Professional Documents
Culture Documents
Sponsors
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
Czech Republic
2006
[Content]
Contents
Part 1
Invited lectures
1
Lectures
10
11
Czech Republic
2006
[Content]
12
13
14
15
16
17
18
19
20
21
22
23
24
Czech Republic
2006
[Content]
25
26
27
28
29
30
31
32
. . . . . . . . . . . 356
. . . . . . . . . . . . . . . . . . 368
Only in Proceedings
33
34
35
36
37
Czech Republic
2006
[Content]
Part 2
Posters
38
39
40
41
42
The Use of a Kenics Static Mixer for the Reaction of TNT with NaOCl
Anthony Bellamy, Victoria Lomax, Kate Read . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
43
44
45
46
47
48
49
50
51
Czech Republic
2006
[Content]
52
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
55
56
57
608
58
59
60
61
Salts OF 5-Nitrotetrazolate
Thomas Klapoetke, Peter Mayer, Kurt Polborn, Jan Wlech . . . . . . . . . . . . . . . . . . . 631
62
63
64
65
Czech Republic
2006
[Content]
66
67
68
69
70
71
72
73
74
The Use of Low Pressure Closed Vessel and Rocket Motor for Measurements
of Burning Rate of Rocket Solid Propellants.
Petr Stojan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 730
75
76
77
78
79
Czech Republic
2006
[Content]
80
81
82
83
84
Abstract
85
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.
Svatopluk
eman
SvatoplukZZeman
Invited lectures
Czech Republic
2006
[Content]
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
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
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,
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
Czech Republic
2006
[Content]
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
14
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
Czech Republic
2006
[Content]
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.
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.
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
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.
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
Impulse density
= mass x velocity = mv
(kgm/s)
= force x time
(kps)
= Ft
Manfred Held
17
3.
Czech Republic
2006
[Content]
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.
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
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.
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).
Manfred Held
19
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.
M radial
-1
-2
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.
20
M radial
-1
-2
kg_m_s cm _kg
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
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.
Manfred Held
21
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.
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
Czech Republic
2006
[Content]
Table 1. Measured initial sled velocities by the blast and fragment loads
Casing
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
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
Manfred Held
23
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
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.
24
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
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
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.
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)
10
Manfred Held
25
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.
IDS,bare[6] = 370(R/W1/3)-1
confined charge:
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
11
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:
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.
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
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
13
10.
Czech Republic
2006
[Content]
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
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.
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
15
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
32
Czech Republic
2006
[Content]
17
18
Manfred Held
Czech Republic
2006
[Content]
33
34
Czech Republic
2006
[Content]
19
20
Manfred Held
Czech Republic
2006
[Content]
35
36
Czech Republic
2006
[Content]
21
22
Manfred Held
Czech Republic
2006
[Content]
37
38
Czech Republic
2006
[Content]
23
24
Manfred Held
Czech Republic
2006
[Content]
39
40
Czech Republic
2006
[Content]
25
26
Manfred Held
Czech Republic
2006
[Content]
41
Czech Republic
2006
[Content]
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
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
43
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.
44
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.
45
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
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
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
47
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
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]
49
Czech Republic
[Content]
N
H
H
N
N
N
2006
N
C
N
H
H2O
Figure 2 DSC curves for BTAw at 5 C min-1 (____); 2 C min-1 (__ __ __); 1 C min-1 ();
..
0.5 min-1 (
).
50
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
51
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.
52
11
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
53
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
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
13
Czech Republic
2006
[Content]
14
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 %
1000
m/e
17
18
28
29
30
42
43
44
100
10
1
50
100
150
200
250
300
350
Temperature/C
56
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 %
-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 14 Example of ARC results for BTAw: pre-dried at 50 C, and run in a closed system of
ambient Ar.
16
57
Czech Republic
2006
[Content]
-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
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
2006
[Content]
59
60
air
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)
100 2
101 2
77.2 0.7
76.2 0.7
87.2 0.7
75 1
84 5
78.1 0.6
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)
19
Czech Republic
2006
[Content]
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.
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.
61
2.
Czech Republic
2006
[Content]
3.
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:
-
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
Czech Republic
2006
[Content]
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
63
Czech Republic
2006
[Content]
4.
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.
broader use of micro and nano technologies to decrease mass, size and cost
of instruments
64
Czech Republic
2006
[Content]
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.
(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 ) =
R
A
ZA
(r)dr
r r
A r
(3)
65
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)
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
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
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
67
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]
Czech Republic
2006
[Content]
---
---
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.
69
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
3.
70
Czech Republic
2006
[Content]
(7)
(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
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
71
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
72
[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]
73
Czech Republic
2006
[Content]
74
Lectures
Czech Republic
2006
[Content]
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.
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
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.
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)
2416
2630
2320
2790
Comp. B
2310
2700
TNT (cast)
2040
2370
Tetryl
2209
2500
Joseph Backofen
77
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)
(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
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
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
1.2
Czech Republic
2006
[Content]
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
Comp.B in Aluminum
0.100
1.000
1.000
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
Czech Republic
2006
[Content]
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
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
Czech Republic
2006
[Content]
0.45
Comp B in Al 2S Cyl
Comp B in Al Sphere
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
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
Czech Republic
2006
[Content]
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.
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
Zinc
0.6
Magnesium
Brass
0.5
Dural
Copper
Dural
0.4
0.2
0.001
0.01
0.1
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
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
Aluminum
Uranium, Lead
Plexiglas, Antimony
Copper
0.10
Iron
0.05
0.00
0.001
0.01
0.1
10
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
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
Czech Republic
2006
[Content]
20.5
23
14.7
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
0.6
0.7
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
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).
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
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
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
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.)
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
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
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.
0.14
0.12
0.08
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.
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
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:
Thoroughly measure the gas-push 2nd stage propulsion out to large expansion ratios,
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
88
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
Czech Republic
2006
[Content]
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.
1.
INTRODUCTION
H2 N
H2 N
Czech Republic
N NO2
N
NH
2.
[Content]
NH NO2
2006
METHODS OF INVESTIGATIONS
3.
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.
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
91
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
H 2N
H 2N
NO2
3
H2N
H2N
Fig. 1.
92
NO2
H2 N
NO2
NH2
H2 N
NO
H2 N
Czech Republic
2006
[Content]
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
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
93
N
H2
Czech Republic
2006
[Content]
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
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
95
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
96
Czech Republic
2006
[Content]
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.
97
Czech Republic
2006
[Content]
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
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.
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
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.
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
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.
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
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
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
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].
Yurii Burov
103
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(
P (Vs ) = exp(
Vs
vf
(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
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.
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]
Yurii Burov
105
[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]
Czech Republic
2006
[Content]
107
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
Czech Republic
2006
[Content]
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
109
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
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
111
Czech Republic
2006
[Content]
112
Czech Republic
2006
[Content]
Velocity
ms
event
-1
mass %
Recovered
13 mm flat spigot
100
Yes def
42
13 mm flat spigot
65
No
92
53
Yes def.
48
As above
45
Yes def
45
As above
36
No smell
96
86
No sl. evi
95
103
yes
52
4 mm round nose
115
No smell
97
Flat nosed
115
yes
30
113
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
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
115
Czech Republic
2006
[Content]
116
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
11
117
Czech Republic
2006
[Content]
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.
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
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.
119
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
Czech Republic
2006
[Content]
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
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.
121
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
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
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
Czech Republic
2006
[Content]
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
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.
125
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]
126
Czech Republic
2006
[Content]
1.
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
127
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).
TABLE 1.
2006
[Content]
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
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
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.
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)
4.
Czech Republic
2006
[Content]
NMR SPECTROSCOPY
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].
131
Czech Republic
2006
[Content]
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.
132
6.
Czech Republic
2006
[Content]
CONCLUSION
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.
133
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
134
Czech Republic
2006
[Content]
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.
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
135
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
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
136
Al-Al2O3
Al-AlB2
Ni
Fe
W
Fig 1.
Czech Republic
Cu-6 % Ni
2006
[Content]
Al-NC
Cu
Cu-45 % Ni
137
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
70
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
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
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.
SFAP-Al2O3 + AP
16
11
14
23
19
35
19
22
28
+m (to 600oC), %
Czech Republic
2006
f)
e)
-5
15
25
DTA, mV
Czech Republic
2006
[Content]
on, o
T, oC
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]
141
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 . . .
Czech Republic
2006
[Content]
INTRODUCTION
1.
2.
DISCUSSION
2.1
We have carried out the synthesis of new energetic d-metal complex perchlorates with
1,5-pentamethylenetetrazole (PMT) as ligand.
143
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
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
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 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.
145
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
20
10
8
6
1
0.1
0.2
0.4
0.6 0.8 1
8 10
20
Pressure, MPa
Fig 2.
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
Czech Republic
2006
[Content]
Burn temperature, K
1451
2329
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
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
2.2
147
Czech Republic
2006
[Content]
-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
148
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.
149
Czech Republic
2006
[Content]
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
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.
151
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
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
1.20
4.83
1.81
7.25
2.28
9.66
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
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
153
Czech Republic
20
2006
[Content]
100
dP/dt (Mpa/ms)
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
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
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)
2006
[Content]
flattened
propellant
12.5
12
11.5
11
10.5
10
9.5
9
8.5
8
spherical
propellant
10
20
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.
155
Czech Republic
2006
[Content]
350
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
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)
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.
157
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.
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
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
159
Czech Republic
2006
[Content]
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.
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
2.
Czech Republic
2006
[Content]
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%.
161
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
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.
NH
H2N
NO2+, -H+
NH2
-NO2+, H+
H2N
NO2
NO2+, -H+
NH2
-NO2+, H+
H+
O2N
N
H
NO2
NH2
H+
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
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.
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
163
Czech Republic
2006
[Content]
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
2.3
0,5
Wg^-1
-80
-60
-40
-20
20 C
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
Czech Republic
Fig 5.
Fig 6.
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.
165
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
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.
167
3.
Czech Republic
2006
[Content]
CONCLUSIONS
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.
Czech Republic
2006
[Content]
169
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.
Czech Republic
2006
[Content]
171
Czech Republic
2006
[Content]
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
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.
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
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
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
174
Czech Republic
2006
[Content]
dIsp
60
NH4NO3
ADNA
50
Be
40
30
N2H5N3O4
Al
20
10
0
Mg
-1000
-500
500
1000
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.
175
Czech Republic
2006
[Content]
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
176
Czech Republic
2006
[Content]
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.
177
Czech Republic
2006
[Content]
Isp
Tc, K Ta,K
HCl+Cl2,
mol/kg
CO,
mol/kg
Cond. phase,
mass.%
38
PA+CH*+Al
AP+CH
7.8
0.8
HMX+ Act**+Al
12
38
ADNA+ CH+Al
38
ADNA+ CH
3.6
N2H5NO3+ CH+Al
254
7.7
38
N2H5NO3+ CH
930
NH4NO3+ CH+Al
7.3
38
NH4NO3+ CH
920
11
38
178
1.6
2910 1770
12
N2H5C(NO2)3+ CH
254
13
NaClO4+ CH +Al
14
NaClO4+ CH
1.7
Czech Republic
[Content]
1.1
10
66
211
3080 2030
0.25
44
15 AP+NaNO3+ CH +Al
0.9
4.7
37.7
16 AP:Na2N2O3+CH+Al
0.98
5.3
37.7
2.0
3070 1600
2006
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
Combustion rate
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
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.
Czech Republic
2006
[Content]
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:
Alexander Lukin
181
1.
Czech Republic
2006
[Content]
INTRODUCTION
2.
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
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
Czech Republic
2006
[Content]
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
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
Czech Republic
2006
[Content]
3.
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
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
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.
h a / uC
(0.010 0.070) mm
188
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
0 =
CuC
( T R T0 )
(2)
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 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
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
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.
Alexander Lukin
191
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.
Czech Republic
2006
[Content]
torch structures on the burning surface can exist steadily in the event that they form a pair of
two torches.
SELF-ORGANIZING
OF THE SPMS
FORMATION OF THE CARBON
MICRO-STRUCTURES
7.
Alexander Lukin
193
Czech Republic
2006
[Content]
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
Czech Republic
2006
[Content]
8.
15
Alexander Lukin
195
Czech Republic
2006
[Content]
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
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
17
Alexander Lukin
197
Czech Republic
2006
[Content]
Nomenclature
a
C
Ea
FL
FST
FTE
g
h
k0
P
q0
198
Czech Republic
2006
[Content]
Q0
R
t
T
T0
TC
Th
TR
TS
uC
u (Ci , j)
Greek Symbols
0
C
Condensed phase;
Gas phase;
Surface;
Parameters of equilibrium state;
REFERENCES
[1]
[2]
[3]
19
Alexander Lukin
199
Czech Republic
2006
[Content]
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]
[9]
[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
Czech Republic
2006
[Content]
21
Alexander Lukin
201
Czech Republic
2006
[Content]
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.
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
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-
Scheme 1. Reaction equations for the synthesis of the methylated 5-aminotetrazole salts
203
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.
204
2.1.3
Czech Republic
2006
[Content]
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
2.1.5
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 %).
205
2.1.7
Czech Republic
2006
[Content]
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
206
2.1.10
Czech Republic
2006
[Content]
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):
207
2.2.1
Czech Republic
2006
[Content]
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
2.2.2
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
208
2.2.3
Czech Republic
2006
[Content]
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
2.2.4
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
209
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
3.
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
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
(ii)
(iii) An insight into their sensitivity was gained by submitting them to standard
BAM tests.
10
211
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
11
Czech Republic
2006
[Content]
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
213
Czech Republic
2006
[Content]
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
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].
215
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)
t 30 0 C = t 90 0C a10
[(90 30 ) / 10 ]
(2)
216
3.
Czech Republic
2006
[Content]
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.
10000
8000
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
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.
217
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.
Denotation
Unit
E (-115C) MPa
MPa
E(25C)
1.
2.
3.
4.
5.
6.
T(EON1)
T(EON2)
E1
T(E1)
C
C
MPa
C
No.
Denotation
Unit
E(-115C)
E(25C)
Wpeak
hpeak
Tg
E1
T(EON1)
T(EON2)
MPa
MPa
C
MPa
C
MPa
C
C
No.
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
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
7000
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
20
40
60
80
100
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
219
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.
500
400
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
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
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
97.13
181.38
30.59
72.68
58.39
91.35
45.64
36.02
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
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).
221
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
[Content]
180
80 0
60
400
2006
500
450
500
Tg
E1
E (-115C)
10
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).
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
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
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
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
223
Czech Republic
2006
[Content]
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
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
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
Fig. 10. Non-isothermal DSC curves of DBR propellants aged for different period of time at 90
C
224
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.
Denotation
Unit
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
Tie
Tp
Tfe
Hr
hp
wp
(d/dt)max
(Tp)
(d/dt)max
max
C/min
No.
Denotation
Unit
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
(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
J/g
W/g
o
%
%
W/(g min)
min
%
min
min
min
%
%
%/min
%/min
225
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
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
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
30.00
2200
1.600
5.50
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
80
3.00
100
20
40
60
80
1800
34.00
1700
1600
100
Fig. 12. Changes of some characteristic points/parameters on DSC curve of DBR propellant
with the ageing time at 90 oC
226
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.
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
227
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**
0.0
4.0
20.0
49.9
89.9
143.8
163.8
t(10 %),
min
t(15 %),
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
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
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
30.00
m (200 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
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
Fig. 15. Changes of some characteristic points/parameters on TGA curve of DBR propellant
with the ageing time at 90 oC
228
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]
[2]
[3]
[4]
[5]
16
229
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
230
Czech Republic
2006
[Content]
Czech Republic
2006
[Content]
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.
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
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
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.
Henric stmark
233
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
234
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
Henric stmark
235
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)
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
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
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
Density (g/cm3)
189
(kcal/mol)
194
Sublimation energy
N4(D2h)
Hf(g) (kcal/mol) a
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
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
Czech Republic
2006
[Content]
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
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
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
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.
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
N
N
Zn 2+
N
N
N
N
N
N
Henric stmark
243
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.
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
Czech Republic
2006
[Content]
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
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
Czech Republic
2006
[Content]
-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)
1.65
Energetic material
Impulse (s)
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
Henric stmark
247
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
[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]
249
[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
Czech Republic
2006
[Content]
AEROBIC BIODEGRADATION OF
MONONITROTOLUENES IN DIFFERENT REACTOR
CONFIGURATION
Jan Paca*1, Martin Halecky1, Rakesh Bajpai**
*
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
251
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.
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.
2.5. 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 =
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 =
253
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 =
3
2
1
0
t (days)
254
-1
S (mg.L )
-1
S (mg.L )
t (days)
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
-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)
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
256
2006
[Content]
Czech Republic
2006
[Content]
257
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
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.
259
Czech Republic
2006
[Content]
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
Czech Republic
2006
[Content]
..
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].
261
Czech Republic
2006
[Content]
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' ..
N
R
OR'
O-
ROR' + N2O
NO2
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
Czech Republic
2006
[Content]
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
+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
N2
H2CO2 CH2O
+NO
CHO
-NO
OH + N2O
+M
H2 O
CO2
Here and at other schemes of decomposition processes the known experimentally reactions and products are
marked by the bold type.
263
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
+M
H2O
+M
+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
[Content]
CH2 N NO 2 CH3Cl
+M
CH3Cl
CH3
.
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
CH3
O2N N
Cl
CH2
CH3Cl + CH2
N NO2
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
_
O
N
O CH CH2
N OH
CH2 CH
+M
+M
+NO
CH
N O
.
OH + N2O O C N N OH CH2
CH2
+M
CO2
N ONO
CH3
+HONO
+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
+NO2 -NO
CO2
+R
CH4
CO
+ONO
CO2 +HNO
+NO2 -NO
CO2
265
Czech Republic
2006
[Content]
N NO2
H3C
H3C
Cl
. ONO
+M
+R
H2C
Cl
Cl
HCl
+M
+R
HCN
H2O
HNO
O C N N OH N2O + OH
+R
+R
N2
+ONO
OH
+M
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
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
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.
267
Czech Republic
2006
[Content]
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.
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)
Czech Republic
2006
[Content]
3.
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
269
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
270
Rating
tII/1,6
1
2
5
8
+2
Czech Republic
2006
[Content]
Rating
tII/1.4
2
3
5
4.
Rating tZ,I/1
0.5
3
2
1
0,5
+0,5
+1
+1
Every selected unit may be classified to one of the five process hazard classes KZP as shown in
Table 6 below.
271
Czech Republic
2006
[Content]
5.
Hazard level
KZP1
KZP2:
2/III
2/II
2/I
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
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
6.
Czech Republic
2006
[Content]
273
7.
Czech Republic
2006
[Content]
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
274
Czech Republic
2006
[Content]
7.2
7.3
275
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)
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
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
277
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
Czech Republic
2006
[Content]
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.
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
279
Czech Republic
2006
[Content]
2.
EXPERIMENTAL
2.1
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
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
280
2.2
Czech Republic
2006
[Content]
4
5
3
6
2
EXAMINER
1
Fig 1.
3.
EXPERIMENTAL RESULTS
281
Czech Republic
2006
[Content]
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
226
2,5
274,5
2
-10
-20
283,7
249,0oC (19,73%)
1,5
egzo
-30
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.
DISCUSSION OF RESULTS
282
Czech Republic
2006
[Content]
-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.
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
283
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
Czech Republic
2006
[Content]
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.
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
(1)
285
Czech Republic
2006
[Content]
(2)
(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.
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
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)
287
Czech Republic
2006
[Content]
4.
288
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
289
Czech Republic
2006
[Content]
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
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
Czech Republic
2006
[Content]
291
Czech Republic
2006
[Content]
292
Czech Republic
2006
[Content]
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.
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
293
Czech Republic
2006
[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
294
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
2.4
2.5
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
295
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 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.
60
W, %
(N2+CO2)/ N2
30
0
0
50 nm
Fig 2.
296
40
Czech Republic
2006
[Content]
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.
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
XRD, volume %
Al Al2O3 AlN Al N5O14
96
4
90
5
~5
91
9
297
Czech Republic
2006
[Content]
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
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
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
5.
Czech Republic
2006
[Content]
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
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
of
299
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.
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, %
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
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.
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]
301
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
CONCLUSION
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
303
Czech Republic
2006
[Content]
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)
304
Czech Republic
2006
[Content]
X X 0
(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.
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.
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.
305
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
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, /
2.
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]
307
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)
- the average of
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) ,
U 12
the value as obtained thereby. Table 3 includes the values obtained for 4DE
and the
ff vv
calculated values of
vv .
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
308
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.
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.
309
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.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Czech Republic
2006
[Content]
311
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
312
R, A
0.5
1.5
2
t, ps
2.5
4
t, ps
3.5
4.5
0.5
1.5
2
t, ps
2.5
3.5
4.5
0
t, ps
0.05
0.1
0.2
0.15
0.25
2006
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
Czech Republic
0.8
1.2
1.4
1.2
1.3
1.4
1.5
1.6
1.7
1.8
Ek, eV
Ek, eV
10
R, A
313
Czech Republic
2006
[Content]
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
314
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.
315
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.
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
Czech Republic
2006
[Content]
100
5
HMX
3
2
10
5
TNT
3
2
1
NTO
5
3
2
2
6 8 1
6 8 10
Pressure, MPa
Fig 1.
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.
317
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.
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.
318
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.
319
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.
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
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
321
Czech Republic
2006
[Content]
5
4
0.00160
0.00200
0.00240
0.00280
1/Ts, K-1
Fig 6.
3.2
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
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
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
323
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
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
12
325
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]
326
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
327
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
Czech Republic
2006
[Content]
Keywords:
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.
329
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)
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
3.1
Czech Republic
2006
[Content]
3.2
331
Czech Republic
2006
[Content]
3.3
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
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
Czech Republic
2006
[Content]
Fig 3. The comparison of simulated and experimental DTA data (heating rate 2 C/min).
333
Czech Republic
2006
[Content]
Fig 5. The simulated time behavior of temperature in the center of the sample during cookoff.
334
Czech Republic
2006
[Content]
335
3.5
Czech Republic
2006
[Content]
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.
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
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.
337
Czech Republic
2006
[Content]
1.
INTRODUCTION
338
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
20
double base
2-NDPA
30, 35, 40
34
double base
DPA
19
triple base
EC
21
double base
DPA + EC
10
porous
DPA
18
3.
EXPERIMENTS
339
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
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
340
4.2
Czech Republic
2006
[Content]
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
131 kJ/mole
135 kJ/mole
137 kJ/mole
137 kJ/mole
141 kJ/mole
4.3
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.
341
Czech Republic
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.
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
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.
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.
343
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
6
4
2
0
-2
0.0
0.5
1.0
1.5
2.0
2.5
t [days] at 89C
Fig 6.
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
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.
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.
345
4.4
Czech Republic
2006
[Content]
(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)
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
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.
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]
5.
DISCUSSION
5.1
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
347
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
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]
348
11
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
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)
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
Fig 14. Shift of 2nd minimum with increasing ageing time at 40, 35 and 30C
(propellant N)
12
349
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
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
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
13
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
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
351
Czech Republic
2006
[Content]
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
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
average
352
15
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
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
353
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.
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]
= 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
354
17
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).
355
Czech Republic
2006
[Content]
* 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
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
3.
Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min
357
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
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
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
N12
C6 125.9
1.336
N2
T2j-c
N1
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
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
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
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
0.191
0.195
0.198
Total charge of
the tetrazine rings
-0.196
-0.054
-0.063
-0.046
-0.043
-0.031
-0.049
-0.029
-0.036
-0.034
-0.012
-0.001
0.074
-0.151
-0.115
0.112
-0.074
-0.093
-0.095
0.102
Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min
Total charge of
the substituents
0.194
0.055
0.063
0.048
0.044
0.030
0.050
0.029
0.035
0.033
0.010
-0.001
-0.074
0.150
0.115
-0.110
0.073
0.095
0.094
-0.102
361
-0.197
-0.203
-0.193
-0.185
-0.188
-0.198
-0.179
-0.193
-0.180
-0.188
-0.184
-0.195
-0.251
-0.228
-0.208
-0.239
-0.201
-0.226
-0.232
-0.245
-0.233
-0.230
-0.191
-0.182
0.329
0.352
0.342
0.542
0.342
0.349
0.552
0.329
0.537
0.325
0.525
0.325
-0.181
-0.195
-0.253
-0.191
-0.236
-0.188
-0.232
-0.172
-0.189
-0.167
-0.190
-0.188
-0.194
-0.201
-0.194
-0.182
-0.189
-0.195
-0.179
-0.189
-0.177
-0.185
-0.183
-0.196
Czech Republic
0.196
0.194
0.190
0.198
0.204
0.201
0.196
0.206
0.204
0.216
0.209
0.212
0.193
0.190
0.193
0.199
0.197
0.193
0.201
0.197
0.203
0.200
0.201
0.196
2006
-0.105
-0.091
-0.123
0.142
-0.071
-0.064
0.127
-0.067
0.165
-0.029
0.187
-0.028
[Content]
0.104
0.091
0.123
-0.143
0.071
0.066
-0.129
0.068
-0.165
0.029
-0.187
0.027
As showed in the Table 2, after the C-N heterocycle substitutes a hydrogen atom
connected with the atom C3 of the tetrazine, the positive charge of C3 atom increases
obviously. The part electrons on the tetrazine rings shifts to the substituent. It indicates that
the C-N heterocycle as the substituent has an electron-absorbing characteristic. The electronabsorbing characteristic of the six-membered C-N heterocyclic substituent would boost up
along with the increase on the number of nitrogen atom. And the electron-absorbing
characteristic of the five-membered C-N heterocyclic substituent is related to not only the
number of nitrogen atom, but also the place of the C-N heterocycle connected with the
tetrazine. When the nitrogen atom of the C-N heterocycle connected with the tetrazine ring,
the electron-absorbing characteristic is stronger than the instance joining with the carbon
atom. As above discussed, the electron-absorbing characteristic of the substituent 42 is
strongest among all the whole system.
