and friction sensitivity, sometimes low thermal stability, high price, etc.). The
first relatively high-enthalpy component that has been widely investigated during last 15
years or so is hexanitrohexaazaisowurtzitane (CL-20, Hf =+185 kcal/kg, d=2.044 [5]).
Propellants containing aluminum, active binder, CL-20, and AP are a bit more effective on
the lowest stage (Isp~258 at Pc:Pa=40:1, d=~2.00; combustion temperature 3970K;
temperature of products leaving nozzle section 2770K) than the similar propellant containing
HMX instead of CL-20. Actually this higher effectiveness is the result of higher density of
CL-20 against HMX.

176

Energetic Characteristics of Solid Composite Propellants and Ways for Energy . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

There are many organic substances (containing mainly high-enthalpy N-heterocyclic


frame with fragments-oxidizers, such as NO2, NNO2, C(NO2)3, etc) that can be used as highenthalpy oxidizers (Hf >500 kcal/kg), e.g. ditrinitromethylfuroxane (DTNMF, Hf=+570
kcal/kg ); 1,2,5-oxodiazolo-[3,4-e]-[1,2,3,4]-tetrazine 4,6-dioxide (ODATDO, Hf=+730
kcal/kg); 5-(3-nitro-furoxan-4-yl)-1H-[1,2,3]triazolo[4,5-c]-furoxan inner salt (NFTAF,
Hf=+760 kcal/kg) [6,7]. Using such oxidizers with appropriate binders its possible to
accomplish Isp values up to 268-270s.
(O2N)3C

NO2
H2C

O2N N
H2C

CH2

O2NN
O2NN

N NO2

N NO2
N

CH2

C(NO2)3

NNO2
N
N

N
O

O2NN

NNO2

O
N

N
O

NO2
N

NO2

HMX

CL-20

DTNMF

ODATDO

NFTAF

Anyway propellants with such new oxidizers as DTNMF, NFTAF or ODATDO will
cost much more than propellants basing on accessible PA and HMX. Moreover, as it was
already remarked, further increasing of energy power of solid composite propellants is
attended with an appreciable aggravation of other properties (impact and friction sensitivity,
thermal stability etc.).
Anyway such high-enthalpy oxidizers are still rather promising, moreover they do not
need metals (see above) and therefore they have no loss of Isp because of biphasic
combustion products while metal-containing SCP have them (we mean here the loss as the
difference between real Isp value and the calculated one). The reason of so-called biphasic
loss is the following: when in combustion products there is a solid phase some extra heat is
carried away from the nozzle section with these solid particles - they are a bit superheated in
comparison with the gas phase because the thermal equilibrium between solid and gas phase
can not be established during so small time when combustion products expand trough
nozzle. Thermodynamic calculation of Isp does not include this kind of thrust loss, therefore
if calculated Isp values of two different formulations are equal, but the first formulation
contains metal while the second does not, it means that other characteristics being equal the
second formulation procures higher real thrust. Actually biphasic loss depends on solid
particles dimension and of the rocket engine type and may average 5% or so.

WAYS FOR FURTHER SOLID COMPOSITE PROPELLANTS


DEVELOPMENT.
Recently the geopolitical situation changed considerably because of the cold war end.
Therefore the creation of cost-to-cost ballistic missiles and their propellants have not such a
great importance any longer while the portion of SCP civil applications grows constantly in
importance.
If we estimate briefly the succeed energy level of SCP that are already produced in
reality it is evident that the most powerful are the following ones:
propellants based on aluminum, HMX (~55%) and active binder - Isp~260 at
Pc:Pa=40:1, d=1.91; combustion temperature (Tc) =3780K; temperature of products leaving
nozzle section (Ta) =2550K - this kind of propellant is the most effective on lower stages
because of high density;

David Lempert, G. Nechiporenko, G. Manelis

177

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

propellants based on aluminum hydride, ADNA and active binder (Isp~276 at


Pc:Pa=40:1, d=1.64; Tc=3550K; Ta=2330K), this kind of propellant is the most effective on
upper stages;
propellants basing on CL-20 (see above) that are developed recently are more effective
on the lowest stage only because of its higher density, on upper stages HMX is more
effective.
SCP basing on new high-enthalpy oxidizers (see above) have rather higher Isp values,
but this kind of propellants is not developed yet, and in any case they will cost much higher
than propellants with PA, HMX, ADNA. Moreover, further increasing of SCP power is
attended with an appreciable aggravation of other properties. For example, application of
high-enthalpy CHNO-substances (>+500 kcal/kg) would decrease the stability; would
degrade impact and friction sensibility, would increase values of exponent in combustion
law U=A*P. Anyway maybe SCP of this type (with high-enthalpy CHNO-substances) can
be used too, but this perspective is too far.
In the case of civil applications one puts a considerable accent on ecological problems.
The HCl effusion ruins greenery and atmospheric ozone coat. The solution of these problems
may be carried by different ways, using chlorine free inorganic (e.g. ADNA, Table 4, #4) or
organic oxidizers, e.g. HMX (#3), RDX (hexogen), as well as using propellants containing
potassium or sodium salts (e.g. sodium nitrate or hyponitrate) together with AP (Table 4,
#15,16) when equivalent amount of alkaline metal is a bit higher than one of chlorine, thus
almost all chlorine undergoes into the most thermodynamically stable alkaline metal chloride
[8].
Table 4. Energetic characteristics and concentration of toxic gases and condensed
phase in combustion products
Propellant

Isp

Tc, K Ta,K

HCl+Cl2,
mol/kg

CO,
mol/kg

Cond. phase,
mass.%

251 1.84 3610 2520

38

PA+CH*+Al

AP+CH

240 1.76 3000 1900

7.8

0.8

HMX+ Act**+Al

257 1.95 3700 2550

12

38

ADNA+ CH+Al

262 1.76 3580 2330

38

ADNA+ CH

249 1.64 2970 1580

3.6

N2H5NO3+ CH+Al

254

7.7

38

N2H5NO3+ CH

930

NH4NO3+ CH+Al

244 1.68 2880 1800

7.3

38

NH4NO3+ CH

205 1.56 1870

920

10 N2H5C(NO2)3+CH+Al 265 1.81 3580 2360

11

38

178

1.6

2910 1770

215 1.53 1950

Energetic Characteristics of Solid Composite Propellants and Ways for Energy . . .

New Trends in Research of Energetic Materials

12

N2H5C(NO2)3+ CH

254

13

NaClO4+ CH +Al

14

NaClO4+ CH

1.7

Czech Republic

[Content]

225 2.05 3650 2610

1.1

10

66

211

3080 2030

0.25

44

15 AP+NaNO3+ CH +Al

230 1.96 3665 2650

0.9

4.7

37.7

16 AP:Na2N2O3+CH+Al

237 1.94 3675 2600

0.98

5.3

37.7

2.0

3070 1600

2006

SOME NEW CIVIL APPLICATIONS OF SOLID COMPOSITE


COMPOSITIONS.
Multiyear experience of chemists in the area of military propellants creation and
production may help to resolve currently some tasks of exceptionally civil application of
energetic materials, i.e. in development of gas-generating compositions for airbag inflators.
In this task the thrust value is no longer necassary, but quite other attributes become crucial.
Moreover, very high energy makes even cardinally worse the characteristics of the toppriority. New propellants must be free of drawbacks of currently used propellants. The main
drawbacks are toxic ingredients, presence of toxic gases (CO, NOx) and solid phase in
combustion products, too high combustion temperature, insufficient stability. By the way
propellants for airbag inflators require much higher thermal stability level than propellants
for solid rocket engines do. Main requirements for compositions for airbags and the
characteristics that SCP can provide differ much (Table 5). Thus, for improve characteristics
of propellants for their application in airbags, we have to decrease their power, that is to
decease considerably combustion temperature of stoichiometric formulation (#1, table 5),
and on the other hand we have to increase the stability and combustion rate.
It was shown that it is possible to decrease combustion temperature of stoichiometric
formulations (this is CaHbNxO2a+0.5b, otherwise combustion products contain either CO or
NOx) by introducing low-enthalpy oxygen-containing groups (-OH, -O-, -COOH, >CO, COO-, -OCO-COO- etc.) into oxidizer containing regular groups-oxidizers (-NO2, -ONO2, NNO2). Later groups provide high oxygen amount in the formulation while the first ones
provide considerable enthalpy decreasing, and consequently Tc decreasing at minimal
oxygen amount decreasing [9].
Table 5. Requirements to propellants for airbags and characteristics that can be
reached with accessible SCP.
#

Characteristics

Element ratio

Condensed phase in
comb.products

Requirement to
propellants for
airbags
2C+0.5H~ =O (atoms
ratio)
Absence of solid, at
least very low amount

Combustion Temperature

~ 2000-2400K

11

David Lempert, G. Nechiporenko, G. Manelis

Characteristics that can


be reached with
accessible SCP.
No requirements
Usually 37 mass % of
Al2O3 if the formulation
contains 20% Al. If there
is no metal in the
formulation no
condensed phase
>2900 K
179

New Trends in Research of Energetic Materials

Combustion rate

Exponent in combustion law


(U=A*P)
Thermal stability
toxic products and hydrogen
content combustion products

6
7

Minimal melt.point of
ingredients
9 sensitivity to impact and friction
10 requirement to dioxines amount

Czech Republic

2006

[Content]

25-30mm/s at 200
atm
~0.5-0.6

~10-20 mm/s
at 200 atm
0.3-0.5

~ 2 weeks at 110C
No toxics, no HCl, no
Cl2.
Low H2 content, very
low CO and NOx
content.
110C

One week at ~60-70C


[CO] =3 mol/kg (even if
Tc~1900K) and higher.
[H2] =0.5-15 mol/kg.
Many HCl if the
formulation contains AP.
80C

Low
No data

Rather low
Accordingly international
ecological requirements<10-10g/m3

High energetic compositions may be used more widely as gas generating compositions
for different goals. Of the most importance are generators producing gases (air, oxygen,
nitrogen and other gases) with given content with a necessary combustion rate.

REFERENCES.
[1]
[2]
[3]

[4]
[5]
[6]
[7]
[8]
[9]

180

Lukyanov O.A., Gorelik V.P., Tartakovskii V.A., Izvestia of Russ.Ac.Sci., ser.chem., 1994,
P.94-97.
Lempert D.B., Dolganova G.P., Nechiporenko H.N.// Hydroxylammonium perchlorate thermal
decomposition on low decomposition level. (1998) J.Appl.Chem. (Russ) v.98, N3, P.365-370.
Lempert D.B., Dolganova G.P., Nechiporenko H.N.// Thermal stability of mixture
hydroxylammonium perchlorate with HMX. (1998). J.Appl.Chem. (Russ) v.98, N10, P.16231628.
Lempert D.B., Nechiporenko H.N., Dolganova G.P.//Chemical Physics (Russ), 1998. V.17, N7,
P.114-120
R.L. Simpson et al.//Propellants, Explosives, Pyrotechnics 22, 249-255 (1997)
Lempert D.B., Nechiporenko H.N. et al. Chemical Physics (Russ), 2003. V.22. N4. P.64-69.
Nechiporenko H.N., Lempert D.B., Soglasnova S.I. Chemical Physics (Russ), 2005. V.24. N3.
P.74-80.
Lempert D.B., Dolganova G.P., Nechiporenko H.N., Stesik L.N.// Chemical Physics (Russ),
1997. V.16. N9. P.91-100.
Lempert D.B., Manelis G.B., Nechiporenko H.N., Shteinberg V.G., Zyuzin I.N.// Principles of
Creation of Smokeless Gas Generating Compositions Free of Toxic Combustion Products. 7th
International Airbag-2004, 2004, Karlsruhe, Germany, P.43.

Energetic Characteristics of Solid Composite Propellants and Ways for Energy . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

MECHANISM OF THE LOW-FREQUENCY NON-ACOUSTIC


INSTABILITY PHENOMENA FROM THE POINT OF VIEW OF THE
CONCEPT OF THE SPATIAL-PERIODIC MICRO-STRUCTURES
EXCITATION AT THE ENERGETIC MATERIALS COMBUSTION
Alexander N. Lukin
Institute of Applied Mechanics, Ural Branch of the Russian Academy of Sciences,
426067, Izhevsk, Udmurtia Republic, Russian Federation

Abstract:
Understanding of the mechanism of the low-frequency non-acoustic instability
phenomena at the solid rocket propellants combustion - one of the most complicated
problems in the solid propulsion physics. Several theories have been proposed
highlighting one or other process as the dominant mechanism, but a unifying theory is
yet to emerge. The present work focuses on one of the critical events, namely the spatialperiodic micro-structures (SPMS) excitation in the evaporated energetic materials (EM)
liquid-viscous layer (LVL). The low-frequency non-acoustic combustion instability
phenomenon, the chuffing phenomenon and the accompanying physico-chemical effects
have received a new explanation within the concept based on the data of optical
visualization of the physico-chemical processes on the EM burning surface. This
concept connected, mainly, with excitation of the synergetic dissipative SPMS in the thin
LVL and on the EM burning surface and determining the burning wave spatial
instability. At heating from above in the thin LVL occurs the thermo-electric convection
excitation, that induce cellular movement and formation of the synergetic SPMS.
Besides the velocity cells, in the LVL arise the electric field cellular structures.
Preliminary results of the investigation show that mechanism of the low-frequency nonacoustic combustion instability of the evaporated EM is connected with the phenomenon
of the existential micro-structures excitation in the EM LVL. Comparison of the scales of
the EM heated up layer, the LVL, the SPMS and the torches (burning cells) on the
burning surface also shows that the torches (cells) are formed from the set of SPMS from the set of thermo-electric convection cells. The experimental data have shown that
torch combustion phenomenon develops irrespective of the properties and structure of
the specific EM. By other words, on the EM burning surface is observed the process of
self-organizing of the dynamic dissipative synergetic SPMS into the torch macrostructures. In this connection, on the burning surface the process of moving (wandering)
of the torches is observed. The torch macro-structures on the EM burning surface can
be considered as independent synergetic structures. Besides, the torch structures on the
burning surface can exist steadily in the event that they form a pair of two torches. This
phenomenon reflects universal law of the SPMS excitation on the EM burning surface.
Within suggested mechanism becomes possible to give a new explanation of the
phenomenon of the end-burning solid propellant charges rhythmic extinctions. An
attempt has been made to indicate future direction of research that can be lead to more
accurate prediction of starting transients in the solid propulsion systems.
Keywords:

evaporated energetic materials, combustion instability, chuffing,


torch macro-structures, ionic fusion with thermo-electric properties,
thermo-electric convection excitation, carbon grid

Alexander Lukin

181

New Trends in Research of Energetic Materials

1.

Czech Republic

2006

[Content]

INTRODUCTION

The problem of combustion instability and anomalies of burning of the energetic


materials (EM) traditionally remains one of actual problems in the theory of combustion. [1]
The principles of the EM combustion anomalies theory were established by outstanding
Russian scientist Ya.B. Zel'dovich. [2]
Prediction of ignition transients and low-frequency non-acoustic combustion instability
in the solid propulsions systems (SPS) has remained a topic of active research for several
decades, yet there appears to be no model that can describe the roles played by all the
complex physico-chemical processes. The development of larger and more sophisticated
SPS have emphasized the need to model the ignition transients accurately as they do not lend
themselves to costly trial and error development techniques. The radical difference in size
and technology of these SPS defy extrapolation of the empirical knowledge gained in the
development of earlier, more conventional SPS.
With regard to the physical mechanisms of the solid rocket propellant combustion, the
combustion zone is quite complicated even under steady state conditions, with multiple
flamelets attached to different parts of the burning surface. The complications are, in part,
due to the statistically random nature of the distribution of the reactants in the condensed
phase, and, in part, due to poor understanding of the combustion zone owing primarily to its
microscopic nature and the hostility of the environment to clear investigation.
Understanding of the mechanism of the low-frequency non-acoustic instability
phenomena at the solid rocket propellants combustion - one of the most complicated
problems in the solid propulsion physics. Several theories have been proposed highlighting
one or other process as the dominant mechanism, but a unifying theory is yet to emerge. The
present work focuses on one of the critical events, namely the spatial-periodic microstructures (SPMS) excitation in the evaporated EM liquid-viscous layer (LVL).

2.

THE LOW-FREQUENCY NON-ACOUSTIC INSTABILITY IS THE


CHARACTERISTIC PROPERTY OF COMBUSTION PROCESS OF
THE SOLID ROCKET PROPELLANTS

SPS with a small characteristic length of the chamber L* may exhibit spontaneous
oscillations in the chamber pressure. If the oscillations occur simultaneously in the bulk of
the chamber, such an instability has been variously called the L*, bulk mode or non-acoustic
type. The phenomenon is typically characterized by the amplifying low-frequency pressure
oscillations of below some 300Hz, leading to the extinction of the SPS. In extreme cases,
when the L* is very small, extinction occurs almost immediately after ignition, followed by a
sequence of periodic pressure build-up and extinction: such a process has been called
chuffing.
The phenomenon of chuffing involves large pressure fluctuations, sometimes with
widely varying frequencies. Although the order of magnitude of chuffing frequencies
suggests a dominant solid-phase energy release mechanism at low pressures, the underlying
process is not convincingly explained yet. Such kind of combustion instability in the
combustion chamber is most frequently observed for the end-burning solid propellant
charges and for channel-shaped propellant grains. According to the data presented in the

182

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

paper [3], for the channel-shaped charges the boundary of low-frequency combustion
instability depends not from the channel volume and the chamber volume, but is determined,
mainly, by the pre-nozzle volume. Therefore, if the volume of the SPS combustion chamber
with end-burning charge coincides with the pre-nozzle volume of the SPS with a channelshaped charge, the boundaries of their low-frequency stability coincide.
In the paper [4] on the basis of large number of the fire stand tests has been made
conclusion that periodic processes - the chuffing processes and the low-frequency nonacoustic combustion instability are the characteristic properties of combustion process of the
solid rocket propellants. Usually, these properties appears in the conditions of low pressure
in the combustion chamber and (or) in the conditions of the low ambient temperatures. As a
matter of fact, for the first time has been suggested assumption that the burning process of
each specific solid propellant can be characterized by the set of own frequencies of
pulsations of the burning surface that appears in the critical burning conditions. And also has
been mentioned the existence of some universal law connected with unstable processes on
the burning surface of the solid propellants. However how it is possible to explain this
property and what physical phenomena appears here not enough clearly till now. Existence
of such problems has been indicated in the recent dissertations. [3, 5] The hypothesis that the
determining mechanism of excitation of the low-frequency combustion instability and
chuffing are the reactions in the gas phase has been suggested in the paper [6]. The similar
hypothesis about existence of the gas-phase oscillatory chemical reactions which excite lowfrequency combustion instability and chuffing has been suggested in 2003, in the dissertation
[5]
. However the papers [7, 4] have convincingly shown that the determining mechanism of
excitation of the chuffing processes and the EM low-frequency combustion instability are the
reactions in the condensed phase (the mechanism of thermal explosion of the condensed
phase), but not the reactions in the gas phase. Frequency of pulsations at the burning process
of the propellant laboratory samples in the air appeared same as at burning of the samples in
the conditions of the nitrogen flow above the propellant burning surface.
At the same time, the heat feedback from the hot gases trapped in the chamber when the
chamber and ambient pressure equalized due to extinction, could not cause re-ignition of the
propellant. Re-ignition after extinction, and hence chuffing, could not be obtained with a
surface reaction model.
For successful solution of the problems, connected with suppression of the SPS
combustion instability are necessary to have detailed understanding about essence of
mechanisms of the EM unstable burning on the new qualitative level. Obviously the
possibilities of understanding of laws of this complex phenomenon are closely connected
with excitation and formation of the cellular micro-structures on the EM burning surface.
Certainly, the new level in understanding of the fundamental laws of the EM lowfrequency non-acoustic combustion instability can be provided on the basis of use of the
newest technologies of visualization of zones of burning of the EM.
Recently with use of new interferometric and shadow technologies of optical
visualization of zones of burning of the EM, the unique results overturning traditional
representations about the mechanisms of cellular-pulsating burning and low-frequency nonacoustic instability at the solid rocket propellants combustion have been received [8 - 10].

Alexander Lukin

183

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Some fragments of optical visualization of the process of burning of the standard


ballistite propellants are shown on the Fig. 1 and Fig. 2.

Fig 1. Sequential images of dynamics of the carbon spatial-periodic micro-structures on the


burning surface of the standard ballistite propellant.

On the Fig. 1 are shown the fragments of the process of throwing away of the carbon
grid, formed at the burning surface.
The experimental data have shown that torch combustion phenomenon develops
irrespective of properties and structure of the specific EM. At realization of this burning
mode the sizes of burning cells are not connected with sizes of components and the sizes and
structure of researched EM. Besides, according to the experimental data, ignition of long
propellant channels occurs in the cellular or in the torch mode. [11]
Also the experimental data has shown that active (reaction-capable) SPMS (the burning
cells) on the burning surface are united into the more large-scale structures - into the torch
macro-structures. By other words, on the EM burning surface is observed the process of selforganizing of the dynamic dissipative synergetic SPMS into the torch macro-structures. The
torch micro-structures on the EM burning surface can be considered as independent
synergetic structures. This phenomenon reflects universal law of the spatial-periodic microstructures excitation on the burning surface of the EM.
In this connection, on the burning surface the process of moving (wandering) of the
torches is observed.

184

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 2. The effect of formation of the pair of torch structures. Sequential images of dynamics
of two torch structures on the burning surface of the standard ballistite propellant.
Fig. 2 shows dynamics of the two torch structures on the burning surface of the standard
ballistite propellant. The processes in each of the burning cells develops independently from
each other. The torch structures on the burning surface can exist steadily in the event that
they form a pair of two torches. This phenomenon is observed also on the shadow photos of
torches on the burning surface.
The Fig. 2 (1) shows a luminescence of gaseous products of reaction in the one torch. At
this time occurs formation of the carbon grid on the burning surface. After formation of the
carbon grid on the burning surface the luminescence of the torch disappears. Further the
process starts to develop on the burning surface and appears luminescence of the carbon grid
on the burning surface (Fig. 2 (2)). The combustion products flowing from the burning
surface throw away a carbon grid from the burning surface ( Fig. 2 (3)). After throwing
away of the carbon grid the luminescence of the torch above the burning surface appears
again ( Fig. 2 (4)).
Separate researches of the torch structures were executed, with use of shadow method,
both for the gun-powder "N" and for standard gun-powders on the basis of nitrocellulose
with the lead oxide. [8 - 10] In the researches were used the gun-powder plates with thickness
of 3 mm (on the course of the light beam) and with width of 12 mm.
Fig. 3 and Fig. 5 contain the information only of two categories. Firstly, this is jet stream
of the gasification products from the burning surface and secondly, the dependence of the
sizes (and accordingly, the numbers) of the burning cells from the pressure.
5

Alexander Lukin

185

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 3. Burning of the ballistite gunpowder "N" in the nitrogen


atmosphere at the pressure 0.4
MPa. Two torch macro-structures
on the burning surface.

Fig 4. Burning of the ballistite gunpowder "N" in the nitrogen


atmosphere at the pressure 0.6
MPa. Four torch macro-structures
on the burning surface.

Fig 5. Burning of the ballistite gunpowder "N" in the nitrogen


atmosphere at the pressure 0.8
MPa. Six torch macro-structures
on the burning surface.

At execution of the researches in the field of high pressures it was impossible to


differentiate the cells on the burning surface. In these conditions on the burning surface there
was so many burning cells, that at transverse passing of the light beam through the several
jets it was impossible to differentiate behavior of the separate burning cell on the
background of others.

3.

EXCITATION OF THE SPATIAL-PERIODIC MICROSTRUCTURES IN THE IONIC FUSION WITH THERMO-ELECTRIC


PROPERTIES ON THE BURNING SURFACE

In accordance with extensive experimental data [12 - 14] burning of the evaporated EM is
accompanied by occurrence of electric conductivity of the burning surface (of the liquidviscous layer) and by chemical ionization of the gas layers adjoining to the burning surface.
186

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In conditions of burning wave where the temperature in the condensed phase increasing by
exponential law, the thin reactionary LVL can be considered as the molten mass with ionic
properties. Concentration of the ions increases with the burning rate and determines electric
conductivity of the burning surface. At burning in conditions of threshold-low pressures,
periodic pulsations of the electric conductivity are observed.
The low-frequency non-acoustic combustion instability phenomenon, the chuffing
phenomenon and the accompanying physical-chemical effects have received a new
explanation within the following universal concept based on the unique data of optical
visualization of physico-chemical processes on the EM burning surface.
This concept connected, mainly, with excitation of the synergetic dissipative SPMS in
the thin LVL and on the EM burning surface and determining the burning wave spatial
instability. At heating from above in the thin LVL occurs the thermo-electric convection
excitation, that induce cellular movement and formation of the synergetic SPMS. [15] Besides
the velocity cells, in the LVL arise the electric field cellular structures.
Fig. 6 schematically shows the distribution of the physico-chemical processes in the thin
LVL in the EM reaction zone.

Fig 6. Thin liquid-viscous layer in the EM reaction zone, having thermo-electric properties
at heating from above.
In accordance with the experimental data, the ratio of the longitudinal and cross-sectional
sizes of the elementary SPMS are the most stable characteristic of the cellular movement
excitation. Change of the boundary conditions practically does not influence this parameter.
The reason of stability of the sizes of the SPMS is connected with the thermo-electric
mechanism of excitation of such structures.

Alexander Lukin

187

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In other words, stability of the SPMS is provided by the electro-magnetic field cellular
structures in the LVL. The experimental data shows the boundaries of the electric field
structures coincide with the boundaries of the convection cells (the structures of velocity).
And on the LVL surface under influence of thermo-electric field is excited the electric
charge.

4.

TWO SCALE OF THE SPATIAL-PERIODIC STRUCTURES ON THE


BURNING SURFACE

Comparison of the scale of zone of the EM cellular-pulsating burning shows that


reaction-capable SPMS excited in the LVL are the initial existential structures that are united
in the set of aggregates. Each of such aggregates represents a torch or the burning cell on the
burning surface.
The burning process localized in each of the SPMS, is supported by the process of selforganizing of the SPMS into the groups. In other words, each of such aggregates, formed
from two and more SPMS with identical properties it is possible to consider as a cluster.

TORCH (CELLULAR) SPATIAL-PERIODIC MICRO-STRUCTURES ON THE


BURNING SURFACE
Scale of the torch (the burning cell) on the burning surface - (10 - 15) h
(Diameter of the torch - (1 - 2) mm)

CARBON GRID (CARBON SPATIAL-PERIODIC MICRO-STRUCTURES)


ON THE BURNING SURFACE
HEATED-UP LAYER
OF THE ENERGETIC
MATERIAL
Reaction Liquid-Viscous Layer
The Thickness of the
Heated-up Layer
(The Conductive Size)

h a / uC
(0.010 0.070) mm

The thickness of the reaction liquid-viscous layer determines


scale of the SPMS (cells) of the thermo-electric convection.
For estimation of the average thickness of the reaction liquidviscous layer (the effective thickness of the zone of electric
conductivity) in the EM heated-up layer can be used the
following expression: S a / ( 10 u C )
The thickness of the reaction liquid-viscous layer
- (0.0010 0.0035) mm

The Inert Heated-up Layer of the Energetic Material

188

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The torch structures on the burning surface forms the non-uniform burning front: on the
burning surface appears the local discontinuity.
In the conditions of low pressures the sizes of the SPMS and the cluster structures
(torches) will increase, but the total number of the torch structures on the burning surface
will decrease. And vice versa, in the conditions of high pressures the sizes of the SPMS and
the cluster structures (torches) will decrease, but the total number of the torch structures on
the burning surface will increase.
Besides for estimation of the S value can be used D.A. Frank-Kamenetskiis
characteristic temperature interval [16]. Let's consider that the main part of reactions takes
place within the limits of the quadruple specified temperature interval:
S =

4 R T S2

(1)

Ea 0

With taking into account equality for the temperature gradient:

0 =

CuC

( T R T0 )

(2)

We can write down:


S =

4 R T S2

E a u C ( TR T0

4 R TS
a

u C E a (T R T0
2

(3)

For example, at T 0 = 293 K and P = 4 MPa, S = 0.0035 mm; and at T 0 = 293 K and P =
10 MPa, S = 0.0022 mm.

5.

THE PHENOMENON OF THE SPATIAL-PERIODIC MICROSTRUCTURES EXCITATION ON THE ENERGETIC MATERIAL


BURNING SURFACE AT THE INFLUENCE OF THE LASER
RADIATION ENERGY

The experimental data have shown that the SPMS develops on the EM burning surface
under influence of the laser radiation energy. Both the experiment, and the theory confirm,
that the SPMS formation is rather universal phenomenon.
In the researches was used radiation of the 2 - laser with power of 60 watt and with
wave length 10.6 micrometers. The time of influence of the laser radiation energy on the
gun-powder surface gradually increased up to the moment of ignition of the burning surface
of the sample. It has allowed to observe development of the micro-structures on the burning
surface in various phases of the process. Fig. 7 shows the image of the burning surface of the
standard ballistite propellant under influence of the laser radiation energy in the nitrogen
atmosphere at the pressure of 0.4 MPa.

Alexander Lukin

189

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 7. The structure of the surface of the sample after influence of laser radiation during 120
milliseconds.
Excitation and formation of the SPMS on the burning surface is accompanied by
separation of the LVL on the local zones with different viscosity and density. Local zones
with the increased viscosity are formed on the boundaries between cellular micro-structures.

The Carbon
Particle

Fig 8. Formation of the carbon micro-structures on the burning surface.


190

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In such zones with increased viscosity occurs formation of the carbon micro-structures
(Fig. 8). Actually, the carbon grid reflects one of sequential stages of development of the
cellular micro-structures on the EM burning surface. In this case, excitation of the thermoelectric convection in the layer is a main structures-forming factor. At further increase of the
duration of influence of the laser radiation energy on the surface of the sample, the carbon
structures will unite with each other and will cover all a greater and greater surface. Fig. 9
show gradual increase in quantity of carbon structures on the burning surface.

Fig 9. Increase in quantity of the carbon micro-structures on the EM burning surface.

Fig 10. Formation of the carbon micro-structures on the burning surface.


11

Alexander Lukin

191

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

When the carbon micro-structures covers the surface of the sample approximately on 60
%, transition from ignition to burning will be observed. Fig. 10 shows the threshold time
moment, after which the surface of the sample will be ignited.
The EM burning surface gradually becomes covered by a carbon grid. During formation
of the carbon grid on the burning surface the burning process is gradually reduced and
eventually the torch luminescence disappears. At this stage of the process the luminescence
of the carbon grid is observed only (Fig. 2, (2) ). Further, the combustion products flowing
from the burning surface throw away a carbon grid from the burning surface Fig. 11.

Carbon Grid

Fig 11. The process of throwing away of the carbon grid, formed at the burning surface
accompanied by extinction of the burning surfaces.
After throwing away of a carbon grid the luminescence of the torch above the burning
surface appears again. Formation of the carbon grid are determined by formation of the
cellular micro-structures on the EM burning surface.

6.

THE PROCESS OF SELF-ORGANIZING OF THE DYNAMIC


DISSIPATIVE SYNERGETIC SPATIAL-PERIODIC MICROSTRUCTURES INTO THE TORCH MACRO-STRUCTURES

On the EM burning surface is observed the process of self-organizing of the dynamic


dissipative synergetic SPMS into the torch macro-structures. In this connection, on the
burning surface the process of moving (wandering) of the torches is observed. Besides, the
192

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

torch structures on the burning surface can exist steadily in the event that they form a pair of
two torches.

The SPMS (Cells) of The


Thermo-Electric Convection

SELF-ORGANIZING
OF THE SPMS
FORMATION OF THE CARBON
MICRO-STRUCTURES

SELF- RGANIZING OF THE SPMS INTO


THE TORCH MACRO -STRUCTURES

7.

THE PHENOMENON OF THE END-BURNING SOLID


PROPELLANT CHARGES RHYTHMIC EXTINCTIONS

Within suggested mechanism becomes possible to give a new explanation of the


phenomenon of the end-burning solid propellant charges rhythmic extinctions. This
phenomenon is observed enough frequently both for small-sized (Fig. 12) and for large-sized
end-burning charges of the gas-generators and SPS (engines for control on a roll channel, the
nose cone engines, having diameter up to 0.4 m.). This anomaly of burning for the first time
was described in details by Dr. Valentin N. Fomenko in 1990 (the Federal Centre of Dualuse Technologies "SOYUZ", Dzerzhinsky, Moscow region, Russia), on the basis of use of
the results of large number of the fire stand tests. The phenomenon is typically observed only
for charges with insulation layer.
13

Alexander Lukin

193

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 12. Sectional view of the solid propellant gas-generator. [17]

Fig 13. Combustion products pressure developnent at the end-burning solid propellant
charges rhythmic extinctions. Curves 1-4 corresponds to the solid propellant rocket
motor charge, manufactured from the standard ballistite propellant NMF-2 220/0400. Curves 5 and 6 corresponds to the gas generator propellant charge,
manufactured from the standard ballistite propellant NDT-OM 60/0-500. Figure
shows curves at different ambient temperatures: 1, 4 293 K; 2, 5 223 K; 3,6 323
K; Curve 4 corresponds to the situation at flame propagation under the charge
armoring cover.
In the conditions of increased pressure, inflow of the inhibitor of burning from the
insulation layer on the end-burning surface will increase viscosity and density of the LVL
and will lead to increasing of sizes of the SPMS and the burning cells. In this situation the
quantity of the burning cells will decrease. Accordingly, more dense and steady carbon grid
on the burning surface will be formed (as in conditions of the low pressures). Periodical
throwing away of the carbon grid, formed at the burning surface causes extinction and the
subsequent flashes that is recorded on the pressure curve.

194

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

It is important to note, that in the majority of theoretical models of the low-frequency


non-acoustic instability phenomenon in the SPS, the processes of excitation of the SPMS on
the burning surface are not taken into account. Thus, suggested mechanism and theoretical
model opens possibilities for understanding the essence of the EM unstable burning
phenomena on the new qualitative level. Such an improved model will be a step towards
detailed interpretation of the large amount of experimental data on L* instability that is
available.

8.

CONTROLLING OF THE ABNORMAL PHYSICO-CHEMICAL


PROCESSES AT THE ENERGETIC MATERIALS BURNING

Actuality of the researches connected with controlling of the abnormal physico-chemical


processes at the EM burning is evidently demonstrated by continuous perfection of the
designing-technological development in the given scientific-technical area. Let us examine
some new Russian technologies intended for effective suppression of the unstable
combustion and abnormal modes of the SPS operation. These technologies are created on the
basis of traditional theoretical ideas about the EM combustion instability (Table 1).
Table 1. Technologies for the combustion instability suppression in the SPS.
Design Schemes and Brief Description
The SPS with acoustic
damper, manufactured
as the cylindrical twolevel chamber. Device
provides
effective
suppression
of
vibrating burning in
the wide frequencies
band. [18]
The SPS with largelengthened casing. The
engine design provides
effective suppression of
longitudinal
pressure
pulsations
(longitudinal
acoustic oscillations) with
low frequencies. [19]
The SPS contains main and additional
chambers, connected by subsonic nozzle. The
engine design provides effective suppression
and stabilization of the low and highfrequency combustion products oscillations at
the ignition system operation and at
appearance of the charge unstable burning. [20]

15

Alexander Lukin

195

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The SPS contains the acoustic cavity of given


sizes, and at the solid propellant charge end
faces are installed the elastic boots. The engine
design provides effective suppression of the
high- and low-frequency pressure oscillations at
appearance of the charge unstable burning. [21]
In the SPS, between the front bottom and
the engine casing the perforated thin
diaphragm is installed, and between the
diaphragm and the front bottom are
located the acoustic cavity. Oscillations
of the front diaphragm provide additional
suppression of the combustion products
acoustic oscillations. [22]

The SPS contains a sectional solid propellant


charge and the acoustic cavities, located
between the charge sections. The engine design provides stability of the internal ballistics
characteristics by suppression of the pressure
oscillations in the wide frequencies band, at
appearance of the charge unstable burning. [23]

Mechanical device installed in the


SPS charge channel: longitudinal
and cross perforated plates and
diaphragms. [5]

Method of ignition provides increased


reliability of the solid propellant rocket
motor with multigrain charge with large
relative elongation in wide temperature
range. [24]

Suppression of the combustion instability can use various mechanical devices installed in
the combustion chamber or in the charge channel: the resonant rods of various forms;
196

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

longitudinal and cross perforated plates and diaphragms; various screens. Other versions of
technologies for suppression of combustion instability are purposeful change or regulation of
form and geometrical characteristics of the combustion chamber; exact tuning of geometrical
characteristics of the chamber or its separation on several volumes; creation of resonant
acoustic cavities in the combustion chamber; installation of damping rings (diaphragms) and
also internal and external acoustic (gas dynamics) dampers (Helmholtz resonators). Besides,
in some cases for suppression of the combustion instability there may be used high-porosity
cellular materials in the form of disks installed in the combustion chamber. Such disks are
usually manufactured by means of powder metallurgy.
At the same time there are deficiencies of these technologies. The efficiency of the
mechanical damper decreases after burning out a portion of the propellant charge, in
connection with change of the cavity sizes, because the mechanical means of suppression of
unstable burning operates in a narrow frequencies band. Also, damper design elements
induce the energy losses due to braking of the combustion products flow. Use of mechanical
dampers leads to increase of the engine inert weight. In some cases there may be damper
destruction and ejection of its separate elements from the engine. Absorbers of resonant type
the acoustic dampers (Helmholtz resonators) operates effectively in a narrow frequencies
band and can induce transition of unstable mode to other frequency or lead to destabilization
of steady burning process in the engine.
The main source of instability is a solid propellant pulsating burning surface. However,
the existing technologies for suppression of the solid propellant combustion instability do
not take into account the influence of the synergetic micro-structures on the propellant
burning surface.
Further progress in this vital area may be connected with understanding of mechanisms
of formation and destruction of the synergetic micro-structures on the EM burning surface.
In the SPS of a new generation with high energy and mass characteristics, the problem
of prevention of development of combustion instability and anomalies of burning have
extreme importance. Solution of this problem can be provided with using of suggested
mechanism and with using of plasma physics technologies for control of the process of
excitation of the SPMS on the burning surface.
For control of the process of the SPMS excitation on the EM burning surface the effect
of acoustic cavitation in the LVL can be used. This effect can be organized in the LVL under
influence of the acoustic field with certain frequency.

9.

CONCLUSION

The low-frequency non-acoustic combustion instability phenomenon, the chuffing


phenomenon and the accompanying physico-chemical effects have received a new
explanation within the concept based on the data of optical visualization of the physicochemical processes on the EM burning surface. On the EM burning surface is observed the
process of self-organizing of the dynamic dissipative synergetic SPMS into the torch macrostructures.
Understanding of the mechanism of the torch combustion phenomenon of the solid
rocket propellants - one of the most complicated problem in the solid propulsion physics.

17

Alexander Lukin

197

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The torch micro-structures on the EM burning surface can be considered as independent


synergetic structures and can exist steadily in the event that they form a pair of two torches.
Excitation of the spatial-periodic micro-structures in the ionic fusion with thermoelectric properties on the burning surface is a main source of development of the synergetic
phenomena in the EM burning zone.
Within suggested mechanism becomes possible to give a new explanation of the
phenomenon of the end-burning solid propellant charges rhythmic extinctions.
The existing technologies for suppression of the solid propellant combustion instability
do not take into account the influence of the synergetic micro-structures on the propellant
burning surface. Further progress in this vital area may be connected with understanding of
mechanisms of formation and destruction of the synergetic micro-structures on the EM
burning surface.
Design community and quality assurance people should continue to give due importance
to the above problem of controlling low-frequency non-acoustic combustion instability of
the energetic materials in practical applications so that their guidelines will be realistic for
safety.
Acknowledgments:
Author would like express deepest gratitude for the extensive experimental data
provided by Prof. Valery S. Abrukov (Chief of the Chair of Thermo-Physics of the Chuvash
State University, Cheboksary, Russia), by Dr. Valery D. Kochakov (Associate Professor of
the Chair of Thermo-Physics of the Chuvash State University, Cheboksary, Russia), by Prof.
Vladimir N. Marshakov (Chief of the Group of Non-Stationary Combustion of the
N.N.Semenov Institute of Chemical Physics of the Russian Academy of Sciences, Moscow,
Russia) and by Prof. Nickolay M. Pivkin (Chief of the Department No. 012 of the ScientificResearch Institute of Polymeric Materials, Perm, Russia).

Nomenclature
a
C
Ea

FL
FST

FTE

g
h
k0

P
q0
198

- Thermal diffusivity of the liquid-viscous layer;


- Specific heat capacity of the condensed phase;
- Activation energy of the condensed phase reactions;
- Buoyancy (lifting) force (the difference between Archimedean force and the
force of gravity), (Rayleigh, 1916);
- Surface tension force (thermo-capillary force), (G.K.A. Pearson, 1958), (the
convection cells arising under influence of this mechanism is known as
Maringoni cells);
- Thermo-electric (electrostatic) force, Coulombs force, (I.V.Ioffe and
E.D.Eidelman, 1976);
- Acceleration due to gravity;
- Characteristic size of fluctuation movement (the thickness of the liquidviscous layer, cross-sectional size of the cell);
- Pre-exponential factor of chemical reaction rate;
- Pressure;
- Heat release distribution due to chemical reactions in the solid phase and
liquid-viscous layer of the cell;
Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

liquid-viscous layer of the cell;


- Heat flux issuing from the gaseous phase (flame) to the surface of the
burning cell;
- Gas constant;
- Time of the process;
- Temperature;
- Initial temperature;

Q0

R
t
T
T0

- Temperature of the cold boundary surface of the liquid-viscous layer;


- Temperature of the hot boundary surface of the liquid-viscous layer (on the
interface of liquid-viscous layer and gasification zone);
- Temperature of beginning of the effective exothermic reactions;

TC

Th

TR

- Temperature on the burning surface;


- Linear burning rate of the energetic material sample;
- Linear burning rate of the energetic material sample in the cell with number
(i, j);
- Space coordinate, perpendicular to the liquid-viscous layer;

TS
uC
u (Ci , j)

Greek Symbols

0
C

- Average thickness of the reaction liquid-viscous layer (the effective thickness


of the zone of electric conductivity);
- Temperature gradient;
- Conductivity;
- Density of the liquid-viscous layer, density of the condensed phase;

Subscripts and Superscripts


c
g
s
0

Condensed phase;
Gas phase;
Surface;
Parameters of equilibrium state;

REFERENCES
[1]

A.M. LIPANOV, A.N.LUKIN: Abnormal Physics-Chemical Processes in the Technical Systems,


Based on Combustion of the Solid Propellants and Gun-powders, The Concept of Development
of Combustion and Explosion as a Field of Scientific-Technical Progress, Ed. by Acad.
A.G.Merzhanov, Publ. House Territoriya, Chernogolovka, p. 124, 2001, (in Russian)

[2]

YA.B. ZEL'DOVICH: About Stability of Burning of Gun-powder in the Semi-Closed Cavity, J. of


Appl. Mech. and Tech. Phys., No. 1 , p. 67-76, 1963, (in Russian)

[3]

I.G.ASSOVSKII: The Critical Phenomena and Non-Stationary Burning of the Energetic


Materials, Dr. of Sci. Thesis (Phys.-Math., Spec. 01.04.17 Chem. Phys., incl. Physics of

19

Alexander Lukin

199

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Combustion and Explosion), N.N.Semenovs Institute of Chemical Physics of the RAS,


Moscow, Russia, p. 360, 2001, (in Russian)
[4]

R.A. YOUNT and T.A.ANGELUS: Chuffing and Non-acoustic Instability Phenomena in Solid
Propellant Rockets, AIAA Journal, vol. 2, No. 7, p. 1307-1313, 1964

[5]

N.M. PELYKH: Non-Stationary Burning of the Solid Propellant Charges and Use of This Effect
in the National Economy, Dr. of Sci. Thesis (Tech., Spec. 01.02.05 Mech. of the Liquid, Gas
and Plasma and 01.04.17 Chem. Phys., incl. Physics of Combustion and Explosion),
Scientific-Research Institute of Polymeric Materials, Perm, Russia, p. 244, 2003, (in Russian)

[6]

C.M. HUGGETT: Combustion Processes, High Speed Aerodynamics and Jep Propulsion,
Princeton University Press, Princeton, N.J., vol. II, Sec. 4, 1956

[7]

G.B. COOK: Some Developments in the Theory of Thermal Explosions, Sixth Symposium
(International) on Combustion, Reinhold Publishing Corp., New York, p. 621-631, 1957

[8]

V.S. ABRUKOV, S.V.ILYIN, V.M.MALTSEV: Interferometric Techniques and Other Optical


Methods in Combustion Research. A New Approach, Optical Techniques in Fluid, Thermal, and
Combustion Flow, San-Diego, USA, Ed. Soyong Stephen Cha and James D. Trolinger, Proc.
Of the SPIE's Int. Symp., vol. 2546, p. 420-426, 1995

[9]

V.S. ABRUKOV, S.V.ILYIN, V.M.MALTSEV, I.V.ANDREEV: Interferometric Technique in


Combustion, Gas Dynamic and Heat Transfer Research. New Results and Technologies. CDROM Proc. of the VSJ-SPIE98 Int. Conference on Optical Technologies and Image Processing
in Fluid, Thermal, and Combustion Flow, Yokohama, JAPAN, AB076, p. 13, 1998

[10] V.S. ABRUKOV, A.V. KHRISTOFOROV, V.YE. NIKONOROV, I.V. ANDREEV: Interferometric and
Shadow Technique in Analysis of Combustion Wave Shaping and Development During Ignition
of Solids, Optical Diagnostics for Fluids/Heat/Combustion and Photo-mechanics for Solids, Ed.
by Soyoung S. Cha, Peter J. Bryanston-Cross, Carolyn R. Mercer, Proc. Of the SPIE's Int.
Symp., USA, Denver, vol. 3783, p. 5, 1999
[11] V.T. KUZNETSOV, V.S. AVERIN: Ignition of Heterogeneous System in the Dead-end Channel
with Variable Cross-Section Area, Proceedings of the Fourth International Workshop on IntraChamber Processes, Combustion and Gas Dynamics of Dispersed Systems, (June 28 - July 3,
2004, Baltic State Technical University, Saint-Petrsburg, Russia), vol. 1, Baltic State Technical
University, Saint-Petrsburg, Russia, p. 52-54, 2004, (in Russian)
[12] YU. S. IVASHENKO, A. L. SADYRIN, and V. L. PAVLENKO: Investigation of the Electric
Conductivity of the Ammonium Perchlorate Burning Surface, J. Physics of Combustion and
Explosion, vol. 22, No. 4, p. 30-33, 1986, (in Russian)
[13] G. V. MELIK-GAIYKAZOV, V. N. MARSHAKOV: Investigation of the Electric Conductivity of
the Ammonium Perchlorate at Burning, J. Physics of Combustion and Explosion, vol. 23, No.
4, p. 15-20, 1987, (in Russian)
[14] G. V. MELIK-GAIYKAZOV: Experimental Investigations of the Gunpowder Non-Stationary
Burning Modes in the Semi-Closed Cavity, Ph.D. Thesis (Phys.-Math. Sci., Spec. 01.04.17
Chem. Phys., incl. Physics of Combustion and Explosion), N.N.Semionov Inst. of Chem. Phys.
RAS, Moscow, p. 147, 1998, (in Russian)

200

Mechanism of Low-Frequency Non-Acoustic Instability Phenomena From the . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

[15] A. N. LUKIN: Concept of the Spatial-Periodic Micro-Structures Excitation at the Evaporated


Energetic Materials Transient Combustion as a Way for Improvement of the Solid Propulsion
Technologies, in: Proc. 8th Int. Seminar New Trends in Research of Energetic Materials, Part
I, University of Pardubice, Czech Republic, p. 267-295, 2005
[16] D. A. FRANK-KAMENETSKII: Diffusion and Heat Transfer in Chemical Kinetics, Publishing
House "Nauka", Moscow, p. 502, 1987, (in Russian)
[17] A. M. LIPANOV, A.YU. LESCHEV: Ignition System for Gas-Generator, Pat. 2 239 474 of
Russia, A 62 C 13/22, Application 2002133993/02, Priority from December 17, 2002, (in
Russian)
[18] N. N. GLUKHAREV ET AL: The Solid Propellant Rocket Motor, Pat. 2 056 519 of Russia, F
02 K 9/08, Application 93035598/06, Priority from July 8, 1993, Registered at March 20,
1996, The Inventions Bulletin (Russia) No. 8, March 20, 1996, (in Russian).
[19] M. I. SOKOLOVSKII ET AL: The Solid Propellant Rocket Motor, Pat. 2 173 783 of Russia, F
02 K 9/26, Application 2000111590/06, Priority from May 10, 2000, Registered at
September 20, 2001, The Inventions Bulletin (Russia) No. 26, September 20, 2001, (in
Russian).
[20] O. G. BORISOV ET AL: The Solid Propellant Rocket Motor, Pat. 2 125 173 of Russia, F 02 K
9/08, Application 98103166/06, Priority from February 20, 1998, Registered at January 20,
1999, The Inventions Bulletin (Russia) No. 2, January 20, 1999, (in Russian).
[21] V. R. ALYAZHEDINOV ET AL: The Solid Propellant Rocket Motor, Pat. 2 147 342 of Russia,
F 02 K 9/26, Application 98119256/06, Priority from October 23, 1998, Registered at April
10, 2000, The Inventions Bulletin (Russia) No. 10, April 10, 2000, (in Russian).
[22] P. I. TRAPEZNIKOV ET AL: The Rocket, Pat. 2 170 910 of Russia, F 42 B 15/00, Application
2000103008/02, Priority from February 10, 2000, Registered at July 07, 2001, The
Inventions Bulletin (Russia) No. 20, July 07, 2001, (in Russian).

21

Alexander Lukin

201

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

STABLE SALTS OF METHYLATED 5-AMINOTETRAZOLES


Michael Gbel*, Konstantin Karaghiosoff*, Thomas M. Klaptke*, Carles Mir*, Jan
M. Welch*
*University of Munich, Butenandstr. 5-13 (D), D-81377 Munich, Germany

Abstract:
Several salts of 1-methyl-5-aminotetrazole (2), 2-methyl-5-aminotetrazole (3) and 1-4dimethyl-5-aminotetrazole (free base of 4) with a variety of anions: nitrate (5, 7, 10),
perchlorate (6, 8, 11), picrate (9,15), azide (12), dinitramide (13) and azotetrazolate (14) as
the dihydrated species have been successfully synthesized in reasonable to good yields and
characterized by analytical (elemental analysis and melting point) and spectroscopic (IR,
Raman and 13C-, 1H-, 15N-NMR) methods. The structures of all compounds have been
determined using X-ray diffraction techniques. All compounds were found to be stable at room
temperature and moisture and air stable. An insight into their sensitivity was gained by either
throwing them into a flame or by submitting them to standardized BAM tests for friction,
impact and electrostatics sensitivity. All perchlorates (6, 8, 11) and the dinitramide (13)
exploded, however the azide (12), surprisingly, and all other compounds subject of this study
remained mainly unmodified after the test.
Keywords:

1.

methylated 5-aminotetrazoles, high energy density material (HEDMs),


crystal structure, vibrational spectroscopy, NMR spectroscopy
(multinuclear)

INTRODUCTION

The synthesis of energetic, non-nuclear materials for possible military application has
been a long term goal in our research group [1,3]. In the last years there has been great
interest in investigating the chemistry of new high energy density materials (HEDMs) with a
high-nitrogen content and very high positive heats of formation (directly attributable to the
large number of inherently energetic N-N and C-N bonds) that show insensitivity to shock,
friction, electrostatic discharge [4] and thermal stimuli. These compounds combust ideally to
give only gaseous products (i.e. CO, CO2, and N2). Most of the energy obtained from
modern energetic materials comes either from oxidation of the carbon backbone (such is the
case for traditional energetic materials as for example HMX or TNT) [5] or from their high
positive heats of formation, as in for tetrazole azide [6].
Recently, the synthesis of new heterocyclic-based energetic, low melting salts were
reported [7]. These materials that are salt-based often possess advantages over nonionic
structures since they normally exhibit very low vapor pressures, eliminating, thus, the risk of
exposure through inhalation, and they generally have higher densities than the nonionic
parents [8].