Table 3. the part stabilization interaction energies E(2) of the tetrazine substituted
by six-membered C-N heterocycle/(Kcal/mol)
T1-a
T1-b
T1-c
T2l-a
362
Donor
Acceptor
E(2)
LP C3
BD* N1-N2
122.24
LP C3
BD* N4-N5
109.61
LP C3
BD* C7-C8
32.67
BD N1-C6
BD* N2-C3
8.76
BD N1-C6
BD* N4-N5
21.38
T2d
T2j-a
Donor
Acceptor
E(2)
LP C3
BD* N4-N5
118.31
LP C3
BD* N1-N2
108.09
LP C3
BD* C7-C12
31.91
LP C3
BD* N1-N2
116.96
LP C3
BD* N4-N5
116.95
BD N2-C3
BD* N1-C6
20.69
LP C3
BD* C7-N8
28.96
BD N2-C3
BD* N4-N5
19.21
BD N1-N2
BD* C3-N4
24.09
BD N4-N5
BD* N1-C6
22.73
BD N1-N2
BD* N5-C6
22.74
BD N4-N5
BD* N2-C3
23.91
BD C3-N4
BD* N1-N2
19.32
BD C7-C12
BD* N2-C3
24.79
LP C6
BD* N1-N2
112.89
LP C6
BD* N4-N5
BD C7-C8
LP* C3
BD C3-N4
BD* N5-C6
20.35
BD N5-C6
BD* N1-N2
21.46
112.89
BD N5-C6
BD* C3-N4
19.10
52.06
BD C7-C12
BD* C3-N4
21.88
BD N1-C6
BD* N2-C3
19.15
BD N1-C6
BD* N4-N5
20.26
T2j-b
T2j-c
LP C3
BD* N1-N2
120.48
LP C3
BD* N5-N4
108.88
BD N2-C3
BD* N1-C6
19.36
LP C3
BD* C7-C8
33.61
BD N2-C3
BD* N4-N5
19.07
BD N1-C6
BD* N2-C3
19.52
BD N4-N5
BD* N1-C6
24.09
BD N1-C6
BD* N4-N5
20.35
T2l-b
BD N4-N5
BD* N2-C3
24.22
BD N2-C3
BD* C7-N12
10.60
Czech Republic
T3
2006
[Content]
BD N2-C3
BD* N1-C6
18.87
BD N2-C3
BD* N4-N5
20.92
BD C7-N12
BD* N2-C3
13.91
BD N4-N5
BD* N1-C6
24.75
BD N1-N2
BD* C3-N4
24.13
BD N4-N5
BD* N2-C3
24.32
BD N1-N2
BD* N5-C6
23.06
BD C3-N4
BD* N1-N2
19.70
TT
BD C3-N2
BD* C6-N1
18.57
BD C3-N2
BD* N4-N5
20.75
BD C3-N4
BD* N5-C6
19.90
BD C6-N1
BD* C3-N2
19.90
BD N5-C6
BD* N1-N2
21.31
BD C6-N1
BD* N4-N5
20.51
BD N5-C6
BD* C3-N4
19.24
BD N4-N5
BD* C3-N2
24.48
BD C7-C8
BD* C3-N4
21.59
BD N4-N5
BD* C6-N1
24.61
In order to study the interaction between the substitutent and the tetrazine ring, the
stabilization interaction energies E(2) are calculated by means of the second-order
perturbation theory. The selected stabilization interaction energies E(2) for the molecules
composed of six-membered C-N heterocycle and tetrazine ring at the B3LYP/6-311G(d,p)
level are summarized in the Table 3, where BD and BD* represent bonding and antibonding
natural bond orbital and LP represents lone pairs. From the Table 3, we find that the
interaction between the bonding and antibonding orbital of the tetrazine ring in the systems
T1-a,T2d,T2j-a and T2j-c becomes weak (their values smaller than 10Kcal/mol are not
included in Table3). But Interaction between the lone pairs of C atom and * antibonding
orbitals of the tetrazine ring becomes strong, E(2) values are more than 100Kcal/mol. And in
the other systems, interaction between the bonding and antibonding orbital has no change
basically. Except for T3 and TT, there exit strong interaction between the two rings. In
addition, the results of NBO analyses show that when five-membered C-N heterocycle
connected the tetrazine ring, there exit strong donor-acceptor interaction between the LP of
N atom in the C-N heterocycle and the CN * antibonding orbitals in the tetrazine ring (see
Table 4). Because of this interaction, the lengths of CN bonds connected the two rings are
less than the lengths of CC bonds. The results agree with the analyses of the structures.
Table 4. the part stabilization interaction energies E(2) of the tetrazine substituted
by five-membered C-N heterocycle/(Kcal/mol)
TP1-1
TP2-1
TI2-1
T123-1
T134-1
T41-1
T42-1
Donor
LP N7
LP N8
LP N9
LP N7
LP N7
LP N9
LP N8
Acceptor
BD* C3-N4
BD* N2-C3
BD*N2-C3
BD* N2-C3
BD* N2-C3
BD* N2-C3
BD* N2-C3
E(2)
56.10
48.97
52.95
45.61
49.74
42.89
32.48
In addition, from the analyses of molecular orbitals, we get much conjugate information
about the molecules of the tetrazine substituted by the single ring, Figure 4 shows the
delocalized occupied orbital in all the molecules. The figure shows that there exit a
delocalized occupied orbital composed of purely 2PZ orbital of all carbon and nitrogen
atoms in the molecules T3,TT and the isomers TP2,TI2,T123, T134, T41 and T42. Although
other molecules have no orbital over the whole molecule, interaction between the two rings
dont destroy the primary conjugation of the tetrazine ring, and the conjugate orbital expands
its range to some atoms of the substituent. As above analyzed, after a hydrogen atom of the
tetrazine is substituted by the C-N heterocycle, there exit strong interaction between the two
rings. This interaction dont destroy the delocalized orbital of the tetrazine, and in some
8
Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min
363
Czech Republic
2006
[Content]
case it can form a delocalized occupied orbital over the whole molecule. This delocalized
orbital is advantage to the stabilization of the whole molecule.
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
TI2-2
TI2-3
TI2-4
T123-1
T123-2
T123-3
T134-1
T134-2
T41-1
T41-2
T42-1
T42-2
Fig 4.
3.3
Fig 5.
The relationship between the total energy and N atom number for C-N
compounds
364
Czech Republic
2006
[Content]
Heat of formation is a critical parameter of energetic materials and the calculation of the
velocity of detonation. It is a key to design and synthesize new energy materials. And highnitrogen energetic materials possess very high positive heat of formation, which attracts
many researchers attention. So the calculation of HOF is a significant matter for theoretical
study and experimental study. In this paper, HOF of the objective compounds are calculated
through atomic scheme. The different methods were adopted to calculate the academic
values of HOF for the single six-membered C-N heterocycle, and these results are compared
with the experimental data, which are illustrated in Table 5. The point energy is calculated at
the B3LYP/6-311+G(2df,2p) level, then it is corrected by the ZPE which is calculated at the
B3LYP/6-311G(d,p) level. Finally we take corrected molecular energy as a parameter to
calculate heat of formation; the results see the last column of Table 5. Form the data of Table
5, we find that the results of this method show a good agreement with the experimental data,
and the most error is 19.4KJ/mol. So we calculated the heat of formation for all the
molecules using the above method discussed. The results see the Table 6.
Table 5. Tthe experimental data and theoretical data of HOF for the single sixmembered C-N heterocycle
Compounds
(1) pyridine
(2d) pyradine
(2j) pyrimidine
(2l) pyridazine
(3) 1,3,5-triazine
(T) 1,2,4,5-tetrazine
B3LYP
Experimental
[16]
data
G3
144.77
196.23
196.65
278.24
225.52
464
143.0
209.0
190.3
285.3
229.0
489.7
6-311G(d,p)
6-311+G(2df,2p)
184.3
237.5
220.4
313.7
249.5
496.8
155.1
204.2
187.6
280.2
213.0
455.5
6-311+G(2df,2p)
(ZPE correction
145.8
196.2
179.5
272.2
206.1
450.1
T1-a
T1-b
T1-c
T2d
T2j-a
T2j-b
T2j-c
T2l-a
T2l-b
T3
TT
Energy
a.u.
-543.5430
-543.5466
-543.5458
-559.5758
-559.5787
-559.5864
-559.5823
-559.5454
-559.5492
-575.6211
-591.5778
fH(298K)
(KJ/mol)
639.1
631.1
632.5
692.4
683.5
667.8
676.4
769.9
761.7
712.9
960.1
TP1-1
TP1-2
TP1-3
TP2-1
TP2-2
TP2-3
TP2-4
TI2-1
TI2-2
TI2-3
TI2-4
Energy
a.u.
-505.4232
-505.4371
-505.4336
-521.4500
-521.4595
-521.4647
-521.4644
-521.4700
-521.4799
-521.4787
-521.4842
fH(298K)
(KJ/mol)
606.8
571.2
580.2
677.1
653.2
641.2
641.4
625.6
599.8
602.4
588.9
T123-1
T123-2
T123-3
T134-1
T134-2
T41-1
T41-2
T42-1
T42-2
Energy
a.u.
-537.4709
-537.4870
-537.4840
-537.4925
-537.5013
-553.4995
-553.5110
-553.4993
-553.5115
fH(298K)
(KJ/mol)
759.8
720.1
727.7
705.9
682.5
825.0
796.0
827.2
797.2
The results (Table 5) show that the different type of bond has different contributions to
HOF, the sequence is NN bonds>CN bonds>CC bonds. For example, the tetrazine (T) has
two NN bonds, the pyridazine(2l) has a NN bond, and the triazine(3) has no NN bond, so
HOF of tetrazine> HOF of pyridazine> HOF of triazine. And the triazine has three CN
bonds, the pyriazine(2d) or the pyrimidine(2j) has two CN bonds and the pyridine(1) has
only one CN bond, so HOF of triazine> HOF of pyriazine or pyrimidine> HOF of pyridine.
When C-N heterocycle as the substituent (see Table 6), the system follow the same rule. For
example, the tetrazole has the most number of NN bonds, so HOF of these isomers are the
most one among all the molecules. In addition, HOF of the isomer is related to the place of
the substituents connected with the tetrazine. When two rings are connected with the CN
10
Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min
365
Czech Republic
2006
[Content]
bond, its HOF is more than that of those molecules which two rings are connected with the
CC bond.
Fig 6.
We take out the HOF of TP1-1, TP2-1, T123-1 and T42-1 (because their HOF are the
maximum in each isomer) and analyze the relationship between the HOF and the number of
N atom in these molecules, the result sees the Figure 8. This curve has a good relativity. The
Figure 6 shows that the HOF would boost up linearly along with the increase on the number
of nitrogen atom when five- membered heterocycle connected the tetrazine ring with CN
bond.
4.
CONCLUSIONS
In this paper, after a hydrogen atom of the tetrazine is substituted by the C-N
heterocycle, we obtain 31 tetrazines C-N heterocyclic derivatives and their optimized
geometries, electronic structures, conjugations, energies, and heats of formation have been
investigated by the DFT methods. The results show:(1) When the nitrogen heterocycle
replaces a hydrogen atom on tetrazine, NN bonds on tetrazine rings are shorter and CN
bonds which are close to the substituent are longer, the other bonds change little. (2) In these
derivatives, the conjugations of the tetrazine rings are not distroyed. The part system forms a
delocalized occupied orbitals composed by 2PZ orbitals of all atoms, which enhances the
stability of the molecule. (3) The C-N heterocycles as the substituent have a strong
absorbing-electron characteristic, and the five- memebered ring 42 (2H-tetrazole) among
them has the strongest absorbing-electron characteristic. (4) There are a good linear
relationship between the energies, HOFs and the number of N atoms, the total molecule
energy decrease about 16a.u when increasing a nitrogen atom. The HOFs would boost up
linearly along with the increase on the number of nitrogen atom when five-membered
heterocycle connected the tetrazine ring with CN bond. 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.
366
11
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
12
A.K. Sikder, Nirmala Sikder. A review of advanced high performance, insensitive and
thermally stable energetic materials emerging for military and space applications. Journal of
Hazardous Materials A112 (2004) 115
H. Feuer and A. T. Nielsen, Nitro Compounds; VCH, NewYork, Weinheim, Cambridge, 1990
A. T. Nielsen, Nitrocarbons; VCH, New York, Weinheim, Cambridge, 1995
J. Khler and R. Meyer, Explosivstoffe, 7. Aufl., Wiley-VCH, Weinheim, 1991
M.-X. Zhang, P. E. Eaton, and R. Gilardi, Hepta- and Octanitrocubanes, Angew. Chem. Int. Ed.
2000, 39, 401
Anton Hammerl, Thomas M. Klaptke, and Peter Schwerdtfeger.Azolylpentazoles as HighEnergy Materials: A Computational Study. Chem. Eur. J. 2003, 9, 5511-5519
Anton Hammerl, Thomas M. Klaptke, Heinrich Nth, et al. Synthesis, Structure, Molecular
orbital and Valence Bond Calculations for Tetrazole Azide, CHN7.Propellants, Explosives,
Pyrotechnics 28 (2003), No.4, 156-173
Parr, R. G.; Yang, W. Density-functional Theory of Atoms and Molecules; Oxford University
Press: New York, 1989.
Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B. 1988, 37, 785.
Carpenter, J. E.; Weinhold, F. J. Mol. Struct. (THEOCHEM) 1988, 169, 41.
Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88, 899.
Foster, J. P.; Weinhold, F. J. Am. Chem. Soc. 1980, 102, 7211.
Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83, 735.
Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,M. A.; Cheeseman, J. R.;
Zakrzewski, V. G.; Montgomery, J. A., Jr.;Stratmann, R. E.; Burant, J. C.; Dapprich, S.;
Millam, J. M.; Daniels, A.D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.;
Cossi,M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.;Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick,D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi,I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; AlLaham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. m.W.;
Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98, Revision A.11;Gaussian, Inc.: Pittsburgh, PA, 2001.
Lias S.G, Libman J.F, Levine R.D. J.Phys.Chem.Ref.Data.1984,13,695
Zhou Yang, Shu Yuan-Jie, Wang Xin, Long Xin-Ping, Tan An-Min
367
Czech Republic
2006
[Content]
Abstract:
Thermal decomposition during the induction period of six dinitrotoluene isomers was examined. To
define the production of contaminations, the liquid-solid phase equilibrium was used (criometry
method). The mean association (aggregation) number of decomposition products was appointed in the
melting temperature of each substance. The presence of macromolecule products of decomposition was
detected.
1. Introduction
The exothermic decomposition of unstable substances can be a reason of the thermal
explosion during different technological operations. The generated heat rate determines
conditions of safe processes. This can be predicted on the bases of reliable kinetic parameters
of the thermal decomposition process. The thermal explosion occurs when the conversion
degree of unstable substance is small. The kinetic parameters for induction period of the
thermal decomposition process are needed for safety estimation. Our investigations were
concentrated on elaborating the methods, which allow analyzing decomposition process for
small conversion degrees of the thermal decomposition reaction. These methods were
developed using substances containing nitro groups like N,2,4,6-tetranitro-N-methylaniline
(TET) and 2,4,6-trinitrotoluene (TNT)
[1-6]
amount of thermal decomposition products. There was also used the method which exploit the
heat effect of partial (initial) decomposition process. It was noticed that the mean value of
decomposition rate during following measurements performed on the same sample in the same
nonisothermal conditions changes significantly in the oscillatory way. These changes were
ascribed to the concentration changes of volatile decomposition products possessing catalytic
368
Czech Republic
2006
[Content]
properties. The changes of volatile products concentration are connected with the release
processes from liquid reaction phase [2,3] .
The first stage of aromatic nitrocompounds decomposition depends on substituents
position. The nitro-group without any substituent in ortho position can separate from the
benzene ring in homolytic way and to form nitrite-group
[7]
position can react with each other forming benzofurazane and benzofuroxane
group in the ortho position to the methyl-group forms anthranil
Rogers
[12]
and Dacons
[13]
[10,11]
[8,9]
. The nitro-
temperatures 300 and 200 C respectively. Rogers have isolated a number of substances, which
were oxidation reaction products or another intramolecular reaction. Dacons have identified 25
substances dissolvable in benzene. One of this was a product of intermolecular reaction, which
was made of two benzene rings bonded by nitrogen-atoms. Additionally Dacons received
insoluble in benzene so called explosive cake, which were observed earlier by Adams [14]. It
is probably a mixture of macromolecule substances. From the presented results follows that
thermal decomposition process is very complicated. In the case of multisubstituted substances
it comes a question about a role of each unstable substituent in the thermal decomposition
process. This information allows to predict the thermal stability of multisubstituted substances
in the thermal explosion aspect (safety).
For better knowledge of the substituents position influence on the thermal stability it is
necessary to compare the thermal decomposition process of substances with a similar structure
homological series. One of these series is dinitrotoluene isomers. We decided to examine
these substances using criometry technique applied earlier to TNT and TET analysis.
For better knowledge of the substituents position influence on the thermal stability it is
necessary to compare thermal decomposition parameters of substances with a similar structure
homological series. One of these series is dinitrotoluene isomers. In the literature there are
not systematic studies for these substances. The onset temperatures values of thermal
decomposition for three isomers: 2,4-dinitrotoluene (2,4-DNT), 2,6-dinitrotoluene (2,6-DNT)
and 3,4-dinitrotoluene (3,4-DNT) have been determined by Duh
DNT and 3,4-DNT) were examined by Ando
[16]
[15]
The aim of this work is to study the thermal decomposition during the induction period and
to define an influence of nitro-groups position in relation to methyl-group on the thermal
369
Czech Republic
2006
[Content]
2. Experimental
Dinitrotoluene isomers were synthesized in the Warsaw Military University of
Technology. Melting temperatures (Tm), melting enthalpies (Hm) and purity (P) determined by
criometry method from DSC measurements were collected in the table 1. Literature values of
melting temperature and enthalpy were placed in the table 1 for comparison.
Table 1. Melting temperatures (Tm), melting enthalpies (Hm) and purity (P) of
dinitrotoluenes.
Isomer
2,3-DNT
2,4-DNT
2,5-DNT
2,6-DNT
3,4-DNT
3,5-DNT
Tm/K
332.7
343.7
324.5
338.3
332.4
366.0
Hm/(kJ/mol)
16.13
21.49
15.67
18.02
18.52
19.79
Tm.(lit)/K
329.7
[17]
340.7
[17]
323.6
[18]
327.4
[17]
329.4
[17]
366.1
[17]
Hm(lit)/(kJ/mol)
P/(%mol)
[17]
99.9
[17]
99.2
17.5
22.2
99.4
[17]
99.9
[17]
99.8
23.8
18.8
99.9
The purity of dinitrotoluenes was investigated by gas chromatography and for all isomers it
was about P=99.9% mass.
Measurements were made on a heat-flux DSC 605 UNIPAN calorimeter. The calibration
was performed with gallium, indium, cadmium, tin, naphthalene and benzoic acid. The metals
used had purities greater than 99.999%, and the organic substances had purities greater than
99.95%. The measurements were made using hermetic aluminium pans with heating rate 2
K/min. The samples were sealed under a reduced pressure of about 1.3 kPa. Measurements
were started from the temperature below melting process to the chosen temperature below the
exothermic decomposition process (there was no visible exothermic effect). Consecutive
measurements were made on the same sample.
370
Czech Republic
2006
[Content]
f =
H m H n
H n
(1)
where: Hm is the melting enthalpy before a thermal decomposition process determined from
the first DSC run, Hn is the melting enthalpy in the consecutive measurement performed on
the same sample.
Criometry method can be used to predict the real mol fraction ( x ) of impurities according
with following equation
R(Tm0 ) 2 1
x
Tm = Tm0
H m F
(2)
where F is the melted fraction of a sample, Tm0 is the melting temperature of ideally pure
substance, Tm is the melting temperature of the sample, Hm is the melting enthalpy. The purity
of a sample is equal to
P = 1 x 100%
Fig. 1 shows DSC curves of the 2,3-DNT melting process in five consecutive runs made on
the same sample. Measurements were leaded to chosen temperature Tf. The increase of the
decomposition products ( x ) follows during the measurement. This increase ( x ) is equal to the
371
Czech Republic
2006
[Content]
DSC Signal
310
330 T/K
350
Fig. 1. DSC curves of the melting process for 5 consecutive runs performed on the same
sample of 2,3-DNT.
difference between the mole fraction of impurities of the sample taken to the investigation and
the mole fraction after the thermal decomposition during DSC measurements. This increase is
equal to real mole fraction of thermal decomposition products connected with the thermal
decomposition of the investigated sample during DSC measurements
x = x i
i
372
Czech Republic
2006
[Content]
f = x = ixi
(3)
where xi is the mole fraction of decomposition product with i number of mers in the
molecules. The size of these associates can be defined by so-called mean association number .
This parameter can be calculated using the equation [19] :
x = f/( - f+ f)
(4)
3.2. Results
In the table 2 are placed the melting enthalpy changes (DHm), the real mole fraction of
thermal products created during thermal decomposition ( x ) and mean association number of
decomposition products . In these DSC measurements the final temperature (Tf) was lower
than the onset temperature of the exothermic process.
Table 2. The final temperature (Tf), the real mole fraction x , the melting enthalpy changes
(DH) and the mean association number () of decomposition products of dinitrotoluenes.
isomer
2,3-DNT
2,4-DNT
2,5-DNT
3,4-DNT
3,5-DNT
2,6-DNT
Number
of run
1
1
2
1
1
2
3
4
1
1
2
3
Tf/K
x*100
547,7
545,8
545,8
561,7
532,8
532,8
532,8
532,8
591,0
562,3
562,3
562,3
2,38
1,38
1,45
1,82
1,07
1,63
1,44
1,01
1,55
0,86
1,78
1,46
DH/
(kJ/mol)
1,978
0,760
0,590
2,619
0,919
0,530
0,960
0,370
3,119
0,720
0,196
1,260
5,73
2,62
3,35
10,83
4,83
3,87
7,56
9,83
4,80
2,19
7,90
5,73
373
Czech Republic
2006
[Content]
Peak intensity
each other.
500
1000
1500
m/z
Fig. 2. Diagram of the MALDI TOF mass spectrum of a 2,5-DNT sample.
The structure of thermal decomposition products depends strongly on the position of nitro-
374
Czech Republic
2006
[Content]
group in relation to methyl-group. Our results indicate that some thermal products have an
ability to form associates (aggregation).
5. Conclusions
Application of the liquid - solid phase equilibrium let us study a decomposition below the
onset temperature of the exothermic decomposition. Values of mean association number which
were obtained in the melting temperature of dinitrotoluene indicates that thermal
decomposition products create molecules with mass many times bigger than DNT molecule.
These values change in oscillatory way during consecutive measurements. The presence of
macromolecules was verified by MALDI TOF analysis.
Literature
[1] A. Ksiczak, T. Ksiczak. Thermochim. Acta, 275 (1996) 27.
[2] A. Ksiczak, T. Ksiczak, T. Zielenkiewicz. J. Thermal Anal., 77(2004) 233-242.
[3] A. Ksiczak, T. Ksiczak. J. Thermal Anal., 60(2000) 25-33.
[4] A. Ksiczak, T. Ksiczak. Thermochim. Acta, 268(1995) 95-104.
[5] A. Ksiczak, T. Ksiczak. J. Thermal Anal.., 41 (1994) 1153-1163.
[6] A. Ksiczak, T. Ksiczak. Thermochim. Acta, 284 (1996) 299-308
[7] Yu. Ya. Maksimov, S. B. Sorochkin, S. V. Titov, Tr. Mosk. Khim. Technol. Inst. im. D. I.
Mendeleeva 1980, 112, 26
[8] D. E. Bliss, S. L. Christian, W. S. J. Wilson, Energ. Mater. 1991, 9, 319
[9] E. K. Fields, S.Meyerson, J. Am. Chem. Soc. 1967, 89, 3224
[10] Tsang, W., Robaugh, D., Mallard, W.G.; J. Phys. Chem. 1986, 90, 5968
[11] He, Y.Z., Cui, J.P., Mallard, W.G., Tsang, W. J. Am. Chem. Soc. 1988, 110, 3754
[12] R.N. Rogers. Anal. Chem., 39 (1967) 730.
[13] J.C. Dacons, H.G. Adolph, M.J. Kamlet. J. Phys. Chem., 74 (1970) 3035.
[14] G.K. Adams, P.R. Rowland, L.A. Wiseman. Ministry of Supply Report A. C. 3982, Great
Britain, 1943.
[15] Y. Duh, C. Lee, C. Hsu, D. Hwang, C. Kao. J. Hazard. Mat., 53 (1997) 183
[16] T. Ando, Y. Fujimoto, S. Morisaki. J. Hazard. Mat., 28(3) (1991)251
[17] H.J. Borchardt, F. Daniels. J. Am. Chem. Soc., 79 (1957) 41.
[18] J.W. Beckmann, J.S. Wilkes, R.R. McGuire. Thermochim. Acta, 19 (1977) 111.
[19] J. Prigogine, R. Defay. Chemical Thermodynamics, London, 1954.
375
Only in Proceedings
Czech Republic
2006
[Content]
Abstract:
The main routes of transformation of
guanidine and dicyandiamide in aqueous
solutions of nitric acid are considered. The nitration of guanidine is shown to be
catalyzed by lower nitrogen oxides. On the basis of studies of the quantitative
regularities of transformation of guanylnitrourea, guanylisocyanate and nitroguanidine,
the probable mechanism of reactions is suggested and the main factors determining the
reactivity of the substrates are addressed.
Keywords: Synthesis, nitration of guanidine and dicyandiamide, nitroguanidine.
1.
INTRODUCTION.
The structural feature of nitroguanidine derivatives associated with the planar structure
of the nitroguanyl moiety stabilized by intramolecular hydrogen bonds and a considerable
delocalization of -electron density, attracts for a long time the attention of researchers to
such compounds as potential objects for the development of insensitive high explosives
(IHE). Nitroguanidine (NQ) itself may be considered as a nitroimine analogue of a
promising IHE, DADNE (FOX-7) [1,2] . In general, NQ can be viewed as one of the amidine
derivatives, incorporated into the structure of a series of such IHE as NTO, DNGU,
DADNE.
NH N
O
N
H
NTO
NO2
H2N
NH2
H2N
NO2
NO2
NO2
DADNE
NH2
N
Nitroguanidine (NQ)
NO2
N
H
N
O
O
N
H
N
NO2
DNGU
377
Czech Republic
2006
[Content]
2.
The study of kinetics of guanidine nitration in aqueous nitric acid has shown that NQ
does not essentially undergo decomposition under such conditions (Fig.1).
, min
Fig 1.
The dependence of the effective rate constants for the nitration of guanidine nitrate on
the nitric acid concentration is given in Fig. 2.
378
Czech Republic
2006
[Content]
HNO3, wt.%
Fig 2.
25
The slopes in eqs. 1 and 2 are close to those observed earlier for the kinetics of nitration
of carbamates and benzene sulfonamides [19], which implies the absence in the course of
nitration of protolytic equilibria and the resulting variations in concentrations of the active
forms of the substrate. Taking into account the high basicity of guanidine (pKBH+=13.6) [15],
we can suppose that during the nitration in aqueous nitric acid, similar to the nitration in
mixed acid [16], the active form of the substrate is its conjugated acid. This conclusion meets
the Schofields criterion for the threshold rate of a bimolecular reaction [17]. Hence, the
guanidine nitration process may be presented as follows:
NH2
H+
NH2
+
H2N C
HN
NH2
NH2
HNO3
NH
H +2
+N C
NH2
O2N
NH2
-2 H+
N
O2N
NH2
379
Czech Republic
2006
[Content]
, min
Fig 3.
The theoretical studies show that the nitrating activity function is rather ambiguous in
nature and depends both on the atom undergoing the attack by an electophile and the
structure of the substrate. This ambiguity is caused by the participation in the nitration
reaction, alongside with the nitronium ion, of some other electrophilic agents. During
nitration most of the highly basic amines undergo protonation which hampers the subsequent
electrophilic attack of the conjugated acids due to large differences in the energies of the
lowest vacant molecular orbital of the acceptor (NO2+) and the highest occupied molecular
orbital of the donor (guanidinium cation). Assuming that the solvation energies of the
nitrating and nitrosating agents are close to each other, it seems plausible that the energy
level of the lowest vacant molecular orbital of the nitrosating agents is considerably
lower[18], which make them sufficiently strong electrophiles. The significance of the
electrostatic contribution in this case is not so marked because of the presence of a partial
positive charge on the reaction center (the nitrogen atom). Besides, the nitration is known to
be sensitive primarily to the charge control whereas the nitrosation is the subject of the
orbital control. Therefore more soft electrophiles should be used to effect nitrosation (the
combination of real charges plays a negligible role; the primary positions determining the
stability of the intermediate occupy the frontier orbital energies, their difference in the
present case being considerably lower compared to that of the nitrating agents). It follows
that the catalytic action of nitrogen oxides in the nitration reaction will develop more
distinctly in the series of amines which undergo nitration in the protonated form.