202

Stable Salts of Methylated 5-Aminotetrazoles

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Salts containing a tetrazole derivative as the cation (which adds a high nitrogen-content
to the molecule) show high densities and interesting energetic properties when combined
with typical energetic anions such as nitrate in 5-aminotetrazolium nitrate [9] or dinitramide
(N(NO2)2) in 1,5-diamino-4-methyltetrazolium dinitramide [10]. Therefore, tetrazoles and
aminotetrazoles, in particular, are prospective materials for the generation of gases, as
blowing agents, solid propellants, and other combustible and thermally decomposing
systems. Based on these results 5-aminotetrazole 1 (with a 82.3% nitrogen-content) was
considered as a valuable starting material for the synthesis of nitrogen-rich HEDMs. The
relative basicity of 1 allows to prepare HEDMs by simple protonation of one of the ringnitrogens with a strong acid such as nitric [10] or perchloric acid [11]. Unfortunately, several
attempts to generate the interesting azide, dinitramide and azotetrazolate parents failed,
apparently due to the higher acidity of 5-aminotetrazole in comparison to the conjugated acid
of the above-mentioned anions. In order to overcome this difficulty and to make more stable
5-aminotetrazole derivatives, introduction of a small alkyl chain was thought to be of use.
Methylation of 1 according to literature procedure [12] with methyl iodide in acetone/water
afforded a mixture of 1-methyl-5-aminotetrazole 2 and 2-methyl-5-aminotetrazole 3 which
can be separated. Salts 5-9 were synthesised and again the methylated 5-aminotetrazoles
turned out to be too acidic to yield the azide, dinitramide or azotetrazolate salts. For this
reason, the quaternization of the tetrazole ring with great excess methyl iodide in acetonitrile
[13]
was carried out to yield dimethyl-5-aminotetrazolium iodide 4 of which a variety of salts
(10-15) with typical energetic anions were prepared by metathesis.
NH2

NH2

Pic-

PicHN

N
N

N
N

9
Picric acid
H2O reflux
NH2

NH2

NH
N

MeI

Acetone/H2O
reflux, 2h

MeI
CH3CN
reflux, 10h

N
N

HX

NH2

HN

AgY
or
AgN(NO2)2*2Py

NH2

N
N

HN

MeOH

NH2

N
N

3
HX

MeOH
reflux
NH2

i) AgClO4
ii) KPicrate
NH2

15

5. X = NO 3-

7. X = NO 3-

6. X = ClO4-

8. X = ClO4-

10. Y = NO 311. Y = ClO412. Y = N3

13. Y = (NO2)2N14. Y = [N4C-N=N-CN4]2-

Scheme 1. Reaction equations for the synthesis of the methylated 5-aminotetrazole salts

Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke, . . .

203

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

EXPERIMENTAL
2.1

Synthesis

Caution! Although no problems came up during the synthesis and handling of all
materials studied in this work, silver azide, silver dinitramide, aminotetrazolaes and their
derivatives are energetic materials and tend to explode under certain conditions. All
HEMDs should be treated with respect and appropriate safety precautions should be taken
at all times. Laboratories and personnel should be properly grounded and safety equipment
such as Kevlar gloves, leather coat, face shield and ear plugs are necessary, specially for
perchlorates 6, 8 and 11 as well as for dinitramide 13, in particular when working with big
amounts.
All chemicals and solvents (analytical grade) were used as supplied by Sigma-Aldrich
Fine Chemicals Inc.. 2-Methyl-5-amino-1H-tetrazole, 1-methyl-5-amino-1H-tetrazole [12],
silver azide and dinitramide [14] as well as disodium azotetrazolate [15] were prepared
according to literature procedure.

2.1.1.

2-Methyl-5-amino-1H-tetrazolium nitrate (5):

Method 1: 2-methyl-5-amino-1H-tetrazole (0.100 g, 1.0 mmol) was dissolved in the


minimum amount of concentrated nitric acid. The clear, colorless mixture was stirred until
all 1-methyl-5-amino-1H-tetrazole had dissolved. After cooling in the fridge overnight,
colourless plates were formed, filtered off cold and washed with diethyl ether until acid
free. No further purification was necessary (0.081 g, 49 %).
Method 2: Alternatively, 2-methyl-5-amino-1H-tetrazole (0.229 g, 2.3 mmol) was
dissolved in little concentrated nitric acid and 2 mL acetonitrile were added. The clear,
colorless mixture was stirred until 2-methyl-5-amino-1H-tetrazole had completely
dissolved and the product was precipitated by addition of 200 mL pentane. The yellow
powder could be washed with pentane and the out-coming white powder recrystallized
from methanol by diffusion of ether to yield clear colourless crystals, which were filtered
off and washed with ether. (0.337 g, 90 %). These crystals were suitable for X-ray
structure determination.
2.1.2

2-Methyl-5-amino-1H-tetrazolium perchlorate (6):

2-methyl-5-amino-1H-tetrazole (0.180 g, 1.8 mmol) was dissolved in little


concentrated perchloric acid. 1 mL ether and 0.5 mL acetone were added to the clear
yellowish solution and ether was diffused into it over 2 days. Little colorless prismic
crystals separated from the mother solution which allowed us to measure the crystal
structure. After filtering and washing with ether the compound required no further
purification (0.267 g, 70 %)

204

Stable Salts of Methylated 5-Aminotetrazoles

New Trends in Research of Energetic Materials

2.1.3

Czech Republic

2006

[Content]

1-Methyl-5-amino-1H-tetrazolium nitrate (7):

To 2 mL concentrated nitric acid were added 1.620 g (10.0 mmol) 1-methyl-5-amino1H-tetrazole. The clear, colorless reaction mixture was stirred until all 1-methyl-5-amino1H-tetrazole had dissolved. The solution was then cooled to 0 C overnight resulting in the
formation of clear colorless crystals, which were filtered off cold and washed until dry and
acid free with diethyl ether. No further purification was necessary and the crystals were
already suitable for X-ray analysis (1.206 g, 73 %).
2.1.4

1-Methyl-5-amino-1H-tetrazolium perchlorate (8):

1-Methyl-5-aminotetrazole (0.142 g, 1.8 mmol) was dissolved with little concentrated


perchloric acid. The mixture was layered with ether and cooled in the fridge overnight to
yield colorless plates which were filtered off and washed with ether. No further purification
was necessary. Alternatively, ether can be diffused into the mixture over 2 days yielding nice
colorless crystals suitable for X-ray analysis (0.204 g, 63 %).

2.1.5

1-Methyl-5-amino-1H-tetrazolium picrate (9):

A solution of 1-methyl-5-aminotetrazole (0.179 g, 1.8 mmol) in 2 mL hot water was


added to a solution of picric acid (0.398 g, 1.7 mmol) in 3 mL hot water. The mixture was
boiled together for a few minutes and let to cool down slowly to room temperature.
Immediately yellow prismic crystals suitable for X-ray structure determination precipitated.
The crystals were filtered off, washed with cold acetone and ether. A second crop could be
obtained from cooling down the mother liquors in the fridge overnight and filtering off cold.
This was treated like the first one. Both crops were pure by EA (0.521 g, 91 %).
2.1.6

1,4-Dimethyl-5-aminotetrazolium nitrate (10):

Silver nitrate (10.878 g, 64.0 mmol) was added to a suspension containing 1,4-dimethyl5-aminotetrazolium iodide (15.436 g, 64.0 mmol) in 50 mL methanol. Immediate
precipitation of yellow silver iodide was observed and the mixture was stirred for 2 hrs.
under exclusion of light. The solid was filtered off and the solvent reduced under high
vacuum to about 30 mL. The solution was divided into 10 equal portions into which diethyl
ether was allowed to diffuse over 2 days yielding clear prismic crystals the structure of
which was determined by single crystal X-ray analysis. No further purification was
necessary (5.912 g, 53 %).

Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke, . . .

205

New Trends in Research of Energetic Materials

2.1.7

Czech Republic

2006

[Content]

1,4-Dimethyl-5-aminotetrazolium perchlorate (11):

To 20 mL methanol were added 4.820 g (20.0 mmol) 1,4-dimethyl-5aminotetrazolium Iodide. The solution was stirred until all solid had dissolved. A second
solution containing 4.140 g anhydrous silver perchlorate weighed out in a dry-box
dissolved in 5 mL methanol was added to the solution of 1,4-dimethyl-5-aminotetrazolium
iodide. White silver iodide precipitated immediately and the temperature of the reaction
mixture rose near to the boiling point of methanol. After stirring for 30 minutes the
solution was filtered, divided into 12 equal portions and recrystallized by diffusion of ether
into the methanol solutions. The recrystallized material was used to measure the X-ray
structure in the solid state. No further purification was necessary (2.812 g, 71 %).
2.1.8

1,4-Dimethyl-5-aminotetrazolium azide (12):

Method 1: To a suspension of 0.200 g (2.5 mmol) potassium azide in 2 mL of


methanol were added 0.430 g (2.0 mmol) of 1,4-dimethyl-5-aminotetrazolium perchlorate.
The mixture (suspension) was covered and stirred for 3 hrs. After 3 hrs. the remaining
white solid was filtered off and the solution was divided into two equal portions. Diethyl
ether was allowed to diffuse into both solutions over 2 days, yielding clear prismic crystals.
No further purification was necessary (0.258 g, 66 %).
Method 2: Alternatively, to a solution of 0.240 g (1.0 mmol) of 1,4-dimethyl-5aminotetrazolium iodide in 3 mL methanol were added 0.200 g (1.33 mmol) silver azide.
This suspension was stirred for 1 hour and then filtered. The remaining solution was
divided into three 1 mL portions into which diethyl ether was allowed to diffuse over
several days yielding again clear prismic crystals suitable for X-ray crystal structure
determination. No further purification was necessary (0.122 g, 78 %).
2.1.9

1,4-Dimethyl-5-aminotetrazolium dinitramide (13):

Method 1: To a suspension of 0.220 g (1.5 mmol) potassium dinitramide in 2 mL


methanol were added 0.210 g (1.0 mmol) 1,4-dimethyl-5-aminotetrazolium perchlorate.
The suspension was stirred for 3 hours then filtered. The resulting clear solution was
divided into two equal portions and diethyl ether was allowed to diffuse into the methanol
solutions overnight yielding clear colorless crystals suitable for X-ray structure
determination (0.220 g, 67 %).
Method 2: Alternatively, to a suspension of 0.370 g (1.0 mmol) of silver bis-pyridine
dinitramide in 2 mL methanol were added 0.240 g (1.0 mmol) of 1,4-dimethyl-5aminotetrazolium iodide. The reaction was stirred under exclusion of light for 30 minutes,
the silver iodide was then removed by filtration and the remaining solution was divided
into two equal fractions and diethyl ether was allowed to diffuse into the solutions
overnight yielding clear, colorless needles (0.157 g, 71 %).

206

Stable Salts of Methylated 5-Aminotetrazoles

New Trends in Research of Energetic Materials

2.1.10

Czech Republic

2006

[Content]

1,4-Dimethyl-5-aminotetrazolium azotetrazolate*2H2O (14):

1,4-Dimethyl-5-aminotetrazolium iodide (0.378 g, 1.6 mmol) was dissolved in the


minimum amount of hot water and was added to a solution of disodium azotetrazolate (0.140
g, 0.7 mmol) in little hot water. The reaction mixture was refluxed for one hour and leave to
slowly cool down to room temperature. A feathery yellow powder precipitated on standing
in the oil bath and the yield could be increased by cooling down the solution in the fridge
(0.235 g, 78 %). X-ray structure determination from some of the material precipitated in the
fridge confirms the presence of two water molecules in the unit cell.
2.1.11

1,4-Dimethyl-5-aminotetrazolium picrate (15):

A fresh solution of 1,4-dimethyl-5-aminotetrazolium perchlorate prepared from reacting


over 1 hour 1,4-dimethyl-5-aminotetrazolium iodide (0.845 g, 3.5 mmol) with silver
perchlorate (0.730 g, 3.5 mmol) in 15 mL methanol was added (after filtering off yellow
silver iodide) to a suspension of orangey potassium picrate prepared from stirring during 30
min. a mixture of picric acid (0.803 g, 3.5 mmol) and potassium hydroxide (0.197 g, 3.5
mmol) in 15 mL methanol. The reaction mixture was stirred under reflux for 3 hours. Yellow
potassium perchlorate was filtered off hot and the reaction mixture left to cool down to room
temperature. Yellow needles precipitated immediately. The solvent was left overnight to
evaporate yielding a yellow crystalline material (0.894 g, 75 %) which could be used to
measure the crystal structure.

2.2

Crystal structures

Crystals were obtained as already mentioned above. X-ray quality crystals were mounted
on a glass fibre, and the X-ray crystallographic data were collected on a Oxford Xcalibur
diffractometer with a CCD area detector with graphite-monochromated Mo K radiation (
= 0.71073 ). The structures were solved by direct methods and refined by means of fullmatrix least-squares procedures using the SHELX-97 program. Fig. 1 to Fig. 4 show the
molecular structure of selected synthesized compounds in the crystalline state (DIAMOND
plot, 50 % probability):

Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke, . . .

207

New Trends in Research of Energetic Materials

2.2.1

Czech Republic

2006

[Content]

2-Methyl-5-amino-1H-tetrazolium nitrate (5):

Table 1. Crystallographic data for 5

2MATN

Formula
Form. weight (g mol-1)
Crystal system
Space-group
a ()
b ()
c ( )
()
()
()
Cell volume (3)
Calc. density (g cm-3)
Z
R1(F>4 (F))
wR2 (all data)

C2H6N6O3
162.11
monoclinic
P 21/n
6.340(1)
8.704(2)
12.223(2)
90
99.88(3)
90
664.6(3)
1.620
4
0.0514
0.1502

Fig. 1. Molecular structure for 5

2.2.2

1-Methyl-5-amino-1H-tetrazolium perchlorate (8):


Table 2. Crystallographic data for 8

1MATP

Formula
Form. weight (g mol-1)
Crystal system
Space-group
a ()
b ()
c ( )
()
()
()
Cell volume (3)
Calc. density (g cm-3)
Z
R1(F>4 (F))
wR2 (all data)

C2H6N5O4Cl
199.55
monoclinic
P 21/n
5.2741(8)
20.696(3)
7.149(1)
90
106.60(1)
90
747.8(3)
1.772
4
0.0535
0.1167

Fig. 2. Molecular structure for 8

208

Stable Salts of Methylated 5-Aminotetrazoles

New Trends in Research of Energetic Materials

2.2.3

Czech Republic

2006

[Content]

1,4-Dimethyl-5-aminotetrazolium dinitramide (13):

Table 3. Crystallographic data for 13

DMATDN

13

Formula
Form. weight (g mol-1)
Crystal system
Space-group
a ()
b ()
c ( )
()
()
()
Cell volume (3)
Calc. density (g cm-3)
Z
R1(F>4 (F))
wR2 (all data)

C3H8N8O4
220.15
orthorhombic
P 2 12 12 1
5.3985(6)
12.926(2
13.336(2)
90
90
90
930.5(2)
1.513
4
0.0622
0.1375

Fig. 3. Molecular structure for 13

2.2.4

1,4-Dimethyl-5-aminotetrazolium picrate (15):


Table 4. Crystallographic data for 15

DMATPic

15

Formula
Form. weight (g mol-1)
Crystal system
Space-group
a ()
b ()
c ( )
()
()
()
Cell volume (3)
Calc. density (g cm-3)
Z
R1(F>4 (F))
wR2 (all data)

C9H10N8O7
342.25
monoclinic
P 21/c
14.862(1)
5.7352(4)
16.839(1)
90
104.96(1)
90
1386.7(2)
1.639
4
0.0649
0.1475

Fig. 4. Molecular structure for 15

Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke, . . .

209

New Trends in Research of Energetic Materials

2.3

Czech Republic

2006

[Content]

Sensitivity data

The following table indicates the results obtained from submitting the studied
compounds to different stimuli. The figures in the boxes indicate the sensitivity obtained
from the minimum amount of weight and height that detonates a compound at least two
times out of six for the drop-hammer (impact sensitivity test, in Joules) and the minimum
force which needs to be applied to detonate a compound two out of four times for the
friction test (determined using a BAM friction tester [16], in Newtons). For the impact tests
about 30-40 mg of the sample were placed in a brass cup and hit with a weight (BAM
Fallhammer) in accordance to the BAM method [17]:
Table 5. Impact, friction, electrostatics and flame sensitivity for the studied compounds*
Impact
Friction
Spark
Flame

5
>30
>360
n.e.
D

6
n.m.
n.m.
n.e.
D

7
>30
>360
n.e.
D

8
2
10
n.e.
D

9
B
>360
n.e.
B

10
>30
>360
n.e.
B

11
6
96
n.e.
D

12
>30
>360
n.e.
B

13
5
>360
n.e.
D

14
n.m.
n.m.
n.e.
D

15
>30
>360
n.e.
B

*n.e. = no explosion, D = deflagration, B = the compound burned, n.m. = not measured

3.

RESULTS AND DISCUSSION

5-aminotetrazole salts are really interesting compounds for the synthesis of HEDMs
due to their high-nitrogen-content and to the fact that they usually exhibit good thermal
stability. In order to increase the nitrogen content and/or the oxygen balance of these salts,
highly energetic anions such as dinitramide, azide, azotetrazolate ([N4C-N=N-CN4]2-),
nitrate, perchlorate or picrate where selected as candidates. Unfortunately, for the first
three anions, from our experience, the acidity of a hydrogen attached to a 5-aminotetrazole
nitrogen makes it impossible to synthesize their 5-aminotetrazole salts unless
quaternization of the tetrazole ring is carried out. In this respect, the synthesis of these salts
is limited to the alkylated derivatives which increases considerably the stability of the salts
but it has the drawback that diminishes the nitrogen-content which would readily combine
to give molecular nitrogen under conditions which initiate the explosion. This relates, in
turn, with the performance of a HEDM.

210

Stable Salts of Methylated 5-Aminotetrazoles

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Several stable salts of the 5-aminotetrazole methylated derivatives (2-4) with a nitrogencontent varying between 34.4 % for tetrazole 9 and 71.8 % for 12 have been prepared in
reasonable to good yields by either protonation of the tetrazole ring or metathesis reactions.
Their sensitivity to different stimuli has been assessed by using the drop-hammer test, a
friction tester, initiating an electrical spark or throwing the compounds into a flame. Most of
them showed to be stable and, in consequence, safe to handle apart from perchlorates 8 and
11 (perchlorate 6 has not been yet tested) which with impact sensitivity values of 2 J and 6 J
turned out to be rather unstable. On the other side, the high sensitivity of 8 to friction (10 N)
is in contrast with a medium friction sensitivity for 11 of 96 N. This just puts into
perspective what would have been expected: the introduction of small alkyl groups on the
tetrazole ring increases the stability of the compounds to stimuli as well as the melting points
of the salts, which are 123 C and 220 C (for 8 and 11 respectively). Regardless of the wellknown explosive character of many azides, compound 12 showed to be really stable and
with a melting point of 186 C (well far-away from its decomposition temperature above 350
C) and a high-nitrogen content (71.8 %) could be considered as a candidate for further
studies. Dinitramide 13, which is quite impact sensitive (5 J) although friction unsensitive
(>360 N), showed much better friction stability than perchlorate 11 (96 N), regardless of the
similar impact sensitivity values (6 J for the perchlorate). The high-melting points of the
prepared salts, well-apart from its decomposition temperature (at least 50 C, only 16 C in
the case of nitrate 7), make the synthesised compounds prospective materials for further
studies.

4.

CONCLUSIONS

From this experimental study the following conclusions can be drawn:


(i)

An easy synthesis for several methylated 5-aminotetrazole salts was found


which yields the compounds in reasonable to good yields.

(ii)

The studied compounds were characterised analytically and by means of


vibrational spectroscopy (IR, Raman), NMR spectroscopy (13C, 1H, 15N) and by
single crystal X-ray diffraction

(iii) An insight into their sensitivity was gained by submitting them to standard
BAM tests.

10

Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke, . . .

211

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REFERENCES
[1] W.-H. WALTHER, T.M. KLAPTKE, AND G. HOLL: Physical Chemical Characterization
and Explosive Properties of Hydrazine Azide, in: Proceedings 29th International Annual
Conference of ICT, June 30 - July 3, 1998, Energetic Materials Production, Processing and
Characterization, Fraunhofer Institut fr Chemische Technologie: Karlsruhe, Germany, 1998,
pp. 134/1
[2] T. M. KLAPTKE, B. KRUMM, G.HOLL, AND M. KAISER: Synthesis, Characterization
and Quantum Chemical Computation of Hexanitrohexaaza-Isowurtzitane, in: Proceedings
30th Int. Annual Conference of ICT, June 30 July 3, 1999, Energetic Materials Modelling
of Phenomena, Experimental Characterization, Environmental Engineering, Fraunhofer
Institut fr Chemische Technologie: Karlsruhe, Germany, 1999, pp. 120/1
[3] T. HABEREDER, A. HAMMERL, G. HOLL, T. M. KLAPTKE, J. KNIZEK, P. MAYER,
AND H. NTH: Methylsubstituted Hydrazinium Azide Derivatives, in: Proceedings 31th Int.
Annual Conference of ICT, June 27 30, 2000, Energetic Materials Analysis, Diagnostics
and Testing, Fraunhofer Institut fr Chemische Technologie: Karlsruhe, Germany, 2000, pp.
150/1-150/7
[4] CHENGFENG YE, JI-CHANG XIAO, BRENDAN TWAMLEY AND JEANNE M.
SHREEVE. Energetic Salts of Azotetrazolate, Iminobis(5-tetrazolate) and 5,5bis(tetrazolate), Chem. Commun., 2005, 2750-2752
[5] J. KHLER AND R. MAYER: Explosivstoffe, 9th ed.; Wiley-VCH: Weinheim, 1998.
[6] A. HAMMERL, T. M. KLAPTKE AND M. WARCHHOLD: Synthesis, Structure,
Molecular Orbital and Valence Bond Calculations for Tetrazole Azide, CHN7, Propellants,
Explosives, Pyrotechnics 2003, 28(4), 163-173
[7] G. DRAKE, T. HAWKINS, A. BRAND, L. HALL, V. A. MCKAY AND I. ISMAIL:
Energetic, Low-Melting Salts of Simple Heterocycles, Propellants, Explosives, Pyrotechnics
2003, 28(4), 174-180. G. W. DRAKE: U. S. Patent 6,509,473, Jan. 21, 2003. G. DRAKE
AND T. HAWKINS: AFOSR Ionic Liquids Workshop, Dulles, VA, October 2002; Air Force
Office of Scientific Research: Arlington, VA. G. DRAKE: AFOSR Ionic Liquids Workshop,
Aberdeen, MD, September 2003; Air Force Office of Scientific Research: Arlington, VA.
[8] HONG XUE, YE GAO, BRENDAN TWAMLEY AND JEANNE M. SHREEVE: New
Energetic Salts Based on Nitrogen-Containing Heterocycles, Chem. Mater. 2005, 17, 191-198
[9] MORITZ VON DENFFER, GERHARD HEEB, THOMAS M. KLAPTKE, GERNOT
KRAMMER, GUNNAR SPIESS AND JAN M. WELCH: Improved Synthesis and X-Ray
Structure of 5-Aminotetrazolium Nitrate, Propellants, Explosvies, Pyrotechnics 2005, 30(3),
191-195
[10] THOMAS M. KLAPTKE, PETER MAYER, AXEL SCHULZ AND JAN J. WEIGAND:
1,5-Diamino-4-Methyltetrazolium Dinitramide, J. Am. Chem. Soc., 2005, 127, 2032-2033
[11] MORITZ VON DENFFER AND THOMAS M. KLAPTKE, unpublished
[12] R. A. HENRY AND W. G. FINNEGAN: Mono-alkylation of Sodium 5-Aminotetrazole in
Aquesous Medium, J. Am. Chem. Soc., 1954, 76, 923-926
[13] THOMAS M. KLAPTKE, KONSTANTIN KARAGHIOSOFF, PETER MAYER,
ALEXANDER PENGER AND JAN M. WELCH: Synthesis and Characterization of 1,4Dimethyl-5-Aminotetrazolium 5-Nitrotetrazolate, Propellants, Explosvies, Pyrotechnics, 2005,
in press

212

Stable Salts of Methylated 5-Aminotetrazoles

11

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

[14] HOW-GHEE ANG, WOLFGANG FRAENK, KONSTANTIN KARAGHIOSOFF, THOMAS


M. KLAPTKE, PETER MAYER, HEINRICH NTH, JOANNA SPROTT AND MARCUS
WARCHHOLD: Synthesis, Characterization and Crystal Structures of Cu, Ag and Pd
Dinitramide salts, Z. Anorg. Allg. Chem., 2002, 628, 2894-2900
[15]
JOHANNES THIELE UND J. T. MARAIS: Tetrazolderivate aus Diazotetrazotsure, Justus
Liebigs Ann. Chem., 1893, 273, 144-160
[16]
Reichel & Partner GmbH, http://www.reichel-partner.de/
[17]
Test methods according to the UN Recommendations on the Transport of Dangerous Goods,
Manual of Tests and Criteria, 4th rev. ed.; United Nations Publication: New York, 2003

Acknowledgments:
Financial support of this work by the University of Munich (LMU),
the Fonds der Chemischen Industrie and the European Research Office (ERO)
of
the
U.S.
Army
Research
Laboratory
(ARL)
under
contract
no.
N 62558-05-C-0027 is gratefully acknowledged.

12

Carles Mir, Michael Gbel, Konstantin Karaghiosoff, Thomas Klaptke, . . .

213

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

APPLICABILITY OF DYNAMIC MECHANICAL AND THERMAL


METHODS IN INVESTIGATION OF AGEING PROCESSES OF DOUBLE
BASED ROCKET PROPELLANTS
Sanja Matei Muani*, Muhamed Sueska*, Sanko Bakija **
*Brodarski Institute, Av. V. Holjevca 20, 10 020 Zagreb, Croatia
**Ministry of Defence, Bauerova 33, 10000 Zagreb, Croatia

Abstract:
Ageing of double based rocket propellants (DBR), as a consequence of chemical
decomposition reaction, as well as physical processes (migration of low molecular
constituents: plasticizers, etc.) has significant influence on their relevant properties (e.g.
decrease of molar mass of nitrocellulose, decrease of stabiliser content, decrease of
specimen mass, change of mechanical and thermal properties, etc.).
The change of relevant properties of DBR propellants limits their safe and reliable service
life. Even more, under certain conditions decomposition of a DBR propellant may become
autocatalytic, which can lead to self-ignition of the DBR propellant. Because of that, it is
very important to find out reliable methods for determination of propellant stability at a
given moment of time, as well as to predict self-ignition probability under given conditions.
In this work we studied dynamic mechanical and thermal properties of DBR propellants
artificially aged for different period of time at 90C, in order to detect and quantify changes
in dynamic mechanical and thermal properties that can be used in the propellants stability
assessment. Dynamic mechanical properties were studied by dynamic mechanical analysis
(DMA), while thermal properties were studied using differential scanning calorimetry
(DSC) and thermogravimetry (TGA).
The obtained results showed that the ageing caused significant changes of DMA, DSC and
TGA curves shape. The changes are quantified by following of some characteristic points
on DMA, DSC and TGA curves. It was found out that the most sensitive
parameters/characteristic points to the ageing process at DMA curves are: storage modulus
at 25 C, storage modulus at softening region, peak width and height on loss modulus curve,
glass transition temperature, loss modulus at viscoelastic region, tan maximum at
viscoelastic region, while the most sensitive parameters at DSC and TGA curves are: peak
height and width, heat of reaction, max. heat flow rate, mass loss at a given time, and rate
of mass loss at a given time.
Keywords:
ageing, DBR propellant, dynamic mechanical analysis, differential
scanning calorimerty, mechanical properties, thermal properties,
termogravimetry

1.

INTRODUCTION

Double based rocket propellants, as well as other materials and living organisms, are subject
to the ageing processes. There are a number of parameters that can influence degradation of

214

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

relevant material properties, such as temperature, time, moisture, light, atmospheric conditions,
stresses produced during production and use of materials, etc.
Nitrocellulose and other nitrate ester (nitro-glycerine, etc.), which are the main ingredients
of the DBR propellants, because of relatively low bound energies (155-163 kJ/mol), and a low
value of activation energy (120-190 kJ/mol) are subject to a slow chemical decomposition even
at the room temperatures [2, 4-9].
The thermal decomposition of nitrocellulose and nitro-glycerine starts with the homolytic
breakdown of the O-NO2 bond of the aliphatic nitrate esters, thus forming nitrogen dioxide and
corresponding alkoxyl radical. The released NO2 radical immediately undergoes consecutive
reactions with other decomposition products or with other propellant ingredients. During this
process NO2 is reduced to NO, N2O, N, HNO2 and HNO3.
Another main decomposition pathway is the neutral to acid hydrolysis of the nitrate esters.
The reaction is catalysed by moisture and residual acids, or by water and acids formed during
decomposition process [10].
The summary reaction of the thermal decomposition is autocatalytic, and accompanied by
the heat generation. The heat released, due to very low propellant conductivity, can accumulate
in the propellant grain, and under certain conditions can lead to the propellant self-ignition [3,
11].
Apart from chemical ageing, DBR propellants are subjected to ageing due to physical
processes such as migration of low molecular constituents (e.g. nitro-glycerine), or crack
formation/propagation that can be initiated by residual stress at rocket grain [4,12] Ageing due
to chemical reactions, as well as due to physical processes, changes their relevant properties
(e.g. nitrocellulose mean molar mass, stabiliser content, specimen mass, mechanical and thermal
properties, ballistic properties, etc.). During use, these changes can result in dangerous failures,
such as explosion of rocket motors. Therefore, it is necessary to find out reliable methods for
determination of propellant stability at a given moment of time, as well as to predict selfignition probability under given conditions.
By now, a number of methods have been used for determination of propellant stability at a
given moment of time. In order to predict propellant stability during some period of time
(service life time), it is necessary to find out the processes which have greatest influence on the
ageing of the investigated propellant, to determine their rates and to quantify them as accurately
as possible [1].
Some quantitative methods based on the measurements of propellants relevant parameters
(such as stabiliser content, decrease of mean molar mass of nitrocellulose, specimen mass loss,
gas formation, heat generation, etc.) can be used to predict the safe service life time [10, 13-15].
In this work we studied dynamic mechanical and thermal behaviour of artificially aged
DBR propellants by dynamic mechanical analysis (DMA), differential scanning calorimetry
(DSC) and thermogravimetry (TGA) in order to check applicability of these methods in
propellants stability studies. The results of our previous investigation of these topics are reported
in papers [1-2, 16-18].

Sanja Muani, Muhamed Sueska, Sanko Baka

215

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

EXPERIMENTAL

The study was performed on the double based rocket propellant containing ~54% of
nitrocellulose, ~35% of nitro-glycerine, ~3% of dietilftalate, ~3% centralite, and ~5% of other
additives. In order to quantify changes of mechanical and thermal properties of the propellant,
caused by the ageing processes, the samples of artificially aged propellant were tested by DMA,
DSC, and TGA analysis.
The propellant was cut up into rectangular shape of the following dimensions: 50 x 10 x 2.5
mm, and then subjected to artificial ageing in glass tubes at 90 C (single-temperature aging).
The samples for DMA, DSC and TGA measurements were taken periodically (table 2-4, 6-7),
while the total ageing time was 82 days.
The dynamic mechanical measurements (DMA) were carried out using TA Instruments
DMA, Model 983. The DMA measurements were carried out using samples of the rectangular
bar shape (50 x 10 x 2.5 mm), while the measuring conditions were:
heating rate:
frequency of an oscillatory load:
amplitude of deformation:
length to thickness ratio (L/T):
temperature range:

2 C/min.
1 Hz
0.2 mm
10
-120C to +80C

The differential scanning calorimetry (DSC) measurements were carried out using TA
Instruments DSC, Model 2910. The DSC measurements were carried out using samples
weighing 1.0 0.2 mg. The samples were tested in aluminium sample pans covered by
perforated cap, with a heating rate of 5 C/min, and under nitrogen atmosphere with flow rate of
50 ml/min.
Isothermal thermogravimetry (TGA) measurements were carried out using TA Instruments
SDT, Model 2960. The TGA measurements were carried out using samples weighing 1.0 0.2
mg. The samples were tested in aluminium sample pans, at 100 C, and under nitrogen
atmosphere with a flow rate of 50 ml/min.
The degree of change of some parameter was calculated using the following equation:
yP =

( P0 Pt )
100
P0

(1)

where: yP degree of change of some parameter, P0 parameter value at the beginning of


the ageing, and Pt parameter value after some time of ageing
The ageing times at 90 C (t90C) were, with assumption that ageing mechanism at 90 and
30C are the same, transformed into corresponding ageing times at 30 C (t30C), in accordance
with the following equation [1]:

t 30 0 C = t 90 0C a10

[(90 30 ) / 10 ]

(2)

where reaction rate accelerating factor (10) is taken to be equal 3.

216

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

3.

Czech Republic

2006

[Content]

RESULTS AND DISCUSSION

Samples of non aged and aged double based rocket propellants were subjected to DMA,
non-isothermal DSC and isothermal TGA tests. For this preliminary investigation, because of
the limitation of the experiment duration time, the selected ageing temperature was 90 C,
which is much higher than the normal storage temperature.

3.1.

Dynamic mechanical properties

Change of storage modulus with ageing


Storage modulus vs. temperature curves of a non aged DBR propellant and the DBR
propellants aged for different period of time at 90 C are given at Fig. 1.

10000





8000 

Flex Storage Modulus (MPa)





6000










 




 



 



 









 



  
 
 
 
 


 


 





 







4000














0
-120

-100

-80

-60

-40







 



2000

Non - aged
Aged at 90C for 2 days
Aged at 90C for 4 days
Aged at 90C for 10 days
Aged at 90C for 25 days
Aged at 90C for 45 days
Aged at 90C for 67 days
Aged at 90C for 82 days

-20











 
 


20

Temperature (C)







 



 



 

40

60

80

100

Universal V3.8B TA Instruments

Fig. 1. Storage modulus curves of a non-aged propellant and the propellants aged for different
period of time at 90 C as a function of temperature
From Fig. 1. is evident that E-T curve was shifted with the ageing time to higher
temperatures, as well as to higher values of modulus. An increase of the storage modulus is
indicator of reduced flexibility of nitrocellulose macromolecules. Reduced flexibility of the
macromolecular chain, in this case is a result of reducing amount of the energetic plasticizer
(nitro-glycerine). The plasticizer migration from the propellant grain interior to the surface, and
their vaporization, causes shortening of distances between macromolecules, and an increase of
intermolecular forces. These processes have significant effect on flexibility of the nitrocellulose
macromolecule units.
In order to quantify changes in the storage modulus, caused by artificial ageing of the DBR
propellant, several characteristic points on E-T curves are analysed, Fig. 2, Tab. 1, 2.

Sanja Muani, Muhamed Sueska, Sanko Baka

217

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig. 2. Typical E-T curve of DBR propellant, along with derivative ET curve
Table 1. Some characteristic points/parameters on DMA curves
No.

Characteristic points on E-T curves


o

Denotation

Unit

E (-115C) MPa
MPa
E(25C)

1.
2.
3.
4.
5.
6.

storage modulus at -115 C,


storage modulus at 25 oC
extrapo. onset temp. at the beginning of transition from glassy to viscoela. stage
extrapo. onset temp. at the end of transition from glassy to viscous stage
storage modulus at max. decreasing rate of storage modulus in softening region
temperature at maximum decreasing rate of storage modulus in softening region

T(EON1)
T(EON2)
E1
T(E1)

C
C
MPa
C

No.

Characteristic points on E-T curves

Denotation

Unit

loss modulus at -115 C


loss modulus at 25 oC
peak width at half height
peak height
glass transition temperature (peak maximum temperature)
loss modulus at viscoelastic region (close to softening point)
extrapo. onset temp. at begining of transition from viscoelastic to viscous stage
extrapo. onset temp. at the end of transition from viscoelastic to viscous stage

E(-115C)
E(25C)
Wpeak
hpeak
Tg
E1
T(EON1)
T(EON2)

MPa
MPa
C
MPa
C
MPa
C
C

No.

Characteristic points on tan -T curves

Denotation

Unit

1.
2.
3.
4.
5.
6.
7.

tan at -115 oC
extrapolated onset temperature at begin of glass transition
tan maximum in viscoelastic region
extrapolated onset temperature at end of glass transition
tan at -115 oC
extrapo. onset temp. at begining of transition from viscoelastic to viscous stage
extrapo. onset temp. at the end of transition from viscoelastic to viscous stage

tan (-115 C)

1.
2.
3.
4.
5.
6.
7.
8.

218

T(tan ON1)

tanmax
T(tanON2)
tan (25C)
T(tanON3)
T(tanON4)

C
C
C
C

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In Tabs. 2-4 and 6-7, the values of the characteristic points at the DMA, DSC and TGA
curves according to time of ageing at 90 C are shown. Also, it is specified degree of change,
and measurement uncertainty for every characteristic point.

Table 2. Values of some characteristic points on E-T curves of DRB propellants aged for
different period of time at 90C
Ageing time
at 90 C,
days

Corresponding
ageing time at
30C, years**

E (-115C)
MPa

T(EON1)
C

T(EON2)
C

E(25C)
MPa

E1
MPa

T(E1)
C

Non aged
2
4
10
25
45
67
82

Non aged
4
8
20
50
90
134
164

8248
8063
7997
8961
8043
7655
7327
7540

-67.18
-69.48
-67.99
-68.13
-67,65
-71.27
-77.07
-91.37

-16.43
-15.67
-16.46
-16.40
-14.84
-6.60
-

1213
1097
1048
1194
1181
1580
2260
3044

572.4
582.6
493.5
513.5
621.1
790.1
1248.0
1438.0

42.42
39.90
42.50
43.70
41.80
43.70
47.20
60.70

8.59

36.00

105.48

132.28

151.22

34.89

8249 369

-67 1.51

-16.43 0.60

1218 97.8

572,4 26.8

42.42 1.79

Degree of change (after 82 days


of ageing at 90C) / % *
Mean value experimental
standard deviation

E at maximum decreasing rate of storage


modulus in softening region, E1, MPa

7000

Extrapolated onset temperature at the end of


Storage modulus at 25C
the transition from glassy to viscous stage,
(E(25
C)),
MPa
(T(E
ON2)),C

*- Calculated according to Eq. 1, **- Calculated according to Eq. 2

3500

6000

3000
peak ve

5000
4000

2500
2000

3000

1500

2000

1000

1000

500

E1
E(25C)

0
0

20

40

60

80

0
100

0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
0

Ageing time at 90C, days

20

40

60

80

100

Ageing tine at 90C, days

Fig. 3. Changes of some characteristic points on E-T curves of DBR propellants aged for
different period of time at 90 C
From Tab. 2 and Fig. 3 it is evident that some of characteristic points at E-T curve
(E(25C), T(EON2, E1) are change significantly with the time of ageing. For example, degree
of change of storage modulus at the maximum decreasing rate in the softening region, E1, after
82 days of ageing at 90 C is more than 150%.
The changes of storage modulus at the maximum decreasing rate in the softening region,
E1, becomes higher than measurement uncertainty limits (Tab. 2) after 10 days of ageing at 90
C (i.e. after ~20 years of ageing at 30 oC), while the changes of other parameters become

Sanja Muani, Muhamed Sueska, Sanko Baka

219

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

higher than measuring uncertainty limits after 25 days of ageing at 90oC (i.e. after ~50 years of
ageing at 30 oC). This means that, on the basis of the E-T curve measurements, a reliable
conclusion on DBR propellant stability, i.e. its age, may be set out only after this period of
ageing.
The accelerated change of all characteristic points at the E-T curves begins after 45 days of
ageing at 90 C. This is related to the intensive degradation of nitrocellulose chain.

Change of loss modulus with ageing


From the E-T curves of DBR propellants aged for different period of time at 90C (Fig. 4)
is evident that ageing causes significant quantitative and qualitative changes on the E-T curves.
For example, peak height at the glass transition region decrease (from ~ 420 MPa for non-aged
propellant to ~180 MPa for propellant aged 82 days at 90 C), peak width increases, and the
complete E-T curve is shifted to higher temperatures.
Very intensive changes on the E-T curve begin after 45 days of ageing at 90 C. After that
time of ageing, begins intensive decrease of peak height, and very intensive increase of the peak
width, that after 67 days of ageing even resulted with the splits of the main peak into two
separate peaks.

500

400

Flex Loss Modulus (MPa)














 
 




 


300

200

100 







0
-120

Non - aged
Aged at 90C for 2 days
Aged at 90C for 4 days
Aged at 90C for 10 days
Aged at 90C for 25 days
Aged at 90C for 45 days
Aged at 90C for 67 days
Aged at 90C for 82 days









 




















 



 









 
 



 







 




 
 











  







 





 

  





 


 
 


 








   
  
 

-100

-80

-60

-40

-20

Temperature (C)

20

40

60

80

100

Universal V3.8B TA Instruments

Fig. 4. Loss modulus curves of non-aged propellant and the propellants aged for different period
of time at 90 C as a function of temperature
From Fig. 4 and Tab. 3 is visible that the temperature of the peak maximum at the glass
transition region, Tg, shifts to the lower temperatures, while at the same time, the temperature of
the peak maximum at the viscoelastic region increases.
Increase of the peak width and presence of two peaks at the E-T curves, after 67 days of
ageing at 90 C, indicates an increase in the heterogeneity of system. This is probably a
consequence of the very intensive decomposition reaction of low - molecular components of the
DBR propellant (nitro-glycerine, stabiliser, etc.), as well as the splits of the nitrocellulose
macromolecules.

220

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In order to quantify these changes in the loss modulus, caused by artificial ageing of DBR
propellant at 90 C, several characteristic points on the E-T curves were analysed, Fig. 5, Tab.
1, 3.

Fig. 5.Typical E-T curve of DBR propellant, along with derivative E-T curve
Table 3. Values of some characteristic points on E-T curves of DRB propellant aged for
different period of time at 90 C
Ageing time
at 90 C,
days

Corresponding
ageing time at
30C, years**

E(-115C)
MPa

Wpeak
C

Tg
C

hpeak
MPa

E(25C)
MPa

E1
MPa

T(EON1)
C

T(EON2)
C

Non aged
2
4
10
25
45

Non aged
4
8
20
50
90

64.0
63.4
63.7
70.6
78.3
76.5

59.28
57.45
57.16
57.84
61.09
78.52

111.7
112.1
112.1
120.2
127.5
155.2

44.76
43.68
44.56
45.08
44.91
46.51

62.90
63.49
62.68
63.61
64.84
68.83

134

102.8

140.68

180.9

180.7

54.26

78.92

82

164

126.2

166.8

421.68
401.30
399.40
431.00
378.20
329.10
229.0
161.7*
179.9
115.2*

123.4
126.1
130.2
136.0
147.1
162.7

67

-41.98
-43.12
-41.62
-42.00
-39.90
-41.79
-45.59
-11.50*
-51.89
4.70*

223.0

213.7

69.53

91.54

Degree of change (after 82 days


of ageing at 90C) / % *

97.13

181.38

30.59

72.68

58.39

91.35

45.64

36.02

Mean value experimental


standard deviation

64.02
6.7

59.28
1.12

-41.98
0.72

421.68
18.47

123.4
6,7

111.7
5.1

44.76

62.90

1.80

2.05

*- Calculated according to Eq. 1, **- Calculated according to Eq. 2

It is evident from Tab. 3 and Fig. 6 that characteristic values at the E-T curves (loss
modulus at -115 oC, peak width at half height, peak height, loss modulus at viscoelastic
region, etc.) are sensitive to ageing processes. The changes of majority of parameters become
higher than measuring uncertainty limits after 10 days of ageing at 90 oC (i.e. after ~20 years of
ageing at 30 oC).

Sanja Muani, Muhamed Sueska, Sanko Baka

221

New Trends in Research of Energetic Materials

450

400
350

peak ve

-10
-20
-30
-40

300
250
200
150
100

-50

50

-60

20

40

350
300
250
200
150
100
hpeak

160
140
120
100
80
60
40

peak ve

20

50
Wpeak

0
100

20
100

40
60
80
Aging time at 90C, days

Ageing time at 90C, days

[Content]

180

80 0

60

400

2006

Peak width (W peak ), C

500

450

Loss modulus at viscoelastic region


(E1), MPa

500

Loss modulus at -115C, MPa

Tg
E1
E (-115C)

10

Peak height (hpeak ), MPa

Temperature of glass transition


(Tg), C

20

Czech Republic

Fig. 6. Changes of some characteristic points on E-T curves of DBR propellant aged different
period of time at 90C
From Tab. 3 and Fig. 6 can be seen that a significant change of characteristic parameters
begins after 45 days of ageing at 90 C. The change of the DBR propellants properties after 45
days of ageing was especially influenced at characteristic parameters related with softening
point (T(EON1), T(EON2)), which before that time of ageing were almost constant. Intensive
increase of softening point after 45 days of ageing is probably the result of increased
intermolecular forces, as a consequence of decreased amount of plasticizer (nitro-glycerine) and
increased heterogenity of the system (nitrocellulose chain breakdown).

Change of tan with ageing


Tan vs. temperature curves of a non-aged the DBR propellant and DBR propellants aged
for different period of time at 90 C are given in Fig. 7.

0.6









Non - aged
Aged at 90C for 2 days
Aged at 90C for 4 days
Aged at 90C for 10 days
Aged at 90C for 25 days
Aged at 90C for 45 days
Aged at 90C for 67 days
Aged at 90C for 82 days

0.4





Tan Delta














0.2









0.0 
-120






 








  




-100

-80


 
  
 

 














  




 












 






-60

-40

-20

Temperature (C)

20










 












40

60

80

100

Universal V3.8B TA Instruments

Fig. 7. Tan -T curves of non-aged DBR propellant aged for different period of time at 90 C
From the tanT curves of DBR propellants aged for different period of time at 90C,
shown in Fig. 7, is obvious that the shape of tan -T curves changes considerably with the

222

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

ageing. At temperatures below the glass transition region, the value of tan remains almost
unchanged, while in the glass transition region, viscoelastic region, and in the softening
temperature region the value of tan decreases significantly. From Fig. 7, it can be seen that
with the ageing time the tan- T curve shifts to higher temperatures. All these changes,
especially increased softening point, are the indication that by the ageing the amount of
plasticizer in the propellant decreases.
In order to quantify these changes in the tan, several characteristic points on tan-T curves
were analysed, Fig. 8, Tab. 1, 4.

Fig. 8. Typical tan -T curve of DBR propellant, along with derivative tan-T curve
Table 4. Values of some characteristic points on tan -T curves of DRB propellant aged for
different period of time at 90 C
Ageing
time at 90
C, days

Corresponding
ageing time at
30C, years**

Non aged
2
4
10
25
45
67
82

Non aged
4
8
20
50
90
134
164

Degree of change (after 82


days of ageing at 90C) / % *
Mean value experimental
standard deviation

tan

T(tan

T(tan

(-115C)

ON1)

ON2)

MPa
0.0078
0.0079
0.0080
0.0079
0.0097
0.0103
0.0139
0.0167

C
-76.90
-75.09
-76.37
-76.38
-75.34
-78.43
-94.46
-98.21

C
-36.38
-36.87
-36.89
-36.12
-35.29
-35.10
-45.47
-56.67

113.74

27.72

104.7

0.0078
-76.90
-36.38
0.001
3.04
1.09
*- Calculated according to Eq. 1, **- Calculated according to Eq. 2

tan max
MPa
0.1074
0.1059
0.1074
0.1035
0.1024
0.0909
0.0604
0.0481

tan
(25 C)

MPa
0.1164
0.1150
0.1242
0.1139
0.1246
0.1030
0.0800
0.0733

T(tan ON3)
C
35.50
32.99
34.22
35.28
35.45
36.61
43.18
55.15

T
(tan ON4)

C
57.32
55.22
54.71
55.29
56.41
59.19
70.16
89.16

55.21

24.71

43.25

59.19

0.1074
0.0025

0.1164
0.0145

35.50
1.98

57.32
2.52

Although, changes of the characteristic parameters at the tanT curves are evident
(Figs. 8 and 9, Tab. 4), for the majority of the parameters, they still remain within measurement

10

Sanja Muani, Muhamed Sueska, Sanko Baka

223

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

uncertainty up to 45 days of ageing at 90 C (i.e. ~ 90 years of ageing at 30 C). Exceptions is


aximum in the viscoelastic region (tanmax), which seems to be very sensitive to
the tan m
ageing. The values of the tan maximum in viscoelastic region, reaches the measurement
uncertainty limits after 10 days of ageing at 90 C (i.e. after ~20 years of ageing at 30 oC), so
that parameters should be considered to one of the parameters that can be used for prediction of
the DBR propellant life time.

0,08
0,06
0,04
0,02
0
0

20

0,019

0,017

-10

0,015

peak ve

0,013

at -115C

0,1

10

-30

0,011

Tan

Extrapolated onset temperature at end of


glass transition T(tan ON2), C

Tan maximum in viscoelastic


region, tan max

0,12

-40

0,009

-20

peak ve

-50

tand max

0,007

tand (-115C)

-60

0,005
400
6020
80 40 100 60
Ageing time at 90C, daysAgeing time at 90C, days

80

100

Fig. 9. Changes of some characteristic points on tan-T curves of DBR propellant aged for
different period of time at 90 C

3.2.

Thermal properties

Change of non-isothermal DSC curves with ageing


Non-isothermal DSC curves of DBR propellants aged for different period of time at 90 C
are given in Fig. 10.
6

Heat Flow (W/g)

Non-aged
Aged at 90C for 4 days
Aged at 90C for 10 days
Aged at 90C for 45 days
Aged at 90C for 67 days
Aged at 90C for 72 days
Aged at 90C for 82 days















 




  




 
























 








 









 






















 











 










 








 









 









 

 






 







 






              


 


 



      
 

 


 
 

 

 

 

 

 



 




 





 






 



 


 
 


 




























 



-1
100
Exo Up







125

150

175

200

Temperature (C)

225

250

275

300

Universal V3.8B TA Instrum ents

Fig. 10. Non-isothermal DSC curves of DBR propellants aged for different period of time at 90
C

224

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig. 11. Typical non-isothermal DSC curve of DBR propellant, along with derivative heat flow
curve and heat flow curve running integral (heating rate 5 oC/min)
Table 5. Some characteristic points/parameters on DSC and TGA curves

12

No.

Characteristic points on DSC curves

Denotation

Unit

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

extrapolated peak onset temperature


peak maximum temperature
extrapolated peak end-set temperature
heat of reaction
peak height
peak width at half-height
conversion at maximum heat flow rate
conversion at peak maximum
maximum heat flow rate
maximum actual heating rate

Tie
Tp
Tfe
Hr
hp
wp
(d/dt)max
(Tp)
(d/dt)max

max

C/min

No.

Characteristic points on TGA curves

Denotation

Unit

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

maximum decomposition rate


time to reach maximum decomposition rate
degree of decomposition at maximal decomposition rate
time to reach 10% sample mass loss
time to reach 15% sample mass loss
time to reach 20% sample mass loss
mass loss after 60 min
mass loss after 200 min
rate of decomposition after 10 min
rate of decomposition after 20 min

(dm/dt)max
t [(dm/dt)max]
m [(dm/dt)max]
t (10 %)
t (15 %)
t (20 %)
m (60 min)
m (200 min)
(dm/dt)10 min
(dm/dt)20 min

%/min

Sanja Muani, Muhamed Sueska, Sanko Baka

J/g
W/g
o

%
%
W/(g min)

min
%
min
min
min
%
%
%/min
%/min

225

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

From Fig. 10, is clear that ageing has caused significant changes of the DSC curves shape
and position. For example, DSC peak height increases, peak width decreases, peak onset
temperature increases, heat of reaction increases, maximum heat flow rate increases, etc. In
order to quantify these changes in the DSC curves, several characteristic points on DSC curves
were analysed, Fig. 11, Tab. 5, 6.