Consequently we can anticipate the catalytic activity of nitrogen oxides for guanidine
nitration, which, as it was mentioned above, in acidic media exists exclusively in the
protonated form. Indeed, the introduction of nitrogen oxides accelerated the rate of nitration
of guanidine nitrate many times (Fig. 4).
380
Czech Republic
2006
[Content]
, min
Fig 5.
381
NH2
Czech Republic
2006
[Content]
NH2
kn
H2N C
O2N
NH2
NH2
NQ
kn1
N2O4
NH2
ON
N
NH2
NsQ
W=kn[C+(NH2)3]+ kn1[NsQ]
Scheme 2. The reaction of catalytic nitration of guanidine .
It is necessary for the catalytic effect to manifest itself that both the rate of guanidine
nitrosation to form the N-nitroso derivative and the rate of its subsequent transformation be
considerably higher compared to that for the direct guanidine nitration.
The enhanced activity of N-nitrosoguanidine in comparison with the starting amine in
the acidic nitration (kn1>kn) implies that the reactive form of the substrate is the covalent Nnitrosoguanidine (NsQ) molecule. It follows from the above that the introduction of the
nitroso group contributes to the decrease in its basicity to a greater extent than to the
lowering of its reactivity in the electrophilic substitution. Such behavior agrees well with the
ambident nature of the nitroso group, its negative inductive effect being compensated for by
its positive mesomeric effect. Hence the nucleophilic nature of N-nitrosoguanidine is
displayed more distinctly than its basicity, which allows one to assume the presence of
sufficiently soft acceptors such as HNO3 molecule, as nitrating agents.
The treatment of a preliminarily prepared N-nitrosoguanidine with nitric acid also
invariably resulted in the formation of NQ in a quantitative yield. The kinetics of Nnitrosoguanidine nitration turned out to be identical to that of guanidine nitration in the
presence of an equivalent amount of nitrogen oxides. Hence we may conclude that the rate of
denitrosation of the N-nitroso compound is rather high because the concentration of NsQ is
linearly related to that of the nitrosating agents (Fig.4).
Addition of sodium nitrite to the nitric acid solution during the nitrating process causes
the immediate increase in the formation of NQ (Fig.6). Hence we can conclude that the
nitrosation stage is a quickly established equilibrium, which is shifted to the side of the
guanidinium cation.
382
Czech Republic
2006
[Content]
, min.
Fig 6.
Generally, two possible routes for NsQ nitrolysis may be imagined (Scheme 3).
NH2
N2O4
-H+
NH2
ON
H2N C
N C
H
NH2
NH2
NH2
ON
NH2
ON
NH2
N
N NO2
H
HNO3
HNO3
-NO
NH2
-NO
N
O2N
Scheme 3.
NH2
ON
NH2
+
N C
NH2
O2N
By analogy with the guanidine nitration, the most probable appears the routeA because
N-nitro-N-nitrosoguanidine is an unstable intermediate in acidic media and undergoes a
prompt denitrosation to produce the final product. The route B (nitrolysis), however, can
not be completely excluded because the attack of the nitrating agent on the nitrogen atom of
the nitrosamine group effectively facilitates the heterolysis of N-NO bond.
In accordance with the conclusions drawn as a result of the kinetic study regarding a
significant acceleration of the reaction of NQ formation under the action of nitrogen oxides,
a preparative study of the nitration of guanidine nitrate with concentrated nitric acid with
added N2O4 has been carried out.
7
383
Czech Republic
2006
[Content]
guanidine
0.15
0.12
1.7
0.1
0.1
1.5
1.2
Maximum yield
of NQ, %
Time
( ),
necessary
to
attain
the
maximum
yield, hours.
45
90
1.5
30
96
45
85
60
81
45
85
20
96
20
93
It follows from the data listed in Table 1 that nitration at temperatures over 20-300C in the
presence of N2O4 is inefficient which is apparently caused by the low concentration of
nitrogen oxides in the liquid phase at elevated temperatures due to their high volatility.
Reducing further the excess of nitric acid over the substrate concentration deteriorates the
results of nitration because of a noticeable decrease in the solubility of nitrogen oxides in the
acid phase.
Hence the suggested procedure, as distinct from the existing methods, is featured by high
efficiency and an essential absence of wastes and by-products.
As an alternative method for the preparation of NQ, an approach based on dicyandiamide
(DCDA) as starting material should be considered. This multistep procedure involves the
following stages:
-
Reasoning from the general theoretic conceptions and taking into account a strong electronaccepting character of the cyano group comparable to that of the nitro group, it is believed
that DCDA, similar to NQ which is not subjected to the double nitration in aqueous solutions
of nitric acid, would be inert in such media. It follows that the process of its nitration will
have a stepwise character involving on the first stage the DCDA hydrolysis to form
guanylurea and consequently its subsequent nitration. The overall rate of this process should
not exceed the rate of nitration of guanylurea itself. However, in the course of kinetic studies
of DCDA nitration it turned out that the combined process of hydrolysis-nitration took place
much more readily compared to the direct nitration of guanylurea, its product of hydrolysis.
So the apparent first-order rate constant for the formation of guanylurea from DCDA in 83%
nitric acid is 17 times as high as the corresponding constant for the production of NQ from
guanylurea. This discrepancy can be rationalized in terms of an assumption that the
384
Czech Republic
2006
[Content]
hydrolysis of the cyano group and nitration take place simultaneously, i.e. that in this case
either the nitric acid molecule or its protonated form play the role of the nitro group carrier.
At the same time the sole hydrolysis product is guanidine. So it may be concluded that in
this N-nitro compound the nitro group is attached to the urea fragment. Thus the combined
process of hydrolysis-nitration can be depicted by the following scheme involving the
transfer of the nitro group from guanidine to the urea moiety (Scheme 4).
NH2
-H+
NH2
NH2
H2N
C
NH
N
H
NH2
H2N
+ NH2
H
+
N C
NH2
N
N
HNO3
HN
O
N
H2N
O
H
H
+ N NO
2
C
NH
NO2
HN
OH
385
Czech Republic
2006
[Content]
HNO3, wt.%
.
Fig.7 Dependence of the apparent nitration rate constants on HNO3
concentration at 25 for: 1-urea; 2-guanylurea; 3-DCDA. The
corresponding linear correlation equations are as follows:
1. log Kapp =(0.190.01) [HNO3] wt.% (19.40.9)
2. log Kapp =(0.120.01) [HNO3] wt.% (13.51.2)
3. log Kapp =(0.0770.005) [HNO3] wt.% (8.80.5)
Formally the nature of the nitrating agent for the nitration of DCDA and other amides is
identical but there exist some basic differences: whereas in the process of amides nitration
the formation of cyclic 6-membered transition state takes place by means of an
intramolecular hydrogen bond, in the case of DCDA the main feature is the ambident nature
of the reagent. On the one hand, the nitrating agent acts as electrophile, but on the other it
plays the role of a nucleophilic partner and as such assists in the cyano group hydrolysis. It
follows from the above that the nucleophilic assistance presupposes the primary participation
of nitric acid as a potential carrier of the nitro group, an additional polarization of N-O bond
taking place under the action of the carbon atom of the cyano group.
It should be mentioned that as the nitric acid concentration drops under 70%, the yield
of guanylnitrourea from DCDA decreases whereas the formation of guanylurea prevails
because the corresponding rate of nitration is slower. With time the yield of guanylnitrourea
approaches the quantitative one, i.e. guanylurea is slowly nitrated to the final product
(Fig.8).
.
386
Czech Republic
2006
[Content]
, h.
Fig. 8 The kinetic curve for the accumulation of guanylnitrourea in the course of
DCDA interaction with 75% nitric acid at 25 .
+
NH2NO2
H+
NHNO2
Nu
Nu
+
R
+ NH2NO2
Nu
R
O
11
387
Czech Republic
2006
[Content]
HNO3
wt.%.
Fig.9 Dependence of the logarithms of the pseudo first-order rate constants for the
decomposition of N-nitrocompounds on the nitric acid concentration at 25oC (1nitrourea, 2-guanylnitrourea). The correlation equations are as follows:
for nitrourea:
for guanylurea:
Both processes are subject to acid catalysis which is evidenced by the increase in the
decomposition rate with the nitric acid concentration, though for nitrourea a more flattened
dependence of log Kapp vs. [HNO3] wt.% is observed.
A conclusion may be drawn from the above reasoning that different mechanisms for
the decomposition of these N-nitro compounds are realized (Scheme 6). The decomposition
of nitrourea is accomplished with the assistance of the water molecule at the rate-limiting
stage and, consequently, as the concentration of nitric acid increases, the content of the
reactive protonated form of the N-nitro compounds raises as well, with the simultaneous
decrease in the activity of available nucleophiles in the reaction system. It gives rise to the
fact that the dependence of the decomposition rate on the nitric acid concentration has a
lower slope compared to that for guanylurea where the nucleophile does not take part in the
rate-limiting stage whereas the substrate undergoes a monomolecular decomposition.
NO2 2H+
NH
H2N
NO2
NH2
+
H3N
O
Scheme 6 .
H2O
+
H3NCOOH
NH2NO2
H+
Czech Republic
2006
[Content]
Under the conditions of kinetic experiments the formation of NQ is not observed even
over a continued period of time ( several days) whereas the dilution of the reaction mass with
water down to 85-86% of HNO3 results in the emergence of NQ (Scheme 8).
H2N
H2N
2H+
O2N
NH
N
O
O2N
H2N
H2N
NH
+
C N C
H O
H2N
Scheme 7.
+ NH2NO2
N
H2N
H2N
H2O
H2N
H+
H2N
+
N
H2N
H2N
HNO3
H2N
H2N
C NH2
H2N
Scheme 8.
OH
-CO2
C NH2
H2N
NO2
N
H2N
The stability of guanylisocyanate formed during the nitration of DCDA, besides being a
function of concentration, is significantly affected by the molar excess of HNO3. When the
weight ratio HNO3 : DCDA is lowered under 15, the only isolable product is NQ. It follows
that not only water but also guanylnitrourea can act as a nucleophile in this reaction. This
conclusion is further supported by the results of preparative experiments which made it
possible, depending on the weight ratio HNO3 : DCDA at the constant nitric acid
concentration (taking into account its dilution during the decomposition of guanylnitrourea),
to record the preferable formation either of guanylisocyanate or NQ.
Thus, depending on the experimental conditions (concentrations, the weight ratio of
reactants, temperature), the rate-limiting stage of the reaction may be either the formation of
guanylisocyanate or its decomposition. In relatively diluted solutions of nitric acid (80-90%),
irrespective of the substrate concentration, its complete conversion takes place. The rate
13
389
Czech Republic
2006
[Content]
3. CONCLUSION.
The nitration of guanidine is shown to be catalyzed by lower nitrogen oxides. On the
basis of studies of quantitative regularities of transformation of guanylnitrourea,
guanylisocyanate and nitroguanidine, the probable mechanism of reactions is suggested and
the main factors determining the reactivity of the substrates are addressed. An effective and
environmentally friendly method for NQ preparation amenable to scale-up, has been worked
out.
Acknowledgments
The authors would like to thank the International Science and Technology Center for
giving financial support to this project.
4. EXPERIMENTAL.
The rates of the formation and decomposition of N-nitro compounds were monitored by
means of an aliquot method. A weighed sample of the starting amine was placed into a
temperature controlled cell containing 20-30 ml of aqueous nitric acid. In fixed time
intervals the aliquots of the reaction mass (0.5-0.8 ml) were quenched in 50-100 ml of water
or in a 1-2 molar solution of sodium acetate. The resulting solution was
spectrophotometrically analyzed for the concentration of N-nitro compound. The 265 nm
wavelength was used as an analytical wave for guanylnitrourea and nitroguanidine. Because
of a low spectral resolution the measuments were carried out at two different pH values,
namely 5-6 and 10-11. The molar extinction coefficient was as follows:
for GNU:
= 5-6
= 13500 l.mol-1cm-1
pH= 3-11
= 13500 l.mol-1cm-1
390
Czech Republic
2006
[Content]
,where D1, D2 stand for optical densities at 265 nm at 5-6 and 10-11, respectively.
As a reference solution we used the solution of the quenched aliquot of initial nitric
acid. During the study of nitration in the presence of nitrogen oxides a portion of sodium
nitrite was introduced into the cell before adding the sample of the reagent.
Synthesis of the nitrate of guanylisocyanate.
Guanylnitrourea (1 g) was dissolved at 5-10oC in 25 ml of 100% nitric acid and left at
10 C for 3 hours. The nitric acid was removed in vacuo at the water bath temperature under
20oC. The residue was dried in a vacuum desiccator. The product was obtained in 95% yield
(0.95 g).
o
REFERENCES.
1. Karlsson, S.; stmark, H; Eldsater, C.; Carlsson,T.; Bergman, H.; Wallin, S.; Pettersson, A.
Detonation and Properties of FOX-7 and formulations containing FOX-7.; San Diego, 2002, p
5.
2. Lochert, I. FOX-7- New Insensitive Explosive DSTO, 2001
3. Y. Tang : The effect of NG and TATB to the Insensitivity of HMX; 33 nd International Annual
Conference of ICT, Karsruhe, Germany,2002p.1-7.
4.
D. Powala.; A. Orzechowski .;A.Maranda; J. Nowaczewski. Spherical Nitroguanidine as
component of high explosives.; 7. International Seminar of Pardubice, Czech Republic, p.606612, 2004.
5. Pat. WO 200015584
6. Y. Yamato, A study of nitroguanidine based propellant. The Proceedings of 1997 Autumn
Symposium of Japan Explosives Society (1997), p.53.
7. Pat. EP 992473
8. A.A. Astratev, D.V. Dashko, L.L. Kuznetsov: Synthesis and some properties of 1,2Dinitroguanidine.: Russ. J.Gen.Chem.2003 v.39 4,p.537-548 (in Russian).
9. Yu Yongzhong.; Su Zhuang; Duan Baoru; Chen Fubo: Synthesis of Polynitrocompounds from
Nitroguanidine .; Propellants, Explosives, Pyrotechnics 14,p.150-152, 1989.
10. Davis T.L.: Nitroguanidine L.; J.Org.Syntheses.1927, 7,p.68-69.
11. Davis T.L. Elderfild R.C.: Alkyl-nitroguanidines. Rearrangement and Preparation by nitration;
J.Am. Chem.Soc.1933, 55, p. 731-740.
12. Pat. USA. 288290,1959
13. German Pat .1142860,1963
14. German Pat .2745774,1979
15. Angual S.J., Warbwiton: The basic strengths of methylated guanidines.; J.Chem.Soc.1951,
p.2492.
16. Williams.G., Simkins R.J.J.: The nitration of guanidine in sulfuric acid. II.. Kinetics of Nnitration of the guanidinium ion and denitration of nitroguanidine.; J.Chem.Soc.1953, p.13861392.
17. Moodie R.B., Schofield. K., Willinmson M.J.: Nitro compounds / Pergamon Press. Lon. 1964.
18. A.I.Titov, Nitration of aromatic and unsaturated compounds. Uspekhy Khim. (Advances of
Chemistry), 1958, v.27, p.845-890 (in Russian)
15
391
Czech Republic
2006
[Content]
392
Czech Republic
2006
[Content]
1.
INTRODUCTION
The FAE type munitions, contrary to the conventional munitions, have only fuel as
combustible, and the required oxidizer comes from air by dispersing the fuel into the
surrounding atmosphere. So, dissemination of fuel in air, for all kinds of the FAE type
devices, is necessary, because this process causes to form fuelair mixture. Concentration of
fuel in the fuel-air mixture should be between the lower and the upper explosion limits,
because only in this case, the fuel-air mixture is detonable. The following factors lead to
increase the destructive effects of the detonation of the fuel-air mixtures.
Homogeneity in Fuelair mixture.
Being the concentration of fuel close to its stoichimetric concentration.
Pan-cake shaped of the cloud
1
393
Czech Republic
2006
[Content]
The above objects indicate the importance of the dispersion process in all kinds of the
FAE devices. A typical FAE device consists of a container filled of liquid or finely
powdered mixture and two separate explosive charges. The first charge bursts open the
container and scatters the contents forming an aerosol cloud and then the second charge
detonates the cloud. Dispersion of powdered fuel in air plays a key role on reliability of the
operation of the FAE device. Without designing a suitable system for dissemination of
powdered mixture into the air; achievement on a detonable dust-air mixture is impossible.
Ratio between amounts of fuel and burster charge is called fill-to-burster ratio. It is an
important factor for dispersion process in FAE type devices. Detonable fuel-air cloud is
produced by proper amount of burster charge respect to the amount of fuel which should be
dispersed. Remaly and Ragus researched about fill-to-burster ratio in 1983[1], but they didn't
published their experiments.
2.
The process of powdered substances expansion in air by detonation wave can be divided
into three stages:
1-Particles acceleration
2-Particles deceleration
3-Particles random dispersion
Detonation wave and gases pressure produced by detonation of burster charge,
accelerate fuel particles radially and rupture the wall of cylindrical fuel container .Duration
of this stage is very short about 2 ms. The fuel particles reach their own maximum velocity
u0 at the end of first stage.
When the fuel particles reach maximum velocity u0 and the wall of the fuel container
ruptures, the second stage of fuel dispersion process begins. In this stage, the fuel particles
move radially, but their velocities decrease by air drag force, so that particles radial
movement comes to an end. Therefore, the radius of the dust-air cloud grows in second
stage. After that the fuel particles continue their movement in random direction in third
stage. The uniformity of the dust air cloud is created by the random movement of fuel
particles.
3.
METHOD
394
Czech Republic
2006
[Content]
Fig. 1. Cross section view of test device (with central burster) and its items.
4.
TEST DESIGN
The first step for designing tests was choice of different fill-to-burster ratios as bellow:
1- Fill-to-burster ratio in test No.1 was 100/1
2- Fill-to-burster ratio in test No.2 was 100/2
3- Fill-to-burster ratio in test No.3 was 100/3
The choice of fuel weight is also important, because fuel dispersion process is a
quantitative process. For example, a small cloud is formed by a small FAE device and fuel
particles move a short distance on the average, but a large cloud is formed by a big FAE
device and fuel particles move a long distance contrary to small one. So at least, 4 kg
powdered substance should be used in each test device. Fill-to-burster ratio in tests No.1 and
No. 4 was 100/1, but their scales were different. Aluminum powder with 18 m average
sizes and pressed TNT with 14 mm diameter were used in practical tests. Details of four test
samples are shown in table1.
395
Table 1.
Czech Republic
2006
[Content]
Test No.
The type of
fuel
The weight
of fuel (kg)
1
2
3
4
Al powder
Al powder
Al powder
Al powder
4.050
4.050
4.050
6.750
The size of
fuel
particles
(m)
18
18
18
18
The type of
burster
charge
TNT
TNT
TNT
TNT
The weight
of burster
charge
(g)
40.5
81.0
121.5
67.5
Fill-toburster
ratio
100 / 1
100 / 2
100 / 3
100 / 1
The density of aluminum is 2700 kg/m3, but tapped aluminum powder 18 m density is
1700 kg/m3. Height-to-diameter ratio for test devices was chosen 1.5. Fuel (aluminum
powder) is placed between the walls of the fuel container and burster tube. Fuel placed
volume can be calculated by bellow equation (1)
VF = ( / 4) [ (dFC)2 (dBT)2 ] hFC
(1)
where, dFC and hFC are the diameter and height of fuel container respectively, and dBT is
the diameter of the burster tube. The height of the fuel container and the diameter of the
burster tube respect to the diameter of the fuel container in test devices have ratios as:
hFC = 1.5 dFC
(2)
(3)
If the volume of the required fuel was known, then by using equation (1), (2), and (3),
the amount of dFC , hFC and dBT cloud be determined. The volume of the fuel and the weight
of the fuel have bellow equation:
WF = VF . F
(4)
where, WF , VF and F are the weight , volume , and density of the fuel respectively . The
above equations were used to determine the dimension of the test devices. Table 2 shows the
dimension of test devices which were used in experimental tests.
Table 2. Dimension of test devices
Diameter of
Height of fuel
Diameter of
Volume of
Volume of
fuel container
container
burster tube fuel container
fuel
(mm)
(mm)
(mm)
(cm3)
(cm3)
1
130
195
26
2480*
2380
2
130
195
26
2480*
2380
3
130
195
26
2480*
2380
4
150
225
32
3970*
3790
* The volume of the fuel container should be a little more than the volume of the fuel.
Test No.
5.
TEST ARRANGEMENT
Czech Republic
2006
[Content]
second one is installed 4 m far from it too. Two other ones are installed on the left
symmetrically. Fig.2 shows the arrangement of the test device and distance indicators. The
high-speed camera is set on 1000 frames/sec, and photography begins just stimulation of
detonator by electric source.
6.
TESTS RESULTS
Table 3 shows the radius of dust-air clouds at different times after detonation of burster
charge in tests 1, 2, 3, and 4. Also Table 4 and Table 5 show the average thickness and
volume of dust-air clouds at different times respectively. You can see other tests results in
Table 6.
Fig.3.a, b, c, and d show dust-air cloud at 15 ms after detonation of burster in tests 1, 2,
3, and 4 respectively. The shape of dust-air clouds are pan-cake shaped, but the shape of
cloud in tests 1 and 4 are better. Fortunately, the dust-air clouds exploded in all tests, that
show proper fuel dispersion in air, and the concentration of aluminum powder was in
explosion limits. Explosion of dust-air clouds in tests 2 and 3 were not powerful, but in tests
1 and 4 were powerful. The color of the exploded clouds in tests 2 and 3 were dark, but in
tests 1 and 4 were bright, which indicate that combustion of aluminum powder in tests 2 and
3 was imperfect and in tests 1 and 4 was perfect. In tests1 and 4, the movement of fire was
rapid, uniform and powerful. The burnt effect and shock wave effect on the ground after the
test 4 was manifested. Therefore, fill-to-burster ratio 100/1 is better than other ratio such as
100/2 or 100/3, so the fill-to-burster ratio is close to optimum.
397
Czech Republic
2006
[Content]
398
Time
Test No.
5 ms
10 ms
20 ms
30 ms
40 ms
50 ms
0.55
1.08
2.10
3.00
3.85
4.10
0.90
1.75
3.00
4.10
4.30
1.10
2.15
3.80
4.50
0.55
1.09
2.15
3.15
4.00
4.50
Czech Republic
2006
[Content]
Table 4. Average thickness (m) of dust-air cloud at different times after detonation
of burster charge
Time
Test No.
5 ms
10 ms
20 ms
30 ms
40 ms
50 ms
0.50
0.55
0.70
0.90
1.00
1.10
0.70
0.90
1.05
1.35
1.50
1.60
0.90
1.05
1.30
1.50
1.70
0.60
0.65
0.80
1.00
1.10
1.20
Table 5. Volume (m3) of dust-air cloud at different times after detonation of burster
charge
Time
Test No.
5 ms
10 ms
20 ms
30 ms
40 ms
50 ms
0.47
2.01
9.69
25.43
46.54
58.06
1.78
8.65
29.67
71.26
87.09
3.42
15.24
58.94
95.38
0.57
2.42
11.61
31.16
55.26
76.30
7.
Test No.
The shape of
dust-air cloud
Explosion of
cloud
Power of cloud
explosion
Effects on the
ground
Burnt and
shocked
Excellent
Perfect (%100)
Powerful
Very good
Imperfect (%80)
Weak
No effect
Good
Imperfect (%90)
Weak
No effect
Excellent
Perfect (%100)
Very powerful
Burnt and
shocked
completely
DISCUSSION
Radius of dust-air clouds in Table 3 are used to draw curves of clouds radius versus time
for accomplished tests 1, 2, 3, and 4 in Fig.4. At first, the growth of clouds radius are high,
but then the rate of cloud growth decreases gradually. If burster charge increases, the rate of
cloud expansion increases, but duration of cloud expansion decreases. So, when the fill-toburster ratio changes, also the rate of cloud expansion changes, but the final radius of cloud
changes little.
399
Czech Republic
[Content]
5
4
Series1
Series2
Series3
Cloud radius , m
Cloud radius , m
2006
5
4
Series1
Series4
2
1
20
40
60
80
20
Time , ms
40
60
80
Time , ms
Fig. 4. Fig. 4 The curves of clouds radius vs time. The series 1, 2, 3, and 4 present
the tests No. 1, 2, 3, and 4 respectively
a) The curves which concern to tests No. 1, 2, and 3
b) b) The curves which, concern to tests No. 1, and 4.
When burster is detonated, at first, the volume of cloud is minimal, but then it grows
gradually. When the volume of cloud is minimal, its average fuel concentration is maximal.
So, if the cloud grows, its volume increases, and its average fuel concentration decreases.
When the fuel concentration is close to the stoichiometric concentration, explosion power of
fuel-air cloud would be at most level. Whatever the fuel concentration becomes far from the
stoichiometric concentration, the explosion power would be lower. If the fuel concentration
is more than the upper explosion limit or less than the lower explosion, in other words, if the
fuel concentration is out of explosion limits, the fuel-air cloud can not explode. Therefore,
delay time between detonation of burster charge and second charge is very important. When
the second charge is detonated to initiate the cloud explosion, the concentration of the fuel
should be close to its stoichiometric concentration in air.
90
120
3
Cloud volume , m
80
Series1
60
Series2
Series3
40
3
Cloud volume , m
80
100
70
60
50
Series1
40
Series4
30
20
10
20
0
0
0
0
20
40
60
80
20
40
60
80
Time , ms
Time , ms
Fig. 5. The curves of cloud volume vs time. Series 1, 2, 3, and 4 denote to tests
No. 1, 2, 3, and 4 respectively.
The volumes of clouds are presented in Table 5. Also Fig.5 shows the curves of the
clouds volume versus time after detonation of burster in tests No. 1, 2, 3, and 4. The cloud
volume VCloud in which, aluminum powder concentration receives on its stoichiometric
concentration in air is computed by bellow equations. The stoichiometric concentration of
aluminum powder in air is 330 gr/m3.
CAl =
400
W Al
VCloud
(5)
Czech Republic
2006
[Content]
The weight of fuel (aluminum powder) in test device 1, 2, and 3 were 4.050 kg. So, the
cloud volume in which, the concentration of aluminum powder receives to its stoichiometric
concentration can be computed:
VSt, 1 = VSt, 2 = VSt, 3 = 4.050 / 0.330 = 12.3 m3
Also, the stoichiometric volume of cloud in test No. 4 can be calculated:
VSt, 4 = 6.750 / 0.330 = 20.5 m3
Therefore, when the volume of cloud in tests No. 1, 2, and 3 were 12.3 m3, and in test
No. 4 was 20.5 m3, the average concentrations of aluminum powder have been the
stoichiometric concentration. The times at which, the average concentrations of aluminum
powder in four accomplished practical tests have been the stoichiometric concentration or
close to it, are shown in Table 7.
Table 7. The delay times after burster detonation in which, the experimental tests
reached to stoichiometric concentration.
Test No.
Time
22 ms
11 ms
8 ms
25 ms
The experimental data and photographs of tests (taken by high-speed photography) show
the fill-to-burster ratio 100/1 is closer to the optimum amount than others ratios such as
100/2 and 100/3. In other words, the weight of burster charge respect to the weight of
powder fuel should be %1 or less. Above subject is demonstrated by sufficient reasons:
1- The shapes of the clouds in tests No. 1 and 4 are more pan-cake shaped than others.
The thickness of the clouds in tests No. 1 and 4 is thin which, shows aluminum powder
dispersed radially.
2- The power of cloud explosion in tests 1 and 4 was very high. After the test 4, the test
site was photographed; it shows strong destructive effects on the ground surface.
3- The brightness of cloud explosion in tests 1 and 4 was more than other tests.
4- The colors of clouds after explosion were bright in test 1 and 4, but the colors of
clouds after explosion were dark relatively in tests 2 and 3. The brightness of clouds shows
that all of aluminum powders have combusted completely, but the darkness of clouds after
cloud explosion shows that parts of aluminum powder have combusted imperfectly.
5- The taken film from test 4 shows that dust-air cloud has expanded in two distinct
steps; its reason is that, in first step the dust-air cloud was expanded by detonation wave
produced by burster, and in second step, the cloud was expanded by shock wave produced
by the strong explosion of dust-air cloud. But in other tests especially tests 2 and 3, the
second step of cloud expansion didnt occur.
6- The real double-event FAE devices have two separated charge. The first charge
(burster charge) is used to disperse the fuel in air, and second charge is used to initiate cloud
explosion. The delay time between them is more important. This delay time in real FAE
devices is 20-50 ms [2]. The delay time is a time at which, the fuel concentration in the cloud
is close to stoichiometric concentration. Table 7 shows the time at which, the fuel
concentration is stoichiometry, so these times are proper to the delay time. Appropriate delay
401
Czech Republic
2006
[Content]
times for tests 1, 2, 3, and 4 are 22 ms, 11 ms, 8 ms, and 25 ms respectively. The result is
that the test device 1 and 4 are close to the real types according to their proper delay times.
Therefore, above reasons demonstrate that optimum fill-to-burster ratio was 100/1 in
tests 1 and 4, but their scales were different. If they are investigated exactly, the result will
be that:
1- Optimum fill-to-burster ratio for large FAE devices is 100/1.
2- Optimum fill-to-burster ratio for small FAE devices is 150/1 or 200/1.
Optimum fill-to-burster ratio changes in different scales, its main reason is that not
only the amount of fuel in large scale FAE device is more, but also the average distance
which, is traveled by the fuel particles, is longer. So, if the scale of FAE devices increases,
the fill-to-burster ratio decreases.
8.
CONCLUSION
Fill-to-burster ratio was first subject in this research. The ratios 100/1, 100/2, and 100/3
were chosen in experimental tests. Experiment results show the ratio 100/1 is close to
optimum ratio. Also it is proved that fill-to-burster ratio in large scale devices is a little less
than it in small scale devices, because in large scale device not only amount of fuel
(powdered substance) is more, but also the fuel particles travel longer distance on the
average to from a large cloud. Therefore, optimum fill-to-burster ratio is 200 in small scale,
but it decreases to 100 in large scale. Two thick aluminum or brass circular plates are used in
bottom end and upper end of cylindrical fuel container. They are more effective on cloud
shape, because they prevent dispersion of fuel particles longitudinally, but intensify fuel
dispersion radially; so pan-cake shaped cloud is formed.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
402
Remaly R., and Ragus E., "Explosive and cold gas dissemination of particulate matter", Deepak
Publishing Co., Hampton, Virginia, USA, 1983.
Warren L. McCabe, and Julian C. Smith, "Unit operation of chemical engineering", McGrawhill Book Company, New York, USA, 1967.
Ratzel A. C., and Constantineau E. J., "Aerosol cloud generation experiments", Sandia National
Laboratories Albuquerque, NM, 1993.
Gidaspow D., James L. Austing, "The explosive dissemination of particulate pyrotechnics and
explosive powder", Proceeding of the 13th International Pyrotechnics Seminar, Grand Junction,
Colorado, USA, 1988.
David C. Sayles, "Method of generating single-event, unconfined fuel-air detonation", United
States Patents, No 4463680, 1984.
Tulis A. J., "The role of nitro guanidine in the detonation of aluminum powder-air dispersions",
Proceeding of the 13th International Pyrotechnics Seminar, Karlsruhe, Germany, 1985.
Gidaspow D., James L. Austing, "The explosive dissemination of particulate pyrotechnics and
explosive powder", Proceeding of the 13th International Pyrotechnics Seminar, Grand Junction,
Colorado, USA, 1988.
Xiao Shao-qing, and Bai Chun-hua, "Investigation on control fuel dispersion of FAE for
ignition methods of burster", Beijing Institute of technology, China, 2001.
Austing J. L., Selman J. R., "The initiation shock wave in dispersed reactive particle
detonation. (III) Computation of Rankine-Hugoniot curve from real gas equations of state",
10
Czech Republic
2006
[Content]
Proceeding of the 13th International Pyrotechnics Seminar, Grand Junction, Colorado, USA,
1988.
[10] Labbe J. L. D. S., and M. F. Bardon, "Impulsive dissemination of liquid fuel in air", Paper of
M.S. Thesis, Department of Chemical Engineering, Royal Military College of Canada,
Kingston, Ontario, Canada, April 1991.
11
403
Czech Republic
2006
[Content]
1. INTRODUCTION
Fine and ultra-fine powders are actively used in pyrotechnics, explosives and
propellants. The area of investigations includes metallized solid [1], liquid [2], and metallized
gelled propellants [3]. 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 the agglomeration.
Different methods are developed and used for nanosized metal powder production: laser,
electron- beam, Gen-Miller, electrical explosion, plasma-chemical, and others. The subject
of present study is the investigation of Al powders produced by the Gen-Miller method [4] in
the Institute for Energy Problems of Chemical Physics RAS (INEPCP RAS) and by the
method of electrical explosion [5] in the Institute of High Current Electronics SB RAS (IHCE
SB RAS) and by SIBTERMOCHIM Ltd, (Alex powders [6]).
404
Czech Republic
2006
[Content]
The interaction between metal Al powders and water has been investigated in the
temperature region between room temperature and water boiling point.
2. OBJECTS OF INVESTIGATION
Slurries and dispersions prepared on the base of Al nanopowders and distillated water
were investigated. Polyacrylamide was used as gelling agent. Average size of Al particles
was between 50-250 nm. Besides of widely used Al powders with the surface passivated by
aluminum oxide coatings were investigated the experimental powders passivated by coatings
on the base of aluminum carbide (with admixture of aluminum oxide) [7], aluminum nitride
and the mixture of aluminum carbide and aluminum nitride [8]. The last two types of
passivating coatings are reactive coatings. It means that these coatings can participate at high
temperature in oxidation (burning) processes with heat evolution but they can protect metal
Al from oxidation by atmospheric air at near room temperature. Such type of reactive
coatings on the base of Al carbide was proposed in INECP RAS [9] and successively realized
in INECP RAS and in IHCE SB RAS [7]. The Al carbide is capable to form the gasproof
coatings that are stable and can provide the effective protection of a metal core of aluminum
nanoparticles from the oxidation during contact with atmospheric air at room temperature
and natural humidity. The reaction of Al carbide with hot water leads to both: heat and fuel
gas methane evolution:
Al4C3 + 6H2O = 2Al2O3 + 3CH4 +
Two types of aluminum oxide coatings with hydrophobic properties were investigated in
this work: organosilicon origin [10] (type Al1-6, see Tabl.1), organic origin (types Alex and
AKP-1, see Tabl.1).
405
Czech Republic
2006
[Content]
10
8
7
11
5 4 3
Czech Republic
2006
[Content]
metal Al. It decreases with the increase of the sample temperature. The closed crucible
containing investigated sample was used to minimize the water evaporation intensity. The
cover sealed the crucible nonhermetically.
The presence of water evaporation process makes difficult the quantitative interpretation of
experimental results. Only experiments performed at completely the similar conditions
(shape of crucible, rate of sample heating, mass of sample, velocity of inert gas flow and rate
of sampling flow for hydrogen analyzer) gave the comparable results. The experimental
results were the TG, DTG, DTA curves recording by derivatograph and dependence of
partial pressure of hydrogen in composition of gaseous reaction products evolved during the
sample heating recording by hydrogen analyzer. The special calibration experiments were
performed to calibrate the sensitivity of AH-1 to gaseous hydrogen in these measurements.
Fig.2 shows the scheme of using apparatus.
Samples of slurries or dispersions were investigated immediately after preparation
(manual mixing) and after several (1, 2, 3) days of samples exposure at room temperature in
atmosphere of saturated water vapour. Manual mixing was used to exclude mechanical or
ultrasonic activation of oxidation reaction of metal Al.
Fig.2. The scheme of apparatus
for
investigation
of
reactivity of Al powders
with liquid water down
to 130oC and water vapor
up
to
1000oC
at
atmospheric pressure.
1- inert gas (Ar), 2- gas
reducer, 3-flow meters,
4-thermometer,
5-saturator,
6-Derivatograph-C,
7-analyzer of hydrogen,
8-gas pump.
The experimental results allow determining the temperature ranges in which a failure of
the passivating coating continuousness was arose. Such collapse of coating continuousness
gives possible the interplay between the metal aluminum and liquid water accompanied with
both hydrogen and heat evolutions. The fine structure of maxima of heat and hydrogen
evolution permitted made the conclusion about the mechanism of passivating coating
destruction.
4. RESULTS
4.1. Study of evolution of gaseous reaction products for Al/H2O slurry samples
heated linear in atmosphere of saturated water vapour.
Figs.3 compares the temperature dependencies of a volume of gaseous reaction products
(mainly hydrogen) formed in reaction of Al powder/H2O slurries heated linearly in
atmosphere of the saturated water vapor with the heating rate approximately 5
4
407
Czech Republic
2006
[Content]
degree/minute. The duration of heating was 40 min. The samples composition was 30-32
mass percent of Al nanopowder and 68 -70 mass percent of water gel (including 3 % of
polyacrylamide). The surface of all Al powders was passivated using coatings on the base of
aluminum oxide, carbide and nitride. The shape of curves of temperature dependencies
testifies that the reaction between Al particles and gelled water passes at least through two
stages. The start temperature of the first stage of reaction (57-90.5) depends from the
composition of passivating coating.
140
120
Al2-Ox_1
N95
N2(air)
N2_2
Gel
-1
Al-CO2
100
VH2 (cm3)
80
60
40
20
0
-20
20
30
40
50
60
70
80
90
100
T ( C)
The aluminum oxide coatings are more stable with respect to hot water interaction than
the coatings on the base of Al carbide. The hydrophobization of Al particles surface
increases the start temperature of reaction metal core and hot water with respect to the
similar particles without the hydrophobic treatment. The more stable appears the
hydrophobic AKP-1 powder. The second stage of reaction is responsible for hydrogen
evolution at the temperature near 100.
A l- C O 2 - 1
A l- C O 2 - 3
A l- C O 2 - 4
A l- C O 2 - 5
A l- ( C O 2 - C O 2 )
A l- ( N 2 + C O 2 ) / N 2
A l- ( N 2 + C O 2 ) / N 2 ( H 2 O + S t )
H 2O
160
140
120
VH (CM )
100
80
60
40
20
0
-2 0
20
40
60
o
T ( C)
80
100
Fig.4 shows the similar curves for Al powders passivated with coatings on the base of Al
carbide.
408
Czech Republic
2006
[Content]
Dispers
(nm)
Toin(oC)
Tofin(oC)
Almet1
(%)
AlmetO
(%)
Produser
(1)
1*
2*
3
4
5
6**
7
8
9**
10
11
12
13
14
(2)
(3)
(4)
(5)
(6)
Al-CO2-1
72.4+0.1 79.0+1.0 74.1+1.0
Al-CO2-1
72.5+0.2 80.0+1.0 96.0+4.0
100
Al-CO2-3
62.2
70.3
96.5
Al-CO2-4
59.6
69.8
98.8
Al-CO2-5
56.9
70.5
95.0
Al-CO2-5+St
80.0
96.7
85.3
Al-CO2-CO2
160
64.9
72.8
~100
Al(N2CO2)/N2
72.2
81.5
87.0
Al(N2CO2)/N2+St < 200
79.4
86.9
74.0
Al-N2-CO2
72.7
81.9
85.6
Al-N2 air
71.1
78.6
52.3
50-200
Al-N2_2
70.9
77.9
50.4
Al-N2_1
82.4
89.6
43.6
95
71.5
87.6
41.3
(7)
77.2
77.2
77.2
77.2
77.2
77.2
20.0
75.8
75.8
88.0
79.8
88.0
88.0
88.0
15
Al2-OX
88.0
INEPCP
RAS,
Moscow
SIBTER
MOCHIM
Ltd,
256
75.2+0.6 86.0+0.2
58.4+ 6
16
75.0
89.2
100
88.2
17
L-Alex
78.0
90.6
97.3
88.0
18
Al 1-6
143
19
P-1
200
97.5
30.2
(8)
INEPCP
RAS,
Moscow
IHCE
SB RAS,
Tomsk
Tomsk
88.0
INEPCP
RAS,
Moscow
Moscow
* Sample Al-CO2-1 was investigated immediately after slurry preparation (line2) and 45 hours after slurry preparation (line 1).
** water with admixture of stabilizer (inhibitor)
Column (4) temperature of initiation of the first stage of reactions. Column (5)
temperature of finishing of the first stage of reactions. Column (6) percentage of metal Al
reacting with water during experiment (40 min). Column (7) initial percentage of metal Al
in aluminum powder.
Tabl.1 summarizes the data characterizing the reactivity of Al nanopowder / water slurry
samples including the powders covering with:
6
409
Czech Republic
2006
[Content]
The data of tabl.1 show that reactivity of nanopowders passivated with aluminum
carbide coatings is higher than the reactivity of nanopowders passivated with aluminum
nitride and aluminum oxide coatings. The results of gas chromatographic analysis
demonstrate that the amount of Al carbide rises in content of coatings Al-CO2-1 Al-CO2-5
(lines 1-5). It means that the resistance of these coatings to hydrolysis (Toin value) is in direct
proportionality with the amount of admixture of Al oxide in coating content. Percentage of
metal Al reacting with water during the experiments (40 min) for these samples is near
100%.
Sample Al-CO2-1 was investigated immediately after slurry preparation (line 2) and
after 4-5 hours of preliminary exposition of slurry at room temperature in saturated water
vapor (line 1). The distinction of values Almet1 (column 6) for the line1 and for the lines 2
evidences that the reaction between liquid water and Al particles covered with aluminum
carbide is remarkable even at room temperature. The usage of stabilizer (inhibitor) increase
essentially the start temperature of reaction, that decreases the reactivity of Al nanopowder /
water slurry composition (compare lines 5 and 6 for nanopowder Al-CO2-5 producing by
Gen-Miller method and lines 8 and 9 for nanopowder Al(N2CO2)/N2 producing by method
of wire explosion).
The reactivity of coatings on the base of Al oxide is lower than the reactivity of coating
on the base of Al nitride (Toin = ~75oC and ~71oC respectively) with the exception of sample
Al-N2_1 kept during 6 months at normal atmospheric conditions (Toin = ~82.4oC).
Apparently long term contact of this sample with atmospheric air leaded to essential
strengthening of this passivating coating.
Another distinction between Al oxide and Al nitride coatings is the amount of metal Al
reacting during experiments (Almet1). For the powders with coatings on the base of Al
nitride Almet1=~40-50% and for Al oxide coating Almet1=~90% (for powders with the
similar average size of particles). The nature of this phenomena needs in additional
investigation.
The powders passivated by aluminum oxide coatings with hydrophobic properties
appear the higher stability to hydrolysis than the similar powders without hydrophobic
treatment (compare lines 16 and 17 for nanopowders Alex; line 19 for hydrophobic
nanopowder AKP-1 with lines 7-15 for nonhydrophobic powders).
Czech Republic
2006
[Content]
Fig.6 shows the curve of hydrogen evolution in this experiment and demonstrates also
shoulder and maximum of hydrogen partial pressure in the similar temperature region. It
means that the passivating coating for this sample began to collapse at temperatures higher
90oC and this collapse passes through two stages. The collapse of passivating coating is
accompanied by the hydrogen evolution. The value of temperature of main maximum of heat
and hydrogen evolution is 108oC. It is higher the boiling temperature of distillated water
because the polyacrylamide gelling agent was used in this sample. For the dispersions of Al
nanopowders in distillated water the collapse temperature is lower 100oC.
108
0,7
0,6
0,5
0,4
Al-CO2, Fresh
0,3
0,2
97.5
0,1
0,0
-0,1
9
10
11
12
13
14
15
16
17
18
19
Time (min)
Fig.5. TG, DTA, T curves for Al / H2O slurry Fig.6. Time (temperature) dependence of
linear heating in inert argon flow up to
hydrogen evolution for Al / H2O
700oC. Passivating coating reaches the
slurry linear heating in inert argon
collapse at temperatures 97 and 107oC
flow up to 700oC. The hydrogen
(maxima of DTA curve). Al
evolution starts at temperature
nanopowder was passivated with
97.5oC. Hydrogen concentration
aluminum carbide coating (sample Alreaches maximum at 108oC. Al
CO2-1).
nanopowder was passivated with
aluminum carbide coating (sample
Al-CO2-1).
The presence of fine structure (several maximum or shoulders) on the curves of heat and
hydrogen evolution lets to propose the existence of several types of active surface centers
hot points (defects of structure or composition of passivating coatings or something like
that) responsible for the destruction of passivating coatings contacting with water. These
hot points can exist on the surface of dry coatings or can be the result of the coating
interaction with water during the slurries or dispersions preparation and storage. The heating
of nanopowders in aqueous medium increases the rate of hot point formation. This process
results in the collapse of surface films which make possible the oxidation of metal Al
accompanied by heat and hydrogen evolution. Hot points are different for the passivating
coatings having the different origin or composition and for Al particles having the different
manner of production.
Contact of Al particles with water during several hours even at room temperature
changes the properties of passivating coatings and results also in decrease of metal content in
the powder.
Fig.7. shows the time (temperature) dependences of hydrogen evolution for Al powder
heated in fresh prepared water slurry composition (black curve) and for the same sample
8
411
Czech Republic
2006
[Content]
after 1 day ageing at room temperature (red curve) in atmosphere of saturated water vapor.
The ageing decreases the temperatures characterizing the collapse of passivating coating and
decrease the amount of evolving hydrogen. The origin of passivating coating is Al oxide.
121
0,08
ASD-4, fresh
ASD-4, old
113 C
0,06
0,04
0,02
99 C
160
0,00
150
9
10
11
12
13
14
15
16
Time (min)
77 C
85 C
0,5
0,5
Al-CO2-5, fresh
0,4
0,4
Al-CO2-5, old
0,3
0,2
o
0,1
90 C
58 C
68 C
0,3
0,2
82 C
70.7 C
58 C
33 C
0,1
0,0
0,0
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Time (min)
10
12
14
16
18
20
22
24
26
28
Time (min)
412
Czech Republic
2006
[Content]
Figs.9 and 10 demonstrate the time dependences of Al metal contents for Al nanopowder
Alex ageing in water slurry composition. Fig.9 shows data for hydrophobic Alex in
distillated water slurry. Even 3-4 days contact of nanopowder with distillated water
decreases essentially its metal content. Fig.10 shows the dependence of metal content of
nonhydrophobic Alex in distillated water slurry + inhibitor. It is clear that admixture of
inhibitor increases the value of Alex powders life time by a factor of 5. The metal Al
content was measured by means of X-ray structural analysis. It is necessary to note that the
stability of hydrophobic Alex with respect to contact with liquid water (without inhibitor)
is substantially higher (on the factor 3) the same characteristic of nonhydrophobic Alex
powder.
Al (meta)%l
Al (metal) %
100
105
100
90
95
80
90
85
70
80
60
75
50
70
65
40
60
55
30
0
10
Time (days)
50
0
9 10 11 12 13 14 15 16 17 18 19 20
Time (days)
Fig.9. Time dependence (days) of metal Al Fig.10. Time dependence (days) of metal Al
content of Alex (nonhydrophobic) /
content of Alex (hydrophobic) / H2O
slurry sample (distilleted H2O)
H2O slurry sample (distilleted H2O +
ageing at room temperature in
inhibitor) ageing at room temperature
saturated water vapor atmosphere.
in saturated water vapor atmosphere.
0,8
0,4
PH (mm Hg/gr)
0,6
0,2
0,0
20
30
40
50
60
70
80
90
o
T ( C)
10
413
Czech Republic
2006
[Content]
CONCLUSION
1. The interaction between Al nanopowders producing by different ways and distillated
water in slurry and dispersion compositions was investigated for the temperature region
between room temperature and water boiling point. The surface of Al particles was
passivated with coatings having the various origin and composition: Al oxide, Al nitride, Al
carbide and its mixture. Al oxide coatings were nonhydrophobic and hydrophobic.
2. The special technique was developed for comparable express analysis characterizing
the activity of different types of nanopowders with respect to water or water slurries.
3. More stable to the storage in water slurry or in water dispersion composition are
nanopowders passivated hydrophobic coatings on the base of Al oxide. The powders covered
with reactive Al carbide coatings are more active with respect to water or water slurries.
4. The manners of stabilization Al nanopowders in water slurry and water dispersion
composition at near room temperature were recognized.
Acknowledgment:
This work was performed with financial support by INTAS (grant 03-53-5203) and by RFBR
(grant 05-08-17928).
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
414
Czech Republic
2006
[Content]
36th International Annual Conference of ICT & 32nd International Pyrotechnics Seminar, ICT,
Karlsruhe, Germany, June 28 - July 1, p 110(1-11), 2005.
[8] VALENTIN SEDOI, VICTOR IVANOV, YURI IVANOV, MIRSLAN OSMONOLIEV,
MIKHAIL LARICHEV, AND OLGA LARICHEVA: Synthesis by the exploding wire method
and characterization of nanoaluminum powder with inorganic coating, Pros. of 36th
International Annual Conference of ICT & 32nd International Pyrotechnics Seminar, ICT,
Karlsruhe, Germany, June 28 - July 1, p 98 (1-11), 2005.
[9] M.N.LARITCHEV, A.N.JIGATCH, I.O.LEIPUNSKY, M.L.KUSKOV, A.A.PSHECHENKOV: Air
oxidation of aluminum nano particles. Influence of size and surface coating of particles on the
mechanism of aluminium oxidation process, Pros. of 9 International Workshop on Combustion
and Propulsion Novel energetic materials and applications, September 14-18, 2003, La
Spezia, Italy.
[10] JIGATCH A.N., LEIPUNSKY I.O., KUSKOV M.L., PSHECHENKOV P.A., BEREZKINA N.G.,
LARITCHEV M.N., KRASOVSKY V.G.: Synthesis of coatings on the surface of ultra fine
aluminum particles, Khimicheskaya Fizika (in Russian)//, 2002, v.21, N4, P.72-78.
[11] Hydrogen Analyzer (AH-1, AlfaBassence), Russia: http://www.alfabassens.ru
12
415
Czech Republic
2006
[Content]
Abstract:
The aim of this work was to apply multiwalled carbon nanotubes (MWCNTs) as a
catalyst support in composite Co/TiO2/C electrocatalysts for hydrogen evolution. In
order to elucidate the effect of MWCNTs, corresponding catalysts deposited on
traditional carbon material Vulcan XC-72 were prepared as well.
Structural characterization was performed by SEM, IR and XPS method. It is shown that
intrinsic catalytic activity of catalysts deposited on both MWCNTs and Vulcan XC-72
are almost the same.
It was shown that the surface area of the catalysts deposited on MWCNTs is twice
higher than the one on classical Vulcan XC-72 substrate.
Electrochemical characterization was performed by cyclic voltammetry and steady-state
galvanostatic methods in alkaline solution (3,5 M KOH). Introduction of MWCNTs was
shown to increase considerably catalysts activity. The overpotential (at i = 20 mAcm-2)
is for 45 mV lower than corresponding one with Vulcan XC-72.
These results are valuable per se and also are good indicators for further improvement
of this type of catalysts.
Key words:
composite catalyst, multiwalled carbon nanotubes (MWCNTs),
hydrogen evolution reaction (HER), Co, TiO2
1.
INTRODUCTION
416
Czech Republic
2006
[Content]
It is well known that there are two approaches to improve the catalysts: i) increase of
intrinsic activity by rising of interaction between hypo and hyper d-phases and ii) developing
of real surface area of the electrode.
In the previous works[3,4], the effect of increasing of hypo-hyper d-interaction was
studied involving anatase instead amorphous TiO2 as a hypo phase. The aim of this work is
to determine the effect of developing catalysts surface area by employing multiwalled
carbon nanotubes (MWCNTs) as a carbon substrate instead Vulcan XC-72.
Iijimas discovery in 1991[5] has initiated intensive research on carbon nanotubes
(CNTs). Their unique and superior physical, mechanical and electrical properties allow wide
range of application as e.g. hydrogen or energy storage, electrode material in electrochemical
capacitors, field emission electron source, fillers in polymer composites etc. The
extraordinary conductivity, mechanical stability and inner geometry make as well CNTs
promising material as catalyst support replacing traditionally carbon blacks.
2.
EXPERIMENTAL
The catalysts were prepared by sol-gel procedure described elsewhere[3,6]. Two catalysts
of the same composition (10%Co + 18%TiO2) were prepared, but on different carbon
substrate (MWCNTs and Vulcan XC-72, resp.). The first one was deposited on Vulcan XC72 as a catalyst support, while the second one on multiwalled carbon nanotubes (MWCNTs).
Structural analysis was preformed by scanning electron microscopy (SEM), x-ray
photoelectron spectroscopy (XPS) and Fourie transformed infrared spectroscopy (FTIR).
The real vs. geometric surface area ratio was determined by cyclic voltammetry as
described elsewhere[4]. Polarization characteristics were measured by steady-state
galvanostatic method using gas-diffusion working electrodes[4] in alkaline solution (3,5 M
KOH). Reference electrode was Hg/HgO, while as counter electrode platinum wire was
used.
3.
In Fig 1, SEM photographies of both catalysts are shown. The catalysts particles
deposited onto Vulcan XC-72 are of a spherical shape and grouped in clusters sized of 100200 nm, so that good adherence was achieved. There is a number of holes between the
aggregates on the catalysts surface, higher specific surface area, and consequently higher
catalytic activity is created due to these features.
The catalysts particles grafted on MWCNTs are grouped in smaller clusters than those
on Vulcan XC-72. This causes holes between particles to appear. Due to intrinsic
geometrical features of MWCNTs (empty cylinders with higher developed surface area as
compared to the spherical particles of Vulcan XC-72), they possess inner holes so that inner
porosity of MWCNTs is considerably higher than the one of Vulcan XC-72. This is
important for better dispersion of other catalysts components of the on the surface,
especially of active catalytic centers metallic particles.
SEM analysis is not able to give any information on the nature of bonding (mechanical
or chemical) of different particles into the catalysts clusters. This information can be
obtained by further spectroscopic investigation, by e.g. infrared spectroscopy or X-ray
photoelectron spectroscopy (XPS), as shown bellow.