Table 6. Values of some characteristic points on DSC curves for DBR propellants aged for
different period of time at 90oC
Ageing
time at
90 oC,
days
Non aged
2
4
10
45
60
67
72
82

Tfe,
o
C

Hr,
J/g

193.71
196.11
196.80
194.45
195.97
195.34
191.43
195.46
195.51

224.76
224.74
225.08
222.97
221.45
224.83
222.57
225.24
225.21

4.25

0.93

0.20

172.16
3.07

194.94
2.00

220.71
7.38

Corresponding
ageing time
at 30 oC, year**

Tie,
o
C

Tp,
o
C

Non aged
4
8
20
90
119
134
144
164

168.38
168.42
167.98
169.76
172.71
171.21
173.81
175.15
175.55

Degree of change (after 82


days of ageing at 90C) / %
*
Mean value experimental
standard deviation

hp,
W/g

wp,
o
C

1731
1795
1873
1910
1946
1901
1980
2178
2093

3.82
4.09
4.33
4.57
4.90
4.88
5.46
5.32
5.15

31.91
30.51
30.28
27.53
28.61
28.46
25.46
28.16
27.88

20.91

34.82

12.63

1810
252

4.62
1.13

27.54
5.16

(d/dt)max,
%

(d/dt)max,
(Tp),
%

W/(g min)

26.57
26.59
26.39
27.27
20.44
28.25
28.50
29.82
31.61

46.25
46.08
46.55
45.92
45.33
45.67
46.54
42.96
44.67

0.888
0.884
0.922
1.087
1.536
1.296
1.511
1.489
1.814

18.97

3.42

104.28

24.84
5.04

45.86
3.97

1.2
0.38

*- Calculated according to Eq. 1, **- Calculated according to Eq. 2,

From Tab. 6 and Figs. 10, 12, it is visible that the changes are more intensive during the first
15 days of ageing at 90 oC (i.e. ~30 years of ageing at 30 oC), almost constant between 15 and
45 days of ageing, and then intensive again after that period. According to our findings intensive
changes in the first 15 days of ageing at 90 oC are primarily due to nitro-glycerine evaporation
and decomposition, while changes after 45 days of ageing are connected with intensive
degradation of nitrocellulose (chain breakdown) [2].
2300

2.000
6.00

1.800

wp
hp

5.00

28.00

4.50

26.00

4.00

24.00

3.50

(d /dt )max ,%min

30.00

2200

1.600

5.50

Peak height, W/g

Peak width, C

32.00

1.400

2100

1.200

2000

1.000
1900

0.800
0.600
0.400

(df/dt)max

0.200

Hr

0.000
22.00
0

20

40

60

Ageing time at 90 o C, days

80

3.00
100

20

40

60

80

1800

Reaction heat, J/g

34.00

1700
1600
100

Ageing time at 90 C, days

Fig. 12. Changes of some characteristic points/parameters on DSC curve of DBR propellant
with the ageing time at 90 oC

226

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Although changes of the DSC curves shape with the ageing, and consequently characteristic
points on the DSC curves, are evident they still remain within the measuring uncertainty limits
( ) up to 67 days of ageing at 90 oC (i.e. ~135 years of ageing at 30 oC). This means that, on
the basis of non-isothermal DSC tests, a reliable conclusion on DBR propellant stability, i.e. its
age, may be set out only after this period of ageing.

Change of isothermal TGA curves with ageing


Isothermal TGA curves of DBR propellants aged for different period of time at 90 oC are
given in Fig. 13.


100 





Weight (%)

95

90

85

80


 




   
     




        

           

  




 

 





 


 
 

  



 
  

 
   
  
    




     
  


   




    

   
  




 



 













 



 

 


 


  

  


  

 
   

 
 



Non-aged

 

 

 Aged at 90C for 2 days
 
  
Aged at 90C for 10 days
  
  
Aged at 90C for 25 days
 
 Aged at 90C for 45 days
 Aged at 90C for 72 days
 Aged at 90C for 82 days

75
0

100

200

Time (min)

300

400
Universal V3.8B TA Instruments

Fig. 13. Isothermal TGA curves of DBR propellants aged for different period of time at 90C
In order to quantify changes in the TGA curves caused by artificial ageing of the DBR
propellant at 90 C, several characteristic points on TGA curves were analysed (Fig. 14, Tabs. 5,
7).

Fig. 14. Typical isothermal TGA and derivative TGA curves of NC propellant

14

Sanja Muani, Muhamed Sueska, Sanko Baka

227

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 7. Values of some characteristic points on TGA curves for DBR propellants aged for
different period of time at 90 oC
Ageing
time at
90 oC
Nonaged
2
10
25
45
72
82

Corresponding
ageing time
at 30 oC**

(dm/dt)max, t((dm/dt)max), m(dm/dt)max,


%/min
min
%

0.0
4.0
20.0
49.9
89.9
143.8
163.8

Degree of change (after 82


days of ageing at 90C) / %
*
Mean value
experimental standard
deviation

t(10 %),
min

t(15 %),
min

m(60 min), m(200 min),


%
%

(dm/dt)10 min,
%/min

(dm/dt)20 min,
%/min

0.997
1.019
0.986
0.701
0.644
0.497
0.270

2.47
2.47
2.47
2.39
1.84
1.81
2.55

1.55
1.58
1.73
1.03
0.84
0.72
0.60

23.47 66.81 14.51


33.63 118.00 12.37
27.15 82.60 13.66
132.40
7.57
399.20
5.78
4.80
1.77

34.13
16.90
18.22
13.48
8.41
7.18
2.49

0.4312
0.3396
0.3679
0.2228
0.1344
0.1078
0.0301

0.24
0.19
0.21
0.16
0.08
0.08
0.16

72.92

3.24

61.29

1600.89

23.63

87.80

92.70

93.02

33.33

1.248
0.62

2.471
0.42

2.052
1.12

13.865 30.117
9.11
33.42

19.921
5.81

23.636
5.18

0.663
0.27

0.358
0.11

*- Calculated according to Eq. 1, **- Calculated according to Eq. 2

It is evident from Fig. 13 and Tab. 7 that the ageing causes significant changes of TGA
curves. For example, time to reach the same degree of mass loss increases with the ageing, mass
loss after the same time period decreases, the rate of mass loss decreases (Fig. 14), etc. The
changes of majority of parameters become higher than measuring uncertainty limits (Tab. 7)
after 10 days of ageing at 90 oC (i.e. after ~20 years of ageing at 30 oC).
TGA experiments were carried out using opened aluminium sample pans and under 100 oC
isothermal temperature. Under such conditions the sample mass loss is predominantly a
consequence of nitro-glycerin evaporation, and partly degradation of nitro-glycerine and
nitrocellulose. Our previous experiments have shown that mass loss of NC propellant at 100 oC
is less than 2 % after 400 minutes [1], which is almost negligible comparing with mass loss due
to nitro-glycerine evaporation and degradation.

40.00
t(10 %)

400.00

(dm/dt)max
1.000

1.80

m((dm/dt)max

1.60

300.00

25.00

250.00
20.00
200.00
15.00

150.00

10.00

100.00

(dm/dt )max , %/min

30.00

m (200 min), %

t (10 %), min

350.00

2.00

1.200

35.00

m(200 min)

1.20
0.600

0.80

50.00

5.00

0.00

0.00
100

0.000

20

40

60
o

Ageing time at 90 C, day

80

1.00

0.400
0.200

1.40

0.800

0.60

m (dm/dt )max , %

450.00

0.40
0.20
0

20

40

60

80

0.00
100

Ageing time at 90 C, day

Fig. 15. Changes of some characteristic points/parameters on TGA curve of DBR propellant
with the ageing time at 90 oC

228

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The results of isothermal TGA measurements (Fig. 15) clearly explain the conclusion that
the dominant process during artificial ageing at 90 oC is nitro-glycerine evaporation and
degradation (simultaneously). Because of that, the TGA curve of DBR propellant aged for 82
days at 90 oC becomes very close to TGA curve of NC propellant.
Taking into account all DSC and TGA experiments carried out on the DBR propellant, it is
obvious that the ageing causes changes of their thermal properties. However, the changes of
DSC parameters still remains within the measuring uncertainty limits during 67 days of ageing
at 90 oC (i.e. ~135 years of ageing at 30 oC), while the changes of TGA parameters remain
within the measuring uncertainty limits during the first 10 days of ageing at 90 oC (i.e. 20 years
of ageing at 30 oC). This means that TGA is more sensitive than DSC to the ageing of DBR
propellants.

4.

CONCLUSION

The results presented in this paper have shown that artificial single-temperature ageing (at
90 C) of DBR propellants causes significant changes of their mechanical and thermal
properties. Obtained results have confirmed that dynamic mechanical analysis, differential
scanning calorimetry and termogravimetry can be used to follow and to quantify these changes.
Obtained changes on DMA, DSC and TGA curves indicate that ageing processes have two
stages. In the primary stage of ageing dominant processes are migration and evaporation of
nitro-glycerine, while in the secondary stage of ageing a dominant process is intensive
degradation of nitrocellulose polymer chain.
It was found out that the most sensitive parameters/characteristic points to the ageing
process at DMA curves are: storage modulus at 25 C, storage modulus at the softening region,
peak width and height on loss modulus curve, glass transition temperature, loss modulus at the
viscoelastic region, tan maximum in viscoelastic region, extrapolated onset temperature at the
end of the glass transition, while the most sensitive parameters at the DSC and the TGA curves
are: peak height and width, heat of reaction, maximum heat flow rate, mass loss at a given time,
and rate of mass loss at a given time.
However, because of the relatively high measuring uncertainty and limited sensitivity
(especially for DSC measurements) to the age of the propellants, it follows that DMA, DSC and
TGA are not too sensitive techniques to the propellants age, but they still may help in the
propellants stability evaluation. In the future work we will try to found out optimal testing
conditions which will give minimal measurement uncertainty.

5.

REFERENCES

[1]

M. SUESKA, S. MATEI MUANI, M. RAJI LINARI, Applicability of Thermal Methods in


Stability Studies of Homogeneous Propellants, IASPEP Conference, Beijing, China, 2005.
S. MATEI MUANI, Thermal decomposition of double base rocket propellant and its
influence on viscoelestic properties, Masters thesis, University of Zagreb, Faculty of Chemical
Engineering, Zagreb, 2005. (in Croatian)
M. SUESKA, Baruti, MORH, 1996.
TAGER, Physical Chemistry of Polymers, Mir Publishers, Moscow, (1972)
M. A. BOHN, Kinetic Modelling of the Ageing of Gun and Rocket Propellants for the Improved
And Time-Extended prediction of Their Service Lifetime, Proceed., 1998 Life Time Cycles of

[2]

[3]
[4]
[5]

16

Sanja Muani, Muhamed Sueska, Sanko Baka

229

New Trends in Research of Energetic Materials

[6]

[7]
[8]

[9]
[10]
[11]

[12]
[13]

[14]

[15]
[16]

[17]
[18]

230

Czech Republic

2006

[Content]

Energetic Materials, Fullerton, California, USA, 29 March-1 April, (1998)


M.A. BOHN, Methods and Kinetic Models for the Lifetime Assessment of Solid Propellants,
Paper 2 on the 87th Symp. Of the Propulsion and Energetic Panel (PEP) of the AGARD, Service
Life of Solid Propellant Systems , Athens, Greece, May 6-10, (1996), AGARD-ConferenceProceedings 586 (AGARD-CP-586), F-92200 Neuilly sur Seine, France, May (1997).
F. VOLK, G. WUNSCH, Determination of the Decomposition Behaviour of Double Base
Propellants at Low Temperature, Propell. Expl. Pyrot. 10, 181 (1985).
A. PFEIL, H.H. KRAUSE, N. EISENREICH,, The Consequences of Beginning Slow Thermal
Decomposition on the Molecular Weight of Nitrated Cellulose, Thermochim Acta 85, 395
(1985).
F. VOLK, M. A. BOHN, G. WUNSCH, Determination of the Chemical and Mechanical Properties
of Double Base Propellants During Aging, Propell. Expl. Pyrot. 12, 81 (1987).
B. VOGELSANGER, Chemical stability, Compatibility and Self Life of Explosives, Chimia 58,
401-408, 2004
M. A. BOHN, F. VOLK, Prediction of the Lifetime of Propellants by Stabilizer Consumption and
Molar Mass Decrease, Proceed. ADPA Predictive Technology Symp., June 22-24, (1993),
Orlando, FL, US-Army ARDEC, Picatinny Arsenal, New Jersey, USA.
T. MURAYAMA, Dynamic Mechanical Analysis of Polymeric Material, Elsevier Scientific
Publishing Company, Amsterdam, Oxford, New York, (1978)
BOHN M. A, Kinetic Modelling of the Ageing of Gun and Rocket Propellants for the Improved
and Time-Extended Prediction of Their Service Lifetime, Proc of the 4th Life Cycles of Energetic
Materials Conference, March 29 April 1, Fullerton, USA, 1998, p. 1-38.
M. A. BOHN, F. VOLK, Ageing Behaviour of Propellants Investigated by Heat Generation,
Stabiliser Consumption, and Molar Mass Degradation, Propellants, Explos. Pyrotech., No. 17,
(1992), p. 171-178.
R. R. Rogers, J. L., Rogers. Explosive science, Los Alamos, NM, http://home.att.net/~mrogers,
2002.
M. SUESKA, S. MATEI MUANI, M. RAJI LINARI, Determination of Arrhenius kinetic
constants for DB propellants by non-isothermal DSC measurements. Influence of some testing
conditions, Proc. of 6th Int. Seminar New trends in research of energetic materials, Pardubice,
2003, pp. 374-391.
M. SUESKA, M. RAJI, Thermal characterisation of stable and unstable nitrocellulose
propellant, Proc. of 33rd Int. Annual conference of ICT, Karlsruhe; ICT, 2002. 98.1-98.17
M. RAJI LINARI, M. SUESKA, S. MATEI MUANI, Changes of some thermophysical
properties of nitrocellulose propellants during accelerated aging, Proc. of 8th Int. Seminar
New trends in research of energetic materials, Pardubice, Czech Republic, 2005., pp.758-764.

Applicability of Dynamic Mechanical and Thermal Methods in Investigation of . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

HIGH ENERGY DENSITY MATERIALS (HEDM):


OVERVIEW, THEORY AND SYNTHETIC EFFORTS AT FOI
Henric stmark
Department of Energetic Materials
FOI, Swedish Defence Research Agency
Grindsjn Research Centre, 147 25 Tumba, Sweden

Abstract:
This paper presents a literature overview of HEDM. A number of potential HEDMs with
performances theoretically up to more than 10 times HMX have been identified, ranging
from ionic all-nitrogen compounds like N5- and N5+ with suitable counter ions (N5+N5gives a performance of ~1.6 times HMX) to polymeric nitrogen (10 times HMX) and
metallic hydrogen. In between there are for example metastable nitrogen clusters (Nx,
x4 to 60). Calculations show that N4 has approximately 3.2 times the performance of
HMX. Both metallic hydrogen and polymeric nitrogen are high-pressure compounds
(several Mbar) and thus probably hard to synthesize in larger amounts. The nitrogen
clusters offer an interesting alternative or complement in that they are all normal
chemical compounds and they all have high heat of explosion, and probably also high
density (2.0-2.7 g/cm3).
This overview covers conventional chemistry, novel chemistry (all-nitrogen
compounds, new oxidizers in the form of NxOy compounds) and exotic physics (eg.
metallic hydrogen, cold fusion and anti-matter). The time in which these concepts can
be realised ranges from near future to many years ahead and some may possibly never
be made at all or perhaps only in lab scale. The overview is complemented with
theoretical and experimental results from our institute in our effort to synthesize new all
nitrogen compounds, e.g. Pentazolate ion (N5-) and tetraazatetrahedrane (N4(Td) .
Keywords:

1.

HEDM, synthesis, QM, performance, all nitrogen compounds

INTRODUCTION

The three most important reasons for the development of energetic materials for military
purposes today are the needs for increased performance, increased safety and certain tailored
properties. This is important for warhead, propulsion and launch applications. Several factors
influence the development directly or indirectly. The most important are increased use of
insensitive munitions (IM), the need for more precise weapons effects, higher performance,
weapon specific properties, environmental concerns, internationalisation, standardisation,
and increased system integration.
In this paper we will discuss the possibilities to develop materials with performance
considerably better than what is used today, commonly referred to as High Energy Density
Materials (HEDM). The aim of this paper, which is based on the available literature and on
work done at our institute, is to give an overview of the area, to see the possibilities of what
could be achieved with different concepts, soon or in a longer time perspective.
When dealing with new energetic materials a term that has been used in the Energetic
Material community is High Energy Density Materials (HEDM). The limit between
1

Henric stmark

231

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

conventional energetic materials and HEDM has never been defined and may therefore differ
between research groups. In the first European Workshop on High Energy Density
Materials, held at FOI in October 2005, the participants agreed on a common definition for
HEDM as energetic materials better than the best that are in use in systems today. The
baseline will vary with the application and we end up with different reference substances for
different applications. The European definition gives the following baselines; HEDMs are
explosives with higher performance than HMX, solid rocket propellants with higher
performance than AP/Al/HTPB and gun propellants better than JA2 (NC/NG) propellants.
Computational modelling, using mainly quantum chemical methods (numerically solving
the Schrdinger equation at different level of approximation), plays an integral part in the
development of new high energy density materials. This is the case both for the synthesis,
detection and characterization of new all-nitrogen compounds, and in the design and
synthesis of conventional type explosives.
The performance of a substance is readily calculated from its heat of formation and
density; two properties that can be estimated by means of quantum chemical computations.
Apart from performance, some practical aspects such as stability, sensitivity and state of
aggregation are important. In order to make experimental work meaningful it is also essential
to calculate properties that make it possible to detect and identify the material once it has
been made. It is also very useful to use quantum mechanics to evaluate synthesis routes.
This paper is mainly based on a more comprehensive overview of HEDM research,
including chapters on QM and performance calculations High Energy Density Materials
(HEDM) A Literature Survey [1]. This report was a joint effort by many people whose
contribution is hereby acknowledged.

2.

HISTORIC REVIEW

Yesterday, today
The earliest known energetic material was black powder and Greek fire, both with origins
held in some mystery but probably invented about 2000 years ago, Greek fire in ancient
Greece and black powder in China. Black powder was first used only for fireworks and
mainly not for war. The first real use for warfare was probably in the 14th century as
propellant for guns and firearms. It was then in use until the little revolution of the midnineteenth century when the black powder era ended, and when modern-day propellants and
high explosives came into use. The development of energetic materials has since then
proceeded relatively slowly. Most of the explosives found in todays weapons systems are
still TNT, PETN, HMX or RDX, i.e. energetic materials developed in the 19th century or, in
one case, the 1940s (HMX).
Table 1 shows an overview of the history of energetic materials development. The
column for Performance gives values relative to the performance of HMX. The performance
increases about 70% from picric acid (60) to HMX (100) between World War One and
today. This can be compared to the threefold increase at the end of the black powder era.
The slow development during the greater part of the 20th century was partly due to a lack
of understanding of energetic molecular structures rather than a poorly developed synthesis
technique. For instance, scientists believed for a long time that HNB, hexanitrobenzene, was
the optimal high explosive. The relative performance of HNB vs. RDX is 1.23. After
Russian scientists first reported the successful synthesis of this compound, US researchers

232

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

spent 10 years trying to find the synthetic path. This was finally achieved, after which it
became apparent that the substance was unstable and highly sensitive to light.
tomorrow
During the last decade, the development of computers and thus the increased capability
for quantum chemical calculations, as well as a broadened view of the characteristics of
energetic molecules and their applications, have yielded new opportunities for the
development of energetic materials. It was generally believed that useful energetic materials
would be CHNO-substances, with molecules containing carbon, hydrogen, nitrogen and
oxygen. Today, several completely different types of compounds have been added to the list
of possible energetic materials. Among these are caged molecules that contain more energy
per unit of mass and volume, molecular composites where fuel and oxidizer are mixed at a
molecular level, nano-systems of reactive materials and meta-stable compounds. Even very
extreme forms of energy storage, such as nuclear isomers and anti-matter are investigated.
In other words, the last decade has seen an explosive development in the number of
theoretical candidates for useful energetic materials. Some of these have already been
synthesised and produced in useful quantities.
Today, the development of energetic materials follows two distinct lines, one
evolutionary and one revolutionary. There is also the development of reactive materials that
combined with energetic materials can increase the performance of future weapons systems.
Table 1 Historical overview of development of energetic materials.
Invented

In Application

Energetic Material

Energy

Density
3

Performance

(MJ/kg)

(kg/m )

(relative to HMX=100)

1000?

1600-1870

Black Powder

20

1847

1870-2010

Nitroglycerine

6.5

1.6

70

1867

Not in weapons

Dynamite

1.45

50

1771

WW1

Picric Acid

1.6

60

1863

1900-2020

TNT

1.55

60

1899

WW2

RDX

6.6

1.8

90

1894

WW2

PETN

6.3

1.8

80

1940s

1960-

HMX

6.5

1.9

100

1986

2005

CL-20

7.5

2.1

120

2006

3.

HEDM POSSIBLE BY THE USE OF CONVENTIONAL


CHEMISTRY

The evolutionary development of new energetic materials concerns CHNO-substances


prepared by conventional chemistry. The two main driving forces for the development of
explosives are lower sensitivity and increased performance. Until now, new energetic
materials have either been less sensitive with retained performance compared to current
materials or more powerful but also more sensitive (Fig 1). A long-term objective is of
course to merge these two qualities and produce energetic molecules that are both more
3

Henric stmark

233

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

powerful and less sensitive than current materials. Compounds with the performance of CL20 and the sensitivity of TATB would be desirable long-term goals.
To increase performance compared to current explosives the molecules need to contain
particularly energetic structures that give a high heat of formation (e.g. caged molecules,
strained rings, nitramines, furazanes, tetrazines) and/or have high-density functional groups
(e.g. fluorine, cage, urea, guanidine). Computational methods allow quite accurate prediction
of density and heat of formation, which makes it possible to evaluate large numbers of
candidate molecules before their preparation is attempted.
It is more difficult to predict the sensitivity of new energetic materials, which leads to a
more qualitative analysis of the proposed target molecules. Factors that influence sensitivity
are for instance functional groups (e.g. nitrate esters), crystal structure and hydrogen
bonding. It is also possible to influence the sensitivity of explosives by perfecting their
crystal shape. Round, flawless crystals are desired, both because of their easier processability
and their decreased sensitivity. Recently the long studied crystallization of RDX yielded the
product I-RDX [2] which is equal to normal RDX performance-wise, but has improved
sensitivity characteristics towards shock. This breakthrough has sparked an interest for the
preparation of I-HMX [3] and perhaps even I-CL-20.
The most promising new candidate molecules are CL-20, ADN, and FOX-7, which have
all been produced in sufficient quantities to allow them to be used for various applications.
TATB and NTO are two low sensitivity energetic materials that are already in use.
Current
RDX, HMX
TNT

Sensitivity
FOX-7, NTO, TATB

Performance
CL-20

HEDM
Lower sensitivity

Increased
Performance

Fig 1. Lines of development of energetic materials.


A number of potential HEDM molecules have been theoretical or experimental
investigated. In some of the most promising or interesting shown . A short description of
each is also given below.
CL-20 or Hexanitrohexaazaisowurtzitn contains more energy than earlier high
performance energetic materials, e.g. HMX. The purely chemical theoretical increase in
performance is 20%. In shaped charges, because of the higher energy content in the
explosive charge, different materials can be used in the liner yielding an increase in
performance of about 60%. CL-20 is, however, as sensitive as PETN regarding impact and
friction. As for HMX there are several polymorphs of the substance adding further
complexity to its handling. Due to a difficult synthesis the price is also quite high.
Octanitrocubane or ONC was first envisioned by Gilbert in the beginning of the 1980s
as a compound with a perfect oxygen balance and exceptionally high heat of formation and
density [4]. As the compound was a C-Nitro explosive good sensitivity properties were also

234

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

expected [4]. These predictions were also supported by advanced calculations. The
compound was finally prepared in the year 2000 after a very long and laborious synthesis but
proved to be a disappointment with a lower than expected density [4]. It turned out that the
heptanitrocubane was a superior explosive due to its higher density, there are however hopes
to find a higher density polymorph of ONC which might live up to the predictions.
O2N

NO2
N NO
2

O2N
N

O2N

NO2

O2N
NO2
O2 N

NO2

N
O2N

O2 N

NO2

NO2

Octanitrocubane (ONC)

CL-20
O2N

NO2

NO2

NO2

O2N

O2N

O
N

O2N

HHTDD
O

N
N

N
O

NH

NO2

NO2

Tetranitro-tetraaza-cubane
O

NO2

N
N

HN

TTTO

Oxadiazolo-oxadiazole-dioxide

Fig 2. HEDM molecules possible by the use of conventional chemistry


2,4,6,8-Tetranitro-1,3,5,7-tetraaza-cubane a proposed compound with a structure
similar to that of ONC. The proposed high density is derived from the cubane structure
which also contributes to its heat of formation. The compound would also be perfectly
oxygen balanced. The calculated performance is 1.4 times that of HMX.
HHTDD or cis-syn-cis-2,6-Dioxo-1,3,4,5,7,8-hexanitrodecahydro-1H, 5H-diimidazo[4,5b:4,5-e] pyrazine. The dinitrourea moiety confers high density to molecules containing it.
A property, which has been realized and utilized in a whole series of experimental energetic
materials. HHTDD [5] is supposed to be the explosive with the best performance ever
prepared, it lacks however both in sensitivity and hydrolytic stability something it shares
with its close analogue Sorguyl (TNGU) [6]. Our calculations based on experimental density
and heat of formation implies a performance slightly below that for CL-20.
TTTO or [1,2,3,4]Tetrazino[5,6-e][1,2,3,4]tetrazine 1,3,5,7-tetraoxide is a compound
that has not been prepared but which has been evaluated by computational techniques and
found to be extremely interesting. Calculations propose a density of 2.38 g/cm3 and a heat of
5

Henric stmark

235

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

formation of 190 kcal/mole, leading to a relative performance of 2.2 times HMX. This
compound can be seen as a possible performance limit to normal explosives and what could
be achieved by a focused, well funded research program.
1H,4H-[1,2,5]Oxadiazolo[3,4-c][1,2,5]oxadiazole 3,6-dioxide is another proposed
bicyclic heterocycle similar to TTTO. The two nitrogenatoms are N-oxidized and the
compound has a proposed high density (2.3 g/cm3) which would give it excellent
performance as a high explosive (1.7 times HMX).
In Table 2 the calculated densities, heats of formation and performances of the described
compounds are compiled. Data were calculated by the methods described in Reference 1.
The values for ONC were taken directly from the Cheetah 2.0 database.

Density (g/cc)

Heat of formation (kJ/mol)

Detonation velocity (mm/s)

Detonation pressure (GPa)

Energy (HMX=100, V=2.2 V0)

Table 2. Calculated performance of HEDM prepared or possible to prepare by ordinary


chemistry

CL-20

2.04*

393*

10.0

47.8

121

ONC

1.92

381

9.2

37.1

105

Tetranitrotetraaza-cubane

2.02

912.4

10.5

50.1

137

HHTDD

2.07*

153.4*

10.2

46.1

118

TTTO

2.38

794.6

10.9

131

220

Oxadiazolooxadiazoledioxide

2.30

263.8

11.3

71.3

167

Energetic material

* Experimental data
In conclusion, the evolutionary development of energetic materials during the last decade
gives a potential for future increase in systems performance of up to 100%. The foundation
has been laid, but more applied research is needed in order to fill the requirements of specific
weapons systems. This would also make it possible to tailor energetic materials and high
explosives for various applications.

4.

ALL-NITROGEN COMPOUNDS

Polynitrogen compounds have received considerable attention in recent years as potential


candidates for high energy density materials (HEDMs). The fact that most of these have a
very high calculated energy content is a consequence of the large bond energy of 228
kcal/mol of the triple bond in molecular nitrogen (N2) compared to the bond energies of
nitrogen-nitrogen single and double bonds, which typically amounts to 39 and 100 kcal/mol,
respectively. In addition to the favorable energetics of the polynitrogen compounds, they
also have the advantage that the energy releasing decomposition into N2 leads to no
deposition of harmful substances. This makes them particularly suitable for propulsion of
236

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

spacecrafts, since the commonly used propellants of today results in large depositions of
harmful substances in the atmosphere. The special properties of polynitrogen compounds
also make them of great interest for energy storage. The high heat of formation and large
density makes polynitrogens nearly ideal for use as high explosives.
Even though theoretical studies have identified a number of potentially stable structures,
the experimentally observed nitrogen compounds are still few. The azide anion, N3-, was first
synthesized in 1890 by Curtius. Christe and coworkers have since 1999 reported the
preparation and isolation of N5+ together with several different counter ions [7]. A few other
species, such as N3, N3+ and N4+, have been observed only as gaseous or matrix-isolated ions
or radicals. The formation of N5- from arylpentazoles in mass spectrometry experiments has
recently been reported from our lab [8, [9] and from Air Force Research Laboratory (Edwards
Air Force Bas) in the US [10]. The detection of HN5 and a zinc pentazolate salt in solution
has also been reported [11]. These results have however been questioned [12].
Quantum mechanical studies of stability and energy content have been carried out on a
number of nitrogen compounds. For a recent review on the structure and stability of N4 and
N5 systems see Nguyen [13] and references therein. An overview of some calculated
structures and stabilities of N2 to N10 species as well as their spectroscopic characteristics
can be found in a report by Bartlett [14]. Many all-nitrogen structures have also been treated
by Glukhovtsev el al. [15]. Theoretical calculations lay the foundation for ongoing attempts to
synthesise such compounds and increases the possibility of successfully producing one or a
few useful substances and formulations. A possible development in this field is substances
allowing an increase in performance as high explosives of 25 times the values of today. The
impulse of for example N4 is comparable to H2/O2 but has the enormous advantage of a
much higher density which would allow for smaller and thereby lighter rockets.
Quantum mechanical calculations to determine the heat of formation for a number of
poly-nitrogen molecules (Fig 3) have been made at FOI using Gaussian 98 at B3LYP/631G(d) level of theory.

Fig 3. Calculated structures of N4(D2h), N4(Td), N6(D3h), N8(Oh), N10(D5h) and N12(D6h)
The difference between Hf(g) and Hf(s) is the sublimation energy, i.e. the energy
gained by the molecules by forming a solid or liquid instead of gas. The size of this
difference is one indication of a substances inclination to be in solid phase.
The estimation of the state of aggregation for a substance from calculations is a difficult
task. In Table 3 one can see that very little is gained for N4(Td) so this may not be expected
to be in solid state at ambient temperature and pressure. Absolute estimations are very
difficult to make. Use of the methods described by Murray [16] give boiling points. The
accuracy of these calculations varies with substance. For comparison, the calculated boiling
point for hydrazine is 338 K (exp 386 K) and for tetrazole 434 K (exp 429 K mp sublimes).
The trend is clear; N8, N10 and N12 are liquids at room temperature, N10 and N12 may even be
solids. N4 is a gas at room temperature, but would be a liquid or even probably solid at liquid
nitrogen temperature.
Interesting from a point of view of usefulness of new materials are properties such as
energy content (H), density, energy barrier towards dissociation, i.e. stability and
aggregation form. Some of these properties such as aggregation form are more difficult to
estimate than others. From an experimentalists point of view in trying to make these
7

Henric stmark

237

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

molecules it is also necessary to have reasonable synthesis routes, something that can often
also be studied with quantum mechanics. One of the most important questions for an
experimentalist to answer is how to detect the successful synthesis of the molecule. Quantum
chemistry is very useful in supplying answers.
Table 3 Heats of formation from QM calculations (Gaussian 98, hs95fo), densities
calculated with MOLPAK and detonation velocities, pressures and performances (except Isp)
calculated with Cheetah 2.0 (BKWC), Isp calculated with NASA CEA.

Impulse (sec)

Tc (K)

Force (J/g)

121

289

313

408

7563

3932

N4(Td)

202

201

2.3

15.7

125

200

327

422

7795

4212

N6(D3h)

346

342

2.5

17.6

169

285

428

439

8079

5193

N8(Oh)

480

473

2.7

19.0

206

331

498

444

8162

N10(D5h)

551

544

2.7

19.2

208

364

484

433

7971

4628

N12(D6h)

712

703

2.8

19.7

224

407

526

443

8140

Hf(s) (kcal/mol) a

(HMX=100, V/V0=2.2)

15.5

Energy

2.2

Boiling point (K) c

Detonation Pressure (GPa)

Density (g/cm3)

189

(kcal/mol)

194

Sublimation energy

N4(D2h)

Hf(g) (kcal/mol) a

Detonation velocity (km/s)

Molecule

Heat of formation is calculated at B3LYP/6-31G(d) level of theory. This level has been chosen for
comparative purposes and does not give the most accurate values calculated. The differences to heats of
formation calculated at higher levels of theory are however small and do not affect the calculated performance
in an important way.
b
Results by Karle [17] indicate a higher sublimation energy for N4(Td) (5 kcal/mol) which would of course mean
a higher boiling point.
c
The boiling point has been calculated using a lower level of theory, STO-5 using the method described in ref

There are many papers describing quantum mechanical calculations on various allnitrogen molecules. Calculations on larger nitrogen molecules are mainly focused on
geometry and finding stable forms. Little has been done on determining chemical or thermal
stability and performance. For any of these molecules to be of experimental interest, there
must also exist ideas on synthesis routes. All of these compounds have high energies (based
on the number of single or double bonds), but in order to assess their performance, their
density must also be calculated. Only a few, such as N4 and N5-, have been extensively
studied with quantum mechanics and other calculations. In this paper N4(Td) and N5- will
serve as example on how to proceed towards the synthesis of HEDM with the aid of QM
calculations.

238

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

4.1

Czech Republic

2006

[Content]

Tetraazatetrahedrane, N4(Td)

One of the most interesting candidates for a HEDM among the polynitrogen compounds
is tetraazatetrahedrane, N4(Td). Phosphorous (below nitrogen in the periodic table) exists in,
among other forms, the form of P2 as well as P4(Td) (white phosphorous) whereas nitrogen is
only known in the form of a dimer (N2).
Energy content and stability
In 1990, Francl and Chesick [18] calculated the energy of formation of N4 (814 kJ/mol) as
well as its dissociation barrier (315 kJ/mol). They concluded that the calculated barrier was
large enough to suggest a significant lifetime and metastability if N4 could be made and
trapped. The energy release involved (heat of reaction) in the dissociation of N4(Td) into N2
has been calculated in many studies, giving similar results. Lee and Martin [19] give a value
of 763 2 kJ/mol.
The dissociation barrier has also been discussed in many studies with results ranging from
118 to 315 kJ/mol. The height of the dissociation barrier, i.e. the activation energy, can be
calculated with much higher accuracy now than only ten years ago. The first studies used
very simple approximations to estimate the height. Francl and Chesick [18] used the lowest
energy point on the ridge of intersection between the potential energy surface (PES) of
ground state N4 and the PES of two N2 molecules, giving the high value of 315 kJ/mol.
Yarkony [20] used a much more refined method taking into account that the potential energy
surfaces of other states may cross the investigated PES (Fig 4). This resulted in the low value
of 118 kJ/mol. A later study taking in more information about the system but essentially
using the same principle has given a higher value of 214 kJ/mol [21]. Despite the incredibly
high energy content, this computed dissociation barrier indicates that it is sufficiently stable
for most potential applications.

TS

Ea
N4 GS

H
2 N2

Fig 4. Schematic picture of a potential energy curve (blue) for the dissociation of a
substance via a transition state (TS) releasing the energy H. A potential energy curve
(green) from a different state crosses it, effectively lowering the activation energy Ea.
The high energy content of N4(Td) makes it hard to make by traditional methods in
chemical synthesis. All attempts of experimental preparation and detection of N4(Td) have so
far been unsuccessful. Radziszewski and coworkers reported to have detected N4(Td) in 2000
[22]
, but it was later shown that the reported isotope effect on the observed IR-band is not
compatible with N4(Td) [19]. An open chain N4 isomer in the triplet state was recently
detected in mass-spectrometry experiments [23, [24, [25]. However, this species is expected to
be too unstable and also too low in energy to be of any interest as a HEDM.

Henric stmark

239

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Suggested synthesis routes


A number of synthesis routes have been suggested, mainly by our institute, and in some
cases the chances of success using these routes have been assessed by quantum chemical
calculations. Francl and Chesick [18] suggested that prolonged irradiation of liquid nitrogen
with radiation of wavelength less than 140 nm might yield evidence of N4 formation. (Note
that the calculated wavelength here is too high. The energy of the barrier has to be added to
the photon energy which gives ~106 nm.) This suggestion was of course based purely on an
energetic basis. In order to increase the chances of success, more refined arguments and
investigations can be made. In Fig 5 an overview of possible synthetic routs to N4(Td) is
shown.
I

N4(Td)

N2* + N2

N(2D) + N2

III

(alt.)
N(2D)

cyc-N3(C2v)
h

ClN3

N4(Td)

O
t-Bu

t-Bu

VI
N

N3* + cyc-N3(C2v)

NXY + cyc-N3+(D3h)

t-Bu

N N

t-Bu

t-Bu

t-Bu

-N2

IV

N4(Td)

N4(Td)

cyc-N3(C2v)

-Cl

+ cyc-N3(C2v)

t-Bu

N2(A3+u) + N2(A3+u)

II

N N

t-Bu

h
t-Bu

t-Bu

N4(Td)

t-Bu

t-Bu

N
N N

- NX-1Y+1

Actual synthesis
route for
tetratertbutylhedrane
and suggested route
for N4(Td)

X=O,S,Se,NH,..

Fig 5. Proposed routes for synthesis of N4(Td). Excited states are marked by *.
I. The collision of a ground-state N2 with an excited-state N2 to form N4 requires that
their total energy is high enough to form the molecule and reach over its dissociation barrier.
The triplet excited states generally have longer lifetimes than the singlet excited states and
are therefore most interesting for this type of reaction. The result would be a N4 molecule of
triplet character. Calculations [26] show that metastable excited triplet states exist, which
increases the chances of success. However, the electronic structure of ground-state N2 is not
optimal for formation of new N-N bonds. The required excited states can be produced by
excitation of nitrogen in liquid or solid form. Excited nitrogen molecules can also be
obtained by RF (microwave cavity) or electrical discharges (hollow cathode or pulsed nozzle
discharge), and the reaction products can be trapped on a cold surface. Ion bombardment of
solid nitrogen or nitrogen in rare gas matrices is another possibility.
II. Another suggested route is N2(A3+u) + N2(A3+u) N4(Td). This reaction has the
advantages that the orbital occupancy of N2(A3+u) is favorable for formation of new N-N
bonds, and that this state has a long life time (1.9s). The collision of two molecules in an
240

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

excited state is generally very difficult to achieve but there are ideas on how to succeed. Ion
or -particle bombardment of solid nitrogen in argon has been shown to excite nitrogen to
the A3+u state [27]. It is likely that molecules are excited primarily along the trajectory of the
ions or -particles through the matrix. This enhances the chances of two molecules next to
each other being excited to the correct state.
III. On the basis of quantum chemical calculations we have suggested that N4(Td) may be
synthesized from molecular nitrogen and atomic nitrogen in the 2D state in a two step
process [28]. In the first step a cyclic N3 isomer, cyc-N3(C2v) is formed by the addition of N2
to N(1D). Addition of a second N(1D) to N can lead to the formation of N4(Td). Nitrogen
atoms in the 2D state have a long lifetime in cryogenic matrices (37 s) and can be generated
at high concentrations. The main drawback of this route seems to be side reactions. Linear
N3, N3(D) can be formed in the first step and 2 N2 in the second step. The existence of cycN3(C2v) has been confirmed in a recent experimental study, where it was produced by
photolysis of ClN3 [29]. An experiment could be setup where the photolysis products are
deposited in a nitrogen matrix with a high concentration of N(2D). In addition, recent
unpublished computations from our group indicate that a cyclic N3+ isomer, cyc-N3+(D3h),
exists, which can be obtained directly by ionization of cyc-N3(C2v). These two new species
opens up new potential pathways for synthesis of N4(Td) and other polynitrogen allotropes.
IV. It can be noted that both the production of N3 from N(2D) + N2 and ClN3 is likely to
lead to a mixture of N3(D) and cyc-N3(C2v). (A 0.8/0.2 mixture was obtained from ClN3.)
The reaction between N3(D) and cyc-N3(C2v) to form N4(Td) and N2 should be considered.
Preliminary calculations indicate that this reaction could have a too high barrier to be
feasible. However, the reaction between cyc-N3(C2v) and an excited state of N3 may be
barrierless. This could be achieved experimentally by simultaneous application of two laser
sources, one tuned for photolysis of ClN3 and the other to the proper excitation wavelength.
V. Unpublished calculations show that a stable cyclic N3+ isomer, cyc-N3+(D3h), can be
formed by ionization of cyc-N3(C2v). Although it is higher in energy than ground state linear
N3+, it is very stable towards unimolecular decomposition. The reaction between cycN3+(D3h) and N3- should be investigated. It may be possible to deposit both ions into a
matrix. Also reactions between cyc-N3+(D3h) and other neutral and ionic states of N and N3
should be explored by computations to find possible pathways for formation of N4(Td).
VI. Tetratertbutyltetrahedrane can be formed by photolysis of tetra-tertbutylcyclopentadienone in argon matrices [30]. It may be possible to form tetraazatetrahedrane
(N4(Td)) by a similar method starting from tetrazole-5-one. This starting material is not
known from the literature and low level (B3LYP/6-311G) calculations indicate that it is not
stable, but higher level calculations should be performed to test the stability of this and
alternative starting materials. Another option is to use arylpentazoles as starting material.
This type reaction is suitable for mass spectrometric studies with laser-induced
decomposition of suitable precursors.
Detection
The symmetric nature of N4(Td) means that IR spectroscopy alone is insufficient. The
only line available gives no identification capabilities. The difficulties are clearly illustrated
by the questionable assignment of the origin of an IR line as N4(Td) [22], made by
Radszisewski et al. Lee and Martin showed that the experimental isotopic shift upon 15Nlabeling unfortunately does not match the theoretical predictions for N4(Td) [19].

11

Henric stmark

241

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Raman spectroscopy is a much better choice of detection method for N4(Td). It gives a
fingerprint of the molecule with three lines. The detection limit in a matrix has been
determined by combining experimental signal strength and noise levels of N2 in solid
nitrogen with ab initio quantum mechanical calculations of Raman intensities of N2 and
N4(Td) [31]. Using that particular experimental setup 80 ppm is necessary for detection using
8 h measurement time and 1W laser power. This detection limit is too high considering that
no optimization of a signal can be made during the measurement. An improvement of the
experimental setup, for example by using a Raman probe, could lower the detection limit
rendering Raman spectrometry a feasible detection method.
A more sensitive Raman method, Surface Enhanced Raman Spectroscopy (SERS), has
been evaluated for the purpose of N4 detection using white phosphorous, P4(Td) as a model
substance [32, [33]. No SERS effect was observed for P4 which may indicate that SERS is not
a promising detection method for N4.
Detection with Laser-Induced Fluorescence (LIF) would be ideal due to its sensitivity.
However, accurate calculations of excitation energies and geometry optimizations of the
excited states are necessary to determine laser and fluorescence wavelengths. An excited
state of N4(Td) has been identified by EOM-CCSD theory which is 10.89 eV above the
ground state, and which is optically active for one and two-photon absorption [34]. However,
it has not yet been possible to optimize the excited state geometry. A similar study of
rectangular N4, N4(D2h), has successfully given a useful route for LIF detection [35]. The
vertical absorption and emission energies for the first excited state were determined to 1.58
and 1.55 eV respectively (laser and detection wavelengths 785 and 800 nm respectively).
It may be possible to use the previously discussed state at 10.89 eV to detect N4(Td) by
UV absorption at around 114 nm. This is just outside the relatively easily accessed range
where PM-tubes, MgF2 windows and deuterium lamps work so extra attention to the
experimental setup would be necessary.
Mass spectroscopy is another option for N4 detection, but it is difficult to tell whether an
m/z 56 fragment is due to a bound D2h or Td structure or is the result of a weakly bound (N2)2
complex. There is also a need for calculations to determine whether ionized N4(Td) is bound
and stable enough to detect in a mass spectrometer.
Isotopic labeling can be used in conjunction with the above techniques to confirm or rule
out certain origins of spectral lines and mass spectrometric fragments.

4.2

Roadmap to pentaaza HEDM:s ( N5+/N5-/N5+N5-)

The making of High Energy Density Materials with fivemembered nitrogen compounds
starts with the synthesis N5+ or N5-. Recently the novel homoleptic polynitrogen ion, N5+ (Fig
6) was synthesized [7]. This is the first new stable all nitrogen ion in more than 100 years. So
far no useful counter ion has been found. It has been suggested that cyclo-N5- (Fig 7) may be
combined with N5+ to form a N10 salt.

Fig 6. N5+ the first new all-nitrogen


compound in 100 years
242

Fig 7. Pentazolate anion, cyclo-N5-

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The nitrogen pentazole ring system has been studied since 1903, when Hantzsch made the
first attempts to prepare phenylpentazole [36]. The first preparation of substituted pentazoles
was however achieved by Huisgen and Ugi [37] in 1956. The making of pentazolate
compounds starts with decomposition of arylpentazoles (
Fig 8). A breakthrough was achieved when the feasibility of this process was confirmed
in a laser induced mass spectrometry experiment at our institute [8], in which the pentazolate
anion was detected from selective cleavage of p-dimethylaminophenylpentazole.
2N

N
N

Ti
N

N
N

N
N

N
N
R

N
N

N
N

Me

N
N

sol

N
N

N
N

N
N

Me=Fe, Mn, Co..........


N

N
N

Zn 2+

N
N

N
N

N
N

Fig 8. How to make pentazolate HEDMs.


Calculations have shown that the aromatic pentazole anion (N5-) should be stable enough
to isolate, especially if trapped as a metal complex [38, [39, [40, [41, [42, [43]. This complex could,
if a suitable metal were used as the complexing agent, be highly energetic. More important,
it could serve as a N5--containing compound, stable enough for an ion exchange to be
performed. Some metals are believed to stabilize the pentazole ring, both in the
arylpentazole and in pure ionic form. In a recent study [44] to investigate if amination of
halobenzenes is a viable way to isolate the pentazolate anion, the reactivities of halobenzenes
were compared to phenylpentazole. The reaction energetics for phenylpentazole became
favorable after complexation with Zn2+. The study indicated that addition of Zn2+ lowers the
free energy of activation by more than 6 kcal/mol, which may be sufficient to make the
reaction useful for synthesis.
Once a stable complex with N5- has been made, an ion exchange can be attempted in
order to find a suitable counter ion that makes a good HEDM. Calculations show that a salt
with the recently discovered [7] N5+ ion would be highly energetic, with a performance as an
explosive about 1.6 times that of HMX [45], provided of course that it is stable. QM
calculations on the stability of N5+N5- are not conclusive. There are three studies of its
stability, two of which predict that N5+N5- is stable and one that predicts its instability.
The first publication, by Fau et al. [46, [47], concludes that N5+N5- is a stable compound
with a dissociation barrier of 19 kcal/mol and high density (1.9 g/cm3) and heat of formation
13

Henric stmark

243

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

(296 kcal/mol, 1239 kJ/mol). They conclude that its performance as a rocket propellant is
about the same as that of hydrazine but that its high density (about twice that of hydrazine)
will allow for smaller and lighter rockets.
The second publication, by Evangelisti et al. [48], concludes that N5+N5- is stable and that
(N5+N5-)2 is even more stable. They also conclude that the dimer stability suggests the
possibility of a N5+N5- crystal. Worth noticing here is that even though no crystal structure is
calculated, calculations indicate that (N5+N5-)2 has a planar, layered structure. This is very
interesting since it is believed that such structures give less sensitive energetic materials [49].
The third publication, by Dixon et al. [50], concludes that both N5+N5- and N5+N3- are
unstable. They have also experimentally confirmed that N5+N3- really is unstable.
Interestingly enough, despite their negative results, they state that a hypothetical
polynitrogen compound, such as N5+N5-, would be an excellent monopropellant for rocket
propulsion or explosives, if it could be synthesized in a stable form.
The fourth publication by Gagliardi et al. [51] concludes that the reaction to form a
bicyclic form of N10 from free ions proceeds without a barrier. If we, however, allow the
two ions to approach each other along the energetically most favourable path, an ion pair
can be formed that is bound with 117 kcal/mol with respect to the free ions. The ion pair is
24 kcal/mol higher in energy than the bicyclic form.
Calculations using Cheetah 2.0 give Isp=281 s for a HTPB/N5+N5- mixture with 70 vol%
solids.

5.

METASTABLE STATES: POLYMERIC NITROGEN

The natural state (low energy form) of the elements nitrogen and hydrogen is as diatomic
molecules, N2 (triple bond) and H2 (single bond). Much energy can therefore be stored in
nitrogen molecules with single or double bonds. A form of nitrogen with all single bonds
and therefore high energy content is polymeric nitrogen. It also has high density (compare
with carbon in the form of graphite versus diamonds). Theory suggests that if a volume of
fluid N2 is sufficiently squeezed, the nitrogen will give up its diatomic structure and form a
continuous network of single covalent bonds [52]. This chemical structure is similar to that of
its chemically similar group V neighbours phosphorus and arsenic. The energy released
when the densely packed polymeric nitrogen is released to the diatomic state would be
pollution free and enough to qualify it as an interesting future HEDM. Theoretical
calculations first predicted the existence of a metastable polymeric nitrogen phase at the
pressure 65 GPa [52]. In high pressure experiments several new phases of nitrogen have been
found, among them a non-molecular phase that could be recovered to ambient pressure [53].
The most promising cubic gauge structure (cg-N) was predicted to have high energy and
high density compared to -N2.
The theoretically estimated performance is Isp 516s, density 3.9 g/cm3, and the
performance as an explosive is 10.6 times the performance of HMX. This is based on
thermochemical calculations with Cheetah 2.0 made with BKWC equations of state and with
data from Mailhiot [52]. Notice that the equation of state used (BKWC) only is calibrated up
to 40 GPa so this is a very large extrapolation.

244

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 9. Polymeric cg-N-structure: Each nitrogen atom is connected to three neighbours by


three single covalent bonds [54].
A group under Mikhail Eremets has managed to prove its existence [55]. To synthesize it,
a diamond anvil cell was used (Fig 10). It is a robust device with a small (~20 m diameter)
sample chamber within which pressures exceeding that at the centre of the earth (~350 GPa)
have been reached. In this experiment 110 GPa at temperatures above 2000 K was reached
by heating of a laser. The existence of cubic gauge structure was experimentally verified by
X-ray and Raman measurements.

Fig 10. A cross-section of the sample arrangement. The heated nitrogen sample is coloured
blue and untransformed molecular nitrogen is red. The laser beam is absorbed by a black,
thin, boron plate resting on c-BN pieces that thermally insulate the plate from the bottom
anvil. The sample squeezed between the anvils is surrounded by the c-BN/epoxy gasket
followed by the metallic (Re) supporting ring [55].
To get a useful substance it needs to have a huge hysteresis enabling it to be recovered to
ambient pressures. In the experiment above the structure was stable down to 42GPa at room
temperature. At low temperatures the pressure could be further released, at 140K it reached
25GPa before the gasket ruptured. Its a typical problem of the current technique and a novel
method of recovering the sample to ambient pressure needs to be developed. Theory predicts
that this structure is metastable at atmospheric pressures [52, [56]. It would then be an analogue
to carbon that is stable as graphite at atmospheric pressure but also metastable as diamond
created at very high pressures and temperature. If cg-N is not metastable at ambient pressure,
it can perhaps be stabilized in compounds with other elements or by introducing impurities
[54]
.

15

Henric stmark

245

New Trends in Research of Energetic Materials

6.

Czech Republic

2006

[Content]

CONCLUSIONS

The available energy in exotic materials such as nuclear isomers and antimatter is orders
of magnitude larger than the energy of HMX (Fig 11). These have very high potential but the
risk of failing is also very high. On a more intermediate level, materials such as all-nitrogen
compounds have high potential and lower (but still high) risk. Other materials such as
reactive materials based on nanotechnology and HEDM by conventional chemistry have
good potential, will probably be accessible in shorter time and involve much less risk of
failure. In Table 4 a summary of potential HEDM is given. An increase of performance of
three to six times the performance of HMX would revolutionize warfare. This can be
compared to the performance increase of three times at the end of the black powder era
which is known to have changed the battlefield enormously.

Fig 11. Energy content of some HEDMs. The figure is from LLNL (internet). The point for
Anti-matter has been added to the figure. The comment in the figure refers to nuclear
isomers.

246

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Detonation pressure (GPa)

Energy (HMX=100, V=2.2 V0)

Energy (HMX=100, V=40 V0)

-708

7.3

21.2

60

65

226

992

1.65

-63

6.9

19.6

55

63

210

897

1.905

74.75

9.1

38.5

105

103

266

1397

2.04

393

10

47.8

121

116

273

1380

1.9

1239

12.1

62

162

156

347

~2000

2.3

761

15.5

121

308

288

422

3700

N60

1.97

6780

12.3

65

161

150

331

2296

Polymeric nitrogen

3.9

290

30

660

1060

-a

513

-a

2.38

795

10.9

131

220

-a

288

-a

Density (g/cc)

Force (J/g)

Detonation velocity (mm/s)

1.65

Energetic material

Impulse (s)

Heat of formation (kJ/mol)

Table 4. Comparison of performance for common high explosives, high nitrogen compounds
and some theoretical HEDMs, calculations using Cheetah 2.0 with BKWC EOS.

Reference compounds:
NC, Nitrocellulose
O 2N

NO 2

TNT

O2N
NO 2
N

O 2N

NO 2

HMX

NO 2

O2N

NO2
N NO
2

O2N
N

CL-20 O N
2

NO2

HEDM:
N5+N5N

N4(Td)

O
N
N

N
O

TTTO
a

17

N
N

N
O

Calculations do not converge.

Henric stmark

247

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]

[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]

248

WALLIN, S., STMARK, H., WINGBORG, N. et al. (2004) High Energy Density Materials
(HEDM) - A Literature Survey (Stockholm, Swedish Defence Research Agency).
FRECHE, A., AVILES, J., DONNIO, L. & SPYCKERELLE, C. (2000) Insensitive RDX, I-RDX,
Paper presented at the Insensitive Munitions and Energetic Materials Technology
Symposium, San Antonio.
FRECHE, A., LECUME, S., SPYCKERELLE, C., AVILES, J. & DONNIO, L. (2003) Insensitive
Nitramines, Paper presented at the Europyro 8e Congrs International de Pyrotechnic, SaintMalo.
EATON, ZHANG, M.-X., GILARDI, R. et al.: Octanitrocubane: A New Nitrocarbon,
Propellants, Explosives, Pyrotechnics, 27, p. 1-6.2002
VEDACHALAM, M., RAMAKRISHNAN, V. T., BOYER, J. H. et al.: Facile Synthesis and
Nitration of cis-syn-cis-2,6-dioxo-1,3,4,5,7,8-hexanitrodekahydro-1H,5H-diimidazo[4,5b:4',5'-e]pyrazin, Journal of Organic Chemistry, 56, p. 3413-3419.1991
PAGORIA, P. F., LEE, G. S., MITCHELL, A. R. & SCHMIDT, R. D.: Thermochimica Acta, 384,
p. 187-204.2002
CHRISTE, K. O., WILSON, W. W., SHEEHY, J. A. & BOATZ, J. A.: N5+: A Novel Homoleptic
Polynitrogen Ion as a High Energy Density Material, Angewandte Chemie International
Edition, 38, p. 2004-2009.1999
STMARK, H., WALLIN, S., BRINCK, T. et al.: Detection of pentazolate anion (cyclo-N5-)
from laser ionization and decomposition of solid p-dimethylaminophenylpentazole.,
Chemical Physics Letters, p.2003
HAHMA, A., HOLMBERG, E., HORE, N. et al. (2002) Synthesis and Characterization of
Phenylpentazoles, Paper presented at the 33rd International Annual Conference of ICT,
Karlsruhe.
VIJ, A., PAVLOVICH, J. G., WILSON, W., VIJ, V. & CHRISTE, K. O.: Experimental Detection
of the Pentaazacyclopentadienide (Pentazolate) Anion, cyclo-N5-, Angewandte Chemie
International Edition, 41, p. 3051-3053.2002
BUTLER, R. N., STEPHENS, J. C. & BURKE, L. A.: First generation of pentazole (HN5,
pentazolic acid), the final azole, and a zinc pentazolate salt in solution: A new N-dearylation
of 1-(p-methoxyphenyl) pyrazoles, a 2-(p-methoxyphenyl) tetrazole and application of the
methodology to 1-(p-methoxyphenyl) pentazole, Chemical Communications, p. 10161017.2003
SCHROER, T., HAIGES, R., SCHNEIDER, S. & CHRISTE, K. O.: The race for the first generation
of the pentazolate anion in solution is far from over, Chemical Communications, p. 16071609.2005
NGUYEN, M. T.: Polynitrogen compounds 1. Structure and stability of N4 and N5 systems,
Coordination Chemistry Reviews, 244, p. 93-113.2003
BARTLETT, R. J. Structure and Stability of Polynitrogen Molecules and Their Spectroscopic
Characteristics.
GLUKHOVTSEV, M. N., JIAO, H. & SCHLEYER, P. V. R.: Besides N2, What Is the Most Stable
Molecule Composed Only of Nitrogen Atoms? Inorganic Chemistry, 35, p. 7124-7133.1996
MURRAY, J. S., LANE, P., BRINCK, T. et al.: Relationships of Critical Constants and Boiling
Points to Computed Molecular Surface Properties, Journal of Physical Chemistry, 97, p.
9369-9373.1993
KARLE, J. (2000) A discussion of azacubane and some peripheral remarks about other N8
compounds, Paper presented at the HEDM contractors meeting, Park City, Utah.
FRANCL, M. M. & CHESICK, J. P.: The N4 Molecule and Its Metastability, Journal of Physical
Chemistry, 94, p. 526-528.1990
LEE, T. J. & MARTIN, J. M. L.: An accurate quartic force field, fundamental frequencies, and
binding energy for the high energy density material TdN4, Chemical Physics Letters, 357, p.
319.2002

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

[20]
[21]

[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]

[41]

19

Czech Republic

2006

[Content]

YARKONY, D. R.: Theoretical Studies of Spin-Forbidden Radiationless Decay in Polyatomic


Systems: Insights from Recently Developed Computational Methods, Journal of the American
Chemical Society, 114, p. 5406-5411.1992
BRINCK, T., BITTEREROVA, M. & STMARK, H. (2003) Electronic structure calculations as a
tool in the quest for experimental verification of N4, in: Politzer, P. A. & Murray, J. S. (Eds.)
Energetic Materials, Part 1: Decomposition, Crystal and Molecular Properties, pp. 421-439
(Elsevier).
ZHENG, J. P., WALUK, J., SPANGET-LARSEN, J., BLAKE, D. M. & RADZISZEWSKI, J. G.:
Tetrazete (N4). Can it be prepared and observed? Chemical Physics Letters, 328, p. 227233.2000
CACACE, F.: From N2 and O2 to N4 and O4: Pneumatic Chemistry in the 21st Century,
Chemistry - A European Journal, 8, p. 3838-3847.2002
CACACE, F., PETRIS, G. D. & TROIANI, A.: Experimental Detection of Tetranitrogen, Science,
295, p. 480-481.2002
NGUYEN, M. T., NGUYEN, T. L., MEBEL, A. M. & FLAMMANG, R.: Azido-Nitrene is
Probably the N4 Molecule Observed in Mass Spectrometric Experiments, Journal of Physical
Chemistry A, 107, p. 5452-5460.2003
BITTEREROV, M., BRINCK, T. & STMARK, H.: Theoretical study of the triplet N4 potential
energy surface, Journal of Physical Chemistry A, 104, p. 11999-12005.2000
SCULLMAN, R., WALLIN, S., LAUBERSHEIMER, F. & REHM, O.: Luminescence from Matrices
of Nitrogen and Argon Excited by -Particles, Physica Scripta, 52, p. 172-177.1995
BITTEREROV, M., STMARK, H. & BRINCK, T.: A theoretical study of the azide(N3) doublet
states. A new route to tetraazatetrahedrane (N4); N+N3->N4, Journal of Chemical Physics,
116, p. 9740.2002
HANSEN, N. & WODTKE, A. M.: Velocity Map Ion Imaging of Chlorine Azide Photolysis:
Evidence for Photolytic Production of Cyclic-N3, Journal of Physical Chemistry A, 107, p.
10608-10614.2003
MAIER, G., PFRIEM, S., SCHFER, U. & MATUSCH, R.: Angewandte Chemie, International
Edition, England, 17, p. 520.1978
STMARK, H., LAUNILA, O., WALLIN, S. & TRYMAN, R.: On the possibility of detecting
tetraazaterahedrane (N4) in liquid or solid nitrogen by Fourier transform Raman
spectroscopy, Journal of Raman Spectroscopy, 32, p. 195-199.2001
STMARK, H., BITTEREROV, M., BRINCK, T. et al. (2002) The N4 project. Annual report for
the year 2001, pp. 69 (Swedish Defence Research Agency).
STMARK, H., WALLIN, S., HORE, N. & LAUNILA, O.: Raman spectra of P4 at low
temperatures, Journal of Chemical Physics, 119, p. 5918-5922.2003
BITTEREROV, M., BRINCK, T. & STMARK, H.: Theoretecal study of the singlet
electronically excited states of N4, Chemical Physics Letters, 340, p. 597-603.2001
BITTEREROV, M., STMARK, H. & BRINCK, T.: Ab initio study of the ground state and the
first excited state of the rectangular (D2h) N4 molecule, Chemical Physics Letters, 347, p.
220-228.2001
HANTZSCH, A.: Ber. d. Chem. Ges., 36, p. 2056.1903
HUISGEN, R. & UGI, I.: Zur Lsung eines klassischen Problems der organischen StickstoffChemie, Angewandte Chemie, 68, p. 705-706.1956
NGUYEN, M. T., MCGINN, M. A., HEGARTY, A. F. & ELGURO, J.: Can The Pentazole Anion
(N5-) be Isolated and/or Trapped in Metal Complexes, Polyhedron, 4, p. 1721-1726.1985
GLUKHOVTSEV, M. N., SCHLEYER, P. V. R. & MAERKER, C.: Pentaaza- and
Pentaphosphacyclopentadienide Anions and Their Lithium and Sodium Derivatives:
Structures and Stabilities, Journal of Physical Chemistry, 97, p. 8200-8206.1993
BURKE, L. A., BUTLER, R. N. & STEPHENS, J. C.: Theoretical characterization of pentazole
anion with metal counter ions. Calculated and experimetnal 15N shifts of aryldiazonium, azide and -pentazole systems, Journal of the Chemical Society, Faraday Transactions 2, p.
1679-1684.2001
LEIN, M., FRUNZKE, J., TIMOSHKIN, A. & FRENKING, G.: Iron Bispentazole Fe(5-N5)2, a
Theoretically Predicted High-Energy Compound: Structure, Bonding Analysis, Metal Henric stmark

249

New Trends in Research of Energetic Materials

[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]

[51]
[52]
[53]
[54]
[55]
[56]

250

Czech Republic

2006

[Content]

Ligand Bond Strength and a Comparison with the Isoelectronic Ferrocene, Chemistry - A
European Journal, 7, p. 4155-4163.2001
STRAKA, M. & PYYKK, P.: One Metal and Forty Nitrogens. Ab Initio Predictions for
Possible New High-Energy Pentazolides, Inorganic Chemistry, 42, p. 8241-8249.2003
TSIPIS, A. C. & CHAVIARA, A. T.: Structure, Energetics, and Bonding of First Row
Transition Metal Pentazolato Complexes: A DFT Study, Inorganic Chemistry, 43, p. 12731286.2004
CARLQVIST, P., STMARK, H. & BRINCK, T.: Computational Study of the Amination of
Halobenzenes and Phenylpentazole. A Viable Route to Isolate the Pentazolate Anion?
Journal of Organic Chemistry, 69, p. 3222-3225.2004
STMARK, H., GOEDE, P., WALLIN, S. et al. (2004) Development of energetic materials over
the time span 2004-2025, Special forecast (Stockholm, Swedish Defence Research Agency,
FOI).
FAU, S., WILSON, K. J. & BARTLETT, R. J.: On the Stability of N5+N5-, Journal of Physical
Chemistry, 106, p. 4639-4644.2002
FAU, S., WILSON, K. J. & BARTLETT, R. J.: Correction: On the Stability of N5+N5-, Journal
of Physical Chemistry, 108, p. 236.2004
EVANGELISTI, S. & LEININGER, T.: Ionic nitrogen clusters, Journal of Molecular Structure
(Theochem), 621, p. 43-50.2003
BEMM, U. & STMARK, H.: Structure of 1,1-diamino-2,2-dinitroethylene: A novel Energetic
Material with Infinite Layers in Two Dimensions, Acta Crystalography, C54, p. 19971999.1998
DIXON, D. A., FELLER, D., CHRISTE, K. O. et al.: Enthalpies of Formation of Gas-Phase
N3,N3-,N5+, and N5- from Ab Initio Molecular Orbital Theory, Stability Predictions for N5+N3and N5+N5-, and Experimental Evidence for the Instability of N5+N3-, Journal of the American
Chemical Society, 126, p. 834-843.2004
GAGLIARDI, L., ORLANDI, G., EVANGELISTI, S. & ROOS, B. O.: A theoretical study of the
nitrogen clusters formed from the ions N3-, N5+ and N5-, Journal of Chemical Physics, 114, p.
10733-10737.2001
MAILHIOT, C., YANG, L. H. & MCMAHAN, A. K.: Polymeric nitrogen, Physical Review B,
46, p. 14419-14435.1992
EREMETS, M. I., HEMLEY, R. J., MAO, H.-K. & GREGORYANZ, E.: Semiconducting nonmolecular nitrogen up to 240 GPa and its low-pressure stability, Nature, 411, p. 170174.2001
MCMAHAN, A. & LORENZANA, H. E. (1994) Polymeric Nitrogen: A potential compound to
store energy Energy & Technology Review, LLNL periodical, pp. 10-11.
EREMETS, M. I., ET AL: Single-bonded cubic form of nitrogen, Nature materials, 3, p. 558562.2004
BARBEE III, T. W.: Metastability of atomic phases of nitrogen, Physical Review B, 48, p.
9327-9330.1993

High Energy Density Materials (HEDM): Overview, Theory and Synthetic . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

AEROBIC BIODEGRADATION OF
MONONITROTOLUENES IN DIFFERENT REACTOR
CONFIGURATION
Jan Paca*1, Martin Halecky1, Rakesh Bajpai**
*

Department of Fermentation Chemistry and Bioengineering, Institute of Chemical


Technology, Prague, Technicka 5, 166 28, Czech Republic
Tel: +420-2-24353785, Fax: +420-2-24355051, Email: Jan.Paca@vscht.cz
**
Department of Chemical Engineering, University of Missouri, Columbia, MO 65211, USA
Tel: (573) 882 3708, Fax: (573) 884 4940, Email: bajpair@missouri.edu
1

Presenting and corresponding author

Abstract:
Degradations of nitrotoluenes, individually and in a mixture, were carried out in batch and
continuous aerobic reactors by a defined mixed microbial culture. The degradation rates and
efficiencies of the isomers were evaluated in batch and continuous reactors. The results
demonstrated that all the three-mononitrotoluene isomers were degraded simultaneously and
completely in presence of excess oxygen. The nitro group position on the benzene ring
influenced the degradation rates of the individual MNT isomers in batch systems. In the
continuous biofilm reactor with a sufficient biocatalyst quantity, quality and excessive
oxygen supply rate, the degradation rates of the mononitrotoluenes were almost identical as
long as the compounds were present individually and their loading did not exceed the
capacity of the catabolic master reaction. The microbial composition of the biofilm changed
qualitatively and quantitatively during long-term continuous operation under aerobic nonaseptic conditions. This complex investigation resulted in data that can be applied for the
scale-up procedure for field experiments.
Keywords: Degradation rate and efficiency, individual and mixtures of MNT, interactions in
catabolism, biofilm composition and surface structure.

1. INTRODUCTION
Explosives and other nitrated compounds are widely distributed environmental
contaminants [1]. Nitrophenols and nitrotoluenes are used extensively as raw materials and
intermediates in industry, especially in ammunition plants.
Extensive research has focused on the microbial removal of nitroaromatic compounds
in batch [3] and continuous reactors [4]. Although nitrotoluenes are not easily
biodegradable, several microorganisms with ability to use nitrotoluenes as the sole carbon,
energy and nitrogen sources have been reported [5]. Biodegradation of nitrotoluenes has been
demonstrated under both aerobic and anaerobic conditions [3].

[2]

The objective of this work was to study degradation of the MNTs individually and in a
mixture in batch and continuous reactor systems. The results of biodegradation by a mixed
1

Jan Pca, Martin Halecky, Rakesh Bajpai

251

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

culture are reported here. The mixed culture was isolated in our laboratory from
contaminated soil from an ammunition plant (Explosia Co., Pardubice, Czech Rep.).
Analyses of the microorganisms in the mixed culture in the inocula and in biofilm after four
months of continuous operation are presented. Observation of the packing material and
biofilm surfaces from a packed bed reactor (PBR) were carried out using an electron
microscope during a long-term continuous experiment and these are also reported here.

2. MATERIALS AND METHODS


2.1. Microorganisms
Soil contaminated with a mixture of mononitrotoluenes, 2,4-DNT and 2,6-DNT was
collected from a Czech Ammunition plant in Pardubice, Czech Republic. A mixed culture
was enriched from this contaminated soil and several pure cultures capable of growing on a
mixture of mono- and dinitrotoluenes were isolated from this mixed culture.

2.2. Media, Culture Conditions, and Reactor Systems


A basal salt medium (BSM) containing 3.40 g KH2PO4,
4.30 g K2HPO4, 0.80 g KNO3, 0.34 g MgCl2.6H2O, and 1 mL
trace element solution per liter was used for preparation of
inocula and for the degradation experiments. Batch degradation
tests were performed in triplicate in 500 mL Erlenmeyer flasks
with working volumes of 100 mL at 26 C in a rotary shaker at
120 rpm. All the flasks for batch experiments were wrapped with
an aluminum foil in order to avoid any photocatalytic (abiotic)
degradation. All the experiments were conducted in triplicate.
For continuous degradation studies, a packed bed reactor
(PBR) with an internal diameter of 22 mm and a bed height of
240 mm was used (Figure 1). The cooling jacket of the reactor
was wrapped in an aluminum foil to prevent light exposure. The
packing material was expanded slate with a particle size 4 to 8
mm. The selected defined mixed culture (Table 1) was
immobilized on the packing material. The reactor was operated in
cocurrent up-flow mode. Inlet pH of the medium was 7.2.
Concentrations of individual MNTs in the feed stream to the
packed bed reactor were 8, 15, 20, and 30 mg.L-1.
Figure 1. Schematic of PBR. 1 packing, 2 medium influent, 3 medium effluent, 4 air
inlet, 5 air outlet, 6 tempering jacket, 7 combined electrode for dissolved oxygen
concentration and temperature, 8 sampling port.

2.3. Analytical methods


Sample analyses were conducted by HPLC (System DeltaChrom, Watrex Prague Ltd.,
Prague, Czech Republic) using WATREX column (250x4 mm Nucleosil, 120-5 C18,
Watrex Praha, s.r.o. Czech Republic), a mobile phase of methanol/water (50:50) at 30 C
and a flow rate of 1 mL min-1, and measuring absorbance with a diode array detector (Model
UV 6000 LP, Thermo Separation Products Inc., San Jose, CA, USA) at 230 and 238 nm.
Under these conditions, retention times of 2-NT, 3-NT and 4-NT were 22.2, 26.2, and 23.7
min.
pH of the medium was measured at the end of batch degradations and in the outflow
252

Aerobic Biodegradation of Mononitrotoluenes in Different Reactor Configuration

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

from the PBR using a pH meter, model MFD 77 (INSA Corp., Czech Rep.).
An oxygen probe (Type CSOT44LL, INSA Corp., Czech Rep.) was used to measure
dissolved oxygen concentration (DOC) in the outflow from PBR. The data were collected
and stored in a computer. DOC was maintained above 5 mg.L-1 in PBR effluent by
manipulating airflow rate.

2.4. Microbial analyses


Identification of the individual strains in free cells or in biofilm was performed using
several biochemical tests. Fungi were identified by their morphological characteristics
especially by the shape of their fructification organs. The cells of biofilm, immobilized on
expanded slate, were suspended in BSM by vigorous mixing before any microbial analysis.
Table 1. Identification of the inoculum strains.
G- bacteria
Pseudomonas putida A1
Pseudomonas veronii B1
Chryseobacterium sp. D1
Stenotrophomonas maltophilia D2
G+ bacterium
Paenibacillus glucanolyticus D1/B

2.5. Microscopy
Samples of the packing material with and without biofilm were used for scanning
electron microscopy (SEM). The packing material with biofilm was taken from a depth of 2
cm from the top of PBR after 4 months of operation.

2.6. Calculations
Degradation rate (qi) of individual nitrotoluenes (NT) in a batch process:
S0 S
t
Where, S final NT concentration (mgL-1); S0 starting NT concentration; t time
of the batch process (days).
qi =

The degradation rate of NTs in the packed bed reactor

F ( Sin S out )
VL
Where, qi degradation rate of the individual NT; Sin NT concentration in the
medium influent (mgL-1); Sout NT concentration in the medium effluent (mgL-1); F flow
rate of the medium (mLday-1); VL reactor liquid volume (mL). The removal efficiencies
(RE) of individual NTs, the overall degradation rate of the NT mixture (qw), and hydraulic
retention time (HRT, min) were calculated as
S
RE = (1 out ) 100
Sin
qw = qi
qi =

Jan Pca, Martin Halecky, Rakesh Bajpai

253

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

VL
F
Where F (mLmin-1) is the volumetric flow rate of fluid through PBR, and VL is the
liquid volume in the reactor (mL).
HRT =

Standard deviations of the concentrations in batch experiments were calculated from


parallel replicates and these are shown as bars in Figures 2-3. Degradation rates were
calculated from the experiments using Microsoft Excel statistic program and these exceeded
the reliability limit of 0.9.

3. RESULTS AND DISCUSSION

3.1. Degradation in a Batch System


The mixed culture was able to utilize the mononitrotoluenes (MNTs) as sole sources of
carbon and energy (Figure 2). From Figure 3 it is clear that at least 2-NT and 3-NT are
degraded simultaneously when present together in solution. 4-NT was still degraded the
fastest.
Another important finding throughout all the degradation experiments was that no
dead-end intermediates accumulated in the medium. Since all the analyses were carried out
in UV range using a diode array detector, all of the MNTs were considered degraded to
aliphatic compounds devoid of any unsaturated bonds.
Since a buffered medium was used, pH changes during the degradation of MNTs
individually and in a mixture were very low. The C/N/P ratio in experiments where MNTs
were the only C-, N-, and energy source was about 1/0.17/73.0.

3
2

1
0

t (days)

Figure 2. Degradation of the individual


NTs without any other C-,
energy, and N-sources in the
medium. () 2-NT; () 3-NT;
(); 4-NT.

254

-1

S (mg.L )

-1

S (mg.L )

t (days)

Figure 3. Courses of the individual NT


concentrations during a degradation
of the NTs mixture (without any
other C-, energy, and N-sources in
the medium). () 2-NT; () 3-NT;
(); 4-NT.

Aerobic Biodegradation of Mononitrotoluenes in Different Reactor Configuration

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Differences in the degradation rate of MNTs when present individually and in the
mixture are shown in Table 2. Taking into account the statistical error from the parallel
experiments, it is obvious that regardless to the S0 value, the presence of 4-NT and 2-NT in
the medium strongly suppressed the degradation rate of 3-NT. The degradation rates of 4-NT
and 2-NT remained the same as those reached when present individually in the medium.
From this finding it can be concluded that at such low starting NT concentrations the rates of
the master reaction of catabolism of both 4-NT and 2-NT are not affected by the presence of
the other MNTs in the medium. This behavior is also probably influenced with a large
excess of oxygen in the medium (DOC >5 mgL-1 in the outflow).
Table 2. Differences in the uptake rate of NTs when present individually and in a mixture.
Initial Concentration,
mg.L-1
5.0
4-NT
3-NT
2-NT

q (mg. L-1. day-1)


Individual NTs
Mixture of NTs
36.8 5
38.9 7
12.5 1
4.4 0.5
3.7 0.5
3.6 0.5

3.2. Continuous Degradation in a Packed Bed Reactor

-1

RE (%)

-1

q (mg.L .day )

Figure 4 shows the loading characteristics of individual MNTs in the PBR at a constant
HRT of 88 min. Increasing MNT concentrations in the influent changed the loading. 100%
degradation efficiency for 4-NT, and 3-NT was maintained at inflow concentrations of up to
22 and 20 mgL-1, respectively; the corresponding removal rates were 370 and 320 mg.L1.
day-1. The maximum removal efficiency of 2-NT was above 95 % but never 100% up to
feed concentration of 20 mgL-1 at this HRT. Above these threshold feed concentrations, the
removal efficiencies decreased with increasing concentration (4-NT and 2-NT dropped to
RE of 85 % and that of 3-NT to 95 %) in the experiment. In the feed concentration range
studied, the degradation rate continued to increase with feed concentration; the highest
degradation rates for 4-NT and 2-NT
100
800
was 460 mgL-1day-1 and it was 400
700
mgL-1day-1 for 3-NT. pH of the medium
80
decreased from 7.20 in the inflow to
600
7.03 in outflow with 3-NT and 2-NT;
500
60
the outflow pH remained above 7.15 for
400
4-NT.
40

300
200

20

100
0

0
0

10

4-NT (RE)
4-NT (q)

15

20

-1

25

S (mg.L )
3-NT (RE)
3-NT (q)

30

35

40

2-NT (RE)
2-NT (qs)

Figure 4. Comparison of loading characteristics


for the individual NTs in PBR. () 2-NT; ()
3-NT; () 4-NT; RE empty points; q full
points.
5

Jan Pca, Martin Halecky, Rakesh Bajpai

Measurements of the dissolved


oxygen concentration (DOC) in the
effluent showed that the airflow rate
had to be increased in order to keep
DOC above 5 mgL-1 during the loading
tests. This suggested an increase in
oxygen uptake rate of cells with
increased loading of MNTs.
This
observation supports assertion that the
MNTs were perhaps mineralized, rather
than just transformed into some
incompletely oxidized intermediates.
255

New Trends in Research of Energetic Materials


100

700

85

400

80

300

75

200

70

100

65

0
0

10

15

20

25

100

700

600

90

500

-1

85

400

-1

qw (mg.L .day )

95

RE (%)

-1

500

-1

90

qw (mg.L .day )

600

80

300

75

200

70

100

65

0
0

10

15

20

25

100

700

90

500

-1

600

85

400

-1

95

80

300

75

200

70

100

65

qw (mg.L .day )

RE (%)

95

RE (%)

Czech Republic

0
0

RE 4-NT

10

RE 3-NT

15

20

RE 2-NT

25

qw

Figure 5. Comparison of NT mixtures


degradations in PBR. RE: () 2-NT; () 3NT; () 4-NT; () qw. (a) Loading by 4NT; Sin = 8 mgL-1 of 3-NT and 2-NT each.
(b) Loading by 2-NT; Sin = 8 mgL-1 of 4NT and 3-NT each. (c) Loading by 3-NT;
Sin = 8 mgL-1 of 4-NT and 2-NT each

256

2006

[Content]

Due to changes in the gas hold-up,


it was necessary to correct the
medium flow rate to maintain
HRT constant.
A second set of experiments
focused on loading tests with a
mixture of nitrotoluenes in the
medium, again at the constant
HRT of 88 minutes. The loading
of 4-NT was changed by changing
its inflow concentration in a
medium containing 8 mgL-1 of 2NT and 3-NT. The results of
degradation rate (qw) of 4-NT and
removal efficiencies of all the
MNTs are shown in Fig. 5. The
removal efficiency of 4-NT was
100% in the entire loading range
tested (up 20 mgL-1). Nevertheless,
the removal efficiencies of 2-NT
and 3-NT did not reach 100% even
at the lowest 4-NT concentration
in the medium ( 8 mgL-1). All the
MNTs
were
degraded
simultaneously with no apparent
inhibition of 4-NT degradation by
either of the other MNTs.
Fig. 5 shows the same
parameters for the loading by 2NT in presence of 3-NT and 4-NT
in the packed bed reactor. Under
these conditions also, 100%
removal efficiency of 4-NT was
achieved in the entire loading
range. Although the RE of 3-NT
was 100%, it was similar to the
values reported in Figure 8a with
variable loading of 4-NT. For 2NT, the removal efficiency
dropped with increase in its
loading.
When loading of 3-NT was
changed at constant concentrations
of the other MNTS, the removal
efficiency of 2-NT and 3-NT was
strongly and negatively impacted
(Figure 5). At feed concentration
of 20 mg.L-1, the removal
efficiency of 4-NT also started to
drop.

Aerobic Biodegradation of Mononitrotoluenes in Different Reactor Configuration

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In presence of mixture of MNTs, simultaneous degradation of all the MNTs was


confirmed in the continuously operated reactor too and no dead-end intermediates were
observed in the medium. pH of the medium remained unchanged from inflow to outflow in
these experiments.

3.3. Microbial Analyses


An identification of the strains used for the selected mixed culture inoculum is shown
in Tab. 1. All the selected strains except Stenotrophomonas maltophilia D2 were able to
grow on a mixture of MNTs (10 mgL-1 each) as the sole C-, N-, and energy source. The
Stenotrophomonas maltophilia D2 was involved into the mixed culture as it was found in
almost all mixed cultures degrading a very broad spectrum of aromatic hydrocarbons.
The same microbial identification tests were carried out with resuspended cells taken
from the biofilm on the packing particles after four months of PBR operation. The results are
shown in Table 3. A comparison of Tables 1 and 3 shows that Pseudomonas putida A1 was
the only organism able to survive the strong selection pressure of various loading modes for
over four months of continuous operation and sustained degradation of MNTs.
Table 3. Identification of the strains in biofilm after four months of MNT degradation.
G- bacteria
Pseudomonas putida A1
Unidentified rods PCN 2
Kluyvera cryocrescens PCN 4
Sphingomonas sp. PCN 3
Fungi
Penicillium albicans
Aspergillus oryzae
Since the reactor operation was not aseptic, three other gram-negative bacteria and two
fungi were also found in the reactor. These supposedly were also primary MNT degraders,
i.e. they were able to grow on a mixture of MNTs as the sole C-, N-, and energy source.
To get a better idea of the packing material and the biofilm surface, scanning electron
microscopy was employed. Electron micrograph of a particle with the microbial film after
four months is shown in Figure 6. The biofilm surface is almost homogenous and the
bacterial cells appear to be trapped in some extracellular polymeric material. The cell
attachment to the packing surface is rather strong.

Jan Pca, Martin Halecky, Rakesh Bajpai

257

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Figure 6. SEM of the biofilm surface taken after four months of the reactor performance.

4. CONCLUSIONS
Much higher entire degradation rates of NT mixtures in PBR comparing with that one
from the batch process result from different catalyst composition and concentration in PBR.
Degrading various MNT mixtures in PBR, the metabolic interactions have been
demonstrated (cf Figures 5 and 3). Under conditions of the 4-NT loading the removal rates
of both 2-NT and 3-NT were not affected but their RE was not 100%. The total 4-NT
mineralization follows from a high oxygen supply rate necessary to maintain DOC > 5 mg.L1
and from no presence of any intermediate in the outflow medium. This finding coincides
well with that of Walia et al. [3].
With respect to the fact that 2-NT and 4-NT are precursors of DNT and are significant
contaminants at TNT manufacturing facilities while 3-NT is found only in trace amounts the
RE and degradation activities, that can be reached during NT mixture degradations, could be
high efficient (cf Figure 5).
Optimization of NT mixture degradations in the production wastewater could be
achieved using aerated continuously operated bioreactors with an immobilized enrichment
culture on a convenient packing material. In order to reach the complete oxidation
(mineralization) of NT mixtures from the influent wastewater, the important performance
parameters affecting the process optimization are organic load (OL), hydraulic retention time
(HRT), and dissolved oxygen concentration (DOC).
Tha data can be applied for the scale-up procedure for field experiments.

258

Aerobic Biodegradation of Mononitrotoluenes in Different Reactor Configuration

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REFERENCES
[1] J. C. Spain: Introduction. In: Biodegradation of Nitroaromatic Compounds and
Explosives, pp. 1-5 ( Spain, J. C., Hughes, J. B., Knackmuss, H. J., Eds.), Boca Raton,
FL, Lewis Publishers, 2000
[2] D. F. Carpenter, N. G. McCormick, J. H. Cornell, A. M. Kaplan: Microbial
transformation of 14C-labeled 2,4,6-trinitrotoluene in an activated-sludge system, Appl.
and Environ. Microbiol., 35, pp. 949-954, 1978
[3] S. K. Walia, S. Ali-Sadat, G. R. Chaudry: Influence of nitro group position on
biotransformation of nitrotoluenes in Pseudomonas strain OU83, Pesticide Biochem.
Physiol., 76, pp. 73-81, 2003
[4] C. M. Peres, B. Van Aken, H. Naveau, S. N. Agathos: Continuous degradation of
mixtures of 4-nitrobenzoate and 4-aminobenzoate by immobilized cells of Burkholderia
cepacia strain PB4, Appl. Microbiol. Biotechnol., 52, pp. 440-445, 1999
[5] B. E. Haigler, J. C. Spain: Biodegradation of 4-nitrotoluene by Pseudomonas sp. strain
4NT, Appl. Environ. Microbiol., 59, pp. 2239-2243, 1993
Acknowledgment:
The study was financially supported by the Czech Science Foundation Project
104/04/0686.

Jan Pca, Martin Halecky, Rakesh Bajpai

259

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

POLYFUNCTIONAL N-NITRAMINES STRUCTURE


DIFFERENTIATION AS A BASIS FOR SIMULATION OF THEIR
DECOMPOSITION MECHANISM

Tatyana V. Petukhova1, Victor P. Ivshin1,


Vyacheslav L. Korolev2, and Tatyana S. Pivina2
1

Mari State University, Lenin square 1, Yoshkar-Ola 424000,


Mari El Republic, RUSSIA, E-mail: organic@marsu.ru

Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences,


Leninsky Prospect 47, Moscow 119991, RUSSIA

ABSTRACT
The structural classification of different chemical classes of N-nitramines has been
developed and the differentiation of their types throughout the functional surroundings
has been proposed. Basing on it and using up-to-date experimental data on their
thermal decomposition mechanisms the set of generators for chemical reactions that are
typical during nitramines decomposition process has been elaborated. The original
schemes for different types of N-nitramines thermolysis reactions have been designed
and the initial stage activation energy of it has been calculated by using the B3LYP/631G* level of density functional theory. As the result the most favorable pathways of
compounds decomposition have been displayed.
The suggested methodology for thermochemical processes simulation can be used for
a set of practical problems solution, including the investigations of the mechanisms of
decomposition, ignition, combustion, and detonation of energetic materials as well as
for structure-properties relationships study and for the search of prospect highenergy substances structure.

INTRODUCTION
The systematical investigations of structure and properties relationships for energetic
materials (aliphatic and aromatic nitrocompounds, nitramines and nitroethers, heterocycles
and others) are in progress from the end of the fifties last century. Both individual
compounds and the ascertainment of common features of structure-property relationships
have been investigated. N-nitramines take the first place among energetic materials (EM)
and the basic property of them is the thermostability concerned with the processes of
compounds decomposition.
260

Polyfunctional N-nitramines Structure Differentiation as a Basis for Simulation . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Although a large number of experimental investigations concerning the kinetics of


(EM) thermolysis were performed, the problem of possible pathways prediction of the
decomposition mechanisms remains quite topical. The problem is also of interest, from the
viewpoint of assessment of the capability of potentially energetic compounds, which have
not yet been synthesized. Meanwhile, theoretical methods for the solution of such a problem
are very insufficient at present. In addition, they mostly deal with the interpretation of
experimental data rather than with predictions. In this connection, we have formulated [1,2]
generalized problem: computer generation of possible homolysis pathways for a given
compound on the basis of its structural formula, as well as correction of these pathways on
the basis of some formal rules, which are specified in accordance with the experimental data.
The framed set of rules allows one to differentiate of the thermal decomposition
pathways of nitramines according to its molecular structure.
During energetic materials thermal decomposition process in gas phase thermolysis
reactions can pass practically throughout all known mechanisms: molecular dissociation with
free radicals formation, elimination, rearrangement with the subsequent decay, disclosing of
a cycle, etc. The prevalence one or other way of transformation during decomposition
process depends on molecular structure, first of all [3].
Against amine nitrogen surroundings there are possible to mark out four Nnitrocompounds groups:

..
O2N

..

CH2CH2R

..
O2N

CH2CH2R

CH2X
O2N

II

..
.. X

.. C X
O2N

III

IV

The specimens of the first group (I), where individed pair of amine nitrogen does not
take part in conjugation or other nonvalent orbital interactions [4] are characterized by the
highest thermal stability. Their decomposition begins with NNO2 bond homolysis.
In contrast of the first group of N-nitrocompounds, where are not the effects of
undivided electronic pair, the effects of heminal interaction can be came out in substances of
the group (II). This is a reason of the wide ranged variability in thermal constancy of the
second group members. In the case of symmetric heminal systems of the group (II), when
heteroatoms >N and X are approximately equal in basicity and inductive effect, there is
practically no difference in action with compounds of the group (I), since effects of heminal
interaction are multiply degenerated. In the case of assimetric systems (II) effects of heminal
interaction are fairly appeared. It deeply reduces thermal stability and changes the kinetics of
initial stage of thermolysis [5-7].
In compounds of the third class (III) the amine nitrogen of N-nitro group is directly
bonded with the unsaturated fragment. When there are no steric constraints for pN*C=X
conjunction, the sharp decrease of N-N-bond energy is observed that facilitate its hemolytic
scission or 1,3-sigmatropic shift of nitro group from nitrogen to X-atom.
The group (IV) has been formed by N-nitrocompounds containing amine nitrogen
directly connected with heteroatom. The close contact of heterosubstitute with N-nitro group
is accompanied by the weakening of bonds in N-nitro fragment according to inductive
scheme and scheme of exchange interactions [8,9].

Tatyana Petukhova, Victor Ivshin, Vyacheslav Korolev, Tatyana Pivina

261

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Basic scheme of thermal induced transformations of N-nitro compounds are presented at


the Fig. 1.

R' = Alk

R' ..
N
R

R' = CH2X

R N CH2+ X -

(I)

R'
NO2

RN(NO2)CH2
R' =

C C
C X

N2O CH2O

RNH C C NO2
RN

(III)

(IV)

RX

NO2

(II)
R

NO2

C C NO2

R' = H, C(O)R", C(O)OR",


CH=CH2, SiR3
RN
..
R' = Hal, N R
R

R' ..
N
R

OR'
O-

ROR' + N2O

NO2

Fig. 1. Thermal induced transformations of N-nitro compounds.

METHODS
Mechanism of chemical reactions during the decomposition process we presented as a
topologic structure, which is a function of source substances, intermediate and final
products. It can be invariant for any concrete chemical system. To define a mechanism of
decomposition we used a graph theory and formulated the general approach to the computerassisted modeling of the mechanism of homolytical reactions for nitrocompounds. A set of
chemical transformation generators based on experimental data has been formulated to
follow up the more probable decomposition reactions for compounds of different chemical
classes.
Screening of the most advantageous pathways from the energy standpoint were carried
out by calculations of the initial stage activation energy (B3LYP/6-31G* [11] level of
density functional theory [10]). Calculations were executed using GAUSSIAN 98 /12/ in the
Computational Centre of Zelinsky Institute of Organic Chemistry, Russian Academy of
Sciences.
As the result the most favorable pathways of compounds decomposition have been
displayed.

262

Polyfunctional N-nitramines Structure Differentiation as a Basis for Simulation . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

RESULTS AND DISCUSSION


As a model compound of the first group (I) we studied N, N-dimethylnitramine. Fig.2
represents four possible initial stages of decomposition of corresponding substance: (1) N-N
bond homolysis, (2) HNO2 elimination, (3) nitro-nitrite rearrangement, and (4) homolysis of
the C-N bond. As a result of computer generation of probable N, N-dimethylnitramine
thermolysis pathways we have got a whole series of formally probable reactions. The data
obtained can be used to understand the mechanism of thermal decomposition of
corresponding compound as well as to clarify the formation of nitrogen, protoxide of
nitrogen, H2O, carbon monoxide and carbonic acid as the final decomposition products.
CH3

CH3

CH2

CH2

NO2

+R

NO2

CH3

CH3
+M

+ONO

ONO

CH3

O.

CH3

+NO

N O CH3

+M
+R

+R

CHO HCN
+ONO

+R

. CH2ONO

CH3NO

+M

CH3OH HNO NO CH2O

+ONO

CH3NO2

CH3

CH3

+M

CH4

.
CH3

N OH

CH3
CH3

+NO

+R

+R

NO

+M

N CH3 N N OH + CH2O

+M

H2O HNO

CH2 N OH

CO CO2 + HNO

CO2

.
OH

N OH .
CH3 H2C

N2

CH3NO HCN

-NO

HONO

NO CH2 N CH3
+HONO

CH2

O
+M

+M

CH3

+M

. .
N2O CH2O H C N ONO CH3O . .NO
2

.O

CH3 N CH3

CH3

CH3ONO

CH4

NO2

+HONO

CH3OH
+R

+ONO

+R

H2O HCN O C N N OH

CH3O CO CO2 + HNO


+R

N2

H2CO2 CH2O

+NO

CHO

-NO

OH + N2O
+M

H2 O

CO2

Fig. 2. The fragment of N, N- dimethylnitramine thermal decomposition


scheme.
The calculated activation energy () of homolytic reaction of NN bond cleavage is
180.7 kJ/mol, compared to 336.0 kJ/mol for the CN bond cleavage and to 282.2 kJ/mol for
nitro-nitrite rearrangement reactions. In this connection, one can conclude that two lastnamed processes are unlikely in comparison with the N-N bond homolysis reaction in N, Ndimethyl nitramine. The calculated activation energy () of the reaction of HNO2
elimination is 190.2 kJ/mol (197.8 kJ/mol [13]) that is only for 9.5 kJ/mol higher than for the
homolytic cleavage of NN bond. Therefore the HNO2 elimination can be consider as the
alternative pathway of decomposition for secondary nitramines.
The simulation of the second group (II) nitramines was performed by the example of Nmethyl-N-chlorine methylnitramine. Fig. 3 illustrates the reaction scheme of the compound
decomposition process of the compound.

Here and at other schemes of decomposition processes the known experimentally reactions and products are
marked by the bold type.

Tatyana Petukhova, Victor Ivshin, Vyacheslav Korolev, Tatyana Pivina

263

New Trends in Research of Energetic Materials


Cl CH2

N ONO

CH3
Cl CH2
CH3

CH2Cl

.CH2

Cl

N O

+M

+R

CH3

+M

+M

CH3 N
CH4

HNO

+M

+HONO

+M

CH3NO NO

+NO

+R

N2O + OH

CO

+M

H2O

+ONO

CO2 + HNO

+NO2 -NO

+M

+M

HONO

+R

CO CO 2 + HNO

+M

CO2

H2O

+M

NO

CHO

+M

+NO2 -NO

N2O + OH CO 2

+R

+R

OH

H2O HNO

+R

HCN HONO
+ONO

N2O + OH

CO CO2 + HNO OH

+NO +NO -NO


2

+M

H2O

+M

N 2O H2CO H2C N ONO HCl

+ONO

Cl

CH2 N NO 2

HNO
+NO

CH4

+R

NO

+M

N2 CH3 OH CHO

+M

ONO

CHO HCN

+R

N OH CH2O HCl OH

Cl CH2 N NO 2

CH3

CH2 Cl HONO N2O H2CO H2C N

[Content]

CH2 N NO 2 CH3Cl

+M

CH3Cl

CH3

Cl CH2 N CH3 ONO


N NO 2

.
HCl CH

2006

N NO 2

CH3

Cl CH2

Czech Republic

CO2

+M

H2O

+M

NO
+M

H2O HNO

+NO

N 2O + OH
+M

H2O

Fig. 3. The fragment of N-methyl-N-chlorinemethylnitramine decomposition


scheme.
Considerations of the presented scheme allow the conclusion that under the atom of
chlorine introduction into the molecule of N, N-dimethylnitramine the subsidiary channels
of thermal decomposition arise through the cyclic complex with the subsequent elimination
of methylchloride and homolytic cleavage of the CCl bond. These opportunities provide the
explanation for HCl and H3Cl presence among the final decomposition products (the origin
of which was not quite clear before).
The energy of the NN bond cleavage in N-methyl-N-chlorine methylnitramine is 180.7
kJ/mol, for the CN homolysis the energy is 361.1 kJ/mol, for the CCl cleavage bond
295.7 kJ/mol, and (Ea) for nitro-nitrite rearrangement is 211.6 kJ/mol. Unfortunately in our
calculations the transition state localization of the cyclic activated complex formation was
not a success:

CH3
O2N N

Cl
CH2

CH3Cl + CH2

N NO2

The experimental energy activation of the initial stage of N-methyl-N-chlorine methyl


nitramine thermal decomposition is 128.5 kJ/mol [15].
The reactionary scheme of thermal decomposition process for N-vinyl-N-methyl
nitramine, as one of the third type compounds representatives, is shown in Fig. 4.
Analysis of the results of decomposition processes simulation follows up that under
introducing an unsaturated fragment direct fixed with N-nitro group into a molecule, a set of
rearrangements may be observed which are absent in the case of (I) and (II) groups
compounds. At the same time, the possible thermo decompositions channels and resulting
intermediates increase in number.
264

Polyfunctional N-nitramines Structure Differentiation as a Basis for Simulation . . .

New Trends in Research of Energetic Materials

CH2

CH

. ONO

N CH3

+M

+R

CH2 CH

+R

HONO

CH2

CH2 N

CH2

NO2

+HONO

CH2

CH2O CH2 CH N N OH OH
+M

CHO
+R

N2 CH3 CH O

H2O

+ONO

+R

+M

+R

HNO

HCN

+ONO

CH

CH CH2

CH

+M

CH3

+NO2 -NO

CO2

N2

CHO
+R

CO
+NO2 -NO

CO2

+ONO
+

CO2 HNO

CH3 NH CH CH NO2

NO2

CH3

N O
+
+M

CH2

CH

CH2

CH3 N
CH2 CH
N OH
CH3

CH4 CH2 CH .NO


CH2 HNO

_
O

N
O CH CH2

N OH

CH2 CH

+M

+M

+NO

CH

N O

CH3 CH2 CH N O CH2

.
OH + N2O O C N N OH CH2

CH4 CO CO2 +HNO


H2O

CH2

+M

CO2

N ONO

CH3

+HONO

+NO

CO CO2 +HNO CH3 CHO


+NO2 -NO

.
.NO . ONO CH2 N

[Content]

N NO2

CH2

CH3 N NO2

2006

CH
CH3

CH2

Czech Republic

CH3

+R

OH CH2

CH

CH2

OH
+M

H2O

OH + N2O +M
+M

H2O

H2O

+HONO

CH2O

CH2

N OH

CH3 CH O

N2

CHO
+R

CO

CH

+ONO

CO2 + HNO CH3 CHO

+NO2 -NO

CO2

+R

CH4

CO

+ONO

CO2 +HNO

+NO2 -NO

CO2

Fig. 4. The fragment of N-vinyl-N-methylnitramine decomposition scheme.


The calculated activation energy of the NN bond cleavage in N-vinyl-N-methyl
nitramine is 107.1 kJ/mol. That fits the assumption on the sharp NN bond energy decrease
for the (III) group compounds in comparison with the (I) group ones. The activation energy
of the unsaturated substitute NO shift is 226.2 kJ/mol. In case of CH3 elimination the
energy of CN bond homolysis is 271.8 kJ/mol and is 397.1 kJ/mol in case of vinyl radical
elimination. We failed to localize the nitro-nitrite rearrangement and the 1,3-sigmatropic
shift transition states. Though, the nitro-nitrite rearrangement is not likely since the molecule
has a plain configuration.
As a model of the fourth group (IV) compounds we selected N-methyl-N-chlorineamine.
Fig. 5 represents the scheme of its thermal decomposition process.

Tatyana Petukhova, Victor Ivshin, Vyacheslav Korolev, Tatyana Pivina

265

New Trends in Research of Energetic Materials


Cl

Czech Republic

2006

[Content]

N NO2

H3C

H3C

Cl

. ONO
+M

+R

H2C

Cl

Cl

HCl

+M

+R

HCN

H2O

N2O H2CO H2C N ONO Cl

HNO

O C N N OH N2O + OH

+R

+R

N2
+ONO

OH
+M

CO CO2 + HNO H2O


+NO2 -NO

CO 2

H2O

Cl

CH4 NO
+M

H3C

HNO

Cl

OH

N 2O HO

Cl

O
OCl

N O N O CH OCl
2
3
CH3 N O

CH4

+M

N2O + OH HCl
+M

+M

+M

HOCl

+ONO

H3C N

+M

O
+NO

HCN HONO

CO CO 2 + HNO OH
CO2

N ONO

H3C

+M

CHO

+M +NO2 -NO

CHO

+R

+R

+M

+M

Cl NH NO 2

NO 2 HCl

Cl

CH3

+M

NO

+NO

+HONO

CH2

N NO 2

Cl

+M

+R

HONO

N OH

H2C

+M

H3C N NO2 Cl

NO

+M

H2O HCl

+M

H2O

CH3

CH4

NO

+M

OH

+M

+M

HNO H2O

Cl
+M

HCl

+NO

H2O HNO

+NO

N 2O + OH

N 2O + OH
+M

H2O

+M

H2O

Fig. 5. The fragment of N-methyl-N-chlorinenitramine decomposition scheme.


As one can see the introducing of the heteroatom, directly fixed with the NO2 group into
the N-nitramine molecule, makes the competitive channel decomposition rising through the
C-X bond dissociation.
For the N-methyl-N-chlorinenitramine the activation energy values are: 105.8 kJ/mol
(for NN bond homolysis), 188.8 kJ/mol (for NCl bond homolysis), 302.2 kJ/mol (for CN
bond homolysis), and 233.8 kJ/mol (for the chlorine atom NO shift).
As a whole, the data obtained is evidence of the radical mechanism of N-methyl-Nchlorinenitramine decomposition through the N-N bond homolytic cleavage.

CONCLUSION
Thus, the structural classification of various chemical classes of N-nitramines has been
developed.
The results of thermolysis mechanisms simulations of the compounds have allowed to
predict the whole spectrum of formally possible reactions.
The activation energy calculations of decomposition reactions mark out the thermo
chemically most preferable ones.
As is shown, the change of the amine atom surroundings has an influence upon the
molecular structure that determines the modification of compounds thermostability and
makes for the appearance of subsidiary channels of decomposition.

266

Polyfunctional N-nitramines Structure Differentiation as a Basis for Simulation . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The authors are grateful to Dr. A.A. Porollo for the assistance in this work.