2
417
Czech Republic
2006
a)
[Content]
b)
Ti2p3/2
458,9 eV
Co2p3/2
470
781,5 eV
Ti2p1/2
464,7 eV
465
460
455
Binding energy, eV
Co2p1/2
Sat.
810
797,2 eV
800
Sat.
790
780
770
Binding energy, eV
a)
b)
4+
Fig. 2. XPS spectra of a) Ti (Ti2p1/2 and Ti2p3/2) and b) Co2+ (Co2p1/2 and
Co2p3/2), taken at the catalysts surface
The values of maxima of Co2p1/2 and Co2p3/2 peaks (797,2 and 781,5 eV) indicate
existence of cobalt as Co(OH)2 phase on the catalyst surface. The peaks are spread and with
418
Czech Republic
2006
[Content]
low intensity, what suggest that amount of cobalt into elemental state is considerably lower
than in valence state (Co2+). Based on the determined values of the binding energies (Fig. 2),
it is very probable that, next to Co(OH)2, Co2+ also exists in small amount of CoO and
CoTiO3. This is supported with the satellite peaks on 787,3 and 803,3 eV, which is a
consequence of paramagnetism of CoO phase.
XPS analysis also indicates that some interaction between the hypo and hyper dcomponents takes place.
On the FTIR spectra of investigated catalysts the only band of interest originates by
TiO2[7]. To determine hypo-hyper d-interaction one has to compare values of maximum of
TiO2 band incorporated in the electrocatalysts and maximum of the band from pure TiO2
obtained in identical conditions as that in electrocatalyst. The shift of the maximum value is
almost the same for both compared catalysts. This shift is 55 to 57 cm-1 related to the
maximum of pure TiO2 (Table 1). So, it shows that there is an intrinsic interaction between
hypo oxide and hyper d-metallic phase into the catalysts. The level of hypo-hyper dinteraction is the same in both investigated systems.
Table 1. Change of wave number, double layer capacity, Cdl and ratio of real vs.
geometrical surface area, SR/SG
Catalyst
wave number/cm-1
Cdl/mFcm-2
SR/SG
57
153,7
2560
55
396,00
6600
-1
0.003
1 mVs
-1
2 mVs
0.002
-1
5 mVs
-1
8 mVs
0.001
-1
i, Acm
-2
10 mVs
0.000
-0.001
-0.002
-0.003
-0.22
-0.20
-0.18
-0.16
-0.14
-0.12
E vs. Hg/HgO, V
Fig. 3. Cyclic voltammograms in the potential region of double layer charging for
different scan rates (10%Co + 18%TiO2 + Vulcan XC-72)
419
Czech Republic
2006
[Content]
In order to compare the surface roughness (the ratio of real vs. geometrical surface area,
SR/SG), the capacities of double layer (Cdl) of the matched catalysts were determined by
cyclic voltammetry in the potential region where only charging of double layer occurs. Scan
rates of 1 to 10 mVs-1 were applied. Current density of double layer charging icap was
adopted to be as an average value of anodic and cathodic current densities at potential at
middle of the scanned interval (Fig.3.) Linear dependence of the icap vs. scan rate (
E
) was
t
determined with a slope equal to the value of double layer capacity of the electrode surface:
C dl =
di cap
(1)
E
d
dt
The real vs. geometric surface area ratio was determined as a ratio between Cdl of the
catalyst versus double layer capacity of ideally flat oxide surface (60 Fcm-2) The values of
double layer capacity and the ratio of real versus geometrical surface area of investigated
catalysts are shown in Table 1. As can be seen, the surface area of both catalysts is highly
developed, but the one on MWCNTs is higher 2,5 times (see Table 1). Also, Cdl value of
MWCNTs are almost twice higher than that of Vulcan XC-72 (Cdl(MWCNT) = 331 mFcm-2
versus Cdl(Vulc.XC-72) = 179 mFcm-2), what means that real surface area of MWCNTs is twice
then the surface of Vulcan XC-72 [8]. So, carbon nanotubes are the crucial factor which
contributes for developing of the surface area of the catalyst as a whole.
-50
, mV
-100
-150
-200
-250
-300
-350
-10
10
20
30
40
50
60
70
80
90
i, mAcm-2
420
Czech Republic
2006
[Content]
4.
CONCLUSION
At least three main conclusions could be drawn:
1)
2)
This effect is due to both high developed surface area as result of obtained
nanostructured components of the catalysts and the strong metal-support
interaction between hypo and hyper d- electronic phases, which is equal for both
investigated systems.
3)
ACKNOWLEDGEMENT
This paper has been supported by and carried out within the EU Project
PROMETHEAS PL ICA2-2001-10037 and Project 13-1650/4-02 of Ministry of Education
and Science of R. Macedonia.
The authors are grateful and indebted to Professor Milan Jaki, University of Belgrade,
for his bountiful and extensive help in the literature supply and valuable advices for both
theoretical consideration and practical aspects of catalysts preparation.
Special thank to the staff of the Institute of Electrochemistry and Energy Systems,
Bulgarian Academy of Science, Sofia for the fruitful collaboration and the hospitality in the
Academys laboratories during investigation.
421
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
422
Czech Republic
2006
[Content]
Introduction
The values of the study on the infinitesimal substance under the load of sliding detonation
have been recognized. Many scientists have taken efforts on this aspect and many useful
semi-empirical equations have been presented, for example the Taylor equation deduced by
G.T.Taylor [1] in 1942 and improved by P.C.Chou [2]. The steady-state two-dimension
solution of the sliding detonation products flowing has been discussed in reference [3], and the
performance of the metal plate under the press of the sliding detonation has been studied in
reference [4-8]. However, there are few reports on the deformation of the metal plate attached
to the cover under sliding detonation.
If the cover is settled on the interface between the metal plate surface and the explosive,
and the explosive initiated at one side point, the deformation degree of the plate under sliding
detonation is weaker than that contacts with explosive directly. In the same way as the covers
distributed on the large apex angle liner or ball-type liner homogenously, the EFP with fins
can be produced and the stability and the attack ability of EFP are improved. So the
investigation on the metal performance loaded by sliding shock wave is a far-reaching work.
Since the process of the metal under sliding detonation is very complex, it is difficult to
gain the analysis solution as yet. This paper employs the polar curve to deduce the parameters
of the shock wave. The laws of the shock wave refracting and reflecting between the metal and
the cover has been used to analyze the deformation of the metal plate, which will provide the
theoretical proof for the cover design.
423
Czech Republic
2006
[Content]
In the shaped charge field, the model of planar flyer plate under detonation driving is
usually applied to solving the problem of liner element acceleration. So using shaped charge in
experiment is suitable to prove the affection of the thickness of cover to the deformation of
metal plate.
In this paper we analyzed the influence of the covers on the metal plate under sliding
detonation and examined it through experiments.
Theory and methods
1. Theoretical model of the shock wave parameters
The phase of detonation wave propagating determines the reflected and refracted shock
wave parameters. In most applications, the detonation of explosives usually progresses in its
characteristic form that can be well explained by Chapman-Jouguet (C-J) detonation theory
[9]. In Fig. 1, O is the contact point of the detonation wave with the substance interface, which
moves together with the detonation front wave with the same velocity as the initial particle
velocity; the detonation wave front OA contacts with the cover; OC is the shock wave front in
the cover. The interface between the explosive and the cover inflects to OB. Shock wave front
reaches the interface CC obliquely and the refraction and reflection accompanying with this
process and the shock wave front CD has been shaped. CC inflects to the position of CE. If
wave impedance of substance 2 is larger than that of substance 3, rarefaction wave will be
reflected; otherwise shock wave will be reflected [10].
1.1 Analysis of shock wave parameters.
The state solution of the detonation products is expressed by formula (1).
p = A (1a)
k +1
0
A=
k
0 D12
k +1
(1b)
where 0 is the density of the explosive, k is the polytropic index and D1 is the detonation
velocity of the explosive. The expanding process of the detonation products is in the
Prandtl-Mayer shape. The detonation parameters of section 1 can be expressed in following
formulas (2) and (3) [6].
(3)
where P1 is the pressure after the wave and Ma1 is Mach number. Detonation products in
section 2 expand in Prandtl-Mayer shape. Equation (4) indicates the relationship between the
detonation pressure of 1 and that of 3.
424
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
k 1
k 1
Ma21 + 1
p3 2
=
p 1 k 1
Ma132 + 1
2
(4)
And the detonation polar curve of section 2 has been shown as equation (5) [11].
k +1
k 1
arctan
Ma132 1 arctan Ma132 1
k +1
k 1
2 =
k +1
k +1
arctan
Ma132 1 arctan MaI21 1
1
k
(5)
The substance in the section II0 and II1 obeys mass conservation and momentum
conservation and shock compress empirical laws. The equation (6) is obtained based on
momentum conservation and shock compress empirical laws.
p 3 =
0 D12 sin 2
b2
a2
sin 2 (6)
D1
where a 2 and b2 are the shock compress parameters. Equation (7) can be acquired from
equation (4), (5) and (6).
2
3
M 12 +
(k + 1) 0 sin 2
b 0
2
k 1
a
sin 2 2
D1
k 1
k
2
(7a)
k 1
k +1
k +1
k 1
k 1
arctan
( Ma23 1) arctan Ma23 1
arctan
(Ma21 1) arctan Ma21 1
k +1
k +1
k 1
k 1
(7b)
a2
1
tan 2
D1 sin 2
= arctan
a2 2
tan 2
b2 + b2 1 +
D1 sin 2
The unknown parameters M 3 and 2 can be obtained by using equation (7), and then
p 3 is acquired using equation (6).
425
Czech Republic
2006
[Content]
As Fig. 1, front shock wave in direction of OC impacts the surface of metal plate.
Assuming that the thickness of the metal plate has no change, and that the air friction is
ignorant, the metal plate collapses under the shock p ' 1 .
The pressure loading on the surface of the metal plate influences the plate velocity and
collapse angle. Fig. 2 presents the collapse process of the metal plate under shock. And Fig. 3
shows the acceleration process of the metal plate under shock.
OC denotes the front shock wave profile that impacts the surface of metal plate. Since the
thickness of the explosive is limited, the sparse wave from free surface of the explosive will
make the pressure of the detonation products decreasing. As the detonation pressure p 1
decreases to p '1 , the inflection angle of the metal plate changes from 0 to . When the
detonation pressure approaches to 0, the inflection angle of the metal reaches the maximum
value as shown in equation (8) [12].
p '1
= 1
8
p 1
max
where is inflection angle of the metal, max is the maximum value. Based on equation (4)
and (8), equation (9) can be obtained.
k
k 1
k 1
2
2 Ma 1 + 1
p3 =
)] (9)
[ p 1 (1
max
k 1 Ma132 + 1
2
(10)
426
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
system. The pressure is independent with coordinate system, so the velocity u is equal to
u 30 [13]. According to the Newton second law, the relationship between the collapse angle of
k 1
k 1
Ma21 + 1
d 2
1
=
)] exp( / cos 2 ) (12)
m hm u 30
0 D12 (1
[
dt k 1
k +1
max
Ma132 + 1
2
where m is the density of the metal, hm is the thickness of the plate and u 30 is the
particle velocity. The free expansion angle of the detonation products can be obtained from
Prandtl-Mayer law [14], as shown in
k +1
1)
k 1
(13)
Richter et al have deduced that the expanding angle equals the maximum collapse angle
max [16].
The processing of the metal plate collapse causes the displacement dy in the direction y
can be obtained from equation
k
k 1 2 k 1
Ma1 +1
1
2 d
D2 (1
2mhmu30
sin cos = 2
[
)]exp( / cos2 ) (14)
k 1 2 k + 1 0 1
dy
2
max
Ma13 +1
2
Equations (12) and (14) indicate the influence of detonation parameters, material
properties
and
the
thickness
of
cover
to
the
metal
plate
deformation.
Experiments scheme
The covers were attached to the surface of the large apex angle liner and the central line of
the cover is parallel to the trunk of the liner. The way of the covers distributed on the liner is
shown by Fig.6. Fig. 7 is the section view of the shaped charge. Three experiments had been
conducted. The apex angle of each liner experiment was 125, the wall thickness was 3mm,
and the diameter was 50mm. The height of each shaped charge was 55mm. In experiment
No.1the charge density was 1.1g/cm3; in experiment No.2 and No.3 the shaped charge density
was 1.4g/cm3. And the cases of the shaped charges were PVC. The covers were all made by
aluminum LY12 2024 . And the shape of the covers was same. The number of the covers
used in each experiment was four. The thickness of the covers was different. No.1 the
thickness of the covers was 1mm, and No.2 was 2.5mm, and No.3 was 5mm. The recovery
equipment included an iron barrel and leak-tight materials and water and protection plate.
5
427
Czech Republic
2006
[Content]
Fig.7 describes the recovery equipment. In three experiments each shaped charge was
detonated on the center point and the standoff of the each experiment is 50mm, as shown in
Fig. 7. The EFP with fins can be recovered in water.
EFP fins have been presented by picture8. The experimental results show that the liner
element attached to the cover has less deformation than the element directly contacted with the
explosive has. The state of the fins of each EFP changes with the variety of the thickness of the
covers.
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
weaken more. The thicker the covers are the deformation degree of the liner element attached
to covers is less and deformation comparison between liner element attached to covers and
that without covers is more obvious. So shaped charges constructed by the same liner but
different thickness of the covers produce EFP with different fins.
Conclusions
We have shown that the thickness of the covers is a factor on the fins of the EFP. The
cover material property is another factor that affects the deformation of the liner. From former
theory analysis, the bigger is the attenuation coefficient, the smaller is detonation pressure
loaded on the liner element and final deformation. We can further perform experiments to
examine the influence of the material to the deformation of EFP fins.
Reference:
[1]
[2]
Chou P C,Carleone J, Hirsch E, el al. Improved formulas for velocity, acceleration, and
projection angle of explosively driven liners [J].Propellants, Explosive, Pyrotechnics, 1983, 8:
175-183.
[3]
Shao Binghuang, Zhang Dengxia, Li Guohao. The movement of the metal plate under sliding
detonation [J]. Explosion and Impact, 1986 6 1 24-29.
[4]
[5]
[6]
[7]
[8]
Shushan Wang, Yanmei Zhang. Initiation of the cover explosive under sliding detonation [J].
Acta Armamentarii, 1999 20 3 233-237.
[9]
Fickett, W., Davis W.C., Detonation, University of California Press, Berkeley,CA, 1979, 14.
[10] Shouzhong Zhang. Explosion and Impact dynamics [M]. Beijing: Ordnance industry publication,
1993.
[11] The compiling group of the Explosion and Explosion Effect, Explosion and Explosion Effect (A)
[M]. Beijing: National defense publication, 1979.
[12] Zheming Zhen, Zhensheng Yang. Explosion machining [M]. Beijing:
publication, 1981, 342-343.
National defense
[13] Chengsen Qin, Yi Liu. The projection angle equation of flyer plate under detonation driving [J].
Explosion and Impact, 2005 25 1 1-5.
[14] Chengwei Sun. Application detonation physics [M]. Beijing: National defense publication,
2000:610-612.
429
Czech Republic
2006
[Content]
[15] Taylor G I. Analysis of the explosion of a long cylindrical bomb detonated at one end [A].
Scientific papers of G. I. Taylor, Vol III[C]. Cambridge; Cambridge University Press, 1963: 277
- 286.
[16] Richter H. On the theory of shaped charges: Motion of thin liners of plastic material on the
surface of a plane explosive (in German) [ R] . Note Technique ISL n6a/ 48 , 1948.
[17] Chou P C, Carleone J, Hirsch E, et al. Improved formulas for velocity, acceleration, and
projection angle of explosively driven liners [J] . Propellants, Explosives, Pyrotechnics, 1983,
[8]175- 183.
[18] Binghuang SHAO , Dengxia ZHANG , Guohao L I , et al. Motion of the flyer plate under
glancing detonation[J ] .Explosion and Shock Waves , 1985 ,5 (3),1-12.
Table1 The known parameters of explosive, cover and metal plate
D1
(mm/s
7.8
a2
3.0
2.785
5.328
30
b2
(g/cm3) km/s
(g/cm3)
1.4
a3
b3
(g/cm3)
1.338
8.93
mm-1
3.91 1.51
0.0583
430
p1 GPa
1 g/cm
2 (rad)
2 (rad)
15.47
3.21
0.067
0.658
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
Figure Captions
Figure 1. D1- detonation velocity; OA-detonation front wave; OC-shock front wave in the
cover; C-intersection point of the shock wave with cover and the metal plate; 2 - angle
between OC and tangential direction surface of the cover; OB- interface between explosive
and inflecting cover surface; 2 - angle between OB and original surface of cover;
CC-original surface of the metal plate; CD- shock wave in metal plate; 3-angle between CC
and the CD; 3- angle between CC and the CE.
Figure 2.
explosive and cover; S-arc length of the plate; dS-infinitesimal arc length; -collapse angle
of the plate; d -infinitesimal collapse angle; R-radius of collapse round; P-pressure on the
metal plate.
Figure 3. O-original point moving with shock wave; y-direction of the normal line of primal
metal plate surface; u-vector velocity of point O; n-normal line of the curved metal plate
surface; v-vector velocity of the metal plate in static ordinate; q-metal particle velocity in
moving coordinate system.
Figure 4 . Figure a is front view of liner; figure b is top view of liner. 1-liner ; 2-cover;3-
Figure 6 The distance between the shaped charge and the recovery equipment is 50mm.
Figure 7. Picture (a) is experiment result of liner with 1mm thick covers; picture (b) is
experiment result of liner wiht5mm thick covers; picture(c) is experiment results of liner with
2.5mm thick covers. Parts of fins far from the central line of EFP are the deformation results of
the liner elements under covers.
Figure 8. Line1. Calculation result of without cover metal plate; Line 2.Calculation result of
metal plate with 2.5mm thick covers; Line3. Calculation result of metal plate with 3mm thick
covers; Line4 Calculation result of metal plate with 5mm thick coves.
Figure 9. Line1. Calculation result of metal plate without cover; Line2. Calculation result of
metal plate with 2.5mm thick cover; Line3. Calculation result of mental plate with 1mm thick
cover.
431
Czech Republic
2006
[Content]
Figure 1 Sliding detonation loaded on the cover and the metal plate
u11
u0=D1
u13
explosive
cover
u20=D1
o'
u21
B
u22
C
2
3
3 u
31
u30=D1
30 metal plate
C'
E
(31)
D
explosive
D1
o'
u20=D1
cover
C
S
R
dS
d
u
c
n
q
u
432
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
(a)
1
(b)
11
433
Czech Republic
2006
[Content]
water
barrel
inside lining
protect plate
434
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
(a)
(b)
(c)
Figure 8. The relationship of the time and bent angle of the red copper attached by
the Covers with different thickness
13
435
Czech Republic
2006
[Content]
Figure 9. Relationship of time and displace towards y of red copper with different thick covers
436
Sliding Shock Wave Causing Deformation of the Metal Plate Attached to the . . .
Czech Republic
2006
[Content]
437
Posters
Czech Republic
2006
[Content]
Abstract:
The paper deals with numerical 3D simulation, through the code LS-DYNA, of the
effects caused by coffee powder air explosion on a coffee mill cover. The procedure
applied follows strongly ATEX directives. From the analysis, it is concluded that the
current cover coffee mill design is not appropriate, the cover sealing failure occurs at
time around only 17 ms after the explosion when the loading overpressure was only
10% of the maximum overpressure value. On the base of this analysis, it has been
concluded that the 3D numerical simulations using LS-DYNA have been a very powerful
and useful tool for the design of a new coffee mill covers. However, in more, the
analysis performed illustrates quite general approach that can be used successfully for
many industrial explosion accidents.
Keywords:
1.
INTRODUCTION
1)
2)
451
Czech Republic
2006
[Content]
ATEX 137 Directive coming into force also in Czech Republic. Among others ATEX
137 requires employers to put mitigation measures in place. These measures should be
consistent with the risk assessment and appropriate to the nature of the activity or operation
and include:
The contribution concerns the description of LS-DYNA calculations used for design of
an explosion resistant coffee mill.
2.
PROBLEM DESCRIPTION
Potential explosive mixture, coffee powder air, exists in coffee mill industrial
machines. The best mitigation measures against this explosion are measures that contain the
explosion within the machine frame housing so that no explosion effects enter the
surrounding space.
However, usually the frame housing is enough robust but has various manipulation
handling openings covered by housing covers. Generally, these covers are weak parts of a
coffee mill and must be designed to withstand loading caused by possible coffee powder
air explosion.
In this paper, these problems are solved through two consecutive steps:
During the first step, experimental research is conducted to establish the parameters
of the coffee powder air explosion, especially the curve: overpressure respect to the
time.
In the second step, the covers response to this explosion curve is computed. When
the cover cannot contain the explosion inside the machine, a new covers design must
be proposed and verified.
In the following, the explosion curve used in this paper is introduced and a typical
housing cover is described. A typical explosion curve is introduced in fig 1. This curve was
obtained by means of experiments performed in the University of Pardubice, Institute of
Energetic Materials, Pardubice and it was used for the analysis presented.
452
Fig 1.
Czech Republic
2006
[Content]
The housing cover analyzed consists of the cap, two rotation locks, and a hinge
connected, see figs 2 and 3. The lamella is inserted behind the frame housing and the hinge
is connected to the cap through seven rivets. The hinge has a rear inclined part that is
supported by the angling edge (45 deg) of the frame housing.
Fig 2.
453
Fig 3.
Czech Republic
2006
[Content]
A detail of the lock is introduced on fig 4. The lock consists of a lamella, nut, shaft and
front covering, see fig 4.
lamella
shaft
nut
covering
Fig 4.
454
3.
Czech Republic
2006
[Content]
Due to the cover symmetry, only one half of real cover was modeled. For the cover and
lock materials, the Johnson-Cook constitutive model was used. For numerical simulations,
the 3D code LS-DYNA [1] with Lagrangian computational mesh was used.
The cover cap was divided into two parts: loaded part and non-loaded part (outside the
opening), see fig 5. The fig 6 presents the numerical mesh used shell elements of
Belytschko-Tsay type were used (5 integration points across the shell thickness 4 mm).
Similarly, in fig 7, shell elements of the same type for connected hinge are introduced
(thickness 2.5 mm).
In fig 8, solid elements (brick elements) used for the lock nut, shaft, and covering are
illustrated.
In fig 9, shell elements used for the rotation formed lamella (thickness 1 mm) are
introduced. The same shell element type was used as for the cap.
Fig 5.
455
456
Czech Republic
Fig 6.
Fig 7.
2006
[Content]
Fig 8.
Czech Republic
Fig 9.
2006
[Content]
For the cover construction, two materials were used: hard aluminum cap, connected
hinge, steel lock with lamella. For both materials, the constitutive model of Johnson-Cook
was used:
Y = (A + B n )(1.0 + C ln ( * ))(1.0 T *m )
T* =
(1)
(T Tr )
(2)
(Tmelt Tr )
Where Y is the yielding stress, is the plastic strain, * is the strain rate, T is the current
temperature, Tr is the initial temperature and Tmelt is the melting point.
The equation of state has the form:
Peqs = K ln
0
(3)
A = 360 MPa
Tr = 300 K
B = 317 MPa
Tmelt = 775 K
C = 0.019
0 = 2700 kg/m3
n = 0.289
m = 1.00
K = 60 GPa
457
A = 200 MPa
Tr = 300 K
B = 184 MPa
Tmelt = 1800 K
C = 0.006
0 = 7870 kg/m3
n = 0.30
m = 0.53
Czech Republic
2006
[Content]
K = 175 GPa
For numerical simulations, it was necessary to respect all possible contacts between
materials.
The cap part outside the housing opening can enter into contact with the outer part of the
frame housing. This contact was simulated as a contact of the cap with a stonewall defined
by the external part of the housing. Similarly, the contact of the connected hinge with the
angling housing edge was simulated through a stonewall.
The hinge connection with the cap through the rivets was simulated by the contact no.6
of the LS-DYNA (tied hinge nodes to the cap surface).
The contact lamella housing was modeled through a stonewall.
The contact covering-cap was simulated through the contact no.2 of LS-DYNA (tied cap
nodes to the covering surface).
4.
RESULTS
The numerical simulation was performed only up to the real process time of 30 ms,
because even at this time the cover deformations were so high that the cover could not be
able to fulfill its function.
The computation was performed using mm/kg/ms unit system, thus the pressure is in
GPa and forces are in kN.
The dynamics of the cover sealing failure was as follows. It is well described by figs 10
to 13 that illustrate displacement isolines in (in mm) z coordinate, i.e. the displacement into
the outer coffee mill space.
Fig 10.
458
Fig 11.
Fig 12.
Displacement at time 21 ms
Czech Republic
Fig 13.
2006
[Content]
Displacement at time 30 ms
At first, (figs 10 and 11) the displacement is higher for the upper cap part due to the
lamellas deformation.
However, the connected hinge in the lower cap part is sliding along the angling housing
edge. At time around 17 ms, it is outside the frame housing, the lower cap part is free and is
rotating around the lamellas (figs 12 and 13). The maximum displacement is around 77 mm,
i.e. completely outside the frame housing.
The stresses of the cap are below the yielding stress and their distributions are in figs 14
to 17. Initially, stress concentrations near locks and rivets are evident. Then, the stress is
concentrated naturally near locks.
Fig 14.
Stress at time 7 ms
Fig 15.
Stress at time 14 ms
459
Fig 16.
Stress at time 21 ms
Czech Republic
Fig 17.
2006
[Content]
Stress at time 30 ms
As mentioned before, the cap and the hinge did not show plastic strains. On the other
hand, the locks suffered a very small plastic strains, see figs 18 to 21. The maximum plastic
strain at time 30 ms is around 0.0008.
Fig 18.
460
Fig 19.
10
Fig 20.
Czech Republic
Fig 21.
2006
[Content]
As regards the velocities along the coordinate z for individual cover parts, the fig 22
presents their histories: 1 loaded cap, 2 non-loaded cap, 3 hinge 4 rotating lamella.
The hinge has the maximum velocity because it is placed in the lower cover part: up to
10 m/s at time 30 ms. Then, the cap parts follow with the velocity around 4 m/s at time 30
ms. The lamella has the velocity around 2 m/s because it serves as a joint.
11
461
Czech Republic
2006
[Content]
In fig 24, contact force time history for the second mentioned contact is presented. This
contact force is a little lowered after time 17 ms because the cover is rotating around the
lamellas. Then, the force is increasing due to higher loading pressure (at this time however
this loading is not realistic, the pressure must be decreased due to the cover sealing failure).
5.
CONCLUSIONS
The cover dynamic analysis has shown that the current cover coffee mill design was not
appropriate, the cover sealing failure occurred at time around only 17 ms when the loading
overpressure was only 10% of the maximum overpressure value.
For this reason, a new cover design was proposed that would enable to mitigate the
explosion inside the coffee mill: the cover cap made of steel (thickness around 10 mm) and
462
12
Czech Republic
2006
[Content]
provided in the lower part by a rotating welded joint and in the upper part connected with the
robust housing through two welded bolts with female screw (made of steel, too).
The analysis performed illustrates quite general approach that can be used successfully
for many industrial explosion accidents.
The work was performed as a part of the MSM Project no. 0021627501.