REFERENCES
[1]
[2]
[3]

[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]

A.A. POROLLO, D.E. LUSHNIKOV, T.S. PIVINA, V.P. IVSHIN, J. Mol. Struct.
(TEOCHEM), Vol. 391, p. 117-124, 1997.
A.A. POROLLO, D.E. LUSHNIKOV, T.S. PIVINA, V.P. IVSHIN, Izvestiya Akademii nauk,
10, p. 1869-1880,1999 (in Russian).
T.S. PIVINA, A.A. POROLLO, T.V. PETUKHOVA, and V.P. IVSHIN, In Proceedings of the
5th International Symposium on Special Topics in Chemical Propulsion (5-ISICP), Italy, p.
343-344, 2000
V.P. IVSHIN, O.A. Yashukova, V.A. Shlyapochnikov, Izvestiya Akademii nauk, 6, . 12951298, 1986.
F.I. DUBOVITSKY, B.L. KORSUNSKY, ADVANCES OF CHEMISTRY, Vol. 50, 1828-1871,
1981.
G.B. MANELIS, G.M. NAZIN, Yu.I. RUBTSOV, and V.A. STRUNIN: Thermal
decomposition and combustion of explosives and powders, Nauka, Moscow, 223 p., 1996.
Yu. SHU, B.L. KORSUNSKY, G.M. NAZIN, ADVANCES OF CHEMISTRY, 2004, Vol. 73, .
321-335, 2004.
M. GRAFF, C. JR GOTZMER, W.E. MCQUISTION J. Org. Chem., 10, . 3827-3830,
1967.
M. GRAFF, C. JR GOTZMER, W.E. MCQUISTION, J. Chem. Eng. Data, 14, 4, . 513514, 1969.
W. KOCH and M.C. HOLTHAUSEN: A Chemistss Guide to Density Functional Theory,
Wiley-VCH, Weinheim, 300 p., 2001.
T. KLARK: Computational chemistry (In Russian), Mir, Moscow, 383 p., 1990
Programme Complex Gaussian-98 of the Computer Centre for Chemical Investigations,
Russian Academy of Sciences.
M.A. JOHNSON, T.N. TRUONG, J. Phys. Chem., Vol. 103, p. 8840-8846, 1999.
B.L. KORSUNSKY, F.I. DUBOVITSKY, V.I. Losenova, Jurnal Physicheskoy Khimii (in
Russ.), 4, p.1159 1162, 1969.

Tatyana Petukhova, Victor Ivshin, Vyacheslav Korolev, Tatyana Pivina

267

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

PROCESS HAZARD EVALUATION FOR EXPLOSIVES


MANUFACTURING USING TEMCLEV-EX PROCEDURES.

Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak

Instytut Przemysu Organicznego


Annopol 6 str. 03-236 Warszawa, Poland
(+48 22) 811-12-31 ext. 298, piotrowski@ipo.waw.pl

Abstract
The paper presents the rules of the process hazard indices formation: technological T,
protection TZAB and total process hazard ZP. The method of dividing the whole
installation into the process hazard classes KZP has been presented. The evaluation
carried out contains so called material part ensued from the dangerous properties of
the media used in process as well as from the process conditions, devices (apparatuses)
used and even work organization and safety management.
Key words:
Process safety, process hazard indices (material, technology,
protection), hazard evaluation and classification.

1.

INTRODUCTION

System called Temclev-Ex offers a rapid hazard assessment method for use in the
industry of explosives and pyrotechnics or in plant design / modernization. The indexes used
in it are primarily concerned with the fire and explosion problems. Toxicity is considered
only as an additional, partial factor. A brief description of the base Temclev System and its
modification Temclev-Ex has been presented at the 7-th Seminar two years ago. This
time we would like to present the most important part of the system, a method of the process
hazard indices and protection indices formation.

2.

EVALUATION OF PROCESS HAZARDS

The detailed work of hazard evaluation and classification is best done by a team, ideally
including personnel familiar with the technology used (plant and process), as well as a
member with experience of using the system details. As it was earlier presented the equation
describing a hazard level of a process unit ZP in Temclev-Ex for manufacturing and
processing explosives presents as follow:
ZPEx = [(ARw + BRz) SEx] [T/Tzab],

268

(1)

Process Hazard Evaluation for Explosives Manufacturing Using TEMCLEV-EX . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

where: A, B the notes of Rz,w parameters,


RW,Z indices describing explosives parameters according to Polish legislation.
T - technological index, SEx - specific material properties index, TZAB- protection index.
A description of so called material part and the first version of material (explosives)
ranking has been given in [1].

3.

PROCESS HAZARD (TECHNOLOGICAL) INDICES

An assumption has been made that processing a substance causes hazard increasing.
Dangerous properties of processed media may be activated in running process. It is because
of the stimuli acting on them at the process conditions like: pressure, temperature, flows,
atmosphere, etc. and apparatuses / devices which can generate strokes, friction, get catalytic /
inhibiting agents, ignition sources, static electricity, emergency states, corrosion and so on.
In the given equation (1) that role plays technological index T. It has been constructed using
numerical notes given for 4 groups of a process characteristics described by equation (2)
given below.
General characteristics of a process (operation) TO. In this group the information
about dangerous material, process/operation type, a way of acting, devices/apparatuses used
and detailed process/operation parameters have been taken into account.
Detailed characteristics of unit operations TF, containing a specific operation menu
for explosives manufacturing including such types of operations as: static (storage,
conditioning), in solvent (filtration, washing, crystallization), in bulk (supplying, emptying,
weighting), in cloud (milling, sieving, granulating, dust removal), pressing (different types,
rolling), in elevated temperature (heating, casting, melting), hand-made, mechanical working
and general unit operations as: transport, mixing, drying, mass transfer operations and others.
Chemical process characteristics TP. In this group all chemical processes have been
divided into 4 types in the respect of thermal effect, system homogeneity, catalytic agent
presence and reaction phase (gaseous, liquid, solid, mono and multi-phase).
Specific hazards characteristics TS. In this group such information have been taken
into account as: plant area, apparatuses height, potential emission sources, corrosion, static
electricity, extremely difficult processes, fire and/or explosion zone presence, specific low
requirements fulfilling and many others.
The groups 2 and 4 above have been changed a lot comparing to the basic Temclev
System because of the specific operation types used in explosives and pyrotechnics
production. The others have been only slightly changed.
T = TO + TF + TP + TS

(2)

The use of protective measures decreases process hazard level. According to equation
(1) resultant note is diminished TZAB note. The index TZAB is very important in the system. It
shows the areas where a hazard should be reduced and helps acting in that direction. In the
range of using protective measures, which may be technical, technological and
organizational, any selected unit should be analyzed also in 4 thematic groups [2]. They
2

Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak

269

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

cover both: specific and general types of measures used in explosives industry for protecting
workers and processes.
Monitoring emergency states KS,
Failures preventive measures ZA,
Emergency systems and control SA,
Organization and safety management DO.
The index TZAB is computed using equation (3) as described above a product of 4
components:
TZAB = KS ZA SA DO

(3)

The components: KS, ZA, SA, DO are computed using equations presented in [3, 8].
The notes for components of equations (2, 3) may be taken from System Manual [9] or
computer program System_Ex, which has been elaborated in details for use in MS Windows
environment.
Some examples of the Evaluating Tables (1 5) used in Manual have been presented
below.
Table 1. Pressing
Unit Operation (physical) - Pressing
Rolling, pressing of product
Pressing in die and pushing out
Two-sided pressing
Operations in elevated temperature, with vibration, long lasting overpressure and/or,
repeatedly, with mould venting, other types every element additionally

Rating tII/1.6
3
4
5
+3

Table 2. . Mixing
Mixing Procedures

Without moving mechanical elements/devices:


Combustible and non-combustible substances
Combustible substances
Combustible substances and oxidizers
Explosives
In case using mechanical elements/devices like: paddle mixer, stirrer,
helical agitator perpetual etc. additionally

270

Rating
tII/1,6
1
2
5
8
+2

Process Hazard Evaluation for Explosives Manufacturing Using TEMCLEV-EX . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 3. Specific conditions


Unit Operation (physical) Specific conditions

Hand-made operations like: packing, covering by plastic foil, a lot of


product forming, dosage etc.) additionally
Operations run in elevated temperatures like: melting, casting, heat
treatment, operations run under lower pressurefor every specific
condition additionally
Mechanical working (cutting, machining, milling, propellant/blowing
charge grinding)

Rating
tII/1.4
2
3
5

Table 4. Cut-off systems


Rating tZIII/1
Cut off system of media flow
Flow blocking outside the unit:
- cut-off valves on a flow-line (pipeline) remote operated from operating
1
room, long distance from unit area
Flow blocking inside the unit (installation):
- hand-operated cut-off valve on the work place
2
- operator activity in the operating room at the unit area
1
- the valves positioned automatically in safety positions
+2
- complete emergency system for process stoppage, flows cut-off and
5
vessels emptying activated with use of one switch or automatically after
exceeding a fixed limitation
Table 5. Process parameters control
Operation / process parameters control
Periodic
Continuous:
Under process operating computer supervising
Steering automatics, self-steering devices or remotely operated systems
meeting requirements of Polish law regulation MGPiPS (Dz. U. Nr163
poz.1577 z 2003 r.)
Manually from operating room
Manually by operator of a device on the work-place
When control includes more than 1 parameter, every next additionally
With possibility to start safety systems - additionally
Recording process parameters and control signals meeting requirements
of Polish law regulation MGPiPS (Dz. U. Nr163 poz.1577 z 2003 r.) additionally

4.

Rating tZ,I/1
0.5
3
2

1
0,5
+0,5
+1
+1

PROCESS UNIT CLASSIFICATION

Every selected unit may be classified to one of the five process hazard classes KZP as shown in
Table 6 below.

Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak

271

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 6. Division into KZP classes according to ZP index value.


Class KZP

5.

Hazard level

KZP1
KZP2:
2/III
2/II
2/I

Catastrophic (Very high)


High:
upper range
center range
lower range

KZP3:
3/III
3/II
3/I

Middle:
upper range
center range
lower range

KZP4:
4/II
4/I
KZP5:
5/II
5/I

Moderate:
upper range
lower range
Low:
upper range
lower range

ZP index values
ZP > 300
200 < ZP 300
270 < ZP 300
230 < ZP 270
200 < ZP 230
100 < ZP 200
170 < ZP 200
130 < ZP 170
100 < ZP 130
50 < ZP 100
75 < ZP 100
50 < ZP 75
0 < ZP 50
25 < ZP 50
0 < ZP 25

PROCESS RISK ASSESSMENT METHOD PRAM

PRAM is the Risk Score method modified into the process conditions. It is based on the
equation (4) shown below. The meaning of all indices and their digital values have been
detailed described in [10]. Process risk level Rp values for PRAM have been given in Table
7. Potential consequences of hazard level (failure) Sp values are connected with Temclev
System through KZP classes.
Rp = Sp x Ep x Pp

(4)

where:
Rp process risk level index,
Sp potential consequences of hazard level (failure) index,
Ep hazard exposure time (process realization type)
Pp hazard (failure) occurrence probability index
Table 7. Process risk level Rp values for PRAM
Rp index
values
Rp < 20
20 Rp < 70
70 Rp < 200
200 Rp < 400
Rp 400

272

Description of risk level and preventive/protective measures


needed
Very low (risk acceptable), no action required
Low (risk acceptable), a need of giving attention to eliminating
potentially accidental events and initial events
Essential, a need of rapid improvement of safety level and lowering
the risk value (risk temporary tolerable, ALARP area)
High, a need of immediate improvement of safety level and lowering
the risk value (risk temporary tolerable, ALARP area)
Very high, (risk not acceptable), consider to stop an operation or use
the immediate actions for significant decrease the risk value (at least to
ALARP area)

Process Hazard Evaluation for Explosives Manufacturing Using TEMCLEV-EX . . .

New Trends in Research of Energetic Materials

6.

Czech Republic

2006

[Content]

ALGORITHM OF PROCESS HAZARD EVALUATION

Fig. 1. Fig. 1. Process hazard evaluation algorithm.

Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak

273

New Trends in Research of Energetic Materials

7.

Czech Republic

2006

[Content]

PROCESS HAZARD AND RISK ASSESSMENT RESULTS

In the frames of this work 7 selected process units in 4 plants have been analyzed and
assessed using Temclev-Ex and PRAM methods and System-Ex software. Some of them the
authors presented below as the examples of elaborated assessment procedures.

7.1

Emulsion explosives. Cartridging unit.

Emulsion explosives are much safer materials than traditional explosives in


manufacturing process as well as in further use. Their every ingredient is not a real
explosive. They become explosives after sensibilization at the very end of production
process.
Their manufacturing (Fig.2, Fot.1) consist mainly on :
Making a mixture of oil and emulsifier,
Making a mixture of wax and paraffin,
Making a solution of oxidizers (ammonium nitrate and sodium nitrate in water),
Mixing everything together in a high-speed mixer adding aluminium powder, glass
microspheres and granulated ammonium nitrate.
A mixture made in such a way is then given into the head of the cartridging machine to
make blasting cartridges. The end of that is a cooling process, drying, labelling and packing.
The whole production process is controlled by a computer. Every disturbance of the process
causes automatically stoppage of the installation.

Fig. 2. Cartridging operation scheme.

274

Process Hazard Evaluation for Explosives Manufacturing Using TEMCLEV-EX . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fot.1. Cartridging of emulsion explosives using Filipack EX 1100.

7.2

Nitroglycerin explosives for mines. Cartridging unit.

Manufacturing of nitroglycerin explosives for mines Metanit S2H consists on mixing


their ingredients: ammonium nitrate, DNT, nitro-esters and sodium chloride in mixers, then
supplying the mass into the cartridging machines where the cartridging process runs
continuously.
The cartridging oparation itself consists on proportioning the mixture by volume into a
core made by paraffin paper, closing both ends by sticking and giving a ready charge onto
the rolls for packing. The process is fully automatic, runs using cartridging machine
Niepmann A63-IV (not modernized).

7.3

Manufacturing of ANFO type explosive. Mixing unit.

Manufacturing of ANFO type explosive is realized in a big rotary mixer (Fot.2).From


its one side a weighted portion of granulated ammonium nitrate is supplied and from its
opposite side an oil is injected by injection nozzle. After the time fixed in a process
instruction the product is ready and is automatically pour out into the uotlet tube from which
falls down into the multiwall paper bags (Fot.3).

Fot.2. Rotary mixer for ANFO type explosive.

Fot.3. Product packing.

Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak

275

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Risk assessment results for: emulsion explosives Emulinit 2 and nitroglycerin mining
explosive Metanit S2H cartridging and ANFO type explosive Saletrol manufacturing
have been shown in Table 8 below.
Table 8. Table 8. Comparison of risk assessment results for cartridging process
(two different explosive types, two different machines) and ANFO
manufacturing.Temclev-Ex and PRAM methods used.
Assessed
process unit

Emulsion
explosive
Emulinit 2
cartridging using
Filipack EX 1100
Nitroglycerin
mining explosive
Metanit S2H
cartridging using
Niepmann A63IV (before
modernization)
ANFO type
explosive
manufacturing
using rotary
mixer

8.

ZP value
and
KZP
Class

Sp index
value

Ep index
value

Pp index
value

Rp index
value

89,25

18
Very low

4/II

(risk acceptable)

162,08

126
Essential

3/II

(risk temporary

tolerable, ALARP
area)

85,9

4/II

18
Very low
(risk acceptable)

FINAL REMARKS AND CONCLUSIONS

The proposition of applying a semi-quantitative method for hazard and risk assessment
of process installations presented in this paper allows to compare the risk levels for different
media processed in given installation or / and for different installations on site.
The method described above allows also the simulations using for example: different
technical solutions (devices, protective measures) and different technological solutions
(process conditions). As a result of such simulations we can find the answers on questions
like: what to do for significant lowering a risk level of analyzed process? or is it
profitable or not?.
On the basis of assessment procedures carried out for selected manufacturing processes
the authors can conclude that:
1. The system Temclev-Ex it is developed TEMCLEV method (Technology & Media
Classification and Evaluation System) elaborated in the Institute of Industrial Organic
Chemistry, which aim was the assessment of explosion and fire hazard level in chemical

276

Process Hazard Evaluation for Explosives Manufacturing Using TEMCLEV-EX . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

industry, especially for installations working with flammable gases and liquids as well as
solids in state of dusts or powders.
2. The results of assessment procedures received during Temclev-Ex system testing,
show the correctness of assumptions taken and the way of computing the indices used in the
system. However, to get more experience and improve the system, the work should be
continued and the system itself implemented into chosen plant of explosives industry.
3. On the base of assessments made the authors may state that Temclev-Ex method
fulfill requirements for correct evaluation of fire and explosion hazards created in different
operations of explosives manufacturing.
4. Elaborated both: the system and the software open a new area on the field of risk
assessment in production of explosives in the Polish industry and create a new quality on the
area of process hazard and risk evaluation tools.

9.

REFERENCES

[1]

Papliski A., Piotrowski T., Maranda A., Frczak M., Saaciski T. Evaluation of Fire and
Explosion Hazard and Risk of Environment Injury for Condensed Explosives by the Method of
Temclev-Ex. Materiay V-th International Conference Scientific Aspects of Armament
Technology, Waplewo, Poland, 6-8 October 2004., 176-177. ISBN 83-921491-0-6
[2] Piotrowski T i inni - Final Report of Project PBZ 038-07, vol.1.: Opracowanie metod i
procedur identyfikacji, oceny i klasyfikacji zagroe procesowych w przemyle chemicznym
dla zapobiegania awariom technologicznym, powodujcym szczeglnie niebezpieczne poary i
wybuchy technologiczne (Elaboration of methods and procedures for identification,
evaluation and classification of process hazards in chemical industry, preventing technological
failures causing extremely dangerous fires and explosions), Biblioteka Instytutu Przemysu
Organicznego, Warszawa, 1998.
[3] Piotrowski T., Hancyk B., Gowiski J., Seweryniak M. Materiay II Kongresu Technologii
Chemicznej, Wrocaw 15 18.09.1997, tom IV, 1798-1808, Dolnolskie Wydawnictwo
Edukacyjne, Wrocaw, 1998.
[4] Piotrowski T., Hancyk B., Gowiski J. - Organika - Prace Nauk. Inst. Przem. Org., 1997-98,
cz. II, 89-99.
[5] Piotrowski T. Przemys Chemiczny 78/12 (1999), 419-421.
[6] Piotrowski T., Hancyk B. Atest Ochrona Pracy 2/2000 (633), 4-7.
[7] Piotrowski T., Haat A. Organika Prace Nauk. Inst. Przem. Org., 1999-2000, 149-163.
[8] Piotrowski T i inni - Final Report of Research Project No 4T09B 06325 vol. I, Adaptacja
systemu Temclev do oceny zagroe poarowo-wybuchowych w procesach produkcji
materiaw wybuchowych, (Adaptation of Temclev System to Fire and Explosion Hazard
Assessment in Manufacturing of Explosives), Biblioteka Instytutu Przemysu Organicznego,
Warszawa, 2005.
[9] Piotrowski T i inni - Final Report of Research Project No 4T09B 06325 vol. II, System
Manual, (Podrcznik Oceny), Biblioteka Instytutu Przemysu Organicznego, Warszawa,
2005.
[10] Piotrowski T., Gowiski J. Technology&Media Classification and Evaluation System
Temclev. System Construction and Examples of Its Practical Use. 11-th International
Symposium Loss Prevention and Safety Promotion in the Process Industries Loss Prevention
2004, Praha, Czech Republic, 31 May 3 June 2004. Paper Full texts, thematic Section A,
1268-1277, PCHE PetroChemEng, Praha 2004, ISBN 80-02-01574-6.

10

Tadeusz Piotrowski, Tomasz Saaciski, Micha Frczak

277

New Trends in Research of Energetic Materials

10.

Czech Republic

2006

[Content]

ACKNOWLEDGEMENT

This project was sponsored by the State Committee for Scientific Research as a grant
number 4 T09B 06325 entitled: Adaptation of Temclev System to Fire and Explosion
Hazard Assessment in Manufacturing of Explosives.

278

Process Hazard Evaluation for Explosives Manufacturing Using TEMCLEV-EX . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

DETERMINATION OF THE THERMAL RESISTANCE OF THE


EMULSION EXPLOSIVES
Tomasz Saaciski*, Waldemar Witkowski*, Andrzej Maranda**,
Johann Kasperski***, Barbara Gobek***, and Henryka Boniuk****
* Instytut Przemysu Organicznego, ul. Annopol 6, 03-236 Warszawa, PL
** Wojskowa Akademia Techniczna, ul. Kaliskiego 2, 00-908 Warszawa, PL
*** BLASTEXPOL Sp. z o. o., 59-145 Chocianowiec (Duninw), PL
**** Politechnika Warszawska, ul. Noakowskiego 3, 00-664 Warszawa, PL

Abstract:
In this paper the results of two methods of determination of temperature
of decomposition of modern variations of emulsion explosives have been compared.
First kind of the results are obtain from a method which has been use in the assessment
of hazards during transportation by public roads. The second source of results
is a thermogravimetry method. The influence of the heating with the rate of 5 oC/min in
the range from 100 oC to 400 oC on the stability of emulsion explosives has been
presented. The differences in the characteristics of decomposition of the emulsion
explosives tested have been indicate.
Keywords:

1.

thermal decomposition, emulsion explosives

INTRODUCTION

The main blasting materials that have been use in the mining since for years,
are explosives which contain highly processed technical grade ammonium nitrate(V). Their
first modification were ammonium nitrate(V)-fuel oil (ANFO) explosives. Ammonium
nitrate(V) is present in the ANFO in porous state [1]. Next modifications were slurry [2] and
emulsion [3] explosives. Especially in the latest years, emulsion explosives became very
popular modification of blasting materials [4]. It comes from their undoubted advantages,
e. g.: practically the lack of sensitivity to mechanical stimuli and ability to automatic
loading. In some practical applications, emulsion explosives get explosives properties after
finishing of the loading them into the blasting bore. It means, that during transportation
by public road that variations of emulsion explosives are not explosives. In such causes,
that threatens concerned with the range of explosives being transported by roads
are decreasing significantly.
Emulsion explosives are multicomponent and two- or three-phase explosive systems.
Detonation parameters as well as the structure of emulsion explosives are determine
by the content and by physical and chemical properties of particular components. Basic
ingredients of emulsion explosives are: oxidants, combustible materials, water
and substances which guarantee appearance of such structure that is able to detonation
process. In emulsion explosives, water creates structure of this material
and this is the component that assure presence of ammonium nitrate(V) in the completely
soluble form. The presence of water in slurry and emulsion explosives determines
1

Tomasz Saaciski, Waldemar Witkowski, Andrzej Maranda, . . .

279

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

the influence of temperature on their ability to detonation in a considerable degree.


Especially, temperatures below zero influence negatively. For example, in the case
of emulsion explosives, if the temperature is below 35 oC, it takes place a precipitation
of ammonium nitrate(V) crystals followed by inversion of its emulsion structure
from water-in-oil into oil-in-water. As a result of this physical processes, emulsion
explosives lose their ability to detonation are no longer explosives.
Even more negatively can act elevated temperatures. In this case, in emulsion explosives
can proceed exothermal reaction that leads to spontaneous and unintended explosion.
This is why, in this paper, we have analyzed the behavior of the emulgites
one of the modification of emulsion explosives - in the elevated temperatures. Within
a framework of the experiments, presented in this paper, the temperature of decomposition
according to the Polish Standard BN-76/6091-08 and to differential thermogravimetry
analysis (DTA) has been determined.

2.

EXPERIMENTAL
2.1

Characteristics of explosives tested

Emulgites were tested during the experiments. Emulgites are a modification of emulsion
explosives. They were manufactured by Blastexpol (Duninw, Poland). General
compositions of the emulgites tested are presented in Table 1.
Table 1. General compositions of the emulsion explosives tested.
Component

Name of explosives
Emulgite
42GP

Emulgite
LWC

Emulgite LWC Emulgite LWC


AN20
ALAN3

Content of the component [%]


Oxidizers

85,0

89,3

87,1

87,4

Organic phase

6,5

6,0

6,0

4,2

Aluminium powder

5,0

Water

8,5

4,7

3,7

3,2

Modification agent

3,2

Oxidizing phases of the emulgites consists of ammonium and sodium nitrates(V).


Organic phase consists of oil and of emulgent. There were any solid components
in the emulgite 42GP. Additionally, emulgites of the type LWC (Low Water Composition)
incorporated not dissolve granule of ammonium nitrate(V). All the emulgites tested were
make sensitive by a chemical method. Recent results of determination of detonation
parameters of emulgite LWC ALAN3 are presented in [5].

280

Determination of the Thermal Resistance of the Emulsion Explosives

New Trends in Research of Energetic Materials

2.2

Czech Republic

2006

[Content]

Testing methods of sensitivity of explosives on elevated


temperature

Within a framework of the experiments the temperature and character of decomposition


of emulsion explosives, according to the procedure presented in the Polish Standard
BN-76/6091-08 has been determined. The sketch of the experimental set is presented
in Fig. 1. Signal from the thermocouple of the temperature sensor (3) is transmitted
to the data acquisition module ADAM-4018. In all of the tests, three samples (2) of emulsion
explosives, each with the mass of 0,5 g, were tested simultaneously. Samples were heated
in open test tubes (4). The rate of heating was constant in the range from 100 oC to 350 oC
and it was 5 oC/min.
DATA
ACQUISITION
MODULE

4
5

3
6
2
EXAMINER
1

Fig 1.

Scheme of the experimental stand for determination of the temperature


of decomposition, according to the Standard BN-76/6091-08.
Descriptions: 1 heating device; 2 - sample tested; 3 temperature sensor;
4 - test tube; 5 protective glass tube for the sensor (3); 6 Woods alloy;
7 baths.

In DTA, derivatogrametry apparatus PC-MON, made in Hungary, has been used.


Samples of approximately 25 mg were heated in open platinum vessels with a constant rate
of 5 oC/min, in the range of temperatures from 25 oC to 500 oC and in the argon atmosphere.
In the paper, the presentation of the experimental results is restricted to the processes that
appear within the temperature range from 100 oC to 400 oC. It is because, for all the
emulgites tested, there were no evidence of decomposition in the range over 400 oC.

3.

EXPERIMENTAL RESULTS

Results obtained during determination of the temperature of decomposition of emulsion


explosives tested, according to the test procedure presented in the Polish Standard
BN-76/6091-08 and results obtained according to the thermogravimetry differential analysis,
respectively in Table 2 and in Figures 2 4, have been shown. As the evidence
of decomposition of emulsion explosives tested, the occurrence of the white smoke has been
approved.

Tomasz Saaciski, Waldemar Witkowski, Andrzej Maranda, . . .

281

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 2. Characterisation of decomposition processes of emulgites tested.


Emulsion
explosives

Temperature in which, in the surrounding of the


samples tested, occurred [oC]

Emulgite

white matt
deposit

colorless clear
liquid

white smoke

42 GP

117

147

241

LWC

113

148

236

Temperature in which
occurred the
excitation of the
sample [oC]
281
(calm burning out of
some of the samples)*
284
(violent liberation of
nitrogen oxides)

LWC AN20

120

160

LWC ALAN3
133
159
* - the rest of samples no firing until 350 oC.

226

no firing until 350

226

no firing until 350

2,5
274,5
2

-10
-20

283,7

249,0oC (19,73%)

1,5

egzo

-30

Loss of mass TG [%]

DTA

256,1

0,5

-40
0
-50
-0,5
-60

DTA

235,5

-1

-70

-1,5
TG

-80

-2

-90

-2,5

-100

-3
100

200

300

400

Temperature [oC]

Fig 2.

4.

TG-DTA curves of emulgite LWC ALAN3

DISCUSSION OF RESULTS

In support of the results obtained during determination of the temperature


of decomposition (Table 2), one can draw the conclusion, that the influence of heating
on the emulsion explosives tested is pointed out much below their temperature
of decomposition (temperature in which white smoke occurred). Despite the similar content
of particular groups of components (Table 1), there can be different effects accompanied
with the decompositions of the samples (from calm burning out, through violent flameless

282

Determination of the Thermal Resistance of the Emulsion Explosives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Loss of mass TG [%]

-20
o

222,8 C (20,5%)
-40
o

266,4 C (52%)

LWC 20AN

LWC ALAN 3

-60

LWC

42GP
-80

-100
100

200

300

400

Temperature [oC]

Fig 3.

TG curves of emulgites 42GP, LWC, LWC AN20 and LWC ALAN3

DTA

25

20

egzo

15
222,8

42GP

10

LWC

LWC 20AN
5
235,5
0
LWC ALAN 3
-5
100

200

300

400

Temperature [oC]

Fig 4.

DTA curves of emulgites 42GP, LWC, LWC AN20 and LWC ALAN3

- liberation of nitrogen oxides, to lack of visible decomposition effects). Effects that


appeared during decomposition of emulgites LWC do not lead to ignition, unlike
to emulgite 42GP. However ignition of emulgite 42GP did not occurred every time.
Basing on comparisoning of the data presented in Fig. 2 4, one can point out,
that the start of decomposition of emulgite 42GP becomes the earliest. Decomposition
of emulgites LWC begins later than for 42GP and the total exothermic effect is lower.
Between the emulgites tested, decomposition of emulgite LWC AN20 starts as the latest.
In 100 oC, loss of mass of emulgite LWC is approximately 1,5 % and of emulgite 42GP
is approximately 5,2 %. Visual observations founded during liberation of components
from the samples tested (white matt deposit in temperature ca. 120 oC and colorless clear
5

Tomasz Saaciski, Waldemar Witkowski, Andrzej Maranda, . . .

283

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

liquid in temperature ca. 150 oC) were not confirm in the running of TG-DTA curves.
It comes from it, that one of the factors which decide about results of thermal stability testing
of emulsion explosives is the mass of samples. It determine the rate of the process of heat
exchanging between a sample and surroundings.

5.

CONCLUSIONS
1. The heating of emulgites 42GP and LWC is accompanying by various effects.
The effects are observable in both methods by visual observation (character of
decomposition) and by DTA testing (mass losses).
2. The scheme of decomposition obtained in respect to the visual observations of
samples with an order of magnitude of the weight 500 mg differ from results
obtained in DTA method, in which samples were of the weight 25 mg.
3. Emulgites LWC tested decompose much calm and later than emulgite 42GP.
4. Generally, the presence of aluminum powder do not cause differ character of
decomposition of emulgite LWC ALAN3, in comparison to the rest of materials
tested.
5. The emulsion explosives tested are thermally stable in the aspect of their use in the
deep mines.

REFERENCES
[1] V. L. BARON, V. KH. KANTOR: Tekhnika i tekhnologia vzryvnykh rabot v SshA, Izd.
Nedra, Moskva, 1989.
[2] M. A. COOK, H. E. FARNAM: PAT. USA2 930 685, 1960.
[3] H. F. BLUHM: PAT. USA 3 447 978, 1969.
[4] W. THUM: Blasting techniques in the German quarry industry. Proc. of 1st World.
Conf. on Explosives Blasting Technique, Munich, 2000.
[5] J. PASZULA, A. MARANDA, A. PAPLISKI, B. GOBEK, J. KASPERSKI: An analysis of
blast waves parameters and underwater explosion test of emulsion explosives and
dynamites. Proc. of 8th Seminar New Trends in Research of Energetic Materials Part II,
Pardubice, 2005.
Research work financially supported within the state research founds for science within the
years 2005 2007.

284

Determination of the Thermal Resistance of the Emulsion Explosives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REDUCTION OF NITROAROMATIC EXPLOSIVES BY OXYGENINSENSITIVE NAD(P)H:NITROREDUCTASES: IMPLICATIONS FOR


THEIR CYTOTOXICITY AND BIODEGRADATION
Jonas arlauskas*, Lina Miseviien*, Henrikas Nivinskas*, ilvinas Anuseviius*,
Aura Nemeikait-nien**, Evaldas Maldutis***, and Narimantas nas*

* Institute of Biochemistry, Mokslinink 12, LT-08662 Vilnius, Lithuania


** Institute of Immunology of Vilnius University, Molt Pl. 29, LT-08409, Vilnius, Lithuania
*** The General J. emaitis Military Academy of Lithuania, ilo 5a, LT-10322 Vilnius, Lithuania

Abstract:
Nitroaromatic explosives are toxic and mutagenic to humans and other mammalian
species. The first step(s) in the biodegradation/bioremediation of the explosive residues
in soil or groundwater is their reduction by bacterial oxygen-insensitive nitroreductases
to the relatively stable metabolites. Here we analyze the quantitative structure-activity
relationships in the reduction of nitroaromatic explosives and model nitroaromatic
compounds by mammalian DT-diaphorase (NQO1) and Enterobacter cloacae
NAD(P)H:nitroreductase (NR), which perform the four-electron reduction of
nitrogroups to corresponding hydroxylamines, and by Enterobacter cloacae PB2
NADPH: pentaerythritol tetranitrate
reductase (PETNR),
which performs
nitroreduction and reduction of benzene ring with the formation of hydrideMeisenheimer adducts. Our data show that in all the cases the reduction rate of
nitroaromatics mainly depends on the energetics of the charge transfer.
Keywords:

1.

nitroaromatic explosives; enzymatic reduction of; biodegradation of;

INTRODUCTION

Nitroaromatic explosives are toxic and mutagenic to humans and other mammalian
species. Their toxicity is partly attributed to the single-electron enzymatic reduction to
radicals accompanied by the formation of the reactive oxygen species (oxidative stress type
of cytotoxicity).[1,2] Single-electron reduction of nitroaromatics is performed by
flavoenzymes dehydrogenases-electrontransferases, e.g., NADPH:cytochrome P-450
reductase (P-450R, EC 1.6.2.4) or ferredoxin:NADP+ reductase (FNR, EC 1.18.1.2). In this
enzyme group, the flavin adenine dinucleotide (FAD) or flavin mononucleotide (FMN)
cofactors are characterized by the high stability of their neutral (blue) radicals at the
equilibrium.[3] The reduction of nitroaromatic compounds by P-450R and FNR follows an
outer-sphere electron-transfer model, i.e., the reactivity increases with an increase in their
single-electron reduction potential (E17), and is relatively insensitive to their structural
pecularities.[2,3]
Another important mechanism of cytotoxicity of nitroaromatic compounds is their two(four)-electron reduction to nitroso and hydroxylamino metabolites, which covalently bind
to proteins and DNA: [4]
ArNO2 + 2e- + 2H+ ArNO + H2O,
1

Jonas Sarlauskas, Lina Miseviciene, Henrikas Nivinskas, . . .

(1)
285

New Trends in Research of Energetic Materials

Czech Republic

2006

ArNO + 2e- + 2H+ ArNHOH

[Content]

(2)

In mammalian organism, this type of reactions is catalyzed by NAD(P)H:quinone


oxidoreductase (DT-diaphorase, NQO1, EC 1.6.99.2), which is localized mainly in the liver.
Another source of the formation of toxic aromatic hydroxylamines in mammals is the
gastrointestinal tract, where the nitroaromatic compounds are efficiently reduced by the
nitroreductases of Escherichia and Enterobacter sp. Because of the relative stability of the
nitroreduction products under aerobic conditions, these enzymes are named the oxygeninsensitive nitroreductases.
It is important to note that the reactions (1,2) being catalyzed by the oxygen-insensitive
nitroreductases of numerous bacteria, microorganisms, and fungi, also make a first step in
the biodegradation of nitroaromatic explosives in the environment.[5] In parallel, certain
nitroreductases reduce the aromatic ring with the formation of Meisenheimer-type hydride
adducts, which further undergo denitration:[5]
Ar(NO2)3 + H- ArH(=NO2-)(NO2)2
ArH(=NO2-)(NO2)2 + H- ArH(=NO2-)2NO2 NO2- + unidentified products

(3)
(4)

Typically, all the oxygen-insensitive nitroreductases including NQO1 possess the flavin
cofactor with the destabilized anionic (red) semiquinone state at the equilibrium. One may
note that in contrast to the reactions of single-electron reduction of nitroaromatics by
flavoenzymes, the mechanisms and the substrate specificity of their two-(four)-electron
reduction still remain insufficiently understood.
In this paper, we summarize the data of our previous[2,6-10] and ongoing studies on the
mechanisms of two-electron reduction of nitroaromatic explosives and model compounds by
oxygen-insensitive nitroreductases: rat NQO1 and Enterobacter cloacae NAD(P)H:
nitroreductase (NR) which reduce the nitrogroup(s) into hydroxylamine(s) (reactions (1,2)),
and Enterobacter cloacae PB2 NADPH:pentaerythritol tetranitrate (PETN) reductase
(PETNR), which in parallel performs the aromatic ring reduction with the formation of
Meisenheimer-type adducts (reactions (3,4)). The methods of synthesis of nitroaromatic
explosives and model nitroaromatic compounds, purification of NQO1 and NR, enzyme
kinetics and cytotoxicity studies, and quantum mechanical calculations were reported in our
previous papers.[2,7-10] Because the directly determined E17 values are unavailable for a large
number of nitroaromatic explosives, we also used their E17 values calculated from the data of
their single-electron enzymatic reduction (E17(calc.)) (www.bchi.lt/potential.pdf). PETNR was
obtained from Professor Nigel S. Scrutton and Dr. Richard J. Harris (University of Leicester,
UK).

2.

NITROREDUCTASE REACTIONS OF RAT LIVER NQO1

Mammalian NQO1 is a dimeric (2 x 31 kD) enzyme containing one molecule of FAD per
subunit. The midpoint potential of FAD at pH 7.0 is -0.159 V, whereas the potentials of
FAD/FAD-. and FAD-./FADH- couples are -0.200 and -0.118 V, respectively.[11] This
provides ca. 7 % FAD-. stabilization at the equilibrium. Quinone reduction by NQO1
follows a ping-pong scheme and often proceeds at a high maximal rate, > 1000 s-1. The
reactivity of quinones increases with an increase in their single-electron reduction potential
and decreases with an increase in their Van der Waals volume (VdWvol) above 200 3 , due
to the sterical hindrances.[12]
Recent X-ray and computer modeling studies provided
evidence that mono- and bicyclic quinones, an inhibitor of NQO1, dicumarol, and
286

Reduction of Nitroaromatic Explosives by Oxygen-insensitive . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

dihydronicotinamide ring of NADPH bind at the same domain in the active center.[13] These
compounds interact with tyrosine-126, histidine-161, and phenylalanine-106, and form -
complex with the isoalloxazine ring of FAD at a distance of 3.5 by displacement of
tyrosine-128.
Nitroaromatic compounds, although being less reactive than quinones, comprise
another important group of substrates for NQO1. NQO1 plays a considerable role in
mammalian cell cytotoxicity and genetoxicity of nitroaromatic pollutants - the components
of exhaust gases or the explosive residues in the environment. However, the mechanism and
substrate specificity of nitroreductase reactions of NQO1 are still poorly understood. There
is no X-ray data on the complexes of nitroaromatic compounds in the active center of
NQO1, except several computer modeling studies.[14,15]
We examined NQO1-catalyzed reduction of 39 nitroaromatic compounds, including 16
explosive compounds (TNT and its metabolites, tetryl, pentryl, dipentryl, hexyl,
tetranitrocarbazole, nitrobenzimidazolones, nitrobenzofuroxanes, and nitrotriazoles). The
obtained multiparameter regressions show that the reactivity of compounds increases with an
increase in their E17(calc.) with the coefficient log rate constant /E17(calc.) ~ 10 V-1, and the
highest torsion angle between the nitrogroup(s) and the aromatic ring. Another although less
certain factor increasing the reactivity of nitroaromatic compounds, is their affinity to the
oxidized form of NQO1. However, the use of the above parameters resulted in the relatively
uncertain (r2 ~ 0.800) correlations. Another important but poorly understood factor
influencing the reactivity of nitroaromatics is their ability to bind at the different domains in
the active center of NQO1. Our data partly contradict the computer modeling data[14,15]
which suggest that nitroaromatic compounds should bind at the dicumarol/quinone binding
domain, because: i) the entropies of activation (S) of nitroreductase reactions of NQO1 are
much less negative than S of quinone reduction, thus demonstrating the less efficient
electronic coupling of reactants, ii) increase in VdWvol of nitroaromatics up to 250-300 3
does not decrease their reactivity, but increases it, and iii) dicumarol acts as incomplete
inhibitor towards several nitroaromatic substrates, including tetryl. It shows that a number of
nitroaromatic compounds may bind outside the dicumarol/quinone binding domain.
However, dicumarol acts as a complete (linear) inhibitor with respect to 4,5,6,7tetranitrobenzimidazolone, whose reactivity is much higher than expected. This is probably
caused by its binding at the dicumarol/quinone binding domain. Another important aspect of
catalysis of NQO1 is that it catalyzes reductive N-denitration of tetryl in a single-electron
way, with the formation of N-methylpicramide, nitrite, superoxide, and oxidation of excess
NADPH:[2]
e-

Tetryl
O 2-.

3.

[Tetryl]-.
O2

e-, H+
N-Methylpicramide
NO 2

(5)

REACTIONS OF E. CLOACAE NITROREDUCTASE

Enterobacter cloacae NAD(P)H:nitroreductase (NR, EC 1.6.99.7) is a dimeric (2 x 24.5


kD) protein, containing a molecule of FMN per subunit. The midpoint potential of FMN at
pH 7.0 is -0.190 V, with less than 1 % FMN-. stabilization at the equilibrium.[16] E. cloacae
NR shares a significant sequence homology with nitroreductases from Escherichia coli,
Salmonella typhimurium, and FMN reductase from Vibrio fischeri. E. cloacae NR reduces
3

Jonas Sarlauskas, Lina Miseviciene, Henrikas Nivinskas, . . .

287

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

nitroaromatic compounds according to a ping-pong scheme with strong competitive to


NAD(P)H inhibition by the nitroaromatic substrates. It shows that NAD(P)H and
nitroaromatic compounds bind at the same domain. NR reduces nitroaromatic compounds at
high maximal rates, exceeding 100 s-1 or even 1000 s-1 for tetryl and pentryl.[7] The reduction
of polynitroaromatic compounds (tetryl, TNT, dinitrobenzenes) occurs biphasically, with a
fast consumption of 2 moles NAD(P)H per mole of nitrocompound, and subsequent slower
oxidation of further excess NAD(P)H. The second phase of reaction is accompanied by O2
uptake. It shows that polynitroaromatic compounds are first reduced to
monohydroxylamines. According to the studies of E. coli NR, the reduction of nitroso
compounds into hydroxylamines (reaction (2)) is 104 times faster than the reaction (1).[17]
We found that monohydroxylamine metabolites of TNT are further reduced by E. cloacae
NR, but at slower rates than TNT.[8] The reduction of TNT is not accompanied by nitrite
formation. E. cloacae NR and NQO1 reduce tetryl in different ways: NR reduces tetryl to
unidentified product(s) other than N-methylpicramide, nitrite is formed not simultaneously
with the NAD(P)H oxidation, but after a lag-period.
In general, the reactivity of nitroaromatic compounds towards E. cloacae NR increases
with an increase in their electron-accepting potency, which may be expressed either as
E17(calc.), or as the enthalpies of free radical (Hf(ArNO2-.)) or hydride adduct (anionic N,Ndihydroxylamine) formation (Hf(ArN(OH)O-)).[7] The observed parabolic log rate constant
vs. potential dependence is relatively uncertain. However, at present we are unable to
characterize the structural peculiarities of nitroaromatic compounds prerequisite for their
efficient reduction by E. cloacae NR. The X-ray data of E. cloacae and E. coli NR show that
pyrimidine ring of FMN is close to the surface of protein globule.[18] The X-ray data of the
complex of oxidized E. coli NR with nitroaromatic compound nitrofurazone shows that the
compound binds at the isoalloxazine ring of FMN displacing phenylalanine-70 and
phenylalanine-124, but with an unproductive orientation.[17] These factors may impose
flexibility and multiple orientations in the binding of nitroaromatic substrates. It is also
possible that hexyl and hexyl-N-ethylnitrate may bind at several domains of NR, e.g., at the
catalytic and intersubunit domain, because their strong substrate inhibition (Ki 1 M at 200
M NADH) is characterized by the nonlinear (parabolic) Dixon plots. On the other hand, the
reactivity of 4,6-dinitrobenzofuroxane is much lower than one may expect from its E17(calc.)
value.[10] A possible reason for its low reactivity is a loss of aromaticity of benzene ring,
because this compound exists in a form of Meisenheimer-type hydroxy adduct.

4.

REACTIONS OF E. CLOACAE PB2 PETN REDUCTASE

E. cloacae PB2 NADPH:PETN reductase (PETNR) is a monomeric 40 kD protein,


containing a molecule of FMN per subunit. The midpoint potential of FMN at pH 7.0 is
-0.267 V, with less than 1 % FMN-. stabilization at the equilibrium.[19] PETNR belongs to a
structurally separate group of flavoenzymes, the Old Yellow Enzyme (OYE, Saccharomyces
carlsbergensis NADPH:oxidase) family, which also includes Pseudomonas putida M10
morphinone reductase (MR), Pseudomonas fluorescens xenobiotic reductase B (XenRB),
and E. coli N-ethylmaleimide reductase (NEMR).[20] All the enzymes of the OYE family
perform two independent reactions, the reduction of aromatic nitrogroup(s) into
hydroxylamine(s) (reactions (1,2)), and the aromatic ring reduction with the formation of
Meisenheimer-type hydride adducts, which further undergo denitration (reactions (3,4)).
The moles of nitrite formed per mole of TNT consumed vary from 0.25 (NEMR, XenRB)
and 0.18 (PETNR), to 0.03 (OYE, MR).[20] The pathways of nitrite formation during the

288

Reduction of Nitroaromatic Explosives by Oxygen-insensitive . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

reductive degradation of TNT require more thorough studies. It has been proposed that NO2may be liberated during the nonenzymatic reaction between 4-hydroxylamino-DNT and
Meisenheimer-type TNT-hydride diadduct.[21] However, according to the latter studies, the
denitration occurs in the absence of 4-hydroxylamino-DNT, and its rate depends on enzyme
concentration.[20] A specific feature of the OYE family members is that apart from the
reduction of nitroaromatics, they perform the reductive denitration of PETN and glycerol
trinitrate, although with different substrate preference. The transgenic tobacco plants
expressing PETNR perform the biodegradation of nitroaromatic and nitroaliphatic
explosives.[22]
We examined the reduction of 28 nitroaromatic compounds by PETNR, including 20
explosive compounds. For 22 compounds with available E17(calc.) values, there exists an
uncertain linear log rate constant vs. E17(calc.) relationship (r2 = 0.6535) with the coefficient
log rate constant /E17(calc.) ~ 12 V-1. It points to the important role of charge transfer
energetics in the reactivity of compounds. The maximal reduction rates of the most reactive
oxidants, tetryl, pentryl and hexyl, are close to 35 s-1, and their bimolecular reduction rate
constants are above 107 M-1s-1. In contrast to E. coli NR, the reactivity of 4,6dinitrobenzofuroxane also follows this relationship. At present, we cannot identify the
structural peculiarities of nitroaromatics prerequisite for their efficient reduction by PETNR.
According to the X-ray data, the binding of picric acid and TNT to PETN involves their
parallel orientation towards the isoalloxazine of FMN, and the interaction of 1-hydroxy or 1methyl groups with histidines-181,184.[23] This causes their favourable orientation for
hydride transfer with the Meisenheimer-type adduct formation, i.e., the proximity of C-5
carbon to the flavin N-5. However, our data show that the nature of the substituent in 1position of TNT, picramide, N-methylpicramide, 2,4,6-trinitroanizole, or its absence in
2,4,6-trinitrobenzene does not strongly influence the reactivity of nitroaromatics. In contrast,
the maximal reduction rates of tetryl and pentryl are by 20 times higher, and their
bimolecular reduction rate constants by 500 times higher than those of TNT, although the
above compounds possess bulky substituents in 1-position. This may be explained by higher
electron-accepting potency of tetryl and pentryl. The strongest inhibitors of PETNR, hexyl
(Ki < 2 M) and tetranitrocarbazole (Ki = 3.5 M) are structurally different. The presence of
bulky N-ethylnitrate group in hexyl-N-ethylnitrate decreases its affinity to PETNR (Ki = 70
M) and reactivity by 10 times as compared to hexyl. However, the reactivities of hexyl and
pentryl are similar, although pentryl is much weaker inhibitor of PETNR (Ki = 150 M).
This shows that nitroaromatic compounds may bind in the active center of PETNR in
different ways, and that their efficient interaction with histidines-181,184 is not obligatory.
We found that nitrite was formed during the reduction of structurally different compounds,
such as TNT, picramide, tetryl, pentryl, tetranitrobenzimidazolone and tetranitrocarbazole.
The reductive denitration of tetryl did not follow the single-elctron transfer (reaction (5)),
and yielded other than N-methylpicramide reaction product(s). However, the reduction of pdinitrobenzene was not accompanied by the production of nitrite, yielding
phydroxylaminonitrobenzene. We found that this was not due to the structural peculiarities of
p-dinitrobenzene, but due to the more favourable energetics of reduction of nitrogroup
instead of aromatic ring, which was reverse for the compounds forming nitrite.

5.

CONCLUSIONS

The data of this work show that apart from the currently insufficiently understood
structural peculiarities, an increase in the electron-accepting potency of nitroaromatic
explosives, preferably determined in the aqueous medium, increases their reactivity towards
5

Jonas Sarlauskas, Lina Miseviciene, Henrikas Nivinskas, . . .

289

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

two-electron transferring oxygen-insensitive nitroreductases. In general, the mammalian


cell cytotoxicity of nitroaromatic explosives increases with an increase in their E17 or
E17(calc.), [2] except for the amino- and hydroxylamino- metabolites of TNT, which possess
higher than expected cytotoxicity.[8,24,25] In this work, we demonstrate the existence of a
certain parallelism between the reactivity of nitroaromatic explosives towards the singleelectron transferring flavoenzymes and two-electron transferring NQO1. Irrespectively of
this, the positive relationship between the electron-accepting properties of nitroaromatics and
their cytotoxicity should point to the leading role of the oxidative stress-type cytotoxicity,
because the inhibitor of NQO1, dicumarol, equivocally affects the cytotoxicity of
nitroaromatic compounds.[2] Our data also indicate that the order of reactivity of certain
nitroaromatic explosives (nitrotriazoles < TNT < tetryl, pentryl and hexyl) towards two
different bacterial nitroreductases, NR and PETNR, is also in line with their relative
electron-accepting potency. On the other hand, the enzyme specificity strongly influences
the
reduction
rate
of
dinitrobenzofuroxane,
tetranitrobenzimidazolone
and
tetranitrocarbazole. Besides, certain nitroaromatic explosives, e.g., hexyl, may act as a strong
inhibitors of both nitroreductase types, thus impeding their own biodegradation. Because the
nitro- and/or the aromatic ring reduction rates may influence the overall rate of
biodegradation of nitroaromatic explosives,[5] the data presented may provide some
guidelines for the design of specifically adjusted biodegradation systems.
Acknowledgements
This work was supported in part by the Lithuanian State Science and Studies
Foundation. We thank Professor Nigel S. Scrutton and Dr. Richard J. Harris (University of
Leicester, UK) for the fruitful collaboration.