REFERENCES
[1]
[2]
13
463
Czech Republic
2006
[Content]
1.
INTRODUCTION
Conversion high explosives (sensitizer) extracted from ammunition (RDX, HMX and
manufactured phlegmatized RDX and HMX) according to disarmament policy are utilized
for preparing of water gel-like explosives (WGE) for different branches of industry. Gel-like
base (phlegmatizer) of such systems is water solution of ammonium and sodium nitrates that
thickens by poliacrylamide (1%). Solid high explosives are mixed with gel-like base. The
role of water is polyfunctional one: (1) it creates consistence that is convenient for loading of
explosives into a bore-hole, (2) it permits to obtain a solution of sodium and ammonium
nitrates that are able to react in detonation wave in such state most efficiently [1]. At last (3)
water is very effective and inexpensive phlegmatizer.
The main goal of this work is estimation of real explosion hazard of water gel-like
explosives that content high explosives. Estimation of the real explosion hazard of such
systems is important task for users. As usual for solution of the problem it is not enough to
investigate only sensitivity to impact, and complex method including estimation of ability to
burning, detonability and tendency deflagration to detonation transition (DDT) must be
applied.
1
KozakGD@nm.ru
464
2.
Czech Republic
2006
[Content]
EXPERIMENTAL.
2.1
Preparation of WGE.
Content, %
Ammonium nitrate
50
Sodium nitrate
20
Water
30
Polyacrilamide
1*
Potassium phosphate
0,1*
Density, g/cm3
1,41
* in excess of 100%
2.2
Impact sensitivity.
Investigation of impact sensitivity was carried out according the sensitivity test
method[2,3] based on the State Standards GOST 4545-80, and GOST 4545-88 that was
officially approved in Russia. Tools N1 application (Fig.1) with the sample of WGE placed
on anvil of vertical fallhammer machine K-44-II. Falling weight was M = 10 kg, its height of
falling down was H = 25 cm, the mass of the sample was m = 0.05 g. The frequency of
explosions is estimated in 25 runs.
Fig 1.
2
465
466
Czech Republic
2006
[Content]
Fig 2.
Fig 3.
Fig 4.
Fig 5.
2.3
Czech Republic
2006
[Content]
Detonability.
2.4
Burning.
Investigation of water influence on burning of high explosives was begun in work [7],
some later one of us with co-authors investigated burning of WGE in bomb of constant
pressure in a large scale of pressure [8]. This data were utilized at discussion of explosion
hazard of WGE.
2.5
DDT.
As to DDT, the same situation took place: influence of phlegmatizer content (paraffin)
on distance of DDT of RDX, and HMX was thoroughly investigated in one of woks of the
laboratory of B.N. Kondrikov in Mendeleev University [9] and these data was successfully
used in work [10] at examination of explosion hazard questions in industry. DDT distance was
measured in steel closed thick-walled (=13 mm) tubes, canal of them (d=10 mm) was
wholly filled with a sample of explosive. Ignition was carried out from one of the ends of
tube by means of a sample of pyrotechnic composition. When DDT took place the tube
destroyed starting with some distance from igniter. This distance was distance of DDT.
3.
EXPERIMENTAL RESULTS.
The results, that were obtained at investigation of impact sensitivity of WGE on a base
of RDX , are presented in Fig. 6.
100
f, %
90
80
70
60
50
40
30
20
10
0
Fig 6.
10
20
30
40
Cg, %
50
60
70
80
467
Czech Republic
2006
[Content]
20
df , mm
18
16
14
12
10
8
6
4
2
0
20
40
60
80
Cg, %
Fig 7.
468
Czech Republic
2006
[Content]
df , mm
20
18
16
14
12
10
8
6
4
2
10
20
30
40
50
60
70
80
Cg, %
Fig 8.
U, mm / s
30
25
20
15
10
5
P, MP
0
0
Fig 9.
10
20
30
469
Czech Republic
2006
[Content]
explosive
TNT
1
2
3
4
20
TNPh
TNR
15
Hexyl
TNB
Tetryl
DINA
PENT, RDX,
HMX
10
Fig 10. The distance (L) of DDT vs. content of phlegmatizer (paraffin) (1 HMX,
2 RDX, 3 manufactured phlegmatized RDX, 4 - manufactured
phlegmatized HMX). Names of explosives, to that correspond the distance
of DDT of phlegmatized composition, are written near right axiss (TNPh
pycric acid, TNR threenitroresorcinol, Hexyl hexanitridiphenylamine,
TNB threenitrobenzene).
4.
DISCUSSION.
WGE are the systems that fill halfway place between solid and liquid explosives. The
explosion hazard of liquid and solid explosives is estimated at experimental and calculating
investigation of some parameters. It is important to know complex of the parameters: ability
to burning without access of air, detonability, and ability to DDT. Such approach is
necessary condition of industrial safety in the branch of manufacture and treatment of
explosives and dangerous goods. Applying of such approach was demonstrated on example
of organic peroxides in works [11-13].
As one can see, the data of fallhammer machine test does not attract real explosion
hazard of the composition of WGE. It would be seen the lowest one at content of gel-like
phlegmatizer Cg=2.5-5% according to this test (Fig. 6), but investigations of abilities to
detonation, to burning, and to DDT (Fig. 10) contradict to this estimation. All of these
abilities, that determine the real explosion danger, are lower at much more content of gel-like
phlegmatizer. As it was shown at microscope investigation (Fig. 11), the reason of lower
impact sensitivity of WGE at Cg=2.5-5% was in gluing of particles i.e. in enlargement of its
size. As it was described in work [2] impact sensitivity of powdery systems (WGE at
Cg=2.5-5% was powdery, that was slightly wetted by gel-like phlegmatizer) diminished.
470
Czech Republic
2006
[Content]
Growth of Cg changes WGE into gel-like system that can flow, heat, and ignite in tough
conditions of fallhammer machine test, and frequency of explosion enlarges.
In real conditions of WGE employment impact on it leads to free flow, that is not
accompanied by heating. Much more dangerous situation is fortuitous inflammation and
DDT of explosive that occurs much more easy in a case powders than in a case of real WGE
at greater content of Cg.
5.
CONCLUSION.
The attempts to use WGE at gel-like phlegmatizer content Cg=2.5-5%, that are based
on the results only of fallhammer machine test, can lead to dangerous situations and to
incidents. They deprive WGE theirs main advantage over powder explosive: reduced
inflammability, and lower ability to DDT.
ACKNOWLEDGEMENTS
Authors are grateful to Foundation of Basic Research of RF, that supported the
investigation through grant 04-03-32713.
471
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
472
Czech Republic
2006
[Content]
Abstract
The differential scanning calorimetry method was applied last year to compare
explosion hazard of nitrocompounds and untraditional explosives peroxides. In this
work the literature data for c-nitrocompounds heat of melting (Hmelt) were compiled
and analysed. It was found that the rule: (H/T)melt=54.412.6 kJ/mol K is kept for cnitrocompounls, however, there is tendency to correlation (H/T)melt=f(M), where M is
molecular mass. Brilliant linear correlation of these values was found for nnitrocompounds (dimethylnitranine, RDX, and HMX) and o-nitrocompounds at M=75
110. The differential scanning calorimetry method was applied for investigation of heat
of melting of TNT, PETN, and ethylenedinitramine (EDN). Our values of HmeltTNT
coincided with literature data. Firstly received data
HmeltPETNand HmeltEDN
supplemented linear correlation of literature data (H/T)melt=f(M), for nitroesters and
nitramines, correspondingly.
Keywords: differential scanning calorimetry, n-nitrocompounds, o-nitrocompounds, cnitrocompounds, heat of melting.
1.
INTRODUCTION.
The method of differential scanning calorimetry (DSC) was applied in work [1], that was
presented at previous Seminar, to study exo- and endothermic effects, that took place while
heating of benzoil peroxide and cumene hydroperoxide (CH) in comparison with traditional
explosives TNT and PETN and insensitive explosive NTO.
The DSC method [2] is based on simultaneous heating (or cooling) of sample
investigating in comparison with standard one. The given speed was regulated by means of
computer of the arrangement, measurement of the compensating heat flux was supported
temperature of a sample within the limits of the given program. The DSC experimental
curves are dependencies of a heat flux (millijoules per second) versus temperature (in
isothermal runs -versus time). They permit to register the temperature and the heat of
melting, the temperature of very decomposition starting with great accuracy. Unfortunately
the heat of decomposition could not be measured with sufficient accuracy because of heat
dissipation at gas formation. However the heat of fusion of samples this method registers to
a high degree of accuracy. By the way this method is suitable to check of solubility one
substance in another by means of measuring of melting point of a mixture.
KozakGD@nm.ru
473
2.
Czech Republic
2006
[Content]
EXPERIMENTAL.
In this work all experiment were carried out by means of Mettler Toledo DSC 30
calorimeter. Mass of a sample was 1-1.5 mg.
TNT was chosen as there were sure calorimetric data heat of melting (HTNTmelt) in
literature [3,4]. Melting enthalpy of TNT, obtained in our experiment, was coincided precisely
with literature DSC data [3] HTNTmelt = 103 J/g. This coinciding gave an opportunity to
judge about reliability of our measurements. According to earlier sources [4] this value was
92.3 J/g but it was rather close to enthalpy obtained by means of DSC method.
Two another objects of investigation were nitroester PETN and nitramine
ethylenediamine dinitrate (EDNA). The samples of explosives were refined by means of
recrystallization. Melting points of them were measured during DSC experiments.
3.
EXPERIMENTAL RESULTS.
Experimental calorimetric curves for TNT, PETN and EDNA are presented in Fig 1 ac.
There is considerable interval of temperature between endothermic effect of melting and
exothermal effect of decomposition in the case of TNT. This interval is smaller in the case of
PETN, and it is absent in the case of EDNA, that begins to decompose at once after melting.
Experimental calorimetric curves for mixture of EDNA (CEDNA=12.5 %) and TNT is
shown in Fig 1 d. One can see two endothermic effects of melting: the first one corresponds
to melting of TNT, i.e. melting point coincides with melting point of net TNT (Fig 1a).
Second one is closed to melting point of EDNA. Contrary to TNT mixtures with DNT or
PETN [4,5] melting point of TNT mixture with EDNA does not change in comparison with
TNT. In other words, solution of EDNA in TNT does not take place. Properties of this
mixture were investigated in connection with works [5-10], that dealt with investigation of
explosive systems capable to spin pulsating detonation. The heat of explosion of such
systems was Qv=5.3-5.4 MJ/kg, and calculated detonation velocity was D= 7.37.4 km/s.
Calculated Qv and D of the mixture TNT/EDNA 87.5/12.5 had the same values, and spin
pulsating detonation of cast charges of it was observed in experiments. The photograph of
witness-plate is presented in Fig. 2.
Results of measuring of the enthalpy of melting of nitrocompounds and data that were
found in literature are collected in Table 1.
4.
DISCUSSION
The analysis of the references dealing with melting enthalpy calorimetric data
measurements [3,4,11-13] and our data are shown in Fig 35.
It is known [14] that ratio (H/T)melt is the constant value (Rule of Valden), it is
(H/T)melt=54.412.6 kJ/mol K for organic substances. The values (H/T)melt for cnitrocompounds are plotted versus molecular mass (M) in Fig. 3. One can see that this rule is
kept in common for them at M= 60-290, all of the points settle between dash lines that are
the frame 12.6 kJ/mol K , however the tendency to growth of (H/T)melt with molecular
mass arising takes place.
474
10
dH/dt, mW
Czech Republic
10
2006
[Content]
dH/dt, mW
0
0
-5
-5
-10
T,C
-10
50
100
150
200
250
300
350
T,C
400
-15
50
12
15 dH/dt, mW
100
150
200
250
dH/dt, mW
300
10
10
6
4
-5
0
-10
-2
-15
-4
-6
-20
o
-25
60
T, C
80
100
120
140
160
180
200
220
240
-8
50
T, C
100
150
200
250
475
Czech Republic
2006
[Content]
Hmelt
Explosive
M,
Tmelt,
g/mol
J/g
(H/T)melt,
Refe-
J/mol
J/mol K
rence
Nitromethane
61.04
244.8
158.9
9700
39.6
[11]
Tetriyl
287.1
402.6
79.9
22930
56.9
[12]
Pycric
229
394
74.7
17100
43.4
[11]
182
343.3
110.5
20120
58.6
[11]
352
103
23490
66.7
[11]
354
92.2
20920
59.1
[4]
354
103
23430
66.2
Our
acid
4
2,4 DNT
5
TNT
227
data
Nitrobebzene
123.1
278.8
99.2
12200
43.8
[11]
Ethylnitrate
91.07
178.6
93.66
8529
47.76
[15]
Izo-Propyl-
105.09
190.9
96.1
10099
52.9
[15]
nitrate
9
Methylnitrate
77.04
190.2
106.9
8235
43.3
[11]
10
PETN
316
414
171
54036
130.5
Our
data
11
RDX
222
478.2
169.6
37660
78.8
[11]
12
HMX
296.2
553.2
235.9
69873
126
[13]
13
Dimethyl-
90.08
331
43.29
3899
11.78
[15]
150.09
449
190
28512
63.5
Our
nitramine
14
EDNA
data
15
476
DINA
240.13
325.5
208.2
49995
153.6
[11]
Czech Republic
2006
[Content]
100
(H/T)melt, J/mol K
90
80
70
60
50
40
30
20
10
M, g/mol
0
0
50
100
150
200
250
300
350
200
180
(1).
(H/T)melt, J/mol K
15
160
140
120
12
100
80
14
11
60
40
20
0
50
13
100
M, g/mol
150
200
250
300
350
477
Czech Republic
2006
[Content]
DINA (point 15), that is nitramine and nitroester at the same time, does not conform
to correlative dependence (1).
Correlation between (H/T)melt and M in the case of nitroesters : methylnitrate (MN) [11],
ethylnitrate (EN) [15], izo-propylnitrate (izo-PN) [15] and PETN (data of this work) is shown
in Fig. 5. Correlation is brilliant one in spite of M of PETN in 36 times greater than M of,
MN, EN, and izo-PN. Correlative equation is:
(
H/T)melt = 14.6+0.37 M J/mol K
200
(2).
(H/T)melt, J/mol K
180
160
15
140
10
120
100
80
60
8
9
40
7
M, g/mol
20
0
50
100
150
200
250
300
350
400
5.
CONCLUSION.
The correlation between ratio (H/T)melt and molecular mass of nitrocompounds were
found on a base of it measuring by means of differential scanning calorimetry. If in the case
of c-nitrocompounds it is only a tendency to correlation it is quit definite one in a case of
nitramines and it is even greater definite one in a case of nitroesters. It was shown by means
of analyzing of differential calorimetric curves, that ethylenedinitramine (CEDNA= 12.5%) is
insoluble in TNT, spin pulsating detonation of this mixture was observed experimentally, the
heat of explosion of such system was Qv=5.3. MJ/kg, and calculated detonation velocity was
478
Czech Republic
2006
[Content]
D=7.4 km/s. This result justifies the conclusion of works [5,8-10] in that it was shown:
calculated detonation parameters of all of the systems, that were capable to this process, had
such values.
ACKNOWLEDGEMENTS
Authors are grateful to Foundation of Basic Research of RF, that supported the
investigation through grant 04-03-32713. Our students: I.G. Chumakova and T.V.
Masterkova took part in carrying out of some experiments.
REFERENCES
[1]
[9]
KOZAK G. D., LIN Z.: Spin pulsating detonation of cast charges. Trans. of the conf.
(Int). The theory and practice of energetic materials. V. 5, Science press. Beijing-New
York. 2003, pp.404-407.
[10] KOZAK, N.V., KOZAK G.D., LIN Z.: The spin-pulsating regime of detonation in solid and
liquid explosives. Proceedings of the 8th seminar New trends in research of energetic
materials, Univ. Pardubice, Czech. Republic, April, 2005, p.600-605.
[11] DOMALSKI E.S. and HEARING E.D. I. Phys. Chem. Ref. Data, 1996. (Internet.
479
Czech Republic
2006
[Content]
1.
INTRODUCTION
NH
NNO2
ANRTZNQ
N N
NNO2
H2N
2.
SYNTHESIS
ANRTZNQ can be easily and with a high yield synthesized from the 3,5-diamino-1,2,4triazole and 2-methyl-1-nitroisothiourea [4, 5]:
480
H2N
Czech Republic
2006
[Content]
NH2 HNO3
N N
H2 S
NNO2
95%
O4
H2N
H2N
HNO3
SO 4
NNO2
H2N
N
N NH
NH2
CH3S
NH2
H2N
/H 2
O3
HN
NH2
NNO2
N N
NNO2
~100%
H2N
ANRTZNQ
N N
CH3SH
NH
NNO2
91%
H2N
3.
ANRTZNQ is practically insoluble in water and usual organic solvents. It does not allow
to grow good crystals for X-ray analysis. Attempts to grow monocrystal samples of
necessary quality by slow crystallization from solutions of a nitric acid have not resulted to
success.
Therefore we investigated the structure of ANRTZNQ only by calculation methods of
quantum chemistry. In the article the results ab initio HF/6-31G calculations are presented.
The GAMESS package was chosen as the calculation tool. Calculated molecular structure is
shown in Figures 1, bond lengths and angles are represented in Table 1. The calculated
structure of ANRTZNQ corresponds to the structure of other nitrimines, experimentally
studied earlier nitroguanidine derivatives and nitrimino-1,2,4-triazoles [7-9]. As in latter, the
nitrimine fragments and 1,2,4-triazole cycle are plane. Because of delocalization of electron density bond lengths CN, NN and NO have intermediate values between the
values characteristic for corresponding single and double bonds (Table 1). Like other
nitrimines, the planar conformation of ANRTZNQ is stabilized by intramolecular hydrogen
bonds (Figure 1, Table 2).
O1
O2
H5
H1
N4
N3
H4
N6
C2
N5
C1
N2
N1
H3
N9
N7
C3
N8
H2
O3
O4
Fig 1.
481
Czech Republic
2006
[Content]
1.4207
C2N1
1.3577
N6O1
1.2542
N2C1
1.2887
N5N6
1.3570
N6O2
1.2086
C1N3
1.3416
N1C3
1.4012
N8O3
1.2543
C1N4
1.3825
C3N7
1.3080
N8O4
1.2103
N4C2
1.3516
N7N8
1.3603
C2N5
1.3133
C3N9
1.3216
N1N2C1
104.1
N4C2N5
132.0
C3N7N8
120.4
N2C1N3
125.9
C2N5N6
119.3
N5N6O1
120.4
N2C1N4
111.7
C2N1C3
129.3
N5N6O2
116.1
N3C1N4
122.4
N2N1C3
119.9
O1N6O2
123.5
C1N4C2
108.4
N1C3N9
116.7
N7N8O3
121.7
N4C2N1
105.0
N1C3N7
113.1
N7N8O4
115.7
C2N1N2
110.8
N7C3N9
130.2
O3N8O4
122.6
DH
H...A
D...A
DH...A
N4H1...O1
0.995
1.998
2.5774
114.7
N9H2...O3
0.995
1.920
2.5946
122.5
N9H3...N5
0.998
1.945
2.7023
130.6
More detailed calculated structure data of ANRTZNQ (with HF/6-31G and other levels)
will be considered in the other article.
The UV absorption spectra (Figure 2) were recorded on the Shimadzu UV-1601
spectrophotometer. Because of low solubility of ANRTZNQ the maximal concentration of
solution, with which it was possible to write down a spectrum, makes only 4105 M. The
UV spectrum of ANRTZNQ shows an absorption at wavelengths max1 = 213.5 nm, max2 =
271 nm, and max3 = 327 nm. The potassium salt of ANRTZNQ, which is a bit more soluble
in water, shows completely similar peaks. It proves that ANRTZNQ exists in a water
solution as an anion. The absorbtion peak at wavelengths 271 nm is associated with n *
transition in a nitrimine fragment. This is typical for nitroguanidine derivatives, they usually
have intensive maximum at wavelengths 260290 nm, for example, in a nitroguanidine max2
= 265 nm (Figure 2).
482
Czech Republic
2006
[Content]
NQ
A
1.4
Fig 2.
A N R T Z NQ _ P S
1.2
1
0.8
0.6
A NR T Z NQ
0.4
0.2
0
190
210
230
250
270
290
310
330
350
370
390
UV spectrums of
ANRTZNQ (C=4105 M),
its potassium salt
(C=5105 M) and NQ
(C=104 M) in water
solution (thickness of the
UV cell is 1cm).
nm
The infrared spectra (Figure 3) were obtained on the Nicolet IMPACT-400D FTIR
spectrometer with the IR microscope SpectraTech INSPECT IR+. Table 3 summarizes
tentative assignments for some frequencies by analogy with the assignments for
nitroguanidine [10, 11].
1246
0.75
0.70
0.65
0.60
1316
1432
877
0.10
729
682
0.15
774
995
0.20
949
1069
0.25
1474
1660
3432
0.30
1586
3313
0.35
3163
0.40
1627
3370
1687
0.45
3266
Absorbance
0.50
1497
0.55
0.05
0.00
3500
3000
2500
2000
1500
1000
Wavenumbers (cm-1)
Fig 3.
Table 3.
Assignment
1687m, 1660w
(NH2)
1627m, 1586w
(C=N)
1432m , 1246vs
1069w, 1036w
(N-C-N) or rock(NH2)
949w
twist(NH2)
vs = very strong; s = strong; m = medium; w = weak; vw = very weak
4
483
4.
Czech Republic
2006
[Content]
Vt ,
3
cm /g
200oC
320
240
190oC
o
210 C
160
180oC
80
0
0
40
80
120
160
200
240
t, min
The observed rate constants and activation parameters of thermal decomposition are
shown in Table 4, and Arrhenius temperature dependences are shown in Figure 5.
The main products of the thermal decomposition of ANRTZNQ are N2O (m/e = 44),
HNCO (m/e = 43), HCN (m/e = 27, 26) and NH3 (m/e = 17) (Figure 6). There is NO2 (m/e =
46) at products of the thermal decomposition, however its amount is much less in
comparison with N2O ratio of maxima intensities N2O/NO2 = 6:1. The peak with m/e = 30
corresponds to NO. The latter is generated under the action of electron impact from N2O and
484
Czech Republic
2006
[Content]
NO2, but not as a result of secondary oxidation-reduction reaction passing with the assistance
of NO2, because CO and CO2 peaks, which usually accompany the oxidation processes [14],
were not discovered in primary products of thermal decomposition.
Table 4. Kinetic parameters of ANRTZNQ thermal decomposition in a solid phase
Constant rate k105, s1 (at T, C)
180
190
200
210
220
1.82
6.23
16.7
36.1
85.2
Ea, kJ/mol
lgA
175.9
15.59
0.9975
ln k
-6
-8
-1 0
-1 2
2 .0 0
Fig 6.
2 .1 0
1 0 0 0 / T, K -1
2 .2 0
The composition of the thermal decomposition products allows to assume two main
initial ways of ANRTZNQ thermal decomposition:
6
485
H2N
NNO2
NH
H2N
NNO2
N N
N NH
NNO2
NC-NHNO2
ANRTZ
Czech Republic
H2N
NH
2006
[Content]
H2N
N N
N2O
NC-NHNO2
NNO2
NH
N NH
H2N
m/e=187
H2N
ANRTZNQ
2.5
12
ln[(s)]
16
2.0
1.5
1.0
0.5
0
180
Fig 7.
b)
3.0
a)
190
200
210
Temper at ur e, oC
220
2.00
2.04
2.08
2.12
2.16 2.20
1000/ T, K -1
Ratio between the temperature and time to explosion for ANRTZNQ (a) and plot
of ln vs. 1/T (b)
The values of energy of activation (Ea) and pre-exponential factor (A) are found from
this equation. The calculation is based on the following expression:
486
Czech Republic
2006
[Content]
where cp heat capacity (we used approximate values for heat capacity organic
explosives cp = 1.25 J/(gK)); R ideal gas constant, 8.314 J/(molK); Qexpl. heat
decomposition reaction (J/g), identical with heat of explosion (calculated in the chapter 6).
Calculated activation parameters of thermal explosion are represented below:
Texpl.( =5 c) = 205C; T = 184-216C; Ea = 100.7 kJ/mol; lg A = 7.978.03 (average 8.00)
The obtained values of activation parameters of ANRTZNQ and calculated out of them
rate constants visibly differ from obtained in manometric experiments parameters (Table 4).
This fact may be concerned by both large error of activation parameters values from
explosion temperature test and/or greatly discriminate experimental conditions.
5.
IMPACT SENSITIVITY
The sensitivity to impact of ANRTZ was investigated with K-44-II Russian drop-weight
impact machine [17]. Experimental conditions: 5 kg weight drop hammer; the sample of
explosive is placed in the plunger assembly 1 [16], consisting of two steel rollers with
diameter 10 mm and height 12 mm [16]; samples weight is 30 mg. The results of tests are
collected in Table 5, the curves of sensitivity are shown in Figure 8. Each point was obtained
by results of 25 trials. For comparison the data for PETN are given. Samples of RDX with
weight 50 mg are used as a standard for K-44-II apparatus (f = 488% for H = 25 cm).
Table 5.
10
12
14
15
Explosion probability f, %
24
52
88
96
100
From the obtained data it follows that the sensitivity of ANRTZNQ is higher than
PETN, i.e. this compound is the sensitive dangerous high explosive.
f,%
100
PETN
80
ANRTZNQ
60
RD X 50 mg
40
20
Fig 8.
12
16
20
24
H , cm
Explosion probability (f) for ANRTZNQ vs drop height of load (H); for comparison
the data for PETN are given.
487
Czech Republic
2006
[Content]
The best method of the explosive impact sensitivity determination is a method of critical
pressure of initiation [19], however it requires to use complicated apparatus. To estimate
critical parameters of initiation by impact [19] we have tried to use a numerical method [20] and
activation parameters of thermal decomposition wich we obtained (chapter 4 in this paper).
Calculated value of Pcr. for ANRTZNQ is less then 450 MPa if it hypothetical melting point
is above 350C (compound is decomposed in a solid phase without melting). In calculations
we used activation parameters of thermal decomposition from manometric data for solid
phase and approximate value for heat capacity cp = 1.25 J/(gK). If using activation
parameters from explosion temperature test we could not get reasonable values of Pcr..
Experimental and calculation values of Pcr. for PETN is 480 MPa [19, 20].
6.
NH
H2N
NNO2
O2NN
NNO2
ANRTZNQ
NNO2
N
NO2
N N
H2N
RDX
Brutto-formula
C3H5N9O4
C3H6N6O6
Molecular weight
231.13
222.12
Oxygen balance, %
31.1
21.6
Hf , kJ/mol
0
0, g/cm3
Qexpl., MJ/kg
+171
[3, 4]
1.87
4.63
+61.5
[3]
1.6%
1.70
1.80
5.40
1.70 g/cm3
V0, m3/kg
0.7510.6%
0.7621.70 g/cm3
D, km/s
8.733.4%
8.39
8.77
31.4
36.1
3740
3520
2.83
2.83
PCJ, GPa
32.7
TCJ, K
a)
4.9%
1797 3407
3.379.6%
b)
488
7.