REFERENCES
[1]

[2]

[3]
[4]
[5]
[6]

[7]

290

KONG, Q. JIANG and Q. QU: Formation of Superoxide Radical and Hydrogen Peroxide
Enhanced by Trinitrotoluene in Rat Liver, Brain, Kidney, and Testicle in Vitro and Monkey
Liver in Vivo, Biomed. Environ. Sci., 2, 72-77, 1989
N. NAS, A. NEMEIKAIT-NIEN, E. SERGEDIEN, H. NIVINSKAS, .
ANUSEVIIUS and J. ARLAUSKAS: Quantitative Structure-Activity Relationships in
Enzymatic Single-Electron Reduction of Nitroaromatic Explosives: Implications for their
Cytotoxicity, Biochim. Biophys. Acta, 1528, 31-38, 2001
M.V. ORNA and R.P. MASON: Correlation of Kinetic Parameters of Nitroreductase Enzymes
with Redox Properties of Nitroaromatic Compounds, J. Biol. Chem., 264, 12379-12384, 1989
K.H. LEUNG, M. YAO, R. STEARNS and S-H.L. CHIU: Mechanism of Bioactivation and
Covalent Binding of 2,4,6-Trinitrotoluene, Chem.-Biol. Interact., 97, 37-51, 1995
J.C. SPAIN: Biodegradation of nitroaromatic compounds, Annu. Rev. Microbiol., 49, 523-555,
1995
J. ARLAUSKAS, E. DIKANCAIT, A. NEMEIKAIT, . ANUSEVIIUS, H.
NIVINSKAS, J. SEGURA-AGUILAR and N. NAS: Nitrobenzimidazoles as Substrates for
DT-Diaphorase and Redox Cycling Compounds: Their Enzymatic Reactions and Cytotoxicity,
Arch. Biochem. Biophys., 346, 219-229, 1997
H. NIVINSKAS, R.L. KODER, . ANUSEVIIUS, J. ARLAUSKAS, A.-F. MILLER and N.
NAS: Quantitative Structure-Activity Relationships in Two-Electron Reduction of
Nitroaromatic Compounds by Enterobacter Cloacae NAD(P)H:Nitroreductase, Arch.
Biochem. Biophys., 385, 170-178, 2001

Reduction of Nitroaromatic Explosives by Oxygen-insensitive . . .

New Trends in Research of Energetic Materials

[8]

[9]

[10]

[11]
[12]

[13]
[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

Czech Republic

2006

[Content]

J. ARLAUSKAS, A. NEMEIKAIT-NIEN, . ANUSEVIIUS, L. MISEVIIEN, M.


MARTINEZ-JULVEZ, M. MEDINA, C. GOMEZ-MORENO and N. NAS: FlavoenzymeCatalyzed Redox Cycling of Hydroxylamino- and Amino Metabolites of 2,4,6-Trinitrotoluene:
Implications for their Cytotoxicity, Arch. Biochem. Biophys., 425, 184-192, 2004
J. ARLAUSKAS, A. NEMEIKAIT-NIEN, . ANUSEVIIUS, L. MISEVIIEN, A.
MAROZIEN, A. MARKEVIIUS and N. NAS: Enzymatic Redox Properties of Novel
Nitrotetrazole Explosives: Implications for their Toxicity, Z. Naturforsch., 59c, 399-404, 2004
A. NEMEIKAIT-NIEN, J. ARLAUSKAS, L. MISEVIIEN, . ANUSEVIIUS, A.
MAROZIEN and N. NAS: Enzymatic Redox Reactions of the Explosive 4,6Dinitrobenzofuroxan (DNBF): Implications for its Toxic Action, Acta Biochim. Polon., 51,
1081-1086, 2004
G. TEDESCHI, S. CHEN and V. MASSEY: DT-Diaphorase. Redox Potential, Steady-State,
and Rapid Reaction Studies, J. Biol. Chem., 270, 1198-1204, 1995
. ANUSEVIIUS, J. ARLAUSKAS and N. NAS: Two-electron Reduction of Quinones
by Rat Liver NAD(P)H:Quinone Oxidoreductase: Quantitative Structure-activity relationships,
Arch. Biochem. Biophys., 404, 254-262, 2002
M.A. BIANCHET, M. FAIG and L.M. AMZEL: Structure and Mechanism of
NAD(P)H:Quinone Acceptor Oxidoreductase (NQO), Meth. Enzymol., 382B, 144-174, 2004
J.V. SKELLY, M.R. SANDERSON, D.A. SUTER, U. BAUMANN, M.A. READ, D.S.J.
GREGORY, M. BENNETT, S.M. HOBBS and S. NEIDLE: Crystal Structure of Human DTDiaphorase: A Model for Interaction with the Cytotoxic Prodrug 5-(Aziridin-1-yl)-2,4dinitrobenzamide (CB 1954), J. Med. Chem., 42, 4325-4330, 1999
S. CHEN, K. WU, D. ZHANG, M. SHERMAN, R. KNOX and C.S. YANG: Molecular
Characterization of Binding of Substrates and Inhibitors to DT-Diaphorase: Combined
Approach Involving Site-Directed Mutagenesis, Inhibitor-Binding Analysis, and Computer
Modeling, Mol. Pharmacol., 56, 272-278, 1999
R.L. KODER, C.A. HAYNES, M.E. RODGERS, D.W. RODGERS and A.F. MILLER: Flavin
Thermodynamics Explain the Oxygen Insensitivity of Enteric Nitroreductases, Biochemistry,
41, 14197-14205, 2002
P. R. RACE, A.L. LOVERING, R.M. GREEN, A. OSSOR, S.A. WHITE, P.F. SEARLE, C.J.
WRIGHTON and E.I. HYDE: Structural and Mechanistic Studies of Escherichia Coli
Nitroreductase with the Antibiotic nitrofurazone. Reversed Binding Orientations in Different
Redox States of the Enzyme, J. Biol. Chem., 280, 13256-13264, 2005
C.A. HAYNES, R.L. KODER, A.F. MILLER and D.W. RODGERS: Structures of
Nitroreductase in Three States: Effects of Inhibitor Binding and Reduction, J. Biol. Chem.,
277, 11513-11520, 2002
H. KHAN, T. BARNA, R.J. HARRIS, N.C. BRUCE, I. BARSUKOV, A.W. MUNRO, P.C.E.
MOODY and N.S. SRUTTON: Atomic Resolution Structures and Solution Behavior of
Enzyme-Substrate Complexes of Enterobacter cloacae PB2 Pentaerythritol Tetranitrate
Reductase, J. Biol. Chem., 279, 30563-30572, 2004
R.E. WILLIAMS, D.A. RATHBONE, N.S. SCRUTTON and N.C. BRUCE: Biotransformation
of Explosives by the Old Yellow Enzyme Family of Flavoproteins, Appl. Environ. Miocrobiol.,
70, 3566-3574, 2004
J.W. PAK, K.L. KNOKE, D.R. NOGUERA, B.G. FOX and G.H. CHAMBLISS:
Transformation of 2,4,6-Trinitrotoluene by Purified Xenobiotic Reductase B from
Pseudomonas fluorescens I-C, Appl. Environ. Microbiol., 66, 4742-4750, 2000
C.E. FRENCH, S.J. ROSSER, G.J. DAVIES, S. NICKLIN, N.C. BRUCE: Biodegradation of
Explosives by Transgenic Plants Expressing Pentaerythritol Tetranitrate Reductase, Nat.
Biotechnol., 17, 491-494, 1999
H. KHAN, R.J. HARRIS, T. BARNA, D.H. CRAIG, N.S. BRUCE, A.W. MUNRO, P.C.E.
MOODY and N.S. SCRUTTON: Kinetic and Structural Basis of Reactivity of Pentaerythritol

Jonas Sarlauskas, Lina Miseviciene, Henrikas Nivinskas, . . .

291

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Tetranitrate Reductase with NADPH, 2-Cyclohexenone, Nitroesters, and Nitroaromatic


Explosives, J. Biol. Chem., 277, 21906-21912, 2002
[24] B. LACHANCE, P.Y. ROBIDOUX, J. HAWARI, G. AMPLEMAN, S. THIBOUTOT and G.I.
SUNAHARA: Cytotoxic and Genotoxic Effects of Energetic Compounds on Bacterial and
Mammalian Cells in Vitro, Mutat. Res., 444, 25-39, 1999
[25] M.E. HONEYCUTT, A.S. JARVIS and V.A. McFARLAND: Cytotoxicity and Mutagenicity of
2,4,6-Trinitrotoluene and Its Metabolites, Ecotoxicol. Environ. Saf., 35, 282-287, 1996

292

Reduction of Nitroaromatic Explosives by Oxygen-insensitive . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

COMPLEX RESEARCH OF AL NANOPOWDERS PRODUCED BY


THE EXPLODING WIRE METHOD
Valentin Sedoi *, Galina Volkova **, Mirslan Osmonaliev *, Victor Ivanov**,
Ludmila Chemezova *, and Alexander Alekseev **
* Institute of High Current Electronics RAS, 634055, Tomsk, Russia
** Institute of Petroleum Chemistry RAS, 634055, Tomsk, Russia

Abstract:
The impetus for this research came from fundamental and applied questions connected
with specified characteristics of Al nanopowders, theirs controlled syntheses and
passivation, and theirs activity as high energetic ingredients in solid propulsion and in
aluminum-water gels.
Different conditions of powder syntheses by the exploding wire method and powder
passivation were used: syntheses in low-pressure gases (Ar, N2, CO2), passivation in
atmospheric-pressure gases (air, N2, CO2).
Investigations of particles size and morphology have been accomplished with a 125-kV
transmission electron microscope. The specific surface area S has been determined by
BET method. Particles are predominately spherical, BET surface area was of 15 40
m2/g. A composition and surface properties of nanopowders produced under various
conditions were studied by mass-spectroscopy techniques at temperatures of 0 450 0C,
electron and X-ray difraction methods, and by element analysis methods.
To determine activities of the powder samples in the combustion reactions, the
Differential Thermal Analysis, Thermogravimetry, and reaction with water were
conducted. Active metal contents have been determined from amount of impurities and
hydrogen extracting in direct thermohydrolysis of nanopowders by distilled water at 60
0
C. Kinetics of gas evolution was determined therewith and shows two steps. Active
aluminum content is within range of 16 80 %. A composition of adsorbed gases and
dynamics of desorption in vacuum have been investigated too.
Correlation between conditions of nanopowders syntheses, their specific surface and
chemical activity in reaction with water and maximum oxidation degree have been
determined.
Keywords:

1.

nanopowders, characterization, passivation, oxidation, combustion

INTRODUCTION

Different methods are developed and used for nanosized metal powder production: laser,
electron- beam, Gen-Miller, electrical explosion, plasma-chemical, and others. Among them
the exploding wire method has the advantage of direct conversion of electrical energy into
heat with high efficiency [1]. Besides, the pulsed heating provides high and uniform energy
densities, high velocities of expansion and high rates of the substance cooling. As a result,
dimensional, structural, phase, defect, and other energy-saturated states are realized in the
particles formed.
Growing interest in nanopowders is stimulated by the possibility to increase the
inflammability, the burning rate, and the completeness of metal combustion and to decrease
1

Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov, . . .

293

New Trends in Research of Energetic Materials

Czech Republic

2006

the agglomeration. The area of investigations includes metallized solid


metallized gelled propellants [4].

[Content]
[2]

, liquid

[3]

, and

The purpose of the present paper came from fundamental and applied questions
connected with specified characteristics of nanopowders produced by the exploding wire
method, and their activity as high energetic ingredients in solid propulsion [2] and aluminumwater gels [5-8].
The structure of the paper is as follows. Section 2 describes the methods and techniques
of investigation, different apparatus and physical-chemical methods used for analysis of
powders. Section 3 dwells on a size of particles and crystallites. Section 4 is devoted to
constitution of powder samples. Section 5 describes the results of investigation of powder
activity in reactions with air and hot water. Powder fineness effect is described as well.
Section 6 demonstrates powder fineness effect on combustion of Al/gelled water mixtures. It
has been established that the combustion rate sufficiently grows when a high-dispersed
powder is used.

2.

EXPERIMENTAL METHODS

Experimental setups include the installation for powder production and facilities for
characterization of the powder produced.

2.1

Engineering conditions

Engineering conditions for production of powders by the exploding wire method with
given requirements are formulated in the papers [1, 9]. The investigations performed have
shown that the electrical explosion allows synthesizing the powders passivated by oxide or
nitride [9, 10, 11] as well as by oxycarbide and by the layers of absorbed gases [8]. With that, the
activity of powders remains high.
The major characteristic in powder production is the thermal energy introduced into the
wire material. In this study, the energy density was close to the sublimation energy of
aluminum, and the conditions of uniform Joule heating of a wire [9, 12] were used.

2.2

Determination of dispersivity of powders

The analysis of ultra-fine particles is a difficult problem. Measurements only of the


specific surface area of the powder provide accurate information about particle size if the
morphology, smoothness and degree of equiaxialness of the particle are known.
Transmission Electron Microscope (TEM) analysis is capable to give some information on
these properties.
The specific surface area SBET of the samples was measured by the low-temperature
adsorption method (BET).
The particle shape and size distributions were determined using the 125 kV TEM. In
preparation, the suspension powder - industrial alcohol was applied on a slide and dried.
Then a carbon layer was sputtered on the powder-slide using the vacuum deposition system.
The resulted carbon replica was isolated by the use of water solution of gelatin, gelatin was
dissolved, and the carbon replica with the sputtered powder was placed on the copper mesh.
The mesh with the powder was used as the subject for TEM investigations. The mean
diameters D and the standard deviation were calculated from TEM data. Medium sizes of
the powder particles and their size distribution were analyzed with the microphotographs
with (250 300) 103 magnification. The results are given below. The number of measuring

294

Complex Research of Al Nanopowders Produced by the Exploding Wire Method

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

particles is averaged 300-500 depending on the powder size uniformity. The algebraic
number (Ni/N) of particles with the size of the given interval was plotted in histograms.
It is necessary to have in view that this procedure is limited by the particle size of more
than 5 micrometer, because such particles may be lost. So, the results will be distorted. In
this connection, common measurements by TEM and BET methods will be more
informative.
The X-ray diffraction (XRD) method was used to determine mean sizes of regions of
coherent scattering (DRCS) for major phases. The coherent scattering regions (RCS) for
produced powders were determined on the data of roentgen lines broadening at small angles.

2.3

Determination of powder composition

A composition and surface properties of nanopowders produced under various


conditions were studied by mass-spectroscopy techniques at the temperatures of 0 450 0C,
electron and X-ray diffraction methods (XRD).
The phase analysis and determination of the fine crystal structure parameters of powders
have been implemented by means of XRD apparatus DRON-UM1 with filtered copper
radiation K. A relative error in determination of the phase constitution in the powder was
under 10% and in determination of the parameters it was under 0.1%.
Active aluminum mass content, Al %, was determined by measuring the hydrogen
volume, which was formed in reaction of aluminum with solution of sodium hydrate.
The composition of adsorbed gases has been determined by heating of samples in the
vacuum of 10-7 at various temperatures as well as during linear heating with ramp function
of temperature.

2.4

Investigation of powder activity in oxidation reactions

To determine activities of the powder samples in the combustion reactions, the


Differential Thermal Analysis (DTA), Thermogravimetry (TG), and reaction with hot water
were conducted.
DTA and TG were implemented in air with the heating rate of 5 0C/min.
The special gasometric device was used for determination of gas volume extracted
during the reaction of aluminium powder with hot water. Mixtures 2.7 mg of Al and 10 g of
distilled water were placed in the cell thermostated under 60 0C, and the volume of gas
released has been calculated using the bubble recorder.
In reactions with hot water, aluminium oxidized and is transformed to the oxyhydroxide.
Final product has been analysed using BET and TG methods. TG was implemented in air
with the temperature range of 20 1000 0C and heating rate of 10 0C/min.

2.5

Investigation of combustion of nanoaluminum-gelled water


compositions

Experiments were conducted in a close manometric bomb of 5 liter filled with argon at a
constant controlled pressure. Mixtures were placed in quartz glasses of 10 mm diameter and
35 mm length and were ignited by a nichrome spiral. At least 3 tests or more were performed
for one experimental point with the given conditions.
The combustion rate was determined by two photodiodes with the distance between
them of 20 mm. Both photodiodes were coarsened and adapted through two observation slits

Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov, . . .

295

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

to radiation from a combustion wave. A millisecond stopwatch was started by the first
photodiode and stopped by the second one.

3.

THE SIZE OF PARTICLES AND CRYSTALLITES

The following types of powder markings are assumed: the gas used as medium for a
wire explosion is noted first, and the gas used for the powder passivation is noted through
the slant line. For instance, N2/CO2 designation corresponds to the powder produced in
nitrogen and passivated by carbon dioxide etc. The specific surface of powders varies from
16 to 38 m2/g. The particles are predominantly spherical (Figs. 1, 2).
If the particles are spherical, one can determine the mean-surface diameter dS as
d S = 6 / S BET where is the density, for Al = 2.7 g/cm3. These data for some samples are
placed in Table 1 together with the mean diameters D, the standard deviations , determined
from TEM photos, and mean sizes of regions of coherent scattering (DRCS) evaluated from
XRD lines.

60

W, %

N2/CO2

30

0
0

63,6

127,2

190,8

254,4

318

381,6

D, nm

Fig 1.

TEM photo and size distribution of powder sample N2/CO2

60

W, %

(N2+CO2)/ N2

30

0
0

50 nm
Fig 2.

296

40

80 120 160 200 240 280 320


D, nm

TEM photo and size distribution of powder sample (N2+ CO2)/CO2

Complex Research of Al Nanopowders Produced by the Exploding Wire Method

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 1. Size of particles and crystallites.


Sample
N2/air
N2/CO2
N2+CO2/N2

SBET,
m2/g
17.5
21.0
34.3

dS ,
D,
nm
nm
127 40.3
106 40.4
65 32.4

DRCS,
nm
nm
16.4 29
6.1 30
4.7 32.5

Detailed comparison between values D, , and DRCS, shows that the particles of
N2+CO2/N2 sample are monocrystal, and comparison between values dS and D shows that
the particles of this sample are slightly agglomerated. Ideal case, dS = D, may be observed
with oxide or nitride particles because particles of pure metals are easy agglomerated during
preparation and store.

4.

CONSTITUTION OF POWDER SAMPLES

Different conditions of powder synthesis by the exploding wire method and powder
passivation were used: synthesis in low-pressure gases (Ar, N2, CO2), passivation in
atmospheric-pressure gases (air, N2, CO2).
A constitution of gas shells and desorption dynamics were investigated by using the
mass spectrometry method and by measuring the pressure of releasing gases in the interval
of 0 450 C on heating in vacuum at constant temperatures and under the fast (to 50
K/min) linear dynamic heating.
The composition and some properties of powder samples produced and passivated in
various gas mediums are summarized in Table 2.
Table 2. Results of Al nanopowders analysis
Sample
N2/N2
N2/CO2
(N2+ CO2)/N2

S,
m2/g
18.2
21.0
34.3

Al, Water, Sum of gases,


mass %
%
mass %
75,4
6,4
1,7
70,7
0,4
2,9
60.7
1.5
5.2

XRD, volume %
Al Al2O3 AlN Al N5O14
96
4
90
5
~5
91
9

The percentage content of pure aluminum (Al %) was determined by a volumetric


method in which the volume of hydrogen, as a result of reaction of Al with solution of
sodium hydrate, has been measured. The percentage volume contents of Al, Al2O3, AlN, and
Al N5O14 were determined by XRD method.
It was founded that a composition of desorbed gas components is highly changed with
the temperature increase and depends on the surroundings of preparation and passivation.
Tables 3 and 4 give the composition of gases desorbed in vacuum under different
temperatures.
Kinetics of gas evolution shows two steps (Fig. 3).

Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov, . . .

297

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 3. Sample N2/CO2: Composition of desorbed gases under different


temperatures (T0C), volume %
T, 0C
Gas
N2
O2
Ar
CO2
H2 O
NH3

50

100

150

200

250

300

350

400

73.5
8.2
0.0
2.3
16.1
0.0

79.6
8.0
0.0
1.6
11.8
0.0

88.4
5.5
0.02
0.4
5.7
0.0

91.1
4.7
0.01
0.3
3.9
0.0

82.0
8.8
0.0
0.2
6.8
1.2

64.7
5.6
0.0
0.1
23.0
9.7

86.7
4.9
0.0
0.0
6.3
2.1

96.0
1.9
0.0
0.0
2.1
0.0

82.8
15.5
0.0
0.0
1.7
0.0

Table 4. Sample CO2/CO2: Composition of desorbed gases under different


temperatures (T0C), volume %
T, 0C
0

50

100

150

200

250

300

350

400

35.3
11.8
0.0
48.8
4.1

15.6
4.7
0.0
73.3
7.4

26.0
6.4
0.1
57.1
6.4

15.8
6.0
0.2
72.8
5.2

13.3
3.6
0.05
67.0
16.0

9.5
2.6
0.0
53.8
34.1

6.7
1.8
0.0
73.1
18.3

4.9
1.3
0.0
92.1
1.7

4.1
0.8
0.0
92.3
2.8

1,0

0,8

Relative Intensity

Gas
N2
O2
Ar
CO2
H2 O

0,6

0,4

0,2

T, C
0

Fig 3.

100

200

300

400

500

Kinetics of gas evolution

Gas inclusions (air components, water, CO2, et al.) that are the most loosely-bounded
with the surface of particles as well as argon adsorbed after completion of the electrical
explosion, are removed at 100 150 C along with the external particle shell formed in
passivation. The next portion of gases chemically bounded to the surface is desorbed at 250
400 C (Figure 3). The temperature of desorption finishing is about 400 0C.

298

Complex Research of Al Nanopowders Produced by the Exploding Wire Method

New Trends in Research of Energetic Materials

5.

Czech Republic

2006

[Content]

OXIDATION OF POWDERS IN REACTIONS WITH AIR AND


WATER
5.1

DTA and TG

Table 5 demonstrates the data of beginning of the oxidation temperature Tonset during
DTA in air for different powders with the heating rate of 5 0C/min.
Table 5. Comparison of data of oxidation beginning temperature during DTA in air
Sample

Al,

SBET,

D,

Tonset,

m /g

nm

N2/N2

75,4

18.2

44,2

375

N2/air

74

17.5

40,3

400

N2/CO2

70,7

21.0

40,4

425

N2+CO2/ N2

60.8

34.3

32.4

397

CO2/CO2

16,4

38.0

30,8

400

Ar/air

79,8

16.7

54,6

420

It can be seen from Table 5 that Tonset is independent of powder fineness (size varies
twice and purity varies almost 5 times). For all cases the temperature of oxidation beginning
is practically the same: Tonset = (400 25) 0C.
It seems that the temperature of oxidation beginning corresponds to the finishing
temperature of gases desorption. The two processes are connected with each other. Adsorbed
and chemically bonded gases prevent particles from oxidation in air. Oxidation in water
gives another results.

5.2

Oxidation in reaction with hot water

Correlation between conditions of nanopowder synthesis, the specific surface and


chemical activity in reaction with hot water has been determined.
In this study, the temperature of reaction onset in independent experiments on evolution
of the first hydrogen bubbles, a pH change, and some self-heating of the mixture was
registered. Some results are presented in Table 6.
Table 6. Reaction with hot water

No

Sample

1
2
3

N2/air
N2+CO2/ N2
CO2/CO2

SBET
sample,
m2/g
13
34.3
38

of Al
content,
%
74.0
60,8
16.4

,
before/after
reaction
4.96/9.71
5.07/9.41
5.2/8.2

SBET
products,
m2/g
368
342
165

Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov, . . .

of

299

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The Table 6 shows that there exists the direct relation between the amount of active
aluminum content and the specific surface of final product. An ultrafine aluminium
oxyhydroxide with the specific surface SBET = 368 m2/g was produced by oxidation of
nanosized aluminium by hot water.
Fig. 4 demonstrates the hydrogen release during reaction with water at 60 0C.
Volume, mm3
1500
1200
900
600
1

300

0
0

500

1000

1500

2000

2500

Time, s

Fig 4.

Kinetics of hydrogen realise in oxidation of Al samples 1, 2, 3 (Table 6)

It can be seen from Fig. 4 that various types of Al nanopowders oxidized in two stages.
TG analysis of final products shows the aluminum oxyhydroxide dehydration. This result
allows to suppose that the reaction 2Al + 4H2O 2Al OOH + 3H2 is limitative for samples 1
and 2. As for sample 3, Table 6, alumina probably comes together with preudoboehmite in
less alkaline medium (pH 8.2).
The oxidation level of Al powders in reaction with hot water (60 0C), % masses degree
of oxidation, and the temperature of reactions onset, T 0C, are given in Fig 5.

Degree of Oxidation, %

Temperature of teaction onset, C

60
100
55
80
1
2

60

50

45
40
40
20
35
14 16 18 20 22 24 26 28 30 32 34 36 38 40
2

S, m /g

Fig 5.

300

The o degree of oxidation (1), % masses, and the temperature of reactions


onset, T 0C (2)

Complex Research of Al Nanopowders Produced by the Exploding Wire Method

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

An oxidation level in reaction with water is from 36 to 100% for various types of
powders and their specific surface.
The more active is the powder, the lower is the temperature of reaction onset with water
and the higher is the oxidation degree at 60 C. The experimental data on relative chemical
activity show that the value of specific surface is the major characteristic of powder.

6.

COMBUSTION OF NANOALUMINUM / GELLED WATER


COMPOSITIONS
6.1

Preface

The possibility of combustion of the nanoaluminum with water in the mode of a normal
combustion regime has been established in the experimental work [6]. The calculated heat of
combustion of the Al nanopowder stoichiometric mixture with water, recounting to the pure
metal (50/50 %), is 7570 kJ/kg thus attributing this system to high-energy material. Owing to
inert admixtures in the powders, this value is vastly less (by 10 15 %).
In the paper [5] the more detailed data on combustion of the stoichiometric aluminum
mixture with water is given. In particular, the law of combustion rate of the stoichiometric
mixture is: U = 0.183 (p/p0)0.4, cm/s. Also the incompleteness of aluminum combustion is
observed.
Report [8] shows, that, for using the aluminum/water mixture as propellant, the ratio of
40% Al/60 gelled water is preferable. In this case, the completeness of Al combustion
reaches 100 %. The combustion temperature and the combustion rate increase with
increasing the oxidation degree. The combustion rate values under the pressure of 7 MPa for
the aluminum concentrations from 38 to 57 % are from 0.8 to 1.8 cm/s. The temperature of
combustion reaches 2450 0K.
The present study is a continuation of
water 40/60 %.

6.2

[5-8]

, and we used the optimal ratio aluminum/

The combustion rate

Figure 6 demonstrates clearly how the concentration and dispersivity of Al powder in


Al/water slurry influence the combustion rate.
The first curve in figure 6 shows the dependence of combustion rate on pressure for the
sample of Ar/air having the specific surface of 16.7 m2/g (Table 5), in the stoichiometric
50/50 % composition of Al/gelled water.
Investigations confirmed the early-represented data [8] showing that aluminum
decreasing in the composition of 40/60 % increases the completeness of metal combustion
and increases the combustion rate (curve 2). The sample of Al powders was the same
Ar/air.
Curve 3 also corresponds to the mixture of 40/60 %, but the specific surface of
nanoaluminum was of 34.3 m2/g (sample N2+CO2/ N2, Tables 1, 2, and 5). The combustion
rate grew sufficiently and reached 2.4 cm/s under the pressure of 5 8 MPa.
It is important to point out that the dependence of combustion rate on pressure is very
weak (line 3), the exponent in the combustion law = 0.12, and is even independent of
pressure in the area of 5 8 Mpa. That area of pressure is the working range for propellants.
Visual observation of burning surface and analysis of combustion products point to the facts

Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov, . . .

301

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

that burning is more uniform and efficient, and solid products are less agglomerated
consisting of nanosized corundum.
2

U, cm/c

1
2
3

P, MPa

Fig 6.

7.

7 8 9 10

(1) sample Ar/air, SBET= 16.7 m2/g, stoichiometric 50/50 % composition,


(2) sample Ar/air, composition of 40/60 % Al/water, (3) sample
N2+CO2/ N2, SBET =34.3 m2/g, composition of 40/60 % Al/water

CONCLUSION

Because of peculiar chemical features of the Al nanopowders, the oxidation degree of


the powder by water and the temperature of the reaction onset can be used as a criterion to
compare chemical activity of the powders produced under various conditions.
Phase composition and phase structure of powders produced under different conditions
of heating and environment are overwhelmingly important. Stability of phase constitution of
particles as well as stability of their energetic and other characteristics during storage time
and their behavior under external influences are also sufficient.
Inorganic coatings allow producing powders having monocrystal nanosized particles.
Though pure metal content averages out to 75 %, powders demonstrate high energetic
activity in oxidation reactions due to high dispersivity of powders.
Compared to conventional Alex [3], the specific surface of powder produced using
justified engineering conditions is three times higher. This allowed sufficiently improving
characteristics of gelled Al/water compositions, namely, the combustion rate became two
times higher and less agglomeration has been observed.
Adsorbed and chemically bonded gases prevent particles from oxidation in air. The
temperature of oxidation beginning corresponds to the finishing temperature of gases
desorption. Oxidation in water gives another results.
An ultrafine aluminium oxyhydroxide with the specific surface SBET = 368 m2/g was
produced by oxidation of nanosized aluminium by hot water. From the other hand, an
oxidation with water in combustion regime allows producing nanosized corundum.
Important is that the dependence of combustion rate on pressure is very weak, and is even
independent of pressure in the area of 5 8 Mpa.
It has been established that the combustion rate sufficiently grows when a high-dispersed
powder is used.
302

Complex Research of Al Nanopowders Produced by the Exploding Wire . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In conclusion we can say that the exploding wire method is a very promising method of
ultra-fine powders production. It is possible to produce ultra-fine powders with controlled
phase compositions.
Real future success requires the coordinated joint efforts of a wide circle of researchers
and technologists, producers and users.

REFERENCES
[1]

L.I. CHEMEZOVA, V.G. IVANOV, YU.F. IVANOV, M.N. OSMONALIEV, V.S. SEDOI, I.A.
SOBOLEV: Production of Nano-Al Powders by the Exploding Wire Method, Proceedings of
International Conference on Combustion and Detonation, Zeldovich Memorial II, Moscow,
paper WS1-4 pp. 1-58, 2004
[2] L.T. DE LUCA, L. GALFETTI, F. SEVERINI, L. MEDA, G. MARRA, A.B. VOROZHTSOV, V.S.
SEDOI, AND V.A. BABUK: Burning of Nano-Aluminized Composite Rocket Propellants,
Combustion, Explosion, and Shock Waves, Vol. 41, No.6, pp. 680-692, 2005
[3] F. TEPPER AND L. KALEDIN: Combustion Characteristics of Kerosene Containing Alex NanoAluminum (Combustion of Energetic Material, Edited by K.K. Kuo and L.T. De Luca), Beggel
House Inc., New York. Wallingford, pp. 195-206, 2002
[4] B. PALASZEWSKI: Metallized gelled propellants: historical and future developments, (Rocket
Propulsion: Present and Future, L.T. De Luca, Eds.), Grafiche GSS, Bergamo, Italy, pp. 32(119), 2002
[5] V.G. IVANOV, S.N. LEONOV, G.L. SAVINOV, O.V. GLAZKOV, O.V. GAVRILUK: Combustion of
Mixtures of Ultrafine Aluminum With Gel-Like Water, Combustion, Explosion, and Shock
Waves, 30, pp. 569-570, 1994
[6] V.G. IVANOV, O.V. GAVRILYUK, O.V. GLAZKOV, and M.N SAFRONOV: Specific Features of
the Reactions Between Ultrafine Aluminum and Water in a Combustion Regime, Combustion,
Explosion, and Shock Waves, 36, pp. 213-219, 2000
[7] V. SEDOI, V. IVANOV, YU. IVANOV, M. OSMONALIEV, M. LARICHEV, and O. LARICHEVA:
Synthesis by the Exploding Wire Method and Characterization of Nanoaluminum Powders with
tInorganic Coatings, Proceedings of the 36th International Annual Conference of ICT,
Karlsruhe, Federal Republic of Germany, paper 98, pp. 1-11, June 28-July 1, 2005
[8] L. CHEMEZOVA, V. IVANOV, YU. IVANOV, M. OSMONALIEV, V. SEDOI, and I. SOBOLEV:
Complex Research of Al Nanopowders Produced by the Exploding Wire Method, European
Conference for Aero-Space Sciences, July 4-7, 2005, Moscow, pp. 5.3(1-7)
[9] YU. F. IVANOV, M,N. OSMONALIEV, V.S. SEDOI, et al.: Production of Ultra-Fine Powders
and Their Use in High Energetic Compositions, Propellants, Explosives, Pyrotechnics, Vol. 28,
No. 6, pp. 319-333, 2003
[10] V.S. SEDOI, YU.F. IVANOV, and M.N. OSMONALIEV: Study of Ultra-Fine Powders Produced
by the Exploding Wire Method (Novel Energetic Materials and Applications, L.T. De Luca, L.
Galfetti, and R..A. Pesce-Rodrigues, Eds.), Grafiche GSS, Bergamo, Italy, pp. 16 (1-10), 2004
[11] V.S. SEDOI, V.V. VALEVICH: Production of Highly Dispersed Metal Powders by Electrical
Explosion in Reduced-Pressure Nitrogen, Tech. Phys. Lett. Vol. 25, pp. 584-585, 1999.
[12] V.V. VALEVICH, V.S. SEDOI: Ultrafine powders obtaining by the fast electrical explosion,
Russian Physics Journal, No. 6, pp. 70-76, 1998

Acknowledgment:
This work was performed with financial support by INTAS (grant 03-53-5203)

11

Valentin Sedoi, Galina Volkova, Mirslan Osmonaliev, Victor Ivanov, . . .

303

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

AB INITIO MD COMPUTATION OF THE VIBRATIONAL


RELAXATION TIME IN HE MOLECULES
A.A.Selezenev, A.Yu.Aleynikov
RFNC-VNIIEF, Sarov, Russia
Abstract:
Ab Initio MD simulation has been used to compute equilibrium time for the
intramolecular vibrational degrees of freedom in molecules of TNT, RDX, HMX and
PETN. Computation of intramolecular forces field was done utilizing the code
Gaussian with HF/STO-3G method.
Ab Initio MD simulations have been carried out for the decomposition of TNT and RDX
molecules as they collide. For the kinetic energy of molecular collisions, its threshold
values have been found where decomposition of the molecules should occur before the
equilibrium distribution of collision energy into vibrational degrees of freedom.

INTRODUCTION
For kinetic modeling of the HE decomposition by detonation, it is important to find the
conditions that can make direct scission of the chemical bonds occur in HE molecules [1-5].
One possible mechanism of a chemical bond scission in the shock-wave front may be by
direct transfer of a kinetic energy to this bond that is larger than its dissociation energy [1, 4, 5].
Then, if the intramolecular vibrational relaxation time vv is more than the characteristic
time ( X ) of kinetic energy transfer to the bond in question, the energy going to the bond
may be actually higher than the dissociation energy. The occurrence of such mechanism has
been proved to be possible by direct MD simulations of the HE molecule decomposition in
shock-wave front. Thus, the studies using MD simulation [6, 7] have observed for strong
shock waves that the time of chemical bond scission in HE molecules may be of
picoseconds scale. Based on this mechanism, the bond scission essential requirements can
be written as:
Ek > 2 U 0 ,

(1)

X < VV ,

(2)

X < TV ,

(3)

where - E k is the kinetic energy of molecular collision, U 0 - bond dissociation energy,


VV - intramolecular V-V relaxation time, X - characteristic time of kinetic energy
transfer to the bond in question, TV - natural period of the bond ( TV = 2 / 0 ).
Requirement (3) is that the interaction of a colliding molecule and the given vibrational
mode of its partner molecule should be nonadiabatic. There can be efficient transfer of
translational energy to the vibrational mode, provided Massey parameter (Me) is on the order
of or less than unity [6]. In the case of interest, the relation used for Massey parameter is

304

Ab Initio MD Computation of the Vibrational Relaxation Time in HE . . .

New Trends in Research of Energetic Materials


Me =

Czech Republic

2006

[Content]

X X 0

. There is a correlation established between the values in requirements


TV
2

(1)-(3) and HE detonation parameters. From (1)-(3), obtaining such correlation evidently
requires calculations of the HE molecule vibration spectra, bond dissociation energy (BDE)
and V-V relaxation time.

1.

COMPUTING VIBRATIONAL SPECTRA AND NO2 BOND ENERGY


OF HE MOLECULES

Thus, infrared spectra and bond dissociation energy have been calculated for the
molecules of six HE formulations: TNT, RDX, HMX, PETN, TATB and NM. These IR
spectrum and BDE calculations used Gaussian code [7]. The IR spectra were calculated for
preoptimized structure of the molecules. The molecules were optimized in structure and their
spectra calculated by density functional technique (using B3LYP/6-311++g** basis set).
Table 1 shows comparison between the vibration wave numbers of identical fragments
in different HE molecules. The Table gives wave numbers of the vibration modes of
molecules (cm-1) and the molecular fragments to which the vibration modes are related.
Table 1.

Typical vibration frequencies of HE molecules

HE
Fragment\

RDX
1658.71,
1657.96

Nitromethane
1622.88

PETN

TNT

HMX

1747.39,
1735.1,
1729.83

1602.52

1638.88,
1652.77,
1623.38

1589.02
1645.95

905.17,
905.514

918.015,
930.943,
904.622,
996.499
841.648,
844,42,
835.659

1345.04

1285.82,
1303.01,
1310.96,
1245.03,
1339.05

1291.81,
1293.3
1314.36,
1323.37

1369.64,
1379.77
As indicated by the calculations summarized in Table 1, the vibration modes of identical
fragments in different HE molecules have similar frequency values.

Alexander Selezenev, Aleksey Aleynikov

305

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

For accuracy test of the applied numerical method, the calculations of the current study
were compared with the experimental wave number values of some vibration modes in
nitromethane [10]. From this comparison, the calculated wave numbers are found to fit the
experimental values with the relative accuracy of 0.6% or better.
The dissociation energy calculations for -NO2 N-NO2 bonds were also performed
using B3LYP/6-311++g** basis set. The value of bond dissociation energy was calculated as
a difference between the total energy of a molecule in initial state (with allowance for zero
vibration energy) and that of the dissociation products. For RDX molecule, for instance, the
expression for BDE calculation of N-NO2 is
BDE=E (C3H6N6O6) - [E (C3H6N5O4) + E (NO2)].
The BDE calculation was also conducted to find the bond energy dependence on
interatomic distance. The bond energy relation to interatomic distance was approximated by
Morse potential

U (r ) = U 0 [e 2 ( r r0 ) 2e ( r r0 ) ] .
Table 2 shows parameters of Morse potential obtained for R-NO2 bonds in the different
HE molecules. Table 2 also compares the calculated BDE values with the respective
calculations in [11, 12].
Table 2. Morse potential parameters
HE

Bond type

0
U 0 , eV
, 1/ A
(BDE)
TATB
3.8316
2.1483
C NO2
TNT
3.0804
2.2642
C NO2
NM
4.7198
1.5777
C NO2
HMX
1.8937
2.8467
N NO2
RDX
1.7432
2.8351
N NO2
PETN
1.7845
2.6684
O NO2
It is of interest to consider these bond energy
molecules in the shock front of detonation wave:

Ek =

r0 , A

BDE, eV
[11]

BDE, eV
[12]

1.4224
3.35
3.00
1.4484
2.90
2.55
1.5533
1.3799
1.86
1.3922
1.80
1.4082
1.73
values versus the kinetic energy of HE

M U s2
= 0.5182 10 2 U s2 , = M / mu ,
2

where E k is the kinetic energy of molecule, eV, - relative molecular mass, mu atomic mass unit (= 1.6605655 10 27 kg), U s - molecular collision velocity, km/s.
Considering that no more than the kinetic energy of collision [14] can change into
internal energy at molecular collision, Fig 1 makes the comparison with BDE easier as it
gives the value of 0.5Ek as a function of relative collision velocity.

306

Ab Initio MD Computation of the Vibrational Relaxation Time in HE . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

8
-1
-2
6
0.5Ek, eV

-3
-4
4

0
0

0.5

1.5

2.5

3.5

U, /

Fig 1. Kinetic energy of collisions as a function of collision velocity for HE molecules.


(1-PETN, 2-HMX, 3-TNT, 4-RDX).
Comparing the results given in Table 2 and Fig 1, one can see for the collision velocity
of HE molecules at U>2 km/s, that kinetic energy of molecular collision can be larger than
NO2 bond dissociation energy.

2.

COMPUTING INTRAMOLECULAR VIBRATIONAL RELAXATION


TIME IN HE MOLECULES

The following procedure was used for computation of intramolecular V-V relaxation
time in HE molecules. Prior to this, the molecule was structurally optimized using
B3LYP/6-311++g** basis set. For this optimized structure, one of R-NO2 bonds there was
brought out of equilibrium. Thus, initial disturbance of the bond length was specified. The
increase in potential energy of the bond was U 1 (U 1 1 2 eV ) . This was followed by MD
simulation of the energy redistribution by vibration modes of the molecule in time. The
intramolecular force field was calculated using Gaussian code, and SageMD [13] code was
used for MD simulations and data processing. During MD simulation, there were
computations done to find kinetic energy for each atom in molecule and the total kinetic
energy of all atoms (less translational energy of molecule).

A typical kinetic energy total of atoms with time is given in Fig 2 for RDX molecule.
RDX
0.9
0.8
0.7
Ekin [eV]

0.6
0.5
0.4
0.3
0.2
0.1
0
0

10

Tim e [ps]

Fig 2. The kinetic energy of atoms in RDX molecule with time.


4

Alexander Selezenev, Aleksey Aleynikov

307

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

One way to estimate intramolecular V-V relaxation time is as follows. In fact, relative
variance of total kinetic vibration energy can be shown to be inversely proportional to
excited vibration modes of the molecule

DEK
EK

1
,
NK

(4)

where DEK is the dispersion of total kinetic vibration energy, EK

- the average of

total kinetic vibration energy, N K - excited vibration modes of the molecule.


Given relationship EK = U1 / 2 , expression (4) can be written as

4 DE
1

or N K (t ) U12 / 4 DE .
2
U1
NK

(5)

Thus, by dividing the analyzed time interval into subintervals << VV and calculating
the kinetic energy variance on each time interval , a relationship can be obtained for the
number of excited vibration modes versus time. Obviously, relationship (5) should yield the
total of molecular vibration modes for >> VV
U 12

= N K
4DE

>> vv

( N K = 3N 6) ,

where N number of atoms in molecule.


With the above discussed approach, it is possible to make lower estimate of vv
parameter, i.e. to find its minimum value. The actual value of vv could be only larger than

U 12

the value as obtained thereby. Table 3 includes the values obtained for 4DE
and the

ff vv
calculated values of

vv .

Table 3. Calculated vibrational V-V relaxation times


N

NK

U 12

4DE

ff vv

vv , ps

TNT

21

57

59

RDX

21

57

60

HMX

28

78

82

PETN

29

81

82

2.5

Parameter
HE

The MD simulations have found the excitation of different vibration modes in HE


molecules to be essentially non-simultaneous. One might expect (and simulations have

308

Ab Initio MD Computation of the Vibrational Relaxation Time in HE . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

proved it), that the vibration modes at near initially excited mode frequency would have
small relaxation times. Evidently, there would be resonant excitation of this vibration mode.
And conversely, the relaxation time would be larger for the vibration mode whose natural
frequency is much different from that of excited vibration mode. This effect is illustrated
below using TNT molecule as an example (Figs 3-8). Figs 3 and 4 show the bond length and
kinetic vibration energy with time for the initially excited -NO2 bond in TNT molecule.
Figs.5 and 6 show similar data for the -NO2 bond that was initially unexcited.
From the data in Figs 5-6, the vibrations of -NO2 bond are evidently excited without
any delay.
Figs 7-8 provide the bond length and kinetic vibration energy with time for the initially
unexcited C-H bond in TNT molecule.
From the data in Figs 7-8, excitation of - bond occurs with significant delay in time.

3.

MD SIMULATION OF HE MOLECULE COLLISIONS

The simulations of bond scission by collision were conducted for TNT and RDX
molecules. These simulations were performed using PM3 method. Collision geometry was
considered, which was most effective to scission of R NO2 bond (R=C for TNT and R=N
for RDX molecule). Orientation of the molecules was made such that one of R NO2 bonds
would be directed collinearly with the collision. Initial position of TNT molecules in space
and the configurations of molecules after collision are shown in Fig 9.

Fig 9. Fragmentation of TNT molecules after collision at 8.25 km/s velocity.


6

Alexander Selezenev, Aleksey Aleynikov

309

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The simulations have found a threshold velocity of collision - the point where scission of
the bonds, -NO2 and N-NO2, would occur as the molecules collide. For TNT and RDX
molecules, these values were 8.1 and 7.6 km/s, respectively. The value of Massey parameter
for threshold velocity of collision was 0.5.
Thus, as it has been observed by direct numerical MD simulations, the scission of
-NO2 and N-NO2 bonds in the HE molecules colliding with each other can occur, with
probability of near 1, if Massey parameter is ~0.5 (obviously, provided E k > 2 U 0 energy
criterion). The bond dissociation energy data have been used to calculate the particle velocity
U S for Massey parameter to be equal to 0.5. For RDX, HMX and PETN molecules, these
particle velocity values were found to be 7.58, 7.55 and 7.53 km/s, respectively. The
observed values for particle velocity are 2 or 2.5 times the particle velocity of substance at
Neumann peak of strong detonation waves. Given the law of particle velocity doubling upon
the shock wave reaching the free surface, one can expect the analyzed bond scission
mechanism to occur, e.g., when the shock wave is interacting with micropores of the HE
lattice. The high molecule velocities can be also achieved due to dispersion of the particle
velocity at the shock wave front.

CONCLUSION
Infrared spectra and dissociation energy of NO2 bonds have been calculated for six HE
formulations. Calculation of the kinetic energy distribution by intramolecular vibration
modes has been also performed for the HE molecules of interest (V-V-relaxation time).
Molecular dynamics simulation has been carried out to study decompositions of TNT and
RDX molecules by collisions of the molecules with each other. Threshold collision
velocities have been found, when molecular decomposition should occur straight at collision.
The study also includes estimates that have been made for the probability of
decomposition for HE molecules colliding in detonation shock front. One candidate
decomposition behavior, a so-called accumulation mechanism, is discussed in which the
kinetic energy of collision is assumed to have possible location on a molecular bond. The
analysis suggests that the occurrence of such mechanism in the detonation shock front of
condensed HE is limited by the following requirements:
1. Distribution of the kinetic energy of ordered molecular motion into intermolecular
vibrations energy should be much shorter in time than it takes to achieve balance
with the energy of intramolecular vibrations. Then, the relative collision velocity of
molecules would have the same order of magnitude as the HE particle velocity.
2. With energy criterion observed, the bond dissociation probability is substantially
dependent on Massey parameter, which in turn is determined by collision time and
the vibration period of analyzed chemical bond.
The authors would like to express their thanks to Dr. V.N.Lashkov for his support to the
study and valuable discussions in analysis of the results.

REFERENCES
[1]

310

Klimenko V.Yu, Dremin A.N. On the decomposition reaction kinetics in the shock wave
front, In A.G. Merzhanov, editor, Detonation, chemical Physics of combustion and explosion
processes, Russia, Chernogolovka, 1980, Nauka.

Ab Initio MD Computation of the Vibrational Relaxation Time in HE . . .

New Trends in Research of Energetic Materials

[2]
[3]
[4]
[5]

[6]
[7]

[8]
[9]

[10]
[11]

[12]

[13]

[14]

Czech Republic

2006

[Content]

Klimenko V.Yu., Yakoventsev M.A., Dremin A.N. Multiprocessor detonation model,


Khimicheskaya Fizika, 12, 5, 1993, pp.671-680.
Tarver C.M., Chemical Energy in Self-Sustaining Detonation Waves in Condensed
Explosives, Combustion & Flame, Vol.46, 1982, pp. 157-176.
Walker F.E., Physical kinetics, J. Appl. Phys. Vol.63 (11), 1988, pp.5548 5554.
Walker F.E., Quantum Mechanics and Molecular Dynamics Calculations Provide New
Evidence for a Free-Radical Shock Initiation Model, Propellants, Explosives, Pyrotechnics
Vol.7, 2-7, 1982.
Manaa M.R., Fried L.E., Melius C.F., Elstner M. and Frauenheim Th., J. Phys. Chem. A 106,
2002, p.9024.
Alejandro Strachan, Adri C.T. van Duin, Debashis Chakraborty, Siddhard Dasgupta, and
William A. Goddard III, Shock Waves in High-Energy Materials: The initial Chemical Events
in Nitramine RDX, Physical Review Letters, Vol. 91, 9, 2003, pp. 09801-1 09801-4.
Zeldovich Ya.B., Raizer Yu.P. Physics of shock waves and high-temperature hydrodynamic
events. 2nd edition, M.: Nauka, 1976.
Gaussian 03, Revision C.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C. Burant, J. M.
Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G.
A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida,
T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B.
Cross, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R.
Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J.
Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K.
Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S.
Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L.
Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P.
M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez, and J. A. Pople, Gaussian, Inc.,
Wallingford CT, 2004.
Zerilli J.F. and Toton E.T., Shock induced molecular excitation in solids, Physical Review,
B, Vol. 29, 10, 1984, pp. 5891 5902.
Wu C.J. and Fried L.E., First-Principles Study of High Explosive Decomposition Energetics,
Eleventh Symposium (International) on Detonation, Naval Surface Weapons Center NSWC 86194, Albuquerque, NM, 1998, p.490.
Rice B.M, Sahu S., Owens F.J., Density functional calculations of bond dissociation energies
for NO2 scission in some nitroaromatic molecules, Journal of Molecular Structure
(Theochem), Vol.573, 2002, pp.69-72.
Selezenev A.A.; Aleynikov A. Yu.; Gantchuk N.S.; Yermakov P.V.; Labanowski J. K.
Sage MD: Molecular Dynamics Software Package to Study Properties of Materials with
Different Models for Interatomic Interactions Computational Materials Science, Vol. 28,
Issue 2, October 2003, pp. 107-124.
.. . .:
, 1976. (Nikitin E.E. Theory of elementary atomic-molecular behaviors in gases. M.:
Khimia, 1976.)

Alexander Selezenev, Aleksey Aleynikov

311

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

APPENDIX
1.4
2.6

1.2
2.4

2.2

Ek, eV

R, A

2
1.8
1.6
1.4

0.8
0.6
0.4

1.2

0.2
1
0

t, ps

0
0

t, ps

Fig 3. Bond length with time.


(initially excited -NO2 bond in TNT
molecule)

312

Fig 4. Kinetic vibration energy for initially excited


-NO2 bond.

Ab Initio MD Computation of the Vibrational Relaxation Time in HE . . .

R, A

0.5

1.5

2
t, ps

2.5

Alexander Selezenev, Aleksey Aleynikov

4
t, ps

3.5

4.5

0.5

1.5

2
t, ps

2.5

3.5

4.5

Fig 8. Kinetic vibration energy


for initially unexcited -H bond.

0
t, ps

0.05

0.1

0.2

0.15

0.25

Fig 6. Kinetic vibration energy for initially unexcited NO2 bond.

2006

Fig 7. C-H bond length with time


(initially unexcited -H bond).

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Czech Republic

0.8

1.2

1.4

Fig 5. Bond length with time


(initially unexcited -NO2 bond).

1.2

1.3

1.4

1.5

1.6

1.7

1.8

Ek, eV

Ek, eV

10

R, A

New Trends in Research of Energetic Materials


[Content]

313

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

UNUSUAL THERMAL DECOMPOSITION OF NTO: IS IT A RESULT OF


VERY STRONG BONDS OR OTHER REASONS EXIST?
V.P. Sinditskii, S.P. Smirnov, V.Y. Egorshev
Mendeleev University of Chemical Technology
9 Miusskaya Square, 125047, Moscow, Russia

Abstract:
Burn rate characteristics of low sensitive explosive 3-nitro-1,2,4-triazol-5-one (NTO) have
been investigated in pressure interval of 0.1-40 MPa. The temperature distribution in the
combustion waves of NTO has been measured at pressures 0.4-2.1 MPa. Based on burn rate
and thermocouple measurement data, rate constants of NTO decomposition in the molten
layer at 370-425C have been derived from a condensed-phase combustion model
( k = 1.14 1014 exp(39000 / RT ) , s-1), and NTO vapor pressure above liquid phase (lnP = 9914.4/T + 17.12) and solid phase (lnP = -12984.4/T + 22.78) has been calculated.
Taking into account vapor pressure data obtained, decomposition of NTO in the gas phase
at 240-250C has been studied. Comparison of decomposition rate constants in gaseous and
solid state shows that they are close together, therefore, partial decomposition in the gas
can not substantially increase the total rate. High values of the activation energy for solidstate decomposition of NTO are not likely to be connected with submelting effect, because
decomposition occurs at temperatures well below melting point. It has been suggested that
abnormally high activation energy in the interval of 230-270C is stipulated by peculiarities
of NTO transitional process rather than strong bonds in the molecule. In this area, NTO
molecule undergoes isomerization into the aci-form, followed by C3-N2 heterocyclic bond
rupture, resulting in abnormally high value of the observed activation energy.
3-nitro-1,2,4-triazol-5-one (NTO), burn rate, thermal decomposition,
Keywords:
temperature profiles in the combustion waves, mechanism of decomposition

1.

INTRODUCTION

The thermal decomposition of insensitive, high explosive 3-nitro-1,2,4-triazol-5-one (NTO)


was studied widely.[1-12] Various possible mechanisms were suggested, such as a direct scission
by the C-NO2 bond or initial hydrogen transfer with subsequent C-NO2H cleavage. Bimolecular
mechanism producing CO2 and mechanism involving NO2 rearrangement to ONO followed by
direct loss of NO were also reported. The activation energies of NTO ranged from 40.7 kcal/mol
in the interval of 110-140C to 77-88 kcal/mol in the interval of 200-260C. In spite of many
theoretical studies[13-18] proposing that high activation energies are connected with a very strong
C-NO2 bond, one may conclude that the description of the NTO decomposition processes is still
far from perfect. This is particularly obvious from the fact that activation energies of NTO
decomposition in acetone and water solutions (38.4 and 48.8 kcal/mol)[9] are far less than those
in the solid state.

314

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

In order to resolve this large discrepancy in experimental results, smaller values of the
Arrhenius parameters were suggested in work[6] to be a result of predominant or partial
sublimation. Thermogravimetric analysis and differential scanning calorimetry[12] showed that
heating of NTO lead to sublimation competitive with the condensed-phase exothermic
decomposition. Sublimation occurred most readily in open vessel conditions and was described
by low activation energy.
In resent works of Kondrikov et al.,[19-21] the rate constants of NTO initial decomposition
stage were described by two straight lines intersecting at 230C. Within the interval of 230260C, the decomposition rate showed a very high temperature sensitivity (activation energy of
98 kcal/mol), a result observed in many studies. However in the interval of 200-230C, the
activation energy of decomposition was found to be substantially less (42.7 kcal/mol) and was
in a good correlation with data of work[3] in the interval of 110-140C (40.7 kcal/mol).
Experiments in the works[19-21] were conducted in closed glass manometers of the Bourdon
type that did not allow explanation of obtained activation energy by sublimation process. The
behavior at temperatures below 230C, therefore, was attributed to the first-order reaction of
neat crystalline NTO decomposition. In the high temperature area, according Kondrikov the
decomposition includes four macroscopic stages. It comprises fast initial thermal decomposition
stage, first-order decomposition of crystalline NTO, first-order decomposition of NTO in the
liquid or quasi-liquid layer on the surface of the crystals, and an autocatalytic reaction in this
layer. The fast initial decomposition is attributed to reactions of NTO vapors as well as to the
condensed-phase decomposition of a proposed NTO derivative.
It is well known that faster decomposition reactions in liquid and gas phases can distort
results of solid-phase decomposition.[22] However, effect of sample submelting is generally
observed in the vicinity of melting point and is strongly affected by sample purity. During NTO
decomposition a sharp increase in the rate is observed 40 degrees before melting. Additional
purification does not influence this temperature gap. Besides, kinetic parameters describing
NTO decomposition in the liquid phase[21] are quite different from experimental data on NTO
decomposition in solutions.[9] As concerning effect of the gas phase decomposition one may
note that the decomposition kinetics in the interval of 230-260C observed in works[19,20] at ratio
of the sample mass to the volume of the reaction vessel (m/V ratio) of about 1 - 2 mg/cm3 are
similar to kinetics reported in works[8,9] at m/V ratio of about 5 - 10 mg/cm3, suggesting a weak
role of the gas-phase processes in NTO decomposition.
One may conclude, therefore, that the description of NTO decomposition processes in the
interval of 230-270C still remains far from perfect. Data on decomposition of NTO in the gas
phase are lacking so far.
A previous study of NTO combustion[23] showed that the burning rate was governed by the
condensed-phase chemistry at low pressures. This may allow finding NTO decomposition rate
constants in a high-temperature interval of 350-500, provided detailed investigations of
temperature distribution in the combustion wave are conducted. The purpose of the present
work was to investigate mechanism of NTO decomposition in a wide temperature interval and
different phase states of the material. The kinetic results obtained may be hoped to clarify
mechanism of NTO thermolysis.

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

315

New Trends in Research of Energetic Materials

2.

Czech Republic

2006

[Content]

EXPERIMENTAL

Preparation. NTO was synthesized from semicarbazide hydrochloride and formic acid
followed by a mixed acid nitration.[24] Recrystallisation from hot water gave the target NTO as a
white crystalline solid with a melting point of 269C (decomposition).
Decomposition Study. Manometric experiments were carried out in thin-walled glass
manometer of the compensation type (the glass Bourdon gauge) at 240 and 250C. The ratio of
the sample mass to the volume of the reaction vessel was about 0.1 mg/cm3.
Combustion Study. Burning rates were measured in a window constant-pressure bomb of
1.5-liter volume in the 0.1-40 MPa pressure interval. Samples to test were prepared as pressed
cylinders of 0.98 TMD (1.87 g/cm3) confined in transparent acrylic tubes of 7 mm i.d. prior to
pressing, the material was thoroughly comminuted in order to produce samples with a minimum
possible pore size, thus minimizing the possibility of flame propagation between particles.
Temperature profiles in the combustion wave were measured using fine thermocouples. The
thermocouples were welded from 80%W + 20%Re and 95%W + 5%Re wires 25 m thick
followed by rolling in bands 5-7 m thick. The thermocouple signal was recorded with a Pico
ADC 216 digital oscilloscope.
Thermodynamic calculations were performed with the use of the REAL computer
simulation of chemical equilibrium code.[25]

3.

RESULTS AND DISCUSSION


3.1

Burning behavior and flame structure of NTO

NTO in the form of samples pressed into PMMA tubes of 7 mm diameter starts burning at
0.4 MPa with the rate of 0.3 mm/s. At atmospheric pressure NTO can sustain combustion only
when external heating is continuously provided by an electrically heated nichrome wire. In
doing so, a decomposition reaction with formation of gaseous products occurs, but no gas-phase
flame is observed. The gas flame appears at pressures when NTO is capable of sustained
combustion, but again the burning is accompanied by formation of copious yellow smoke and
solid residue inside the tubes. The NTO burning rate-pressure dependence is expressed as rb =
0.727p0.976, mm/s for the pressure interval of 0.4-11.9 MPa, and rb = 1.56p0.667 for the interval of
11.9-40 MPa.
The burning rates of NTO are close to the burning rates of TNT. As compared with HMX,
NTO burns 2 times slower within the whole pressure interval (Fig 1). Seemingly, it is a rather
low burning rate as well as a low pressure exponent (~0.67) at high pressures that are
responsible for low sensitivity of the substance to mechanical and shock stimuli.
The yellow-brown residue formed was shown by IR-spectroscopy to contain NHCOfragments. An analysis of the gaseous combustion products at 0.5 MPa revealed the presence of
NO, N2O, CO and CO2 species.

316

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

100

5
HMX

Burning rate, mm/s

3
2
10

5
TNT

3
2
1

NTO

5
3
2
2

6 8 1

6 8 10

Pressure, MPa
Fig 1.

Comparison of burning rates of NTO, TNT and HMX.

2100

T2

T2

2.1

1900
T2

Temperature, K

1700

T2

1.0

1500

0.5

1300
1100

T1

0.4

T1

T1

900
700
500
300
0.0

2.0

4.0

Distance, mm
Fig 2.

Typical temperature profiles for NTO recorded at 0.4, 0.5, 1.1, and 2.1 MPa.

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

317

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

1000
1.0

0.5

900

Temperature, K

0.4

800

Tc
Ts

700

Tc

600
Tm

500
400
300
-0.8

-0.4

0.0

0.4

0.8

Distance, mm
Fig 3.

Temperature distribution in the condensed phase and adjacent gas phase at


combustion of NTO.

Typical temperature profiles in the combustion wave of NTO measured with fine -shaped
tungsten-rhenium thermocouples at 0.4, 0.5, 1.0 and 2.1 MPa are shown Fig 2 Fig 3. All
experimental data obtained are collected in Fig 4 and Table 1.
2400

Tad

2200

T2

Temperature, K

2000
1800
1600
1400
1200

T1

1000

Tc

800

Ts

600
400
200
3

Pressure, MPa
Fig 4.

Characteristic temperatures in the combustion wave of NTO: surface temperature


(Ts), temperature of carcass (Tc), temperature of first (T1) and second (T2) flame.
Dotted line is NTO boiling temperature.

Thermocouple-aided measurements allow the following typical zones to be distinguished on


NTO temperature profiles (Fig 2 and Fig 3):

318

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

preheated zone in the condensed phase from the initial temperature (To) to the first break
on the profiles at temperatures of 510-540 (237-267), which are close to the melting
point of NTO (270C).[26] For this temperature interval, the thermal diffusivity
coefficient () was calculated from Michelsons distribution to be (1.90.2)10-3 cm2/s;
molten layer from the melting point (Tm) to the surface temperature (Ts), which depth
decreases from 0.27 mm to 0.15 mm when pressure increases from 0.4 to 2.1 MPa.
Michelsons temperature distribution yields the thermal diffusivity coefficient in the
liquid phase as (1.70.2)10-3 cm2/s;
adjacent-to-the surface gas zone of less than 0.1 mm thickness, which includes also a
solid carcass made perhaps of NTO polymeric decomposition products. The temperature
gradient in this zone differs little from gradient in the condensed phase. The carcass
boundary is visible on profiles taken at low pressures (0.4-0.5 MPa) as sharp breaks at
temperatures of 780-820 (507-547).
following is aerosol zone, clearly visible at 0.4-0.5 MPa, in which temperature rises from
carcass decomposition/evaporation temperature (Tc) to the first flame temperature (T1).
luminous flame zone with final temperature T2, which appears at a considerable distance
from the surface (~3 mm) at low pressures and approaches fast to the burning surface as
pressure grows (~0.25 mm at 1 MPa), resulting in profiles with one-stage flame structure
at pressures above 2.1 MPa. Oscillating behavior of temperature traces in this zone is
indicative of condensed particles presented in the flame.
The maximum flame temperature on NTO combustion increases from 1500 to ~2000 K as
pressure rises from 0.4 to 2 MPa, approaching to calculated adiabatic one, which is ranging
around 2200 K. This result may be indication of almost full energy release in NTO combustion
at relatively low pressures.
Table 1. Characteristic temperatures in the combustion wave of NTO (Confidence
intervals are given for three and more parallel runs).
Pressure, MPa

Ts , K

T1, K

T2, K

0.4
0.5
0.6
1.1
1.6
2.1

645
63915
624
67710
673
70018

980
103020
1100
1150
1075
-

1500
166761
1750
1780
1975
198020

Adiabatic flame
temperature, K
2223
2224
2224
2224
2225
2225

The dependence of NTO boiling temperature on pressure was calculated from known
thermodynamic equation:
-RT lnKg = (H-TS) evaporation, where Kg = P,
It was assumed that the surface temperatures fit with boiling points of NTO at corresponding
pressures. At pressure of 0.5 MPa it is 63915 K that has been calculated from seven parallel
runs.

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

319

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

The heat of evaporation of 19.7 kcal/mole was obtained as the difference between heat of
sublimation of 26.4 kcal/mole and heat of melting of 6.7 kcal/mole. Heat of sublimation
calculated from enthalpies of formation of solid (-30.93 kcal/mole[26] and gaseous (-3.2 and 5.7
kcal/mole[27,28]) NTO, and enthalpy of melting of 6.7 kcal/mole calculated from solubility data
of NTO in nitric acid solutions.[29] Unusually high assessment of enthalpy of melting, 92.7
kJ/mole (22.2 kcal/mole), indicated in work[30] is, undoubtedly, an erroneous value.
The dependence of NTO boiling temperature on pressure is presented in Fig 4. Fig 4 shows a
fairly good correlation between experimental surface temperatures in the interval of 0.4-2.1
MPa and calculated dependence of NTO boiling temperature on pressure (lnP = -9914.4/T +
17.12, pressure in atm) based on thermodynamic parameters and selected temperature at 0.5
MPa.
1E+2
5
3
2

1E+1

5
3
2

Pressure, atm

1E+0
5
3
2

Tm

1E-1
5
3
2

1E-2
5
3
2

1E-3
5
3
2

1E-4
0.0012

2
0.0016

0.0020

0.0024

0.0028

1/T, K-1
Fig 5.

Vapor pressure as a function of reciprocal temperature:


1 (crosses and line) NTO surface temperature, 2 (points and line) vapor
pressure above solid NTO, calculated from[6], 3 (crosses) initial pressures in
glass Bourdon gauge in experiments on NTO decomposition.

The temperature dependence of vapor pressure above solid NTO can be obtained as follows.
The sublimation rate under nonisothermal conditions was determined in the interval of 407-423
K in work.[6] It is evident that sublimation rate depends on rate of heat supply, rate of gas-phase
moving off and vapor pressure under solid substances at a given temperature. Only small
amounts of NTO were used in experiments, suggesting that the rate of heat supply was not a
limiting process. Since there were no data on the rate of gas-phase moving off, the kinetics of
evaporation of solid NTO should not be considered as a fundamental independent-of-device
value. Nevertheless, the temperature dependence obtained is characteristic of substances only,

320

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

since the rate of evacuation is temperature independent. The value obtained (25.8 kcal/mol) is
very close to the enthalpy of sublimation, 26.4 kcal/mole, calculated earlier from
thermodynamic data.[26,28] Using this value and taking into account equality of vapor pressure
above liquid and solid phases, one can obtain the temperature dependence of vapor pressure
above solid NTO as following:
lnP = -12984.4/T + 22.78
A good conformation to this expression is nearness of calculated vapor pressures to initial
pressures in glass Bourdon gauge in experiments on NTO decomposition (Fig 5).
The important feature of NTO combustion at low pressures is very low temperature
gradients (~8600 K/cm at 0.5 MPa) just above the burning surface. According to thermocouple
data, the heat transferred back from the gas phase is as low as 10-12 cal/g. This value is
considerably less than the heat required for warming-up the substance to the surface temperature
and melting it (133-136 cal/g). It can be assumed, therefore, that the leading reactions on NTO
combustion at low pressures occur in the condensed phase. After merging gas flames at elevated
pressures, the temperature gradient from the gas substantially increases, the heat flux from the
gas being, however, still insufficient.
Low values of the heat flux from the gas at pressures up to at least 2 MPa allow considering
the condensed-phase chemistry as determining the combustion of NTO. Rate constants of the
dominant combustion reaction in the pressure range of 0.4-10 MPa can be obtained from the
Zeldovich expression[31] for the mass burning rate (m) that is based on a dominant role of the
condensed-phase chemistry:
m=

2 2 Q
c p (Ts T0 + Lm / c p )

RTs2
E / RTs
) Ae
E

The average specific heat, cp, was calculated to be 0.35 cal/gK from experimental data on
thermal diffusivity () as 1.710-3 cm2/s, density of strand as 1.87 g/cm3 and thermal
conductivity of the condensed phase () as 0.00123 cal/cmsK taken from Ref.[32] The heat of
reaction, Q, was taken as 310 cal/g,[12] and the heat of melting, Lm, as 51.5 cal/g.[29]
As seen in Fig 6, kinetics of NTO decomposition derived from the combustion model,
k = 1.14 1014 exp( 39000 / RT ) s-1, do not exhibit very high activation energy, 39 kcal/mol,
and approach decomposition data of solid NTO when extrapolated to the low-temperature
region.
The correctness of these new data can be supported by following considerations. According
to the condensed-phase combustion model, the pressure exponent in the burning rate vs.
pressure dependence is defined by the activation energy of the leading reaction and the heat of
evaporation process, which controls the surface temperature. Consequently, the activation
energy of NTO decomposition reaction, which is the limiting reaction in the condensed phase,
can be found from expression E = 2nLv. Using heat of evaporation as 19.7 kcal/mole, calculated
from thermodynamic data,[26,28] and experimental value of the pressure exponent, n = 0.976, the
activation energy can be calculated as which 38.5 kcal/mole. This value is close to the activation
energy derived from the combustion model. In addition, the activation energy, which

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

321

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Rate Constant, s-1

corresponds to NTO decomposition in melt, is in a good agreement with activation energies of


NTO decomposition in solution,[9] though the rate constants in it are less a little.
1E+4
1E+3
1E+2
1E+1
1E+0
1E-1
1E-2
1E-3
1E-4
1E-5
1E-6
1E-7
1E-8
1E-9
1E-10
1E-11
1E-12
1E-13
0.00120

5
4

0.00160

0.00200

0.00240

0.00280

1/Ts, K-1
Fig 6.

3.2

Comparison of the rate constants in a wide temperature interval: 1 (crosses


and line)- the rate constants of the leading reaction in NTO combustion, 2
(line and dotted line) data of work[3], 3 (rhombs) data of work,[9] 4
(crosses) - data of work,[20] 5(triangles) data of work[1].

Thermal decomposition of NTO in the gas phase

Knowing NTO vapor pressure, experiments have been conducted on decomposition of the
substance in the gas phase. The experiments were carried out by means of the manometric
method using glass Bourdon gauges of ~13 cm3 volume at temperatures of 240 (#221) and
250C (#222) and sample weights of 1.3 and 1.4 mg, respectively. The weights of the samples
were decreased significantly as compared to analogous previous experiments.[19,20] This made it
possible to come to the ratio of the sample mass to the volume of the reaction vessel as little as
0.1 mg/cm3. In doing so, estimated vapor pressure in the vessel would be less than the saturated
vapor pressure of NTO.
The time curves of the decomposition extent for experiments at 240 and 250C are
presented in Fig 7 and Fig. 8 in comparison with decomposition curve at m/V ~ 1 mg/cm3.[19,20]
It was observed during experiments that decomposition is accompanied by formation of
brown transparent film on the internal walls of the vessel. At the same time there was no solid
residue at the bottom of the vessel observed after decomposition, although it was the case for
bigger m/V ratios.

322

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

At 240C pressure in the vessel increases in direct proportion to time, reaching the
calculated value of the vapor pressure of NTO (21 Torr). Then the decomposition rate begins to
decrease. After a half of NTO has been decomposed, a self-acceleration of decomposition is
observed, which is weaker, however, than in experiments with large m/V ratios. The total
decomposition reaction was found to produce 2.85 mole gases per one mole of NTO, as it had
been observed in earlier experiments.[20] In the area after reaching vapor pressure and up to 20%
decomposition, the reaction rate decreases in accordance with the firs-order low and is close to
the rate of decomposition at big m/V ratio (Fig 7, test #97). The firs-order rate constants for
these experiments (#221 and #97) were calculated to be 0.0037 min-1 and 0.0013 min-1,
respectively. Increasing the temperature of experiments did not result in changing the kinetic
pattern: at 250C the self-acceleration in the test with small sample mass (#222) was even
weaker, and the formal rate constant was 0.013 min-1 as compared to 0.007 min-1 for large
sample mass (Fig 8, test #100).
1.0

conversion, %

conversion, %

1.0

1
0.5

#221
#97

2
1
0.5

#222
#100

0.0
0

0.0
0

200

400

40

80

120

160

, min

600

, min

Fig 7.

3.3

Comparison of two kinetic


curves at 240C, 1- m/V = 0.1
mg/cm3, 2 - m/V = 1 mg/cm3.

Comparison of two kinetic


Fig 8.
curves at 250C, 1- m/V = 0.1
mg/cm3, 2 - m/V = 1 mg/cm3.

Mechanism of NTO thermal decomposition

The kinetic data of NTO decomposition collected for a wide temperature interval
comprising solid, liquid, and gaseous phase states of the material, indicate that unusually high
activation energy in 230-270C temperature interval is likely to be connected with some
transitional process rather than with extraordinarily strong C-NO2 bond in the NTO molecule.
In temperature interval of 230-270C, NTO vapor pressure reaches considerable values
(0.03-0.3 atm) and could affect formal kinetic parameters of decomposition. However, the
present study shows that decomposition rate constants in both gaseous and solid states are close
together. Therefore, partial decomposition in the gas can not substantially increase the total rate.
NTO decomposition in the liquid state as well as in solution[9] was shown to proceed much
faster than in the solid state. However, the effect of sample submelting at temperatures well
below melting point, as it was already mentioned above, is improbable. Most likely the
decomposition rate is apt to drastically grow in the 230-270C temperature interval for another
reason. That may be isomerization of NTO into a less stable compound.

10

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

323

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Calculations made in works[3,27,14,18] show that the C-NO2 bond in the NTO molecule is the
weakest one, suggesting homolysis of this bond to be the most probable initial step in
unimolecular decomposition. However, Bruss et al.[1] and then Oxley et al. [9] found out the
deuterium kinetic isotope effects (DKIE), leading them to a conclusion that hydrogen transfer
was involved as a rate-limiting process. Decomposition of NTO was proposed[9] to be initiated
by hydrogen transfer from N4 to O11 of nitro group. This step features concerted rupture of the
N-H bond and formation of O-H bond simultaneously with reorganization of double bonds in
the molecule:
7HN2
3
4

HN

1N

HN

N6

10

N
H8

(1)
HONO

OH

O 11

Calculations conducted in work[18] show that this path resulting in subsequent formation of
HONO has considerably higher barrier than rupture of C-NO2 bond.
At the same time it is reasonable to suggest a structure isomer of NTO, which can be
formed without any bond ruptures as a result of hydrogen shift from N2 to O11:
HN

K1

(2)
O

NO2

N
H

NO2H

N
H

This NTO isomer contains nitro group in the form of quite ordinary aci-form as contrasted
with above-stated product of hydrogen transfer (1). Appearance of yellow color on heating of
NTO samples testifies to formation of aci-NTO. Conversion of nitro group into the aci-form
conjugated with double bond is well known to be usually accompanied by bathochromic shift of
absorption maximum in UV-spectra in visible area.[33]
Calculations based on semi-empirical method AM1 show that the longest bond among C-N
bonds in aci-NTO is the C3-N2 one. Therefore, it may be expected that rupture of this bond will
initiate NTO decomposition:
N

N
H

N2

k2
NO2H

.
O

(3)

.
N
H

NO2H

The radical center at C=O group makes it susceptible to intra or intermolecular attack by
nitro group oxygen, resulting in formation of CO2 at earlier stages of NTO decomposition that
was reported in Ref.[10].
The proposed scheme is in accordance with not only deuterium kinetic isotope effect data,
but also with decomposition of 15N isotopically labeled NTO,[11] which evidenced 54% N2 as
forming from nitrogen of cycle at positions 1 and 2.

324

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

According to the suggested scheme the rate of the initial stage of NTO thermolysis (W) is
determined by the rate constant of the decomposition reaction of aci-isomer (k2) and its
concentration:
W = k2[aci-NTO]
The concentration of aci-NTO depends on the equilibrium constant (K1) of the reversible
reaction (2):
K1 = [aci-NTO]/[NTO]
Using the Arrhenius expression for k2
k2 = A exp(-E2/RT)
and expressing K1 from
-RTlnK1 = H1 - TS
as
K1 = exp(S/R)exp(-H1/RT),
the reaction rate can be written as:
W = k2K1[NTO] = A exp(S/R)exp-((E2+ H1)/RT)[NTO]
In doing so, one may conclude that an increase in temperature leads to increase in not only
the aci-NTO decomposition rate, but also in the aci-NTO concentration. As a result, the
observed value of the activation energy is abnormally high. Actually it is a sum of two values:
the activation energy of aci-NTO decomposition and activation enthalpy of isomerization
reaction, E2 + H1.
Under combustion conditions, the reaction of isomerization (2) is accomplished in the melt,
and the observed activation energy corresponds to that of C3-N2 bond rapture in the cycle (39
kcal/mol). When decomposing in solutions at 220-280C, the ratio of aci-NTO is also high,
resulting in the observed activation energy to be different greatly from that observed for
decomposition in the solid state. Rate constants of NTO decomposition in the gas at 240-250C
are close to those for the condensed phase, that may be explained by predominant
decomposition on the vessel walls.

4.

CONCLUSIONS

Burn rate characteristics of low sensitive explosive 3-nitro-1,2,4-triazol-5-one (NTO) have


been investigated in the pressure interval of 0.1-40 MPa. The temperature distribution in the
combustion wave of NTO has been measured at 0.4-2.1 MPa. An important feature of NTO
combustion at low pressures is very low temperature gradients above the burning surface,
suggesting that the burning rate is governed by the condensed-phase chemistry. Thermocoupleaided measurements in the combustion wave allowed finding the surface temperatures and, as a
result, deriving NTO decomposition reaction rates in the melt at 370-425C as well as NTO
vapor pressure above the liquid and solid phases.

12

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

325

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Taking into account vapor pressure data obtained, decomposition of NTO in the gas phase
has been studied and compared with decomposition in liquid and solid state. At temperatures
above 240-250C, decomposition rates in the gas are close to that in the solid state, whereas
decomposition in the melt and in solutions occurs significantly faster and shows low values of
the activation energy. Nevertheless, high value of the activation energy for solid-state
decomposition of NTO are not likely to be connected with submelting effect, because
decomposition occurs at temperatures well below melting point. It has been suggested that
abnormally high activation energy in the interval of 230-270C is stipulated by peculiarities of
NTO transitional process rather than strong bonds in the molecule. In this area, NTO molecule
undergoes isomerization into the aci-form, followed by C3-N2 heterocyclic bond rupture,
resulting in abnormally high value of the observed activation energy. Actually it is a sum of two
values: the activation energy of aci-NTO decomposition and activation enthalpy of
isomerization reaction.

5.

REFERENCES

[1]

J.A. MENAPACE, J. MARLIN, D. BRUSS, R. DASHER: Photochemical and Thermochemical


Decomposition of 3-Nitro-1,2,4-triazol-5-one and Perdeuterio-3-nitro-1,2,4-triazol-5-one in Neat
and Mixed Systems, J. Phys. Chem., Vol. 95, pp.5509-5517, 1991.
[2] E.F. ROTHGERY, D.E. AUDETTE, R. WEDLICH, D. CSEJKA: The Study of the Thermal
Decomposition of 3-Nitro-1,2,4-triazol-5one (NTO) by DSC, TGA-MS, and ARC, Thermochem.
Acta, Vol. 185, pp. 235-243, 1991.
[3] H. STMARK: Thermal decomposition of NTO, FOA Report D-20178 2.3, National Defense
Research Establishment, Sundbyberg, Sweden, November 1991.
[4] B.C. BEARD, J. SHARMA: Early Decomposition Chemistry of NTO, J. Energ. Mater., Vol.11,
pp.325-344, 1993.
[5] K.V. PRABHAKARAN, S.R. NAIDU, E.M. KURIAN: Spectroscopic and Thermal Analysis Studies on
3-Nitro-1,2,4-Triazole-5-one(NTO), Thermochem. Acta, Vol. 291, pp.199-212, 1994.
[6] G.K. WILLIAMS, T.B. BRILL: Thermal Decomposition of Energetic Materials. 68. Decomposition
and Sublimation Kinetics of NTO and Evaluation of Prior Kinetic Data, J. Phys. Chem., Vol.99,
No.33, pp.12536-12539, 1994.
[7] T.B. BRILL, P.E. GONGWER, G.K. WILLIAMS: Thermal Decomposition of Energetic Materials. 66.
Kinetic Compensation Effects in HMX, RDX, and NTO, J. Phys. Chem., Vol.98, No.47, pp. 1224212247, 1994.
[8] J.C. OXLEY, J.L. SMITH, K.E. YEAGER, E. ROGERS, X.X. DONG: NTO Decomposition Studies, in
Decomposition, Combustion and Detonation Chemistry of Energetic Materials, MRS Symp. Proc.,
Ed: Brill, T.B., Russell, T.P., Tao, W.C. and Wardle, R.B., 1995, 27-30 Nov., Boston,
Massachusetts, USA, Vol. 418, pp.135-142, 1995.
[9] J.C. OXLEY, J.L. SMITH, L.-Z. ZUN, R.L. MCKENNEY: Thermal Decomposition Studies on NTO
and NTO/TNT, J. Phys. Chem., Vol. 99, No.25, pp.10383-10391, 1995.
[10] T.R. BOCHER, D.J. BEARDALL, C.A. WIGHT, L. FAN, T. BURKEY: Thermal Decomposition
Mechanism of NTO, J.Phys.Chem., Vol.100, pp.8802-8806, 1996.
[11] J.C. OXLEY, J.L. SMITH, E. ROGERS, X.X. DONG: NTO Decomposition Products Tracked with
15
N-Labels, J. Phys. Chem., Vol. 101, pp.3531-3536, 1997.

326

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

[12] G.T. LONG, B.A. BREMS, C.A. WIGHT: Thermal Activation of the High Explosive NTO:
Sublimation, Decomposition, and Autocatalysis, J. Phys.Chem. B, Vol.106, No.15, pp. 4022-4026,
2002.
[13] D.C. SORESCU, T.R.L. SUTTON, D.L. THOMPSON, D. BEARDALL, C.A. WIGHT: Theoretical and
Experimental Studies of the Structure and Vibrational Spectra of NTO, J. Mol. Struct., Vol. 384
(2/3), pp. 87-99, 1996.
[14] N.J. HARRIS, K. LAMMERTSMA: Tautomerism, Ionization, and Bond Dissociations of 5-nitro-2,4dihydro-3H-1,2,4-triazolone, J. Am.Chem.Soc., Vol.118, No.34, pp.8048-8055, 1996.
[15] Y.-M. WANG, C. CHEN, S-T. LIN: Theoretical Studies of the NTO Unimolecular Decomposition, J.
Mol. Struct., Vol. 460, No.1/3, pp. 79-102, 1999.
[16] Y. KOHNO, O. TAKAHASHI, K. SAITO: Theoretical Study of Initial Decomposition Process of NTO
Dimer, Phys. Chem. Chem. Phys., Vol. 3, No.14, pp.2742-2746, 2001.
[17] W.-L. YIM, Z.-F. LIU: Application of Ab Initio Molecular Dynamics for A Priori Elucidation of the
Mechanism in Unimolecular Decomposition: The Case of 5-Nitro-2,4-dihydro-3H-1,2,4-triazol-3one (NTO), J. Amer. Chem. Soc., Vol.123, No.10, pp. 2243-2250, 2001.
[18] C. MEREDITH, T.P. RUSSELL, R.C. MOWREY, J.R. MCDONALD: Decomposition of 5-nitro-2,4dihydro-3H-1,2,4-triazol-3-one: Energetic Associated with Several Proposed Initiation Routes, J.
Phys. Chem. A, Vol.102, No.2, pp.471- 477, 1998.
[19] B.N. KONDRIKOV, S.P. SMIRNOV: Mechanisms of Thermal Decomposition and Steady-State
Burning of NTO, Proc. 32nd Inter. Annual Conf. of ICT, Karlsruhe, Germany, 24- 27 June, paper
23, pp.1-14, 2003.
[20] B.N. KONDRIKOV, S.P. SMIRNOV, A.V. MINAKIN, R.M. DOHERTY: Chemical Kinetics of the
Thermal Decomposition of NTO, Propellants, Explosives, Pyrotechnics, Vol. 29(1), pp. 27-33,
2004.
[21] B.N. KONDRIKOV, S.P. SMIRNOV, YA.A. KANIKOVA, R.M. DOHERTY: Chemical Kinetics of the
Macroscopic Stages of NTO Thermal Decomposition, Proc. 35th Inter. Annual Conf. of ICT,
Karlsruhe, Germany, June 29 - July 2, paper 89, pp.1-14, 2004.
[22] G.B. MANELIS, G.M. NAZIN, YU.I. RUBTSOV, V.A. STRUNIN: Thermal Decomposition and
Combustion of Explosives and Propellants, Nauka, Moscow, 223p., 1996, (in Russian).
[23] A.E. FOGELZANG, V.P. SINDITSKII, V.YU. EGORSHEV, V.V. SERUSHKIN, V.I. KOLESOV:
Combustion of 3-Nitro-1,2,4-triazol-5-one and its Salts, Proc. Inter. Conference on Combustion,
Moscow, 12-17 Sept., pp.129-132, 1994.
[24] G.I. CHIPEN, R.P. BOKALDER, V.Y. GRINSHTEIN: 1,2,4-Triazol-3-one and its Nitro and Amino
Derivatives, Chemistry of Heterocyclic Compounds, Vol.2(1), pp.110-116, 1966.
[25] G.B. BELOV: Thermodynamic Analysis of Combustion Products at High Temperature and
Pressure, Propellants, Explosives, Pyrotechnics, Vol. 23, pp. 86-89, 1998.
[26] A. FINCH, P.J. GARDNER, A.J. HEAD, H.S. MAJDI: The Enthalpies of Formation of 1,2,4-triazol-5one and 3-Nitro-1,2,4-triazol-5-one, J.Chem.Thermodinamics, Vol. 23. No.12, pp.1169-1173,
1991.
[27] J.P. RITCHIE: Structures and Energies of the Tautomers and Conjugate Bases of Some 1,2,4Triazolones, J. Org. Chem., Vol. 54, (15), pp.3553-3560, 1989.
[28] P. POLITZER, J.S. MURRAY, M.E. GRICE: Computational Determination of Heats of Formation of
Energetic Compounds, in Decomposition, Combustion and Detonation Chemistry of Energetic
Materials, MRS Symp. Proc., Ed: Brill, T.B., Russell, T.P., Tao, W.C. and Wardle, R.B., 27-30
Nov., Boston, Massachusetts, USA, Vol. 418, pp.55-66, 1995.

14

Valerii Sinditski, Sergei Smirnov, Valerii Egorshev

327

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

[29] V.L. ZBARSKY, A. BASAL, N.V. YUDIN, V.F. ZHILIN: Study on Solubility of 2,4-Dihydro-1,2,4Triazol-3-one and 5-Nitro-2,4- Dihydro-1,2,4-Triazol-3-one in Diluted Acid, Proc. 34th Inter.
Annual Conf. of ICT, June 24 - June 27, Karlsruhe, FRG, paper 139, pp.1-8, 2003.
[30] K.J. KIM, M.J. KIM, J.-M. LEE, S.-H. KIM, H.-S. KIM, B.-S. PARK: Solubility, Density, and
Metastable Zone Width of the 3-Nitro-1,2,4-triazol-5-one + Water System, J. Chem. Eng. Data,
Vol.43(1), pp. 65-68, 1998.
[31] Y.B. ZELDOVICH: Theory of Combustion of Propellants and Explosives, Zh. Eksperimentalnoy i
Teoreticheskoi Fiziki (Russ. J. Exper. and Theor. Physics), Vol.12, No. 11-12, pp.498-524, 1942
(in Russian).
[32] Y. XIE, R.-Z. HU, C.-Q. YANG, G.-F. FENG, J. ZHOU: Studies on the Critical Temperature of
Thermal Explosion for 3-Nitro-1,2,4-Triazol-5-One (NTO) and its Salts, Propellants, Explosives,
Pyrotechnics, Vol.17, pp. 298-302, 1992.
[33] The Chemistry of Nitro and Nitroso Groups. Edited by Feuer H., Interscience Publishers, New
York, part 1, p. 286, 1959.

328

Unusual Thermal Decomposition of NTO: Is It a Result of Very Strong Bonds . . .

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THE PRINCIPAL APPROACH TO THE ESTIMATION


OF COOKOFF TEMPERATURE USING FEM
Jakub eleovsk, and Miloslav Krupka
University of Pardubice, 532-10 Pardubice, CZ
Abstract
The practical approach to the estimation of cookoff temperature using FEM is presented
in this paper. The characterization of explosive material properties (density, thermal
conductivity, specific heat, heat of decomposition, the decomposition kinetics) is
discussed. LS-DYNA3D code is used for the fitting of appropriate kinetic model and for
the simulation of heat transfer during slow cookoff. The results of simulations are
compared to experimental values for the SEMTEX 1A plastic bonded explosive.

Keywords:

FEM, cookoff, simulation, heat transfer

INTRODUCTION

The cookoff tests are used for evaluating of a hazard connected with heating of energetic
materials (especially EIDS Extremely Insensitive Detonating Substances) or ammunition
(LOVA LOw Vulnerability Ammunition). In case of energetic materials the explosive is
placed into a steel confinement and heated. In case of LOVA the whole ammunition object is
tested. The temperature of decomposition and the violence of decomposition reaction are
evaluated.
Because of the cost of experimental procedures, the numerical simulation is often used in
the field of characterization of various loading effect on the material. One of the most
commonly employed numerical techniques is the Finite Element Method FEM. The mean
principle of this method is to divide the body into small elements and to calculate the effect
of applied loads only in discrete points nodes. This creates a large system of equations,
which can then be solved numerically.
Several research groups published results of their effort in the field of cookoff computer
modeling to investigate energetic materials response to thermal heating, in a confined
geometry. Sandia National Laboratory (USA) is working on a multidimensional fully coupled
thermal/chemical/mechanical code to calculate the response of reactive materials[1-5]. They
also tried to develop model for thermally degraded energetic materials[6]. The LLNL (USA)
reported work on a single computer code specialized only for the cookoff response[7-11] of
energetic materials. The cookoff was also simulated in TNO (Netherlands)[12], or in
Australia[13].
One off the most popular FE codes, utilized in the field of explosive effects modeling, is
LS-DYNA3D from Lawrence Livermore National Laboratory (USA). LS-DYNA3D is
intended for the simulations of fast dynamic processes, like car crash tests, detonation, or e.g.
explosive burn. LS-DYNA3D is also used in Department of Energetic Materials, University
of Pardubice for the simulations of detonation processes. The modeling of heat transfer is
also possible.
The complex approach to the simulation of the heat transfer during slow cookoff using
LS-DYNA3D code is presented in this paper. The characterization of required material
properties using laboratory scale tests is also described.

Jakub eleovsk, Miloslav Krupka

329

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THEORETICAL DESCRIPTION

The calculation of the temperature distribution during cookoff can be described as the
transient heat conduction with the heat source. The transient character of the problem is given
by consecutive heating of the confinement and also by the heat generation caused by
decomposition of the explosive. The problem is nonlinear because of exponential dependence
of the heat generation on the temperature.
The equation for transient heat conduction with heat source is given:

T T T
T
+ q& ;
=
+
+
(1)
t
x y z
where C is specific heat, is density, T is temperature, t is time, is thermal
conductivity, x, y, z are cartesian coordinates and q& is a heat source.
C

The source of heat (heat generated by decomposing explosive) can be in the simplest
form expressed by Arrhenius equation (for unit volume):
q& = H r A exp( E / RT ) ;
(2)

where H r is reaction heat, A is preexponential factor, f ( ) is the function of fraction


reacted, E is activation energy and R is universal gas constant.

The equation (1) with the q& expressed from equation (2) can not be solved analytically.
However numerical solution is possible, e.g. using the finite element method. Various
commercial FE codes can be used to solve the eq. (1), we have used the LS-DYNA3D code.
The LS-DYNA code is intended for the simulation of fast dynamic processes primary. It
is also possible to simulate the heat transfer in this code, but there are some limitations. The
LS-DYNA code is not able to calculate the kinetics of chemical reactions. The heat released
by decomposing explosive can be dependent only on the temperature, not on e.g. fraction
reacted. Therefore the only one applicable kinetic model is the zero order reaction ( f ( ) = 1
in eq. (2)).
The characterization of material properties required for the solution of equation (1) and
the numerical modeling is described in experimental part.

EXPERIMENTAL PART

All the measurements and simulations described bellow were carried out with the
SEMTEX 1A plastic bonded explosive. SEMTEX 1A is commercial PETN based high
explosive, manufactured by Explosia, a.s., Czech Republic.
The estimation of cookoff temperature is done in following steps:
1. Determination of material parameters from the equation (1) C, , ;
2. determination of the decomposition kinetics H r , A, E;
3. verification of the kinetic model;
4. numerical simulation of slow cookoff.
330

The principal approach to the estimation of cookoff temperature using FEM

New Trends in Research of Energetic Materials

3.1

Czech Republic

2006

[Content]

Determination of material parameters from the equation (1)

The specific heat C of explosive material can be determined by calorimetric methods. We


used the SETARAM C80 (Setaram, France) heat flow calorimeter. The temperature
increment method was used.
The method of double weighting was used for the density determination. The sample
of explosive was weighed in air, the same sample was weighed in the water and the density
was calculated on the base of Archimedes law.
The thermal conductivity was calculated from the temperature profile inside the
sample subjected to constant ambient temperature. The flat sample (cylinder, 20 mm in
diameter and 4 mm high) with the constant starting temperature was put between two
aluminum blocks tempered to higher temperature. The schema of the apparatus JK-80
(manufactured by OZM Research, Czech Republic) is in the figure 1. The temperature profile
measured in the center of the sample was compared to analytical expression for the
temperature inside the infinite slab and was estimated.

Fig 1. The apparatus for the thermal conductivity measurement.

3.2

Determination of decomposition kinetics

The heat of decomposition H r was measured using the automatic combustion


calorimeter MS-10A (manufactured by Laget, Czech Republic) in nitrogen atmosphere. The
presence of additional substance (nitrocellulose propellant) was essential to initialize the
decomposition in inert atmosphere.
To determine the kinetic parameters A and E, the differential thermal analysis (DTA
551Ex, manufactured by OZM Research, Czech Republic) was used. The isothermal
measurements were carried out at the temperatures of 166, 168, 170 and 172 C with 50 mg
of sample. The activation energy and preexponential factor was determined by isoconverional
method. However the description of decomposition kinetics with the zero order reaction
model is not very good. The values of activation energy and preexponential factor vary with
the fraction reacted (fig. 2).
These kinetic parameters are the key variables in equation (2). Their values have a strong
influence on the amount of heat envolved by the decomposition of the sample. The selection
of a pair of kinetic parameters, which preferably describe the sample decomposition, follows.
3

Jakub eleovsk, Miloslav Krupka

331

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 2. The dependence of A and E on the fraction reacted .

3.3

Verification of kinetic model

As was said, the kinetic parameters A and E vary with the fraction reacted. The
nonisothermal DTA was simulated using simple LS-DYNA model. The various values of A
and E (corresponding to various ) were used for the description of explosive decomposition.
The simulation was carried out with five different heating rates and the results were
compared to experimental DTA curves, measured with DTA 551Ex (manufactured by OZM
Research). The example of such comparison is plotted in figure 3. Based on the least squares
method, the most appropriate model (A and E calculated at = 0.07 ) was chosen.

3.4

Numerical simulation of slow cookoff

The heat transfer during the slow cookoff test is simulated using LS-DYNA3D finite
element code, used FE model is in figure 3. The model contains 332 565 elements of
SOLID164 type. The process was modeled as the transient heat conduction with heat source
eq. (1) with the convection boundary condition. The previously determined material
parameters of explosive were used. The main result of this simulation is the cookoff
temperature.
The configuration of cookoff test, used in the Institute of Energetic Materials, is
described below.

332

The principal approach to the estimation of cookoff temperature using FEM

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 3. The comparison of simulated and experimental DTA data (heating rate 2 C/min).

Fig 4. The FE model of slow cookoff test.

Jakub eleovsk, Miloslav Krupka

333

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 5. The simulated time behavior of temperature in the center of the sample during cookoff.

Fig 6. The simulated temperature distribution before ignition.

334

The principal approach to the estimation of cookoff temperature using FEM

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Fig 7. The schema of cookoff temperature estimation using LS-DYNA3D FE code.

Jakub eleovsk, Miloslav Krupka

335

New Trends in Research of Energetic Materials

3.5

Czech Republic

2006

[Content]

Slow cookoff experiment

The slow cookoff (SCOT 400S apparatus, manufactured by OZM Research, Czech
Republic) test was carried out in accordance with the UN Recommendation for transporting
dangerous goods[14] manual. The only change is the reduction of dimension of the whole set.
The sample is put in the seamless steel tube with inner diameter 21 mm, outer diameter 26.4
mm and length 90 mm. Both sides of the tube were enclosed by steel caps. The whole set is
heated with rate 3.3 C/hr until the decomposition. The temperature in the center of the
sample, the temperature in the outer side of the bomb and the temperature of the air in the
oven is measured during the test.

RESULTS AND DISCUSSION

The average cookoff temperature from three experimental measurements is 136.9 C.


The simulated time temperature behavior in the center of the sample during cookoff is plotted
in figure 5 and the calculated cookoff temperature is 138.0 C (estimated as the onset of the
temperature growth from figure 5). The temperature distribution before the termination the
simulation is in figure 6. The increase of temperature in the center of the sample corresponds
to the initiation.
The agreement between measured and simulated cookoff temperature is quite good. The
small difference (1.1 C) is probably caused by using the zero order reaction model for the
decomposition of the sample.
The sequence of operations is schematically displayed in figure 7. This routing was
successfully used in case of SEMTEX 1A high explosive sample. The explosive material was
completely characterized with laboratory scale experiments and the cookoff temperature was
predicted. Authors believe that this routing can be an instruction for the modeling the heat
transfer during heating of other explosives.

CONCLUSION

The principle of numerical simulation of heat transfer during cookoff was introduced in
this paper. All the material properties of the sample were specified in laboratory scale
experiments. The heat transfer during cookoff was simulated using LS-DYNA finite element
code, and the calculated cookoff temperature is in good agreement with experimental results.
Acknowledgments

This work was created as a part of the project of Ministry of Education, Youth and Sports
of the Czech Republic No. MSM 0021627501.
Authors would like to thank to the OZM Research company for lending the instruments
JK-80 and SCOT 400, and also to Galina Sdovsk, PhD., from Department of Inorganic
Technology, University of Pardubice for the measurement of heat capacity.

336

The principal approach to the estimation of cookoff temperature using FEM

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REFERENCES
[1] M.R. BAER, M.E. KIPP, R.G. SCHMITT, AND M.L. HOBBS. Towards assessing the violence
of reaction during cookoff of confined energetic materials. In JANNAF Combustion
Subcommittee and Propulsion Systems Hazards Subcommittee Joint Meeting, Monterey,
California, USA, November 4-8 1996.
[2] M.L. HOBBS AND M.R. BAER. Multidimensional fully-coupled thermal/chemical/mechanical response of reactive materials. In JANNAF Propulsion Systems Hazards
Subcommittee Meeting, Huntsville, Alabama, USA, October 23-27 1995.
[3] M.R. BAER, M.L. HOBBS, R.J. GROSS, D.K. GARTLING, AND R.E. HOGAN. Modeling
thermal/chemical/mechanical response of reactive materials. In 26th International Annual
Conference ICT, Karlsruhe, Germany, 1995.
[4] M.R. BAER, M.L. HOBBS, R.J. GROSS, AND G.R. SCHMITT. Cookoff of energetic materials.
In 11th International Detonation Symposium, Snowmass, Coloredo, USA, August 31 September 4 1998.
[5] W.W. ERIKSON, R.G. SCHMITT, A.I. ATWOOD, AND P.D. CURRAN. Coupled thermalchemical-mechanical modeling of validation cookoff experiments. In JANNAF 37th
Combustion and 19th Propulsion Systems Hazards Subcommittee Meeting, Monterey,
California, USA, November 2000.
[6] M.L. HOBBS, M.R. BAER, AND R.J. GROSS. A constitutive mechanical model for energetic
materials. In 20th International Pyrotechnic Seminar, Colorado Springs, Colorado, USA,
1994.
[7] A.L. NICHOLS, R. COUCH, R.C. MCCALLEN, I. OTERO, AND R. SHARP. Modelling
thermally driven energetic response of high explosives. In 11th International Detonation
Symposium, Snowmass, Colorado, USA, August 31 - September 4 1998.
[8] M.A. MCCLELLAND, T.D. TRAN, B.J. CUNNINGHAM, R.K. WEESE, AND J.L.
MAIENSCHEIN. Cookoff response of PBXN-109: Material characterization and ALE3D
thermal predictions. In 50th Joint Army-Navy NASA Air Force (JANNAF) Propulsion
Meeting, Salt Lake City, Utah, USA, July 11-13 2001.
[9] A.L. NICHOLS, A. ANDERSON, R. NEELY, AND B. WALLIN. A model for high explosive
cookoff. In 12th Internatioanl Detonation Symposium, San Diego, California, USA, August
11-16 2002.
[10] M.A. MCCLELLAND, J.L. MAIENSTEIN, A.L. NICHOLS, J.F. WARDELL, A.I. ATWOOD, AND
P.O. CURRAN. ALE3D model predications and materials characterization for the cookoff
response of PBXN-109. In Joint Army Navy NASA Air Force 38th Combustions
Subcommittee, 26th Airbreathing Propulsion Subcommittee, 20th Propulsion Systems
Hazards Subcommittee and 2nd Modeling and Simulation Subcommittee Joint Meeting,
Destin, Florida, USA, April 8-12 2002.
[11] J.J. YOH AND M.A. MCCLELLAND. Simulating the thermal response of high explosives on
time scales of days to microseconds. In 13th APS SCCM Conference, Portland, Oregon,
USA, 2003.
[12] G. SCHOLTES, J. MAKKUS, AND N. VAN DER MEER. Measuring and modeling the cook-off
process. In 28th International Annual ICT Conference, Karlsruhe, Germany, 1997.
[13] S.-Y. HO. Modeling of cook-off reaction violence of confined energetic materials. In
V. Yang, T.B. Brill, and W.-Z. Ren, editors, Solid Propellant Chemistry, Combustion, and
Motor Interior Ballistics, pages 591606. Reston, Virginia, USA, 2000.
[14] Recomendations on the transportation of dangerous goods, manual of tests and criteria,
3rd edition, United Nations, New York and Geneva 1999.

Jakub eleovsk, Miloslav Krupka

337

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

LOW TEMPERATURE ACTIVATION ENERGY OF PROPELLANTS


Uldis Ticmanis, Stephan Wilker, Gabriele Pantel
WIWEB ASt Heimerzheim, Groes Cent, 53913 Swisttal (DE)
Abstract
This paper describes a new approach to determine the activation energy of propellant
decomposition reactions which quantifies the temperature dependence of a reaction rate
at temperatures below 50C by using an indirect microcalorimetric method.
Whereas the determination of life time and stability of propellants is usually performed
by artificial ageing at elevated temperatures with a temperature range usually between
70 and 130C only few and scattered data are available for temperatures below 50C.
Normally these activation energies for propellant decomposition reactions are not
known below temperatures of about 50C because (i) the reaction rates are very small
and thus ageing times have to be very long and (ii) because of slow reactions the
chemical change of the propellant becomes very small which in combination with too
short ageing times may lead to wrong estimations and calculations, because the
experimental error is relatively big in comparison with the real change of the
propellant.
To be on the safe side people usually assume a low value of activation energy. This
has the effect that measuring times at high temperatures are much longer than they
should be and that maybe good material is thrown away too early in fear of a thermal
instability.
To overcome this problem we have established a new approach by an indirect microcalorimetric method. This method evaluates the shift of typical points in the heat flow
curve to earlier times after artificial ageing at relatively low (30-70C) temperatures.
From the time shift of a specific point in the heat flow curve (compared to the unaged
material) and the temperature difference between ageing temperature and measuring
temperature activation energies of the whole temperature range between 89C and 30C
are available. We have verified the usability of this approach with six different
nitrocellulose based propellants.
Keywords: propellants, activation energy, microcalorimetry, stabilizer depletion

1.

INTRODUCTION

The determination of life time and stability of propellants is usually performed by


artificial ageing at elevated temperatures. The temperature range usually lies between 90 and
130C for classical stability tests whereas other tests like microcalorimetry, stabilizer
depletion and vacuum stability are usually performed at temperatures between 70 and 100C.
To calculate the real storage time that is covered by a high temperature experiment one must
assume an activation energy which quantifies the temperature dependence of a reaction rate.
Normally these activation energies for propellant decomposition reactions are not known
below temperatures of about 60C because the reaction rates are too small for a direct
measurement even in sensitive calorimeters or for achievement of measurable gas generation
in the vacuum stability test. Also stabilizer depletion is much too slow unless very long
measuring times are realized.