Czech Republic
2006
[Content]
CONCLUSION
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
10
489
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
490
Czech Republic
2006
[Content]
Czech Republic
2006
[Content]
Abstract:
In the standard Shipp-Kaplan batch synthesis of hexanitrostilbene (HNS), the mixing of a
solution of TNT in THF-MeOH with aqueous NaOCl is very exothermic. This causes
considerable problems with scale-up, particularly if the objective is to quench the reaction
at an early stage in order to obtain a good yield of the first intermediate, 2,4,6trinitrobenzyl chloride, with minimum contamination by HNS and HNBB. The use of a
Kenics Static Mixer immersed in ice-water has been used to overcome this problem.
Under these conditions the formation (and precipitation) of HNS is sufficiently delayed to
avoid blocking the mixer and the resultant mixture may be quenched to give trinitrobenzyl
chloride in yield and purity comparable to the material obtained by the Shipp-Kaplan
recipe.
Keywords:
1.
INTRODUCTION
The synthesis of hexanitrostilbene (HNS) by reacting trinitrotoluene (TNT) with NaOCl was first
[1]
[2]
reported by Shipp and Kaplan in 1964 and has been widely studied since . As described by Shipp
and Kaplan, the reaction conditions may be varied so that the isolated product is either HNS (by adding
[1a]
TNT to NaOCl yield normally 40-45%)
, hexanitrobibenzyl (HNBB, by adding NaOCl to TNT
[1b]
yield normally ~80%))
or trinitrobenzyl chloride, the initially formed reaction intermediate (TNBCl,
[1b]
by quenching the reaction mixture in acid a short time after mixing yield normally ~85%)
, (see
reaction scheme in Fig 1). One can envisage that HNS is formed by self condensation of TNBCl
(reaction between TNBCl and its conjugate base B), whilst HNBB is formed by cross condensation of
TNBCl with the conjugate base of TNT (A), the mode of addition generating conditions which favour
one or the other of these two products. In less controlled mixing regimes both products are likely to be
formed.
491
Czech Republic
2006
[Content]
In the batch synthesis of HNS, the reaction occurs in essentially two stages. In stage one (mixing
duration minutes), TNT reacts exothermically with NaOCl to form TNBCl. In stage two (ageing duration ~2h), the TNBCl is allowed to react with the basic medium (NaOCl-NaOH) generating HNS
which separates from the medium. The best selectivity in the reaction is achieved if the temperature on
[2a, 3]
mixing is kept below 15C and the pH during the ageing period is maintained around 10
. Shipp and
[1a]
Kaplan
kept the mixing temperature low by pre-cooling both solutions to 0C, but did not control the
pH during the ageing period.
In scaling up the batch process, one of the major problems is controlling the temperature at the point
of mixing. This is also likely to be a problem if the reaction is performed as a semi-continuous flow
system. Any excess temperature rise on mixing will cause HNS to form prematurely in the mixing tubes
and result in blocking. In order to minimise this problem we have explored the use of a Kenics Static
Mixer immersed in ice-water in the hope that any heat generated on mixing would be rapidly transmitted
through the metal walls of the mixer into the surrounding bath.
2.
EXPERIMENTAL
[4]
The mixer-reaction system employed consisted of a Kenics Static Tube Mixer 37-03-075
[material 316SS, length 7.5 (190mm), OD 3/16 (4.75mm)] with 27 mixing elements (see Fig 2 inset
shows 3 of 27 mixing elements), oriented vertically. The reactant solutions [TNT in THF-MeOH-water
and aqueous NaOCl] were delivered from two syringe pumps [KD Scientific 100 series], via Teflon
tubing [ID 0.030, OD 0.063] to the mixing tee [Anachem U490, material SS, ID 1.00mm, internal
volume 2.2l] and thence into the top end of the static mixer via a short length of SS tube [ID 0.25mm
(0.01), OD 1/16] and a Swagelok reducing union [SS-300-6-1 (3/16 to 1/16), ID 0.05]. The outflow
from the bottom of the static mixer was conveyed by Teflon tubing [ID 2.5mm, OD 4.5mm, length
normally 3m] to the ageing or quenching vessel. Part of the arrangement is shown in Fig 2. Before
commencing a reaction, the whole of the system [inlet tubes (to ensure that the reactant solutions were
delivered at 0C), mixing tee, static mixer, outlet tube] was immersed in ice-water contained in a stirred
Dewar vessel and the inlet tubes were primed to within a few cm of the mixing tee. Because the
combined volume of the static mixer and the outlet tubes was a significant fraction of the total liquid
volume transferred, it was necessary to empty the system after the syringe pumps had delivered their
specified volumes. This was done by replacing both syringes with air-filled syringes and continuing the
delivery at the same rate.
2.1
TNT solution: TNT (2.30g) in THF (18ml), MeOH (9ml), water (3ml) total volume of solution
32ml. The solution (25.0ml) was delivered at 435ml/h; addition time 3.45min. Amount of TNT
delivered 1.80g.
NaOCl solution: the number of equivalents of NaOCl was varied (0.5-1.8) by varying the
concentration of the solution whilst keeping the volume of solution delivered constant at 7.15ml;
delivery rate 124.3ml/h.
The outflow from the system was delivered into a stirred conical flask without cooling. After ageing
for 2h, the product was filtered off, washed with MeOH and then water, and finally dried.
492
The Use of a Kenics Static Mixer for the Reaction of TNT with NaOCl
2.2
Czech Republic
2006
[Content]
In order to give sufficient depth for the glass electrode, these experiments were on twice the scale of
those in (a), but with the same flow rates. All experiments used 1.2 equivalents of NaOCl. After
collecting all of the reaction mixture in a stirred conical flask, the pH of the solution was reduced from
12-13 to the desired value by addition of 25wt% H2SO4 and was then controlled by the addition of 6wt%
NaOH [Radiometer Copenhagen PHM 82 standard pH meter with a TTT titrator]. Product isolation
was as in (a).
2.3
Quenched reactions
The outflow was added to a solution of conc HCl (2.3ml) in water (230ml). The resultant mixture
was stirred for 1h before the solid was isolated and washed with water. The length of the outlet tube was
varied; 1.5m gave a higher TNT content in the product and therefore gave insufficient time for complete
reaction of the TNT.
Reactions on five times the scale (9.43g TNT) were run using peristaltic pumps instead of syringe
pumps; the same flow rates were used.
2.4
Product analysis
HPLC conditions: -Bondapak C-18 10m, 3.9x150mm, mobile phase 52.5% water 47.5%
acetonitrile, flow rate 1.0ml/min. The retention times were: TNT 5.6min, TNBCl 6.4min, HNS 11.6min,
HNBB 13.8min, UNKNOWN 16.4min.
1
H NMR (DMSO-d6) was also used to confirm product identity and distribution :
3.
In the first series of experiments (Table 1), the yield and composition of the product from a fixed
amount of TNT was investigated as the amount of NaOCl was varied from 0.5 to 1.8 equivalents, with
no attempt being made to control the pH in the ageing period. In theory only one equivalent of NaOCl is
required to convert all of the TNT to TNBCl and thence HNS. In practice, the highest material yield does
occur in the region of 1 equivalent of NaOCl, but the product in that case contains a high proportion of
HNBB. The highest proportion of HNS (93-96%) occurs when 1.2-1.4 equivalents of NaOCl are used,
but by this stage the material yield has started to decrease. The proportion of HNBB in the product
steadily decreases from 86% when 0.5 equivalents of NaOCl are used to ~5% when >1.2 equivalents are
used.
493
3.2
Czech Republic
2006
[Content]
pH control
In the second series of experiments (Table 2), all experiments were conducted using 1.2 equivalents
of NaOCl (the amount that gave a low proportion of HNBB but without reducing the material yield
excessively see Table 1), and the pH during the ageing period was controlled by addition of aqueous
H2SO4 and NaOH. Consistently higher material yields were obtained with pH control, compared to the
uncontrolled reaction with 1.2 equivalents of NaOCl. The material yield and composition of the product
was not particularly sensitive to pH in the range 9.5 to 11.5, but slightly purer HNS was obtained in the
[1a]
lower part of this range. This is in agreement with Golding and Hayes observations
from a series of
batch reactions.
An unknown product (<5%) was observed in all of these experiments (and in the uncontrolled
reactions when 1.0-1.2 equivalents of NaOCl were used). There appeared to be a correspondence
between two characteristic 1H NMR signals (an A2X system) and the HPLC peak at RT 16.4min. It is
suggested that this product is either -chlorohexanitrobibenzyl (intermediate C in the reaction scheme,
Fig 1) or more probably its hydrolysis product -hydroxyhexanitrobibenzyl.
3.3
Quenched reactions
One of our main objectives was to be able to prepare TNBCl using a flow reactor. Since the highest
selectivity for HNS formation most probably reflects efficient conversion of TNT to TNBCl, we chose
those reactant proportions that gave the purest HNS viz 1.2-1.4 equivalents of NaOCl, for testing in
quenched reactions (Table 3). Reactions using 1.4 equivalents of NaOCl gave high selectivity (~95%)
and high material yield (~95%) of TNBCl. Indeed these results are similar to those one obtains using the
[1b]
Shipp and Kaplan procedure
viz ~98% purity and 87% material yield. Using these same conditions at
5 times the scale (peristaltic pumps instead of syringe pumps) also gave good results.
4.
CONCLUSIONS
In all experiments where a significant amount of HNS was the final product, the flow system did not
become blocked because further reaction of the intermediate TNBCl, and hence HNS formation, was
delayed by the external cooling until the reaction mixture had left the flow system.
This mixing system gave a convenient way of scaling up the synthesis of HNS and TNBCl on a
laboratory scale, and has the potential for use at even larger scales.
494
The Use of a Kenics Static Mixer for the Reaction of TNT with NaOCl
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
(a) P.GOLDING and G.F.HAYES, A parametric study of the synthesis of 2,2,4,4,6,6hexanitrostilbene from trinitrotoluene and sodium hypochlorite, Propellants, Explosives,
Pyrotechnics, 8, p.35-39, 1983 (b) B.SINGH and H.SINGH, Synthesis of 2,2,4,4,6,6hexanitrostilbene, Defence Science Journal, 31, p.305-308, 1982 (c) B.SINGH and
R.K.MALHOTRA, Hexanitrostilbene and its properties, Defence Science Journal, 33, p.165-176,
1983 (d) E.G.KAYSER, An investigation of the Shipp hexanitrostilbene (HNS) process, J.Energetic
Materials, 1, p.325-348, 1983.
[3]
[4]
From Chemineer Ltd, Cranmer Road, West Meadows, Derby DE21 6XT, UK;
www.chemineer.com
Acknowledgment:
We thank Anjum Agha for HPLC analyses and Eurenco (Sweden) for support.
495
Czech Republic
2006
[Content]
NaOCl
Conc
%wt/vol
Equivalents
4.56
0.5
6.47
7.29
0.6
0.8
9.11
1.0
10.0
10.9
12.8
16.4
1.1
1.2
1.4
1.8
Yield
g*
TNT
TNBCl
HNS
HNBB
Unknown
0.1
2.6
3.3
0.9
0.8
0.7
0.3
0.0
0.0
1.2
3.0
0.0
0.0
0.5
0.5
0.4
0.6
0.0
0.0
2.1
10.6
23.1
9.3
55.9
56.2
67.5
77.1
92.8
95.8
91.0
86.3
73.8
87.4
42.7
41.3
29.8
20.2
6.2
4.2
5.7
0.0
0.0
0.0
0.0
1.2
1.6
1.8
1.0
0.0
0.0
0.60
0.76
0.88
0.74
0.90
0.86
0.80
0.74
0.60
0.35
pH
9.5
10
10.5
11
11.5
TNT
0.4
0.0
0.0
0.0
0.0
0.5
0.6
HPLC analysis
% by peak area
TNBCl
HNS
HNBB
1.2
90.0
3.3
1.1
91.9
4.1
0.9
92.1
3.6
1.3
87.5
7.3
1.0
88.8
6.0
0.8
89.8
4.4
0.5
88.3
6.7
Yield
g*
Unknown
4.1
2.8
3.1
3.9
4.2
4.4
3.8
2.42
1.99
2.31
1.74
1.89
2.14
1.93
496
The Use of a Kenics Static Mixer for the Reaction of TNT with NaOCl
Czech Republic
2006
[Content]
HPLC analysis
% by peak area
Conditions
NaOCl
Exit tube
equivalents
length (m)
With syringe pumps
1.0
3
1.2
1.4
1.5
With peristaltic pumps
1.4
TNT
TNBCl
HNS
6.9
1.4
11.7
5.1
0.3
0.5
1.4
2.6
83.9
98.0
82.9
91.7
99.2
96.1
94.7
95.7
9.2
0.5
5.4
3.2
0.5
1.9
2.0
1.7
3.9
0.2
91.7
97.6
0.8
1.0
Yield
g*
HNBB
1.5
1.9
3.6
1.1
1.54
1.25
1.72
1.85
1.67
2.16
2.20
1.77
10.9**
11.3**
497
NO2
O2N
Czech Republic
2006
[Content]
NO2
CH3
HO
O2N
CH2
NO2
NO2
(A)
(TNT)
NaOCl
NO2
O2N
NO2
CHCl
HO
O2N
NO2
CH2Cl
NO2
(B)
(TNBCl)
O2N
NO2
TNBCl
-Cl-
O2N
CHCl
CH2
NO2
NO2
TNBCl
-Cl-
O2N
CH2
NO2
O2N
NO2
HO
HNS
-HCl
(C)
O2N
CH2
O2N
NO2
(HNBB)
498
The Use of a Kenics Static Mixer for the Reaction of TNT with NaOCl
Czech Republic
2006
[Content]
499
Czech Republic
2006
[Content]
Abstract:
In many laboratories, investigations aimed at replacing diphenylamine (DPA,
propellant stabilizer) with other, less harmful substances are being carried out.
Good effects have been obtained in experiments with triphenylamine (TPA),
which cannot produce owing to its structure carcinogenic nitroso
derivatives. In this work, results of an investigation of the kinetics of TPA
transformation during its nitration with NxOy are presented. The possibility of
application of instrumented thin layer chromatography (TLC) to analyze TPA
and its nitro derivatives is shown. The analytical data obtained are used to
estimate the rate of TPA depletion.
Keywords:
1.
INTRODUCTION
The kinetics of the changes occurring in the currently used stabilizers under influence
of nitrogen oxides emitted in the process of ageing of smokeless propellants is well
known [1]. For example, diphenylamine (DPA) the basic stabilizer of single-base
propellants is initially transformed into N-nitrosodiphenylamine (N-NO-DPA) and next, as
the propellant ageing process advances, 2-nitro-, 4-nitro- and higher nitrated derivatives of
DPA are formed [2]. Since carcinogenic properties of N-NO-DPA are confirmed, its
generation is a drawback for DPA. It should also be mentioned that nitrosamines are formed
during the ageing process of the propellants containing other stabilizers, like centralite,
akardite or 4-nitro-N-methyl aniline, thus they are all suspected of being carcinogenic. This
means that stabilizers which are nowadays used in propellants produce very toxic products.
Taking this into consideration, S. Wilker and coworkers [3] proposed the application of a
new propellant stabilizer triphenylamine (TPA). This compound does not include NH
bonds, so that N-nitroso-compounds cannot be created. The authors have shown that TPA is
an effective stabilizer for nitrocellulose based propellants. Joining this research we
undertook an attempt to develop a methodology of measurements of the reaction kinetics
between TPA and nitrogen oxides. The results of measurements were compared with those
obtained under the same conditions for the well-known stabilizer of smokeless powders
diphenylamine (DPA).
The currently used test methods for the evaluation of propellant stability can be divided
into two categories, depending on the propellant components being analyzed. The first
category includes methods which estimate the degree of decomposition of esters contained
in a representative propellant sample, subjected to heating. The second category includes
500
Czech Republic
2006
[Content]
methods based on an assessment of the stabilizer contents and the determination of the
products of reactions the stabilizer undergoes in the stabilized propellant samples
subjected to natural or accelerated ageing.
Because of the fact that each of the tests is an expensive and time consuming method,
causing especially a lot of trouble while testing new stabilizers, we propose a new, simple,
fast and inexpensive method of evaluation of the stabilizing properties. The method consists
in a direct nitration of a stabilizer adsorbed on a silica gel (so that the stabilizer is dispersed
like in a propellant sample) with nitrogen oxides followed by the qualitative and quantitative
analyses of the formed nitration products.
2.
EXPERIMENTAL
All the analyses were performed by the means of instrumented thin layer
chromatography (TLC). We looked for chromatographic systems and techniques of the
development of chromatograms suitable for separation of the following compounds:
The investigations were also aimed at determination of the analytical parameters of the
compounds, development of a quantitative analysis method and evaluation of the
measurement errors. The samples for analyses were prepared by nitration of the stabilizers
(TPA and DPA) with nitrogen oxides. First, the stabilizers were adsorbed on a silica gel and
then it was subjected to a stream of air/NxOy mixture. The reaction products were eluted with
methylene chloride and analyzed.
The equipment, chemicals and solvents used were: applicator Linomat IV (Camag);
horizontal chromatographic chamber DS (UMCS Lublin); densitometer CS-9000
(Shimadzu); chromatoplates with silica gel 60 F254 HPLC (Merck, No. 1.05548) or silica gel
with chemically bounded octadecyl (Merck, No. 1.05559); solid phase extraction set (J.T.
Baker), stabilizer standards (WIWEB ASt Heimerzheim and Merck), organic solvents (J.T.
Backer and Merck).
Selection of the chromatographic systems. The first step of investigations included
a search for a chromatographic system making the separation of the analytes, their
identification and determination of the analytical parameters possible. The experiments were
conducted using methylene chloride solutions of the standards (TPA, DPA and their nitroderivatives) with concentration of 10 ng/L. Samples of the solutions of 10 L in volume
were applied (spraying technique, band width 4 mm) on the start lines of chromatoplates.
The chromatograms were developed to a length of 510 cm in the horizontal chamber DS.
The spots originating from the nitro-derivatives of TPA and DPA were seen in the visible
light and TPA, DPA and N-NO-DPA were visualized by observation of the fluorescence
quenching. The best separation conditions of the individual groups of analytes (Fig. 1a,b)
were gained in the following chromatographic system:
501
Fig. 1.
Czech Republic
2006
[Content]
Separation of the TPA (a) and DPA derivatives (b), A absorbance in relative units
of the apparatus; z range of the chromatographic band [mm]
Calibration curve,
A = f(c)
Detection
limit
[ng]
200
100
100
100
100
Table 1b.
Analyte
Czech Republic
2006
Calibration curve
Correlation
Range of linearity
coefficient
[ng]
DPA
265
47.7c - 2422
0.990
2500
N-NO-DPA
282
32.7c + 1564
0.988
1800
2-NO2-DPA
430
51.3c + 2047
0.999
2000
4-NO2-DPA
380
0.999
2100
67.5c + 3973
A absorbance, c the mass of an analyte in the chromatographic band [ng]
Fig. 2a.
[Content]
A = f(c)*
Detection
limit
[ng]
150
120
90
80
503
Czech Republic
2006
[Content]
stabilizers (TPA and for comparison DPA) exposed to nitrogen oxides. To this end
a 0.1% (wt.) solution of each stabilizer in methylene chloride was prepared and then it was
mixed with silica gel. The mixture was kept for half an hour at room temperature to reach the
adsorption-desorption equilibrium and retention of the TPA or DPA on the adsorbent. After
that, the mixture was filtered off, the adsorbent was dried and divided into samples of 0.75 g
in mass. The samples were placed into typical SPE cartridges, and from a few cartridges, the
adsorbed stabilizer was eluted with two 1-mL portions of methylene chloride using SPE set.
The masses of eluted stabilizers were determined by analyzing the combined eluates, and in
this way information on the amount of stabilizer that was adsorbed on the silica gel was
obtained.
Through the other cartridges, filled with silica gel containing adsorbed TPA or DPA,
5-mL portions of air/NxOy mixture with increasing concentration of nitrogen oxides were
passed. The stabilizers were then nitrated with a rate depending on the concentration of
NxOy. After finishing the nitration process, remaining stabilizers and their nitration products
were eluted and analyzed. Results of densitometric measurements of the obtained
chromatograms are shown in Figs 2a and 2b.
Fig. 2b.
504
3.
Czech Republic
2006
[Content]
The obtained results of measurements show that the stabilizing properties of TPA are
comparable to that of DPA. This is indicated by the similar rate of depletion of the
stabilizers subjected to nitrogen oxides under the same conditions, Fig. 3.
Fig. 3.
Changes in the amount of unreacted TPA and DPA after their nitration with
nitrogen oxides versus concentration of NxOy in air (c), u relative mass fraction
of an analyte in the sample
seems to be an effective one. As a matter of fact the method cannot be used for a direct
evaluation of the propellant stability, but owing to its simplicity, inexpensiveness and
fastness it is especially preferred for a comparative assessment of the stabilizing performance
of various stabilizers. It makes the observation of direction and kinetics of the changes of
stabilizers possible without the necessity of long-lasting ageing of stabilized samples of
propellants, Fig. 4.
Fig. 4.
Results of nitration of TPA and DPA (adsorbed on silica gel) with nitrogen
oxides, u relative mass fraction of an analyte in the sample, c concentration of
NxOy in air
505
Czech Republic
2006
[Content]
[2]
LUSSIER L.S., GAGNON, H., BOHN M.A., On the chemical reactions of DPA and its
derivatives with NO2 at normal storage temperature condition, Propellants, Explosives,
Pyrotechnics, 25, 117-125 (2000).
[3]
WILKER S., PETRILEK J., SKLDAL J., PANTEL G., STOTTMEISTER L., Stability analyses
of double base propellants containing new stabilizers, Proc. 33rd Int. Ann. Conf. ICT,
109 (112), Karlsruhe, Germany (2002).
506
Czech Republic
2006
[Content]
Abstract:
In this paper some dangerous properties of a composite propellant premix (decomposition
temperature, sensitivity to mechanical stimuli, character of combustion and ability to
detonation) as well as thermal analysis (DTA), burning rate and heat of combustion are
determined. The premix consists of fine-grained ammonium perchlorate, dioctyl adipate and
a binder (PBAN). The results bring to the conclusion that the premix is a dangerous goods.
Keywords:
premix, composite solid rocket propellant, safety, sensitivity and utility
parameters
1.
INTRODUCTION
2.
PERFORMANCE CALCULATIONS
The thermodynamic properties were calculated using ICT Thermodynamic Code. The ICTThermodynamic Code is based on a method developed by the National Aeronautics and Space
Administration (NASA)[4]. Thermodynamic equilibrium was calculated for constant volume
conditions. The calculation of the heat of explosion is of special interest, because the
experimental measurements using a calorimetric bomb are sometimes difficult due to high
temperatures and erosive reaction products. The calculations of physicochemical properties of
the premix are shown in Table 1.
507
Czech Republic
2006
[Content]
Value
Sum Formula:
C
15.2218
27.2935
7.0018
50.2753
Cl
6.5368
3.31
-40.03
Density [g/cm ]
1.562
3.
-2150.13
44.885
3825.7
863.7
508
Czech Republic
2006
DATA
ACQUISITION
MODULE
[Content]
4
5
3
6
2
EXAMINER
1
509
Czech Republic
2006
[Content]
3.2
Value
264
261
367
4174
15.930.12
The adiabatic bomb IKA calorimeter system C 4000 was used to determine the heat of
combustion. Combustion experiments were conducted with an automatically controlled
adiabatic jacket and without oxygen atmosphere (pressure about 400 Pa). The calorimeter was
calibrated by burning standard propellant to determine its effective energy equivalent.
Crawford bomb was used to determine the burning rate of the premix[5]. The measurements
were carried out at temperature 23oC and constant pressure about 40 bar. The results of these
test as well as heat of combustion of the premix are presented in Table 2.
3.3
The BAM Fallhammer and BAM friction apparatus methods, acoording to[8], were used to
measure the impact and friction sensitivity of the premix. The results are shown in table 3.
510
Premix
3.4
Czech Republic
2006
[Content]
Field tests
In Table 3 there are presented the results of the mentioned below tests[8]:
- A.1 test BAM 50/60 steel tube test was used to evaluate the ability to detonation of the
premix initiated by booster; the premix was placed in a steel tube.
- 3 (d) test Small-scale burning test was used to determine the response of the premix to
fire; a sample was placed on a sheet of burning paper.
- 2 (c) (ii) test Internal ignition test was used to determine the violency of explosive
reaction; the premix was placed in a steel tube, closed in both ends and initiated by black
powder.
3.5
4.
CONCLUSIONS
The results presented in the paper have shown that the premix is a kind of a dangerous
goods.
Qualitative results of ability to detonation, ability to ignition and deflagration in
confinement of the tested premix are the similar to the results obtained for ammonium
perchlorate and the tested propellant.
The tested premix characterises smaller impact sensitivity in comparison with pure
ammonium perchlorate and with a hexogen contained propellant. Impact sensitivity of the tested
premix is significantly lower than for fine ammonium perchlorate (fractions below 0.03 mm). It
is probably caused by phlegmatization fine AP grains by the binder.
511
Czech Republic
2006
[Content]
Acknowledgment
This research was supported by the State Committee of Scientific Research through Institute of
Industrial Organic Chemistry, Grant Nr 3 T09B 021 27.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
512
G. P. Sutton and O. Biblarz: Rocket Propulsion elements, John Wiley & Sons, INC, New York,
2001
J. G. Routley: Fire and Explosion at Rocket Fuel Plant Henderson, Nevada, Technical Report
Series, United Fire State Administration, 1988
A. Davenas: Solid Propulsion Technology, Pergamon Press, Oxford, 1993
S.Gordon and B. J. McBride: Computer Program for Calculation of Complex Chemical
Equilibrium Compositions and Applications, NASA Reference Publication. I. Analysis 1311,
October 1994
R. Meyer, J. Khler and A. Homburg: Explosives, Wiley-VCH Verlag GmbH, Weinheim, 2002
T. B. Fedoroff: Encyclopedia of Explosives and Related Items, Picating Arsenal, Dover, NJ, USA,
Vol I, p. XVI, 1960
Polish standard BN-76/6091-08
"Recommendations on the Transport of Dangerous Goods, Manual of Tests and Criteria"
(ST/SG/AC.10/11/Rev.4) United Nations, New York and Geneva, 2003
N. KUBOTA, Propellants and Explosives, Thermochemical Aspects of Combustion, WILEYVCH GmbH, Weinheim, Germany, 2002
Czech Republic
2006
[Content]
Abstract:
Modern solid propellants have found wide application as power sources, from
commercial small gas-generator systems like airbag inflators, to large launch vehicles
in space programs. Complex phenomena, which exist during combustion of solid
propellants, require precise determination of burning rate for ballistic calculations with
exact models of combustion. To predict and/or simulate combustion processes,
knowledge of the thermophysical properties of solid propellants is necessary. Properties
of particular interest are density, specific heat, and thermal conductivity. This paper
reports values of these parameters, which are measured for three different thermoplastic
composite propellants based on ammonium perchlorate as oxidizer and modified PVC
as binder. Thermophysical properties of modified PVC binder have been determined
also. They can be used for prediction of specific heat in other PVC propellant
compositions. Results obtained in this paper may be quite useful in researching
combustion processes of these propellants, analysis and design of their function, and
industrial applications.
Keywords:
1.