338

Low Temperature Activation Energy of Propellants

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

We have established a new approach to solve this problem by an indirect microcalorimetric method. This method evaluates the shift of typical points of the heat flow curve
(maxima, minima, points of inflection) to earlier times after artificial ageing at relatively low
(30-70C) temperatures in an oven. From the time shift of a specific point in the
microcalorimetric curve (e.g. point of inflection, maximum, minimum, thermal runaway)
compared to the unaged material and the temperature difference between ageing temperature
and measuring temperature activation energies of the whole temperature range between 89C
and 30C are available. We have verified the usability of this approach with six different
nitrocellulose based propellants.

2.

PROPELLANTS TESTED

Five types of gun propellants and one rocket propellant (propellant D) were investigated by this low temperature activation energy (LTAE) approach. They are listed (including
their main ingredients) in table 1. Propellants A, K, P and V were taken as received.
Propellant D was cut into small dices (2 - 4 mm length). Propellant N was milled and sieved.
Samples with particle sizes between 1 and 5 mm were taken for this study.
Table 1. Propellants used for the LTAE study
Propellant
abbreviation

Propellant
type

Stabilizer type

Ageing
temperatures [C]

NG content
[%]

Age [y] at
beginning

single base

DPA + EC

30, 35, 40, 50, 70

20

double base

2-NDPA

30, 35, 40

34

double base

DPA

30, 35, 40, 50, 70

19

triple base

EC

30, 35, 40, 50

21

double base

DPA + EC

30, 35, 40, 50, 70

10

porous

DPA

30, 35, 40, 50

18

Explanations: DPA = diphenylamine, EC = ethyl centralite, 2-NDPA = 2-nitro-DPA


No akardite stabilized propellant was taken because their decomposition follows nearly
exactly zero order, which means that no typical points (maxima, minima) could be chosen
for evaluation. Also no TPA stabilized propellant was available in large enough amounts
although the typical shape of their heat generation rate[3] would make them ideal candidates
for this study.

3.

EXPERIMENTS

Microcalorimetric measurements were conducted with a Thermal Activity Monitor


(TAM 2277 and TAM III, Thermometric AB, Sverige). The measurements were performed
in 3 ml glass ampoules at 89C.
Because it was important to repeat the shape of the microcalorimetric curve as precisely
as possible[1] the ampoules were filled with exactly ( 1 mg ) the same amount of propellant.
Because the shape of the curve also depends on the moisture level[2], the filling of the
2

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

339

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

ampoules was performed within one day for one propellant type. Samples for each ageing
step were weighed in triplicate. Two samples were taken for microcalorimetry, the third
sample was used for stabilizer depletion analysis.
Stabilizer depletion was measured by high performance liquid chromatography
(HPLC). The samples were prepared by dissolution in 20 ml of acetonitrile. The
nitrocellulose was precipitated by adding 5 ml of water. The solution was filtered. The
stabilizers were separated and detected with a suitable HPLC system containing a diode
array detector ( = 200-356 nm). For more details see[1].

4.

RESULTS
4.1

Classical iso evaluation from isothermal tests and its limit

In a very long measuring campaign, PB Clermont S.A. (Belgium) has conducted


isothermal microcalorimetry of double base spherical propellants[1]. The measurements were
performed at 89, 80, 70, 60, and 50C. The latter measurement is still running after now
2300 days of continuous measurement. Fig. 1 shows this important experiment. The lines
with slope -1 cut the heat flow curves at the same point for any temperature. This means that
the activation energy is nearly independent of the reaction degree and the temperature.
Overall the value for Ea is around 137 3 kJ/mole.
HFC OF FRESH SAMPLES TLP K6210-13 LOT 225 AT DIFFERENT TEMPERATURES

1.0E-03

P (W/g)

T = 89C
1.0E-04

T = 80C
1.0E-05

T = 70C
T = 60C
1.0E-06

T = 50C
1.0E-07
0.1

Fig 1.

10

100

1000

TIME (DAYS ) 10000

Microcalorimetry of propellant K at five different temperatures

Nevertheless, a continuous measurement at temperatures below 50C is hard to realize.


There are two reasons for that: (i) the heat generation rate is too low to be able to distinguish
it from base line noise and (ii) the time to reach a significant level of decomposition is much
too long to be realizable.

340

Low Temperature Activation Energy of Propellants

New Trends in Research of Energetic Materials

4.2

Czech Republic

2006

[Content]

Precursor tests at 70C

The propellants A, K and P were taken in a preliminary test which was performed to
demonstrate the suitability of the whole project using an ageing temperature of 70C. All
three propellants were subjected to the LTAE test. In addition the activation energy of all
three samples was estimated from continuous measurements at 89C and 70C and
subsequent iso evaluation. Table 2 shows the results of the calculations (for details see
chapter 4.4).
Table 2. Kinetic evaluation of precursor tests at 70C and comparison of activation
energy values with isothermal runs
Propellant

Ea from LTAE (89-70C)

Ea from isothermal runs (89-70C)

131 kJ/mole

135 kJ/mole

144 kJ/mole (from poi);


137 kJ/mole (from maximum)

137 kJ/mole

137 kJ/mole

141 kJ/mole

Explanation: Ea = activation energy, poi = point of inflection


The values of Ea do not differ significantly from each other. Therefore this preliminary
test was declared as being successful and the real LTAE was begun (see chapter 4.3).

4.3

LTAE studies at temperatures between 30 and 50C

After the successful preliminary test we started the ageing programme at 30, 35, 40 and
50C. As ageing times are extremely long, this programme began five years ago and is still
far away from being finished. As the stability of the propellants is different and the different
stabilizers which are responsible for the typical shape of the microcalorimetric curve we did
not set up an overall ageing plan but we decided to take out samples when we assumed that a
significant change would have happened. In the following sub-chapters the features of the
different propellants are presented in alphabetical order.
To be able to make a correct assumption of the heat generation rate of the unaged
material a five-fold determination of all samples were made. For evaluation the average
position of the individual points was taken. In addition, to control the ageing conditions all
ampoules were weighed after artificial ageing. If the weight loss was significantly higher
than that of the parallel samples we have excluded this sample from evaluation because it
was assumed that the seal has not been airtight all the time.

4.3.1.
Propellant A
This single base propellant caused some trouble due to instabilities in the past and was
therefore chosen to be one of the candidates for our LTAE test. Its microcalorimetric curve is
not very spectacular the only special point that can be evaluated is a point of inflection
during the increase into the second level of decomposition. To find the corresponding time
properly the first derivative of the heat generation rate was calculated and evaluated. Figure
2 shows typical derivated curves after ageing at 35C as an example.

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

341

New Trends in Research of Energetic Materials

Czech Republic

1269 d at 35C 1038 d at 35C 757 d at 35C

2006

[Content]

370 d at 35C

1.0
unaged
dP/dt [nW/(g*s)]]

0.8

0.6

0.4

0.2

0.0
0

t [days] at 89C

Fig 2.

Derivated heat generation rates of propellant A after ageing at 35C

4.3.2.
Propellant D
The only rocket propellant in this test also showed only one specific point of the
microcalorimetric curve its second maximum. It coincides with the complete decomposition of the stabilizer 2-NDPA and appears after ca. 10 days at 89C in unaged material.
Figure 3 shows typical results after ageing at 40C as an example.
150
610 d at 40C

140

762 d at 40C

130

1176 d at 40C

120
P [W/g]

351 d at 40C
unaged

1340 d at 40C

110
100
90
80
70
60
50
0

10

12

t [days] at 89C

Fig 3.

342

Heat generation rates of propellant D after ageing at 40C. Please note the
depletion of the first maximum which is almost zero after 1340 d at 40C

Low Temperature Activation Energy of Propellants

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

4.3.3.
Propellant K
This propellant is probably the best investigated propellant concerning stability[1]. As
fig. 1 shows it has a first minimum, a second maximum and, in-between them, a point of
inflection. The point of inflection and the second maximum were chosen for evaluation,
because the first minimum is very flat (and therefore difficult to find) and it appears close
after the first maximum which belongs to oxidation reactions of the propellant with the air
around the grains in the closed ampoule. If once all oxygen is used up the first maximum
disappears and thus also no first minimum is present anymore. The depletion of the first
maximum can be seen in all propellants except for propellant N (this propellant does not
have one). Figure 4 shows typical results after ageing at 40C as an example.
300

1091 d at 40C

838 d at 40C

250

P [W/g]

200
unaged
155 d at 40C
150

332 d at 40C
478 d at 40C
623 d at 40C

100

687 d at 40C
50
0
0.0

0.5

1.0

1.5

2.0

t [days] at 89C

Fig 4.

Heat generation rates of propellant K after ageing at 40C. The heat


generation rate of the 2nd maximum remains nearly constant.

4.3.4.
Propellant N
This propellant has a very interesting heat generation rate curve, from which three points
were selected for evaluation: A sharp minimum after 0.75 days (the chemical reason for it is
unknown but it appears in all propellants containing nitroguanidine), a first maximum and a
second minimum. Typical results after ageing are shown in figure 5.

4.3.5.
Propellant P
This propellant is chemically comparable to propellant K with the main difference that
the presence of EC changes the heat generation rate curve in such a way that no second
maximum is observed. So, like with propellant A only the point of inflection could be
evaluated. Similar to propellant A is also the reason for the choice of this propellant for the
LTAE study: There were problems with instability with this type of propellant before this
study started. Typical derivated heat flow curves are shown in figure 6.

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

343

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

160
140
P [W/g]

120
100
unaged
476 d at 30C
836 d at 30C
1133 d at 30C
1503 d at 30C

80
60
40
20
0
0

t [days] at 89C

Fig 5.

Heat generation rates of propellant N after ageing at 30C. Note that the
heat generation rate is reduced with longer ageing time.

10

unaged
1499 d

dP/dt [nW/(g*s)]

1248 d

1352 d

980 d

1085 d

830 d 695 d 584 d 332 d 151 d

6
4
2
0
-2
0.0

0.5

1.0

1.5

2.0

2.5

t [days] at 89C

Fig 6.

Derivated heat generation rates of propellant P after ageing at 40C. The


maximum decreases with ageing time.

4.3.6.
Propellant V
This propellant may be the most interesting one of the chosen candidates, as it shows a
thermal runaway at 89C after 10-11 days. The pattern of the first part is also very useful for
LTAE evaluation as a first minimum and a second maximum (and two points of inflection
in-between them) could be evaluated. The time to reach thermal runaway was also evaluated,
knowing that this may not be reached by natural ageing (according to [4] 10-11 days at 89C
are equivalent to 26 - 29 years at 25C). Nevertheless, a pre-ageing was suspected to make
this propellant less stable and so there was a likeliness that this thermal runaway could be
shifted to significantly earlier times. Figure 7 shows again as an example heat flow

344

Low Temperature Activation Energy of Propellants

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

curves after ageing at 40C, whereas figure 8 depicts derivated microcalorimetric curves
after ageing at 35C.

230
210
980 d

633 d

158 d

190
1143 d

P [W/g]

170

828 d

345 d

unaged

150
130
110
90
70
50
0.0

0.5

1.0

1.5

2.0

2.5

3.0

t [days] at 89C

Fig 7.

Heat generation rates of propellant V after ageing at 40C


1205 d 1027 d 828 d

645 d

376 d

unaged

dP/dt [nW/(g*s)].

0.5
0
-0.5
-1
-1.5
-2
0.0

0.5

1.0

1.5

2.0

2.5

t [days] at 89C

Fig 8.

Derivated heat generation rates of propellant V after ageing at 35C

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

345

New Trends in Research of Energetic Materials

4.4

Czech Republic

2006

[Content]

Evaluation of activation energies

All microcalorimetric values were transferred into EXCEL. To guarantee a perfect


comparison between all curves they were recorded with the same frequency (each 300 sec).
They were numerically derivated, averaged over 15 data points.
The calculation of the activation energy is performed by linear regression of all suitable
data points from
td = a * te .

(eq. 1)

td = time difference of a certain point in the heat low curve between the aged and the
unaged material [d]
te = ageing time at low temperatures [d].
The slope a is

a =e

1
Ea 1
*

R T d T e

(eq. 2)

Ea = activation energy [kJ/mol]


R = gas constant (0.00831447 J/(Kmol))
Td = measuring temperature (362,15 K)
Te = ageing temperature [K]
The activation energy is then calculated from eq. 3:
Ea =

R * ln a
1 1

Td Te

(eq. 3)

Also the activation energy within the temperature range of ageing (50-30C) was
calculated. This was performed by replacing the single slope (from the td versus te evaluation
in eq. 3 by the ration of two different slopes (from the evaluation of one te versus another)
and Td by the higher of both ageing temperatures.
One of the most interesting questions was whether a change in activation energy would
occur at temperatures around 50-60C as indicated by many authors[5] and as installed in
STANAG 4582[4].

4.5

Stabilizer depletion

Stabilizer depletion was conducted on most of the aged samples. The reason to do this
was to control whether the ageing characteristic which could be observed by microcalorimetry would also be detectable by changes in the stabilizer content. The second idea behind
this part of the work was to check the distribution of stabilizer depletion products which
could give us information about the ageing conditions (closed/open ampoules).
The stabilizer depletion was also evaluated by the kinetic model which was set up during
the development of AOP-48, Edition 2[6].
The following figures give two examples of the results of stabilizer depletion on various
propellants.

346

Low Temperature Activation Energy of Propellants

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

1,8
1,7
2-NDPA [%]

1,6
30C
35C
40C

1,5
1,4
1,3
1,2
1,1
1,0
0

500

1000

1500

2000

t [days]

Fig 9.

2-NDPA depletion of propellant D at 30, 35 and 40C

0,45
0,40

DPA [%]

0,35
0,30
0,25
0,20
0,15

30C

0,10

35C

0,05

40C

0,00
0

500

1000

1500

t [days]

Fig 10. DPA depletion of propellant K at 30, 35 and 40C

5.

DISCUSSION
5.1

Shift of characteristic points in the microcalorimetric curve

The shift of characteristic points in the microcalorimetric curve was presented in some
examples in figures 2 to 8. To give a better overview on all the results the most important
figures are presented in figures 11 to 17. They are sorted in alphabetical order.

5.1.1.
Propellant A
This propellant shows a very unusual behaviour. Within the first part of the ageing the shift
of the point of inflection is nearly proportional to the ageing time. If the ageing time
increases, also the shift increases. The main reason for this behaviour is a change in the type
of the point of inflection. If these points of inflection do not represent the same reaction
degree through all curves then the evaluation becomes impossible. Propellant A seems to be

10

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

347

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

one of these examples, although no sign of this behaviour could be detected at 70C. Fig. 11
gives one example of this non-linear behaviour.
180,000
160,000
140,000
t [s]

120,000
100,000
80,000
60,000
40,000
20,000
0
0

500

1000

1500

2000

t [days] at 35C

Fig 11. Non-linear shift of point of inflection of propellant A at 35C

5.1.2.
Propellant D
Although not enough sample material was available to run a complete programme with
all temperatures the evaluation clearly revealed a linear correlation between shift of the
maximum and ageing time although the series at 35C shows some scattering (see fig. 12).
40C
35C
30C
Linear (40C)
Linear (35C)
Linear (30C)

140,000
120,000
100,000

t [s]

80,000
60,000
40,000
20,000
0
0

400

800

1200

1600

ageing [days]

Fig 12. Shift of maximum heat generation rates in propellant D

348

Low Temperature Activation Energy of Propellants

11

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

5.1.3.
Propellant K
Both points show a nearly perfectly linear correlation between shift and ageing time for
all ageing temperatures. Figure 13 shows one example.
70,000
point of inflection

60,000

2nd maximum
Linear (point of inflection)

t [s]

50,000

Linear (2nd maximum)

40,000
30,000
20,000
10,000
0
0

400

800

1200

1600

t [days] at 35C

Fig 13. Shift of point of inflection and maximum heat generation rate (propellant K)

shift of minimum [days] .

5.1.4.
Propellant N
All selected points showed that they could be perfectly evaluated. Only the first sharp
minimum became smaller with ageing and disappeared after ca. 1000 days at 40C. The
correlation between shift and ageing time turned out to be almost linear. Also the ratio of the
slopes were nearly identical when 2 ageing temperatures are compared. The evaluation of the
2nd minimum is presented in figure 14.
0.9

40C

0.8
0.7
0.6
0.5

35C

0.4
0.3
0.2

30C

0.1
0.0
0

400

800

1200

1600

ageing time [days]

Fig 14. Shift of 2nd minimum with increasing ageing time at 40, 35 and 30C
(propellant N)
12

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

349

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

t [s]

5.1.5.
Propellant P
The relation between ageing time and shift of the point of inflection was almost linear. It
is presented in figure 15.
160,000
140,000
120,000
100,000
80,000
60,000
40,000
20,000
0

40C

50C

35C

30C
0

400

800

1200

1600

ageing time [days]

Fig 15. Shift of point of inflection as a function of ageing at 4 temperatures


(propellant P)

5.1.6.
Propellant V
The evaluation was divided into two parts: The first part covers the shift of the first
maxima, minima and two inflection points whereas the second part deals with the
autocatalytic decomposition at the end of the measurement.
The four different parameters of the first part of the microcalorimetric curve work pretty
well together. The shift of all four points (first minimum, first point of inflection, second
maximum and second point of inflection) is nearly linear with time, see fig. 16.

difference [sec] .

60000

1. Min

50000

1. POI
2. Max

40000

2. POI

30000
20000
10000
0
0

500

1000

1500

ageing time at 35C [d]

Fig 16. Shift of times for first minimum, first point of inflection, second maximum
and second point of inflection after ageing at 35C (propellant V)
350

Low Temperature Activation Energy of Propellants

13

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

This linear behaviour is not true for the thermal runaway. The shift in time of its
beginning (a value of 1600 W/g was taken for comparison) moves to later times after a preageing of more than 600 days. This is shown in figure 17 (later autocatalysis = negative time
shift in this figure). The reason for this behaviour is unknown. The samples aged at 35C and
30C show the same tendency although they have not yet reached the decomposition degree
which is equivalent to 600 days at 40C.
1.0

t autocatalysis [d] .

0.5
0.0
-0.5 0

400

800

1200

1600

-1.0
-1.5
-2.0
-2.5
-3.0
-3.5
ageing at 40C [days]

Fig 17. Shift of the runaway reaction after ageing (negative times = later autocatalysis), propellant V

5.2

Activation energies between 89C and 30C as well as within the


low temperature area

The activation energies were calculated between 89C and the ageing temperature. In a
second step the activation energies between 50C and 30C were calculated as described in
chapter 4.4.
Table 3 summarises the results of the calculation of activation energies between 89C
and the storage temperature. Table 4 deals with the activation energies within the low
temperatures. Propellant A was excluded from this calculation due to its non-constant
behaviour (see fig. 11). How well the slopes fit onto a straight line is shown in fig. 18.
These data, although they are somehow scattered show that a change in activation
energy from the standard value (130-140 kJ/mole) which exists for most of the propellants
between 70 and 90C takes place at temperatures probably around 50C. If the 50C values
are more elaborate we will take them into account to get a clearer picture of the temperature
range where this change is located. No activation energy value below 80 kJ/mole was found
which is good in a sense that the test time / temperature scheme in STANAG 4582 still is a
very conservative approach and does not need to be changed.
With a more exact knowledge of decomposition activation energies it will be easy to
precisely predict the behaviour of propellants at normal or slightly elevated temperatures.

14

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

351

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 3. Activation energies of propellants D, K, N, P, and V between 89C and


ageing temperatures. All values in kJ/mole.
Propellant

point

Ea 89-50

Ea 89-40

Ea 89-35

Ea 89-30

maximum

129.9 3.9

127.5 4.3

129.7 5.8

poi

136.1 0.8

134.7 1.5

132.8 1.8

131.3 1.3

nd

2 maximum

134.7 0.8

134.8 1.6

133.2 1.7

131.4 1.5

maximum

154.2 3.7

141.1 2.3

137.9 3.0

142.1 4.8

2 minimum

135.4 2.4

133.3 4.1

132.4 5.1

poi

129.7 2.8

132.2 1.6

128.8 0.8

124.7 0.8

st

137

141.7 3.5

137.1 2.0

134.9 1.7

st

129

135.8 2.8

132.3 1.6

130.5 2.9

nd

128

134.1 2.6

131.7 1.5

130.6 2.2

nd

129

135.3 3.9

131.8 2.3

131.1 2.3

135.5 3.6

132.6 3.2

131.9 4.4

nd

1 minimum
1 poi
2 maximum
2 poi

average

all points, all 134.7


propellants
poi = point of inflection

- = too few data points to make a good estimation

Table 4. Activation energies of propellants D, K, N, P, and V between 40C and


30C. All values in kJ/mole.
Propellant

point

Ea 50-30

Ea 40-30

Ea 40-35

Ea 35-30

maximum

120

124

116

poi

123

115

112

118

nd

2 maximum

125

116

118

114

maximum

125

139*

109

169*

2 minimum

117

99

134

poi

116

95

98

91

st

100

89

111

st

104

104

104

nd

106

103

109

nd

103

84

120

nd

1 minimum
1 poi
2 maximum
2 poi

all points, all 109 9


104 12
113 12
propellants
Note: Some data for 50C are not yet available (the ageing has just been started at 50C).
The data with * were excluded from averaging.

average

352

Low Temperature Activation Energy of Propellants

15

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

100000

y = 6.1843E+23e
2
R = 9.9916E-01

-1.5758E+01x

y = 7.8020E+23e
2
R = 9.9949E-01

-1.5830E+01x

a [s/d]

10000

1000
poi
2. max
100

Exponentiell (2. max)


Exponentiell (poi)

10
2.75

2.85

2.95

3.05

3.15

3.25

3.35

1000/T

Fig 18. slope a (see eq. 2) as a function of temperature in an Arrhenius plot. The
points nearly lie on a straight line (propellant K)
100000

a [s/d]

10000

1000

100

10
2.70

2.80

2.90

3.00

3.10

3.20

3.30

3.40

1000/T

Fig 19. slope a (see eq. 2) as a function of temperature in an Arrhenius plot. The
slope changes significantly at around 40C (propellant P)

5.3

Stabilizer depletion

There were no detailed studies of stabilizer depletion chemistry performed during this
study as the ageing conditions were exactly the same throughout the whole project. The only
interest was focused on the activation energy of stabilizer depletion. This is important as
many papers in literature refer to a change of activation energy below 50C which is
16

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

353

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

supported by stabilizer depletion analysis. We therefore took the chance to use these samples
to compare directly microcalorimetry with stabilizer depletion. The results of the
calculations are presented in table 5.

Table 5. Activation energies of propellants D, K, N, P, and V between 40C and


30C (stabilizer depletion). All values in kJ/mole.
Propellant

stabilizer Ea 40/35/30 AgeKin

Ea 40-35 zero order

Ea 35-30 zero order

DPA

128

132

2-NDPA

123

68

96

DPA

111

105

107

EC

DPA

111

103

73

DPA
AgeKin = calculated by the AgeKin programme of armasuisse[7]

zero order = evaluation by simple application of a zero order kinetics


-

= no useful value available (too few data points / too much scattering)

As table 5 shows the values have a much broader scattering than the ones delivered by
microcalorimetry but in principle (except for D propellant 0 order) they lie in the same range
than the ones calculated by LTAE data. Data handling also showed that stabilizer depletion
is less reliable than heat generation, because there were several examples with increasing
stabilizer content after further ageing which makes evaluation more difficult. Probably the
ageing does not only occur in the oven at 30-50C but also soon thereafter in the laboratory.
Microcalorimetry though seems to be much more precise than HPLC.

6.

CONCLUSION

An indirect method to measure activation energies of propellant decomposition reactions


at temperatures below 50C was presented in this paper. It is reliable and it works well if
propellants with a very typical heat generation rate shape are present and if a typical point in
a heat flow curve remains an iso point throughout the ageing.
The calculated data for the activation energy lie in the range of 100-125 kJ/mole (for 40 /
30C) which is significantly lower than the usual value of about 130-140 kJ/mole but much
higher than the (estimated) 80 kJ/mole which is the low temperature part of STANAG 4582.
The extrapolation of artificial ageing at temperatures like 70 or 89C to room temperature is
possible if the activation energies gained by this LTAE approach are used. Although the
propellants used for this study differed a lot they more or less show the same behaviour
concerning activation energies which shows that nitrocellulose itself is responsible for the
activation energy of decomposition of propellants.
It must be noted that all Ea values discussed in this paper refer to the early stages of
propellant decomposition ( from 0 to max. 0,05). It can not be assumed that these Ea values
are valid for the main decomposition reaction as well.

354

Low Temperature Activation Energy of Propellants

17

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

REFERENCES
[1]

[2]

[3]

[4]
[5]

[6]
[7]

18

PIERRE GUILLAUME, MAURICETTE RAT, GABRIELE PANTEL, STEPHAN WILKER, Heat flow
calorimetry of propellants Effects of sample preparation and measuring conditions, Propellants, Explosives, Pyrotechnics 26, 51-57, (2001). See also A.H. HEEMSKERK, Stability of
Nitrocellulose Propellants, Symp. Chem. Probl. Connected Stabil. Explos. 8, 137-148 (1988).
DAN S. ELLISON, ANTON CHIN, Understanding the true driving force in the auto-ignition of
single and double base propellants, Proc. Int. Pyrotech. Sem. 28, 245-261 (2001); ANTON
CHIN, DAN S. ELLISON, Mechanistic approach to study the moisture and acidity effect on the
stability of single and double base propellants, Proc. Int. Pyrotech. Sem. 28, 173-186 (2001).
STEPHAN WILKER, GERHARD HEEB, BEAT VOGELSANGER, JAN PETRLEK, JAN SKLDAL,
Triphenylamine a new stabilizer for nitrocellulose based propellants Part I : Chemical
stability studies, Propellants, Explosives, Pyrotechnics 32, 2006, in press.
STANAG 4582, Explosives, Nitrocellulose Based Propellants, Stability Test Procedures and
Requirements Using Heat Flow Calorimetry, 1st Edition, March 2004.
JIM DODDS, GAYNOR KAVANAGH, DAVID TOD, RICHARD WHITE, Bulk Ageing Effects in
double base propellants, Int. Annu. Conf. ICT 30, 81 (1999); Y.-J. YIM, B.-N. RYOO, Y.-C.
PARK, Stabilizer Consumption by Accelerated Aging of PEG/RDX Propellant, Int. Annu.
Conf. ICT 29, 45 (1998); ANDR SOPRANETTI, H.U. REICH, Comparison of high-temperature
stability tests with studies of stabilizers in single-base propellants, Symp.Chem.Probl.
Connected Stabil.Explos. 5, 459-477 (1979); JIM DODDS, Current UK Procedures for
Propellant Shelf-Life Determinations, Symp.Chem.Probl.Connected Stabil.Explos. 9, 247-258
(1992); THOMAS B. BRILL, P.E. GONGWER, Thermal Decomposition of Energetic Materials
69. Analysis of the kinetics of nitrocellulose at 50-500C, Propellants, Explosives,
Pyrotechnics 22, 38-44 (1997); MAX FREY, Mikrokalorimetrische Untersuchungen an Nitratestertreibstoffen, Symp.Chem. Probl.Connected Stabil.Explos. 8, 15-30 (1988); MANFRED A.
BOHN, Comparison of Surveillance Methods for Gun Propellants - Performances and
Limitations, Proc. Finnex Seminar (2002).
AOP-48, Ed. 2, Nitrocellulose based propellants - Stability test procedures and requirements
using stabilizer depletion, CNG Draft 8 (2005).
AgeKin Software, programmed by armasuisse for NATO-AC/326 SG I to support AOP-48,
Version 1.0 (2005).

Stephan Wilker, Uldis Ticmanis, Gabriele Pantel

355

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

THEORETICAL STUDY OF THE TETRAZINES C-N


HETEROCYCLIC DERIVATIVES
Zhou Yang1 Shu Yuan-Jie1* Wang Xin2** Long Xin-Ping1 Tian An-Min2
(1 Institute of Chemical Materials, China Academy of Engineering Physics, Mianyang 621900)
(2 Faculty of chemistry, Sichuan University, Chengdu 610064)
Abstract: 31 tetrazines C-N heterocyclic derivatives have been investigated by density
functional theory. Their optimized geometry structures, electronic structures,
conjugation, energies and heats of formation (HOF) were calculated and predicted at
the B3LYP/6-311G(d,p) level. The results show that most of the species keep a planar
structure and exit considerable conjugation over the whole molecule, which enhances
the stability of these derivatives. There is a good linear relationship between the energy,
HOF and N atoms in these species. Our study shows most of the species have high
HOFs and relative stabilities. The HOFs of these compounds are between 571.2KJ/mol
and 827.2KJ/mol, so they may be potential candidates for energetic materials.
Keywords: tetrazine, C-N heterocycle, theoretical study, heat of formation

* E-mail: syjfree@sohu.com
** E-mail: wangxin@scu.edu.cn
This work was supported by the Foundation of NASF (NO.10576030)

1.

INTRODUCTION

Energetic materials are used extensively both for civil and military applications. In order
to meet the requirements of future military and space applications, there are continuous
researches on how to develop new energetic materials with higher explosive performance
and enhanced insensitivity to thermal or shock insults than the existing ones. In the past,
high explosive performance has always been a prime requirement in the field of research and
development of explosives and common explosives hexahydro-1,3,5-trinitro-1,3,5-triazine
(RDX), octahydro-1,3,5,7-tetranitro-1,3,5,7-etrazocine (HMX) were considered adequate for
all weapon applications. But in modern ordnance, because many catastrophic explosions
results from unintentional initiation of munition by either impact or shock, aboard ships,
aircraft carriers, and munition trains. The explosives with the good thermal stability, impact
and shock insensitivity, and better performance are strongly required. However, these
demands are somewhat mutually exclusive. The explosives with good thermal stability and
impact insensitivity usually exhibit poor explosive performance and vice versa. Therefore,
the foremost objective at the stage of synthesis of new explosives consists of finding the
molecule with both good energy capability and optimal safety (reduced vulnerability, shock
and impact insensitivity) to those in current use[1].
Modern high-energy-density materials (HEDM) derive most of their energies either (i)
from oxidation of the carbon backbone, as with traditional energetic materials [2,3]or (ii) from
their very high positive heat of formation. Examples for the first class are traditional
explosives such as TNT, RDX and HMX [4]. Modern nitro compounds such as CL-20 or the
recently reported hepta- and octanitrocubanes[5] possess very high densities and have
enhanced energies due to substantial cage strain. And the second class of compounds is the
high-nitrogen energetic materials, which derive most of their energy from their very high
positive heats of formation[6]. In recent years, high-nitrogen energetic organic compounds
356

Theoretical Study of the Tetrazines C-N Heterocyclic Derivatives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

have attracted great attentions because of their novel properties, such as high density, high
positive heat of formation and high thermal stability. The large positive heats of formation of
nitrogen-rich compounds make them promising targets for the synthesis of highly energetic
materials, as illustrated by the recent examples of 3,3-azobis (6-amino-1,2,4,5-tetrazine) and
hydrazinium N,N-azobis tetrazolate[7]. Among the synthesized high-nitrogen C-N
compounds, the tetrazine derivatives are one kind of potential and luminous candidates for
high performance insensitive explosive. However, there are few theoretical studies on the
tetrazine derivatives.
To help find new stable nitrogen-rich compounds and obtain a more comprehensive
picture of the stability of such high-energy compounds, in this paper we studied the structure
and stability of a large number of substituted tetrazine compounds by density functional
theory (DFT).

2.

COMPUTATIONAL DETAILS

Theoretical calculations were performed to investigate 31 substituted tetrazine


compounds by density functional theory (DFT)[8] with hybrid B3LYP function[9,10]. The 6311G(d,p) basis set was used for all geometry optimizations. The self-consistent field (SCF)
convergence criterion is set to 10-8. The harmonic vibrational frequencies were calculated to
obtain the zero point energies (ZPE) and verify if it is a minimum on potential energy
surface (PES). The natural bond orbital (NBO) [11,12,13,14] analysis has been carried out using
the optimized geometries. The thermodynamics in a vacuum was computed by correcting the
differences in electronic energy to enthalpies at 298 K upon inclusion of zero-point energy
and thermal corrections. All these calculations were carried out using the Gaussian 98
program [15].

3.

RESULTS AND DISCUSSION


3.1

Single five and six-membered C-N heterocyclic compounds

3.1.1. Geometry analyses


Firstly, we optimized geometries of tetrazine and triazine with different base sets to find
a suitable basis set. And the results were compared with the experimental data, which were
illustrated in table 1. The calculated bond lengths with the 6-311G(d,p) base set is close to
experimental values, and the calculation is relatively inexpensive. Thus, we use 6-311G(d,p)
basis set for all the five and six-membered C-N heterocyclic compounds. The optimized
structures were displayed in Figure 1.
All the molecules have a rigid plane structures, and bond lengths of these single rings
have a averaged tend. In the five-membered C-N heterocyclic molecules the shortest C-C
bond length is 1.370, the longest is1.424. These bonds are between the standard C-C
(1.54) and C=C (1.34). And the C-N bond lengths are between 1.306 and 1.379. The
shortest is approximate to the standard value (1.30) of C=N and the longest is less than
normal C-N bond length (1.47). The N-N bond lengths are between 1.286 and 1.384 .
However in the six-membered C-N heterocyclic molecules, the C-N bond lengths (1.333
~1.336) have smaller change. The N-N bond length is 1.322 in the molecule T, and
1.333 in the molecule 2l. These values were between the standard value (1.460) of N-N
and the standard value (1.25) of N=N. And the C-C bond lengths (about 1.39) are near to
the standard C=C bond.

Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min

357

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

Table 1. The experimental data and theoretical data of 1,2,4,5-tetrazine and 1,3,5triazine
1,2,4,5-tetrazine 1,3,5-triazine
B3LYP

NN

CN

CN

6-311G(d,p)

1.322

1.336

1.334

6-311G++(d,p)

1.321

1.336

1.334

cc-pvdz

1.324

1.340

1.338

cc-pvtz

1.318

1.333

1.331

aug-cc-pvdz

1.323

1.340

1.338

aug-cc-pvtz

1.317

1.333

1.331

1.321

1.334

1.338

1.37
0

1 .3
96

2l

Fig 1.

114.1
N5

.9

11
5

2d

C6

2.3

N4

1.335

.3

1 09

115
.9

11
8

1
C2 .335
127.2
N3

12

3.7

.4

1.
33

N1

C4
1
C5 1.39

116.5

2j

N2

N3

C4

C3

.0
122
C5

4
0.
11
C5

134

1.394

N1

C6 126.4
C6

1.35

1.41

1.35
5

12

4.8

11

1 .3
11

113
.

10 7
.7

106.0

C2

N4

.4

11 N1
1.333

1.322

107

33
N2

N1

4.
4
N1

36

1.3

4
9.

N1

C2

C3
118.5
92
C4 1.3

C5

10

1.3

123.7 C6

C3

8 10
C5
1.373

6 N3 1.3
84
30

123

C6

34 C2
1.3
125.9

5
3.

1.36

10
5.5 06.2
1
N3 1.306 N4
42

1.381
11 C5
6 .9

8.

C4

1.5

C6

N1 1.336
117.2

1
C4

10

1.
C2

11

1.32

41

1
5.
7.
2 10
N3
C4
1.379

I2

1
7 N1 .32
5
32
1. 100.6
N2
C5

1.394

.7

8 .4
8.5
10
N1
C5
1.346

C4

.9
112

105

1.35

10

10

10

P1
N3 1.36
86
5
1.2 111.3
N2
N4

.1

1.379
P2

N2 1.3
4
99
1.2 107.0 9
N3
N1

1.36

C3

C3

C5

7
37 110.7

4.5

66
1.3

1.424

10
6

.0

C4

.4

7
10

1.37

8
1.35

C2

N1
12
1.3 105.4
C2

7 N1 1.32
34
104.2 9
C5

1.

1.
N2

2
11

74 N1
1.3 109.8
C5

111
.6

Experiment

[16,17]

N5

C3

6.8
11
N4

the theoretical data of the bond lengths and bond angles for C-N
heterocyclic compounds

3.1.2. Conjugation of the single rings


From the Figure 1it can be seen that the average trend of bond lengths for triazine is
the most obvious, of which all bond lengths are 1.334In addition, the stabilization
interaction energies E(2), are calculated by means of the second-order perturbation theory. In
the NBO analysis, E(2) is used to describe the delocalization trend of electrons from the
donor bond to the acceptor bond. Through the analyses, we find that there exist strong
donor-acceptor interactions within all the single rings. Interaction energies E(2) (38.23
kcal/mol) between bonding orbital and * antibonding orbital in the triazine ring is the
maximum within all the systems, and E(2) (11.38Kcal/mol) between BD N2-N3->BD* N4C5 in the 41 molecule is the minimum, where BD and BD* represent bonding and
antibonding natural bond orbital. As discussed above, we find that the triazine molecule
358

Theoretical Study of the Tetrazines C-N Heterocyclic Derivatives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

contains the best conjugation system, which means this structure is stable. For the tetrazine
interaction energies E(2) (24.51kcal/mol) between bonding orbital and * antibonding
orbital also shows that the tetrazine is a stable molecule, too. Through molecular orbital
analysis, we find a delocalized occupied orbital composed purely of 2PZ orbitals of all
carbon and nitrogen atoms of the ring in all derivatives.

3.2

Substitute of the single ring


3.2.1 Geometry analyses

11
8.6

1.3
37

121
.9

4
1.33

.8
121

6.2
11

12
2.2

1.3
36

0
1.4
0

7.8

46

1 .3

1.3

11
7.1

91

11
6.0

11

1.3
36

.2
122

1.3
35

7.2
11

35

9.8
11

1 .3

02

1.3
3

11
8.0

6
122
.

1 .4

1.3
45

117
.5

6 .3

11

1.3
34

.2
11
4

.8
125

34

40
1.3

.0
117

3
1.3
3

12
3.5

117
.

36
1.3

12
3.8

87
1.3

1.3
44

1.3

41
1.3

.0
11
8

11
7.8

C7 126.0
11
6.9

C11

N12

.0

N9
117

11
7.1

1.4
0

1.3
38

1.3

7
1.3
3

1.3
44

36

11
7.2

12

11
8.5

46
1.3

11
6.8

3. 9

1.3
98

117
.9

.1
117

7.2
11

117
.0
117
.4

36

39

1.333

126.2 C10

.0

1. 3

N12

34

114.2N10
5.8
12

N4

1.3

C9

C7 125.8
34
1.3

36

12
3.5

8.0
11

34
7

11
7.2

95
1.3

1.

5
1.3
3

8.9
11

11
7.8

8.1
11

11
7.5

1.3
46

34
1 .3

1.3
36

35
1. 3
1.3
35

95

1. 3

35
1.3

4 .2

1.495

1.333

11

125.8 C3

1.319

N8

40
1 .3

N2

7.0
11

1 .3

94
1.3

N11

117

TT

7.0

6 .9

N4

11

36
1.3

11

39

1.488

1.325

T3
N8

1.3

36
1 .3

Fig 2.

1.319

N5

N11

1.331

1.319

C6 126.1

119.3 N10

T2l-b
N2

C3

N5

33
1.3

C12

N1

C9

9.8
11

N4

7.0
11

1.319

C7 116.6
04
1 .4

1.316

C6

45

7.7
11

36

N1

1.3

7.0
11

1.3

N5

1.393

C8

1.477

1.3

N12

T2l-a

N2

124.5 C3

C6 125.9

123.5C10
9.5

N4

C9

11

1.315

6. 8

1.316

37

N5

11

C11

1. 3

1.331

1.380

C8

124.8 C3 1.487 C7 123.2


7.1
11

N12

C6 125.9

1.336

N2

T2j-c

N1

C12 1.332 N11

42
1.3

115.7N10

37
1.3

N4
1.317

1.317

N1
1.335

7.1
12

37

7.2
11

1.3

C7 121.8
1.491
1.344
1.337

C9

6.7
11

.7

1.396

124.8 C3

N5

127.1C10
6.2
11

1.390

C8

6.4
11

6
11

C6 125.9

N9

T2j-b

N2
1.342

2.2

N4

12

1.315

1.332

C7 116.2
00

N5

1.473

1 .4

C11

124.3 C3
7 .8
11

36

1.317

N11

1.334

C8

46
1 .3

7.1
11

1.3

1.334

N2

T2j-a

N1

96
1.3

C12

6.4
11

N4

1.315

N1

C6 125.8

116.6C10
2 .2
12

N12

91
1.3

6.0

35

N4

C9

C7 127.0
1 .3

7.2
11

35

1.319

1.334

N8

11

41
1.3

.0
117

1.3

N5

1.495

121.8C10

T2d

N2

125.5 C3

C6 126.1

C9

C7 121.5
3

1.316

1.487

0
1.4

3.9

8.
11

35
1.3

12

1.319

7.6
11

124.7 C3

1.329

N8

.6

.5

98
1.3

7 .9

N5

C11

1.391

N2

T1-c

N1

123.6C10

C12 1.333 N11

6
11

7.4

37

8
11

46

11

36

C12

N4

C6 125.9

117.1N10

C7 118.2

1.479

1.317

N1

11

C9

1.3

1.3

.1
117

1.3

1.316

1.391

C8

124.2 C3

C6 125.8

N5

N2

C9

T1-b

45
1 .3

1.316

C7 118.0
03

N4

T1-a

N1

1.474

1 .4

7. 9

1.315

.7
118

124.0 C3
11

36

N5

C11

47

.1

1.3

.6
123

1.331

117

C6 125.8

1.387

C8

N2

1.3

9
1.3

8.4

7.3

N12

1.316

N1

C9

118.3C10

C7 123.1
39
1 .3

11

36
1.3

N4

1.490

11

124.4 C3

1.317

1.390

C8
45
1.3

6 .8
11

C6 125.8

N5

N2

1.4
00

1.317

N1

N11

the molecular structures and parameters of the tetrazine substituted by sixmembered C-N heterocycle

Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min

359

4
1.3
5

51

1.3

1.355

5
1.3
3

49
1.3

1.381
1.375

1.
37
1

1.
38
7

1.3
3

1.396

1.3

38

7
1.3
3
1.358

1.3
71

1. 3

35
1.321

1.3
35

N9
C8

T134-1

1.3
36

C3

C11
1.462 1
.3
18

C6

42

1.377

N9

N2

1. 3

N5

1.316

N4

T41-2

N7

1.339

N8
1.295

1.316

1.3
51

4
38

1.3
50

1.3
45
35
1.3

N10 1.35

N9
5

1
1.3
6

1.469

43
1.3

96
N10

N4

T42-2

N9

1.333

37
1. 3
6
1.3
3

6
1.365

1.3
3

37
1.3

1.381

1.387
1.363

96
1 .3

1. 3
43

1.

40
1.3

1.373
1.308
1.315

0
1.3
7

46
1.3

1.2

N7

N4

38

C11

1.3
34

1.4
31

34

8
1.3
4

35
1.3

1.336

1.3
80

7
1.3
4

1 .3
46

1.388

5
1.3
3

1.424

1 .3
80

1
1.3
4
8
1.3
4

1.
39

1.435

53

1.3
40

96
1.3

1.3
36

5
1.3
3
38
1 .3
37
1. 3
39
1.3

34

1.390

42
1.3

1.3

1.316

1.3

36

36

1. 3

N9
N8 .284
1

C11

4
1.3

C3

1. 3

C3

3
1.3

1.317

T123-1

1.315
N1
N2

N5

63
1.3

N7
77
1 .3

37
1 .3

42

6
N5

1.399

N4

N1

71
1.3

1. 3

51

N10
92

C11
N7

1
C11 .363
C10

C3

1.315

1
N9 .345
C10

1.
N10 300

N2

C6

N8
C9 1.347

TI2-2

C6

N10

N8 1.275

T41-1

1.314

N4

C6

1
C11 .302

N7

1.402

1.317

3
1.3

1.3

T42-1

1. 2

64

N7

C8

1.454

1.316
N1
N2

N5

N8
N9
1
1.29

3
1.

N4

1.317

1.3
4

N7

36
1.3

36

Fig 3.

C11

N8

N9

N4

C8
1.355

1
C10

T123-3

C3

N1

C11

N2

3
1.3

1.3

1.412

1.458

C6

N5
N7 1.316

81
.3

36

1.316

N10
N9 1.369

N2

N4

1.318

N1

C11

T134-2

C3

1.317

N5

3
1.3

N4

C6

N5

C3

1
N7 .357

5
31
1.

1.318

C6
35

C3 1.458

C8

N8

1.314
N1
N2

1.3

60

45

N1

N7 336
1.

T123-2

1.3

39
1.3

1.314

N9

C11

N2

C6

N5

N9

1.

1.448

C101.356
82

1. 3

1.317

N7

82

35
1 .3

N5
N4
1.316

1
C11 .364

C10

N5
N4
1.316

48
1. 3

4
1.3

1. 3

C3

C3

TI2-4

1.313
N1
N2

N1

1.441

TI2-3

C6

C6

N5

1.3

N4

C3

34

C8
N7 .306
1

79

1.318

C6
1 .3

C11
1.3

34

N5

1.315
N1
N2

38
1 .3

51

1.3

1.457

N9

52

1. 3

1
C10 .367

C3

N7
C8 1.299

1.311
N1
N2

1.3

1.313
N1
N2

C6

C11

TI2-1

TP2-4

1.32
2

TP2-3

C10 1.360

N4

N4

22
1.3

1.334

1.313

88

N5

1.314

38
1.3

N7

N5

N9

1.386

1.3

C3

C11
1
42
1.
C10
C3
1.449

48
1.3

C6

N7
C8 1.361

50
1.3

TP2-2

N2

C6

N8
N7 .335
1

92
1 .3

N8

N5
N4
1.315

C11

1.316
N2
N1

37

63
1.3

3
1.3

1.448

N5
N4
1.316

1.3

C3

1.4
8C10 05
C11

1.462

N1

1.316

7
1.33

8
1.3
C9

C3

41
1.3

4
1.3

C6

C9
1.448 1
.3
87

N5
N4
1.312

1
C10 .374
19
C9
1 .4
36
1.3

N2

C6

C11
N7 1.317

TP2-1

1.316
N2
N1

N8

1.392

1.314
N1

C3

TP1-3

1.34

3
1.3

N5
N4
1.316

1.3
66
C10

37
1 .3

C3

C9

4
36
1.

42
1.3

C6

C6

TP1-2

1.315
N2

C11
N7 60
1.3

1
C10 .369
C11

N2

1.3

78

TP1-1

N1

1.439 C8

N5
N4
1.314

1.316

N1

C10

3
1.3

C3

C9 1.410

1.3

C11

N4

C6
35

C10
N5

N2

[Content]

51

N7

1.381

1.311

2006

1. 3

N1
C9

1.3

C3

C6

1
C8 .363

54
1.3

5
1.34

1.314
N1
N2

Czech Republic

1. 3

New Trends in Research of Energetic Materials

N8
N7 .318
1

the molecular structures and parameters of the tetrazine substituted by fivemembered C-N heterocycle

As some investigations shown, many organic molecules exhibit much different property
when its group substituted. As discussed above, the conjugation of traizne is the best, but its
nitrogen content is no more than that of tetrazine, and at present many known high-nitrogen
energetic materials take the tetrazine as the parent structure because of its better conjugation
and high-nitrogen content. So we discuss the structures and the electronic structures of these
derivatives coming from C-N heterocycle directly substitute a hydrogen atom of the
tetrazine, and we gained the thirty-one molecules aggregately. At first, we optimize the
geometry of all the molecules at the B3LYP/6-311G(d,p) level. The results see the Figure 2
and Figure 3. As shown in the two Figures, two rings of the molecules T2j-a, T2l-a, T3, TT
360

Theoretical Study of the Tetrazines C-N Heterocyclic Derivatives

New Trends in Research of Energetic Materials

Czech Republic

2006

[Content]

and T42-1 are not in a plane, their dihedral angels are 37.3,16.8,41.2 and 24.9, and the
other molecules have a plane structure. We fixed these non-plane molecules at a plane, then
optimize their structures, at last we find a imaginary frequency and higher molecular energy.
The results show that the non-plane structure of these molecules is the most stable.
From the two Figures, we can know that after a hydrogen atom of the tetrazine is
substituted, the NN bond lengths become shorter, the CN bond lengths adjacent to the
substituent become longer and the CN bond lengths far from them have no change basically.
From the Figure 2, we can find that for the same substituent the different place connected
with the tetrazine has a different influence on the C3-C7 bond length between the two rings,
such as these molecules T2j-a,T3-a and TT, their C3-C7 bond lengths are 1.495,1.495
and 1.488, which are more than that of T1-a,T2d,T2j-c and T2l-a, because a hydrogen
atom of the subsituent is closer to one nitrogen atom of the tetrazine in the latter; the other
molecules have two hydrogen atoms closer to the nitrogen atoms of the tetrazine, their C3C7 bond lengths are the shortest correspondingly. This trend is more obvious in the isomers
T2j-a,b,c. The distance between the hydrogen and the nitrogen of tetrazine ring is between
2.447~2.531, there maybe exist the weak interaction, but there are no the bond critical
point between the two atoms from the results of AIM analyses, namely there dont form the
hydrogen bonding. As the Figure 3 shown, the C-N bonds of the subsituent connected the
tetrazine in the molecules TP1-1,TP2-1,TI2-1,T123-1,T134-1, T41-1and T42-1 are close to
1.40, the shortest is 1.381 and the longest is 1.412; the C-C bonds of the other
molecules are more than 1.40 (between 1.439~1.469). In a word, these bonds are
between the standard single bond and double bond.
3.2.2 Electronic structures and conjugation
The net charge of all the atoms is calculated by the NBO method. A hydrogen atom of
the tetrazine ring is taken as a substituent and the C-N heterocycle is taken as a whole, the
total charge of all the substituents are calculated. Table 2 illustrates the NBO charge of seven
atoms on the tetrazine rings and the total charge of the tetrazine ring and the subsitituent.
Table 2. The atoms NBO charge distribution of the tetrazine ring and the total
charge of two rings/(a.u.)

T
T1-a
T1-b
T1-c
T2d
T2j-a
T2j-b
T2j-c
T2l-a
T2l-b
T3
TT
TP1-1
TP1-2
TP1-3
TP2-1
TP2-2
TP2-3
TP2-4
TI2-1

N1

N2

C3

N4

N5

C6

-0.205
-0.200
-0.193
-0.192
-0.195
-0.202
-0.189
-0.196
-0.198
-0.189
-0.197
-0.192
-0.188
-0.196
-0.201
-0.188
-0.202
-0.197
-0.193
-0.184

-0.205
-0.229
-0.226
-0.217
-0.181
-0.187
-0.220
-0.221
-0.228
-0.216
-0.182
-0.182
-0.243
-0.215
-0.228
-0.245
-0.223
-0.222
-0.206
-0.238

0.215
0.355
0.372
0.370
0.348
0.342
0.367
0.347
0.352
0.364
0.332
0.329
0.559
0.341
0.371
0.546
0.353
0.368
0.345
0.555

-0.205
-0.180
-0.220
-0.217
-0.223
-0.187
-0.220
-0.174
-0.174
-0.212
-0.182
-0.182
-0.243
-0.258
-0.230
-0.199
-0.192
-0.227
-0.240
-0.235

-0.205
-0.198
-0.192
-0.192
-0.196
-0.202
-0.189
-0.194
-0.193
-0.189
-0.197
-0.192
-0.188
-0.196
-0.202
-0.187
-0.200
-0.198
-0.193
-0.183

0.215
0.206
0.202
0.207
0.210
0.213
0.206
0.214
0.211
0.211
0.219
0.221
0.182
0.183
0.186
0.189
0.199
0.192
0.197
0.189

0.194
0.192
0.194
0.195
0.194
0.192
0.196
0.195
0.194
0.197
0.195
0.197
0.195
0.190
0.189
0.196
0.191<