INTRODUCTION
A solid rocket propellant is a complex and stable mixture of different oxidizing and
reducing ingredients. When the solid rocket propellant is ignited, it burns in a controlled
manner during which it is transformed into high-temperature low molecular-weight gases at
the grain surface. These gases provide the kinetic energy, which is released through the
exhaust nozzle of the rocket motor. During combustion, the burning surface regresses
perpendicularly to itself in parallel layers. The rate of regression is called the burning rate.
Burning rate as well as specific impulse, density, suitable physical properties, and safety
characteristics are usually the most important propellant properties which are necessary in
the design of new solid propellant rocket motor.
Numerous kinds of chemicals are used in making solid propellants compositions, to
obtain optimized combustion characteristics for different purposes. Presented here are the
results of an experimental investigation of thermophysical properties of thermoplastic solid
composite propellants based on polyvinyl chloride-PVC binder: specific heat, and thermal
conductivity. Like other composite propellants, the examined propellants basically consist of
inorganic solids (ammonium perchlorate-AP as oxidizer, aluminum-Al as metallic fuel),
uniformly dispersed in a continuous matrix of organic thermoplastic elastomer binder, which
are based on polyvinyl chloride-PVC. As these parameters are necessary in heat transfer
analysis of the burning of solid propellants, the results obtained make possible better
understanding of the combustion phenomena and can be further used in models of combustion.
1
513
2.
Czech Republic
2006
[Content]
THEORY
3.
EXPERIMENTS
This paper reports values of thermophysical parameters: specific heat and thermal
conductivity, which are measured for three different propellant compositions. These
compositions represent three main groups in the examined class of propellant based on
modified polyvinyl chloride binder: 1) low energy propellant denoted by A, 2) medium
energy propellant denoted by B and 3) high energy propellant denoted by C. Formulations of
these propellants are given in Table 1. The other compositions in this class are modifications
of the above, depending upon the specific requirements for each rocket motor.
Heat flow into a substance induces many physical and chemical changes, which can help
to identify and characterize a sample. Differential scanning calorimetry (DSC) monitors heat
effects associated with phase transformations and chemical reactions as a function of
temperature. In a DSC the difference in heat flow to the sample and a reference (inert
material) at the same temperature, is recorded as a function of temperature. The temperature
of both the sample and reference are increased at a constant rate. A flow of nitrogen gas is
maintained over the samples to create a reproducible and dry atmosphere, and eliminates air
oxidation of the samples at high temperatures. Since the DSC is conducted at constant
pressure, heat flow is equivalent to enthalpy changes, so measurements are used to determine
enthalpies of phase change, phase and glass transition temperatures, and heat capacities.
Specific heat of propellant formulations has been determined using Du Pont Model 910
DSC. Sapphire (Al2O3) was used as reference material, as it is a standard material of known
variation of specific heat with temperature. The sample is sealed into a small aluminum pan.
The specific heat of a test specimen was determined by comparing the thermal lag between
sample and reference material under identical conditions at the temperature of interest.
514
Czech Republic
2006
[Content]
comp. B
comp. C
AP [%]
69.00
75.00
69.75
Al [%]
0.50
0.50
10.00
30.50
24.50
20.25
1650
1680
1760
Density- [kg/m ]
2
1
5
3
5
70
1. Test sample
4. Cooling system
2. Cover plates
5. Copper heater
3. End insulator
Fig 1.
Temperature sensors
515
4.
Czech Republic
2006
[Content]
RESULTS
4.1
exothermic
3
2
1
0
-1
-2
-3
-4
-5
-6
-7
0
50
100
150
200
250
300
350
400
TEMPERATURE ( C)
Fig 2.
516
Czech Republic
2006
[Content]
2200
2100
2000
1900
1800
1700
1600
1500
1400
1300
1200
1100
1000
0
50
100
150
200
250
300
350
400
450
500
TEMPERATURE ( C)
Fig 3.
(1)
where Tref=300 K, while the following is valid in the interval from 240 to 450 oC:
ccub, AP = 1167.6 + 1.713 (T Tref ) [J/kg K]
(2)
As the gradients are very similar in both previous equations connected with crystalline
phases, we suggest the following expression which can replace them, and also is valid in the
temperature range from 20 to 450 oC:
c AP = (1093.3 + A) + 1.718 (T Tref ) [J/kg K]
(3)
where A=0 for T515 K (242 oC), A=74.28 for 515<T<723 K (242<t<450 oC). This
relationship yields for the average value of specific heat of ammonium perchlorate in the
range from 20-450 oC the value cAPav=1486.7 J/kg K. This average value also can be used in
propellant burning diagnostics.
Specific heat of pure AP is widely reported in the literature. For orthorhombic crystal
structure is given the following relationship [5] which is valid up to 240 oC:
cort , AP = 577.5 + 1.72 T + 4.72 10 6 T 2 [J/kg K]
(4)
The following relationship is valid for cubic structure (in the interval from 240 to 450 oC):
ccub, AP = 725.4 + 1.5 T + 161.48 10 6 T 2 [J/kg K]
(5)
Data obtained using expressions (4) and (5) are very similar to data reported as JANAF data
and our measurements, but are more complicate than expression (3).
The simplest relationship found in the literature [6] is in relative agreement with the
experimental values in the range from 300 K (27 oC) to above 700 K (430 oC):
5
517
Czech Republic
2006
[Content]
(6)
where Tref=300 K, but it is obvious that the above relationship has not compare well with
experimental data. It is seen in Fig. 3 that equations (3), or (4) and (5) are in better
agreement with experimental data than equation (6).
For aluminum, a metal fuel commonly used in rocket propulsion, there is only one solid
phase of interest, melting at 933 K (660 oC) with a latent heat [7] of ls=397200 J/kg. Densities
in the solid and liquid phases are, respectively, c,Al=2.702 kg/m3 and l,Al=2.380 kg/m3. The
following linear law [8] is valid for the specific heat of the solid phase:
cc , Al = 895.2 + 0.498 (T Tref ) ,
(7)
at Tref=300 K, while for the liquid-phase specific heat one has cl,Al = 1176.5 J/kgK.
Thermophysical properties of binders are difficult to evaluate, because too many
different ingredients exists in compositions. Its effect is tempered by the fact that its mass
fraction seldom goes between 16-30% of the propellant total mass. The binder which has
been used in the propellants under test mainly consists of polyvinyl chloride (PVC) and
dioctyl phthalate (DOP) as plasticizer. Samples of binder weighing around 8 mg were
examined in closed pan, and the measured DSC curve is shown in Figure 4. Samples were
heated at temperature scan rate of 10 K/min in a nitrogen atmosphere (flow 10 ml/min) at
ambient pressure. The curve has shown initially a very slow endothermic effect up to around
240 oC, where a first small peak arises. This part of the curve is due to melting of gelled
PVC binder. As the temperature increases the endothermic effect continues up to 272 oC,
after which the exothermic effect begins. Exothermic effect has two peaks: first at 289.5 oC,
which is completed at 291 oC, and a second at 306 oC.
Thermal degradation of the examined binder began after a temperature of 241 oC, and at
beginning is connected with vaporization of plasticizer. Transformation to exothermal effect
1.0
0.9
0.8
0.7
0.5
exothermic
0.6
0.4
0.3
0.2
0.1
0.0
-0.1
-0.2
-0.3
-0.4
-0.5
0
50
100
150
200
250
300
350
400
TEMPERATURE ( C)
Fig 4.
518
Czech Republic
2006
[Content]
3000
cbinder
binder-equation (8)
2800
2600
2400
2200
2000
1800
1600
1400
1200
0
25
50
75
100
125
150
175
200
225
250
275
300
TEMPERATURE [ C]
Fig 5.
at temperature 272 oC is probably due to its degradation. This process shows maximum rate
at the first exothermic peak, and is finished soon after this. Second exothermic peak is
connected only with decomposition of PVC in binder and its dehydrochlorination, i.e.
elimination of chlorine from PVC which is released as hydrogen chloride (HCl).
Dehydrochlorination of the PVC shows a maximum rate at the second exothermic peak, and
appears to be complete at around 350 C.
DSC examination of binder which is used in rocket propellants has shown that the
internal structure of thermoplastics, basically consists of PVC, and is strongly affected by
addition of plasticizer and stabilizer. Changes in thermal history of plasticized PVC binder
are evident, as it has not any endothermic peak up to 240 oC. Plasticizer-free PVC (as
amorphous polymer) typically has a glass transition with a relaxation peak at temperature Tg
of about 80 oC. Addition of plasticizer to PVC has lowered the glass-transition temperature
to lower values. Also, it is obvious that stability to heat is greatly improved by the addition
of stabilizers, as liberation of HCl is displaced to higher temperatures in comparison with
pure PVC.
Specific heat results are shown in Fig. 5 in the temperature range from 40 to 260 oC. On
the basis of these values, the temperature dependence of specific heat for the binder can be
expressed with the following linear relationships respectively, where Tref=300 K:
cbin = 1527.8 + 4.96 (T Tref ) [J/kg K]
4.2
(8)
Specific heat of propellants has been measured using DSC in a temperature range from
293 (20 oC) to 383 K (110 oC), and the results obtained are shown in Fig. 6. Samples
weighing around 40 mg were closed in aluminum pans and heated in a flowing atmosphere
of nitrogen (flow 50 ml/min) at a heating rate of 10 oC/min and ambient pressure.
519
Czech Republic
2006
[Content]
1450
1400
1350
1300
1250
propellant A-measured
propellant A-equation (9)
propellant B-measured
propellant B-equation (10)
propellant C-measured
propellant C-equation (11)
1200
1150
1100
10
20
30
40
50
60
70
80
90
100 110
120
TEMPERATURE [ C]
Fig 6.
4.3
(9)
(10)
(11)
(12)
where Tref=300 K, while for PVC[9] the following expression is accepted to 368 K (95 oC):
PVC = 0.1636 + 1.056 10 4 (T Tref ) 8.92 10 7 (T Tref ) 2 [W/(m K)]
(13)
These values of thermal conductivity are shown in Fig. 7. For aluminum the following linear
expression can be adequate for limited purposes [7], where Tref=300 K:
c, Al = 203.4 + 0.116 (T Tref ) [W/(m K)]
(14)
We can see that aluminum is a much better thermal conductor than AP and binder.
520
Czech Republic
2006
[Content]
0.55
AP
PVC
0.50
0.45
0.40
0.35
0.20
0.15
0.10
-50
-25
25
50
75
100
125
150
175
200
225
250
TEMPERATURE [ C]
Fig 7.
0.6
0.5
0.4
0.3
propellant A
propellant B
propellant C
0.2
0.1
0.0
10
20
30
40
50
60
70
80
90
TEMPERATURE [ C]
Fig 8.
521
Czech Republic
2006
[Content]
Results have shown that thermal conductivity of examined propellants decreases with
increasing temperature from ambient value. On the basis of values and character of curves,
as the thermal conductivities of aluminum and PVC are increasing functions of temperature
(which is opposite for AP), we can say that thermal conductivity of propellants is basically
controlled by the amount of AP in a formulation. As we have expected, the presence of
aluminum in composition C, as a much better conductor, has increased thermal conductivity
in comparison with composition A (which have same amount of binder), but not to a level to
be expected from values in equation (12). If we compare propellants A and B, where
propellant B has less percentage of binder in formulation, we can see that thermal
conductivity of binder is less than thermal conductivity of AP and values indicate that it is
close to the thermal conductivity of pure PVC. Unfortunately, analysis has shown that
thermal conductivity of modified PVC binder cant be determined exactly from preceding
results, as we have excessive disagreement among results.
5.
DISCUSSION
[7]
can be
(15)
where ox is the oxygen mass, me is the metallic fuel mass, while the binder mass fraction
includes all other propellant ingredients. The previous equation is a good approximation if
the energy of binding is small compared with the sum of the energies associated with the
homogeneous pockets of ingredients; this may not be true for the smallest AP particle sizes.
Expression for linear increase of specific heat of HTPB binder with temperature can be
found in literature [6]. Assuming that the same form of expression is valid for examined PVC
binder, and at the basis of experimental results of three previous compositions and the
known value of specific heat for AP and Al, we have determined specific heat of binder in
each composition, and have obtained the following relationships, where Tref=300 K:
1) Specific heat of PVC binder in composition A:
cbin , A = 1487.7 + 4.13(T Tref ) [J/kg K]
(16)
(17)
(18)
We have obtained the following general expression for specific heat of PVC binder, by
averaging preceding relationships for PVC binders in compositions A, B and C:
cbin , AV = 1566.1 + 4.78(T Tref ) [J/kg K]
(19)
Using this linear expression, we have calculated values of specific heat, and compared
with previously measured data for binder (Eq. 8). Results of this comparison are shown on
Fig. 9, and we can see that values of specific heat of PVC binder obtained from both
equations are approximately equal.
522
Czech Republic
2006
[Content]
3200
equation (16)
equation (17)
equation (18)
average value, equation (19)
equation (8)
3000
2800
2600
2400
2200
2000
1800
1600
1400
0
25
50
75
100
125
150
175
200
225
250
275
300
TEMPERATURE [ C]
Fig 9.
5.4
Additional ballistic properties of the considered propellants: the thermal diffusivity, and
thermal responsivity were computed on the basis of the preceding measured results for
density, specific heat, and thermal conductivity. All of these are valid only for condensed
phase (AP cubic structure, before AP crystalline transition).
The thermal diffusivity- of the condensed phase for the tested composite propellants
are computed according to:
=
(20)
and values at various temperatures are collected in Table 2, and shown in Fig 10. All results
are valid at ambient pressure, and possible pressure effect has not been considered.
Table 2. Thermal diffusivity of examined propellants.
x 107 [m2/s]
11
Temperature [oC]
comp. A
comp. B
comp. C
20
2.148
2.505
2.262
50
1.786
1.953
2.009
80
1.497
1.73
1.746
100
1.415
1.597
1.542
523
Czech Republic
2006
[Content]
The preceding results show that thermal diffusivity was found to decrease linearly for
temperature increasing from ambient value. Also, as the average thermal diffusivity of pure
AP is 1.89x10-7 m2/s, thermal diffusivity of propellants is controlled basically by APs
presence. If we compare compositions A and C, which have the same amount of binder in
compositions, it is obvious that the presence of Al as a good thermal conductor perceivably
increases the average thermal diffusivity. If we compare compositions A and B, we can see
the effect of PVC binder; it diminishes thermal diffusivity in the orthorhombic AP region.
By definition, thermal responsivity- of the condensed phase is:
= c
(21)
and the values, calculated for the three composite propellants, are given in Table 3. These
values are also shown in Fig. 10.
Table 3. Thermal responsivity of examined propellants.
comp. B
comp. C
20
912.2
1002.1
974.2
50
887.2
940.3
980.6
80
858.1
930.3
965.9
100
863.6
926.4
942.9
1050
3.1
1000
3.3
2.9
2.7
950
900
2.5
850
THERMAL RESPONSIVITY
propellant A
propellant B
propellant C
2.3
2.1
800
750
1.9
THERMAL DIFFUSIVITY
propellant A
propellant B
propellant C
1.7
1.5
700
650
1.3
0
10
20
30
40
50
60
70
80
90
100
110
2 1/2
comp. A
Temperature [oC]
600
120
TEMPERATURE [ C]
Czech Republic
2006
[Content]
6.
CONCLUSION
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
13
525
Czech Republic
2006
[Content]
526
1.
Czech Republic
2006
[Content]
INTRODUCTION
The secondary nitramines still are a source of new power stuffs and consequently the
researches of relation of thermal stability of these connections from their constitution prolong to
be actual. It is now considered demonstrated, that the decomposition secondary nitramines starts
with a homolytic bond cleavage N-N [1]. Offered in a number of papers the of reacting scheme
passing through a cyclical transient state, have not received an admission. The main reason for
the benefit of homolytic bond cleavage N-N is served by high value of the pre-exponential
factor, equal 1014,5 -1, reference for reacting separation of radicals. Was established, that there
are two kinds of the nitramine groups - flat and pyramid-shaped, it was showed, that flat groups
more strongly pyramid-shaped [2]. These deduction was followed in the late paper [3]. During
synthesis of molecules, loaded the large number of nitro groups, obtained molecules, in which
one cannot be attributed to known - flat, or pyramid-shaped frames. So, in a molecule of the
trans-1,4,5,8-tetranitro-1,4,5,8-tetraazadecalin (TND) because of a steric repulsion of oxygen
adjacent of nitro groups there is a turn around of the bond N-N. Let's call such of nitramine
groups as twist-groups. One molecule of the TND contains two pair of such twist-groups. It was
possible to suspect, that the decay TND will drive on the same gear, as for earlier studied
secondary nitramines the pre-exponential factor of reacting will be peer 1014,3 -1, and
statistical factor = 4, and the activation energy, equal strength of the bond N-N will be reduced
that nitramine group are deformed by turn around of the bond N-N. This supposition has
appeared fair only half.
2.
The thermal decomposition of the TND was studied by a manometer method in a solid
state, in solution in m-dinitrobenzeneand, and in gas. The reaction rate in solution and in gas is
described by an equation of the first order. In solution in a temperature range 150-1800 lgAliq =
11,151 (-1), liq=24,32 kkal / mol, in gas the constant rate was gauged only at temperature
1800 and it has compounded kg = 110-3 c-1. In a solid state of the constant rate were calculated
on an equation of the first order on initial velocity and V = 425cm 3 / g . The rate in a solid state
by a composite mode depends on degree of fullness of a vessel, and for obtaining activation
parameters the relation V/m was extrapolated on zero. The following values of activation
parameters of the decomposition of the TND in a solid state are obtained: lgAs = 8,141 (-1),
s = 24,23 kkal / mol, the retardation effect of the crystalline lattice (REL) at 180 0 is peer
kliq/ks = 103.
The experimentally measured value of activation energy of the decomposition of TND is
lower than reference for the decomposition of the secondary nitramines almost twice. This is
very large value. The nitramine group is rather conservative to an inductive effect of groups. So,
in 2,4,6-(NO2)3C6H2N(NO2)Me the benzene ring reduces D (C-N) all on 5 kkal / mol, while for
bonds -, -, C-Hal the value of this decrease is peer 15-20 kkal / mol [1]. The relevant
value has a geometrical constitution of the nitramine group. 1,3-Dinitro-1,3-diazacyclopentane
and 1,3,5-trinitro-1,3,5-triazacyclohexane (RDX) have till two pyramid-shaped nitramines
group, and their decomposition rate in vapors with inhibitor on the order is higher, than one for
dimethylnitramine, diethylnitramine, 1,3,5,7-tetranitro-1,3,5,7-tetraazacyclooctane (HMX) and
527
Czech Republic
2006
[Content]
other connections having only the flat nitramine groups. D (N-N) in the flat nitramine groups is
peer 40,5 kkal / mol, and in pyramid-shaped - 37,7 kkal / mol 2. The distinction makes 2,8 kkal /
mol. In the twist-groups of the TND the activation energy is reduced as contrasted to flat
nitramine groups on 16,2 kkal / mol. Even more radical changes have taken place with the value
pre-exponential factor.
For all studied before secondary nitramines average the value of the pre-exponential factor
in calculation on one group is equal 1014,5 -1. In TND the pre-exponential factor in calculation
on one nitramine group is peer 1010,65 -1. Such low value of the pre-exponential factor can
indicate that the transient complex has cyclical frame.
The data, obtained in activity, demonstrate, that the thermal decomposition of the
secondary nitramines with twist-groups considerably differs from decomposition studied before
compounds with flat and pyramid-shaped nitramine groups and demands further research.
3.
[1]
[2]
[3]
528
REFERENCES
G. B. Manelis, G. M. Nazin, Yu. I Rubtsov, V. A. Strunin Thermal decomposition and combustion
of exolosives and pouders. M.: Nauka, 1996. 223p.
Yu. M. Burov, G. M. Nazin - Kinetics and catalysis, 1982, V. 23, p. 12-17.
J. C. Oxley, A. B. Koob, R. Szekers,W. Zheng J Phys. Chem., 1994. V98, p. 7004 7008.
Czech Republic
2006
[Content]
1.
INTRODUCTION
Military explosives should make ammunition to work safely and properly and must not
worsen the ammunition quality even during peace time long-term storage where at any rate
certain reserves should be kept. Therefore the military explosives and ammunition must be
subjected to very rigorous supervision and demands relating to its durability, stability and
also safety. High explosives, as chemical substances blend, have got their durability limited
by several factors among that there are chemical composition, manufacturing type, purity,
storage conditions, ammunition application and related construction material compatibility.
Under these factors, there might be a gradual change of chemical and chemical-physical
properties of an explosive. Of course, also the sensitivity characteristic can be subject of
change resulting to safety and function degradation of ammunition in which is explosives
built in. Therefore there is a must to observe not only quality status but also quality
characteristic of high explosives and to eliminate problems that might have been occurred
during life-in service of ammunition with explosive built in. These problems can be as
follows: limited stability and lifetime, incompatibility with construction material, tendency
for dangerous changes of properties during ageing, too large a sensitivity increase etc.
An important part of military explosive evaluation is determination of ageing influence
on explosive properties while being determine its lifetime. The artificial ageing can be a
proper way how to observe ammunition lifetime process. It is the process of quick ageing
simulating a nature explosive ageing in ammunition system. Usually, it is realized by
temperature conditioning at higher temperatures that speed up the physical-chemical
processes which under normal storage condition would be running for several years in
explosive. Main goal of the process is that during several weeks to bring the explosive to the
status in which explosive might has been after couple of years of ageing. Artificial ageing
method is used for prediction of the changes of stability, sensitivity, mechanical and
operation explosive properties and obviously of their lifetime.
Long-time period physical and chemical stability of military explosives and related
material compatibility are involved between basic requirements of military explosives. The
requirements must be applied for each military explosive munitions-application.
The term of Explosive stability means an ability to keep up quality status and so not to
be a subject to any of chemical-physical changes that would tend to undesired changes under
529
Czech Republic
2006
[Content]
normal condition. There must not be changes of explosive characteristics or unexpected selfignition or even explosion. Long-term stability observation during long-run storage has its
reasonable meaning mainly from point of view of keeping a manipulation safety, storability
and explosive functionality.[1]
2.
Properties study of plastic explosives on the base of hexogen during artificial ageing has
pointed out on changes in physical and detonation characteristics of the explosive. Artificial
ageing conditions were as follows: 30 and 90 days at temperatures 120C, 180C and 195
C, 5 7,5 hours of dynamic and isothermal heating. There was a result of the tested sample
that at the isothermal heating the critical temperatures of fast decomposition, ignition and
explosion was 1520 C less than at dynamic heating.[2]
Japan authors was studying ageing of 6 high explosives PBX types for several years,
evaluated various tests during the ageing, for example detonation rate, explosion under
water, friction sensitivity, hammer impact sensitivity, DTA, TGA, VST. Most of explosives
tested were confirmed as well-power even after ageing process.[3]
American and Canadian scientists detected dynamic-mechanical properties of hexogen
explosives of PBX bound with Kraton (hydrogenated butadiene-styrene polymer) and
plasticized wax explosives X-0287 and oil explosive X-0298. Explosive samples were
ageing at T=23, 60 and 74 C for 3 years. At X-0298 explosive there was at fast ageing no
worsening of mechanical properties but molecular mass was at decrease by ageing increase
temperature because of irregular chain split. In by-wax-plasticized explosive (X-0287)
related to kinetic degradation of Kraton a significant difference was found over and under
wax melting point. This way created model predicts durability of both explosives as equals
as 20,5 years under the worst military conditions ever would be (74 C).[4]
Ageing process study of I-RDX (so called insensitive RDX) and cast PBX explosives on
base of I-RDX proved that artificial ageing process does not affect low impact sensitivity of
PBX composites. This low sensitivity is conditional to I-RDX content what means a big
contribution for insensitive ammunition development. I-RDX is the hexogen type produced
by special re-crystallization process, patented by EURENCO fy.[5]
Study of sensitivity changes of explosives wax-moderated after artificial ageing (113
days,70C) proved an increase of wax-moderated explosive sensitivity to impact, friction
and electrostatic spark in spite of their thermal characteristic (DTA, VST) remained
unchanged. These changes have occurred because of explosive crystal exposure under high
temperature ageing used. Although the temperature of 70 C is not extreme, it is
inappropriate for observing the sensitivity changes of wax-moderated explosives after
artificial ageing.[6]
3.
EXPERIMENTAL
Chemical and thermal stability, impact sensitivity, friction sensitivity and detonation
velocity were studied on the hexogen-based plastic explosive samples during artificial
ageing under various conditions. When comparing, there were except of plastic explosive
samples used also 2 wax-moderated explosive samples. The goal was to find the limits under
that the parameters are still subject of changes and assess proper ageing conditions.
Used explosives:
530
Czech Republic
2006
[Content]
Test to perform
VST
DTA
Temperature of ignition
Impact sensitivity
Friction sensitivity
Detonation velocity
Procedure by
Stanag 4556
TP-SHVV-1080-00-ZH,
Stanag 4491
Stanag 4489,4487
AOP-7
RDX
THE
RDX, TNT
AOa
HMX
NOTES:
VST
DTA
DMDNB
marking agent
RDX
hexogen
HMX
oktogen
TNT
- trinitrotoluene
531
Czech Republic
2006
[Content]
Number of days
50
65
20, 40, 60
Package type
Glass NATO-test tube
Al-foil + sealed PE-foil
sealed PE foil
Glass NATO-test tube
Al-foil + sealed PE-foil
sealed PE foil
NOTES:
Al aluminum
PE - polyethylene
3.1
Chemical stability test of explosives was performed by Stabil 16-Ex vacuum stability
test equipment at constant temperature 100C for 40 hours. The test was done using all
samples after each ageing interval at both temperatures. Tab 4 enlists the results of not-aged
samples and also of samples after ageing completion.
Thermal stability was done using Differential thermal analyses method by DTA-550-Ex
apparatus and by determination temperature of ignition with all samples after ageing at both
temperatures. Heating rate at the two tests was 5C/min.Tab 20 enlists the review of thermal
stability results measured before/after ageing at both temperatures.
Table 4. VST (cm3) results
Ageing type
Sample
Comp
CHe1/97
Comp
CHe2/00
Comp CHa
Comp CHc
Comp C4
AOa
532
Non-ageing
0,526
0,287
0,141
0,176
0,119
0,111
0,188
0,192
0,759
0,822
0,333
0,245
0,427
0,673
0,129
0,132
0,410
0,613
0,115
0,186
0,468
0,471
0,151
0,130
1,664
0,484
0,129
0,195
1,468
0,481
0,148
0,272
1,071
0,677
0,149
0,215
Czech Republic
2006
[Content]
Non-ageing
Aged in tube
Mp St Max TI Mp St Max
204 198 218 209
201 208 213
204 198 216
197 194 214 210
202 197 214
203 199 216
204 210 215 210
204 190 217
205 196 217
207 211 216 212
206 213 219
206 214 220
198 210 213 209
201 213 216
204 214 218
181 243 264 258
189 253 272
190 252 272
TI
Aged in PE foil
Mp St Max TI
210
208
211
213
212
213
214
213
212
211
260
262
NOTES:
Mp melting point
St start of thermal decomposition
Max maximal temperature of thermal decomposition
TI temperature of ignition
Results of the tests have proved that the tested plastic explosives and AOa waxmoderant are chemical and thermal stabile after artificial ageing had been applied.
3.2
Sample
CHe1/97 non-aged
Aged in tube
Aged in Al+PE foil
Aged in PE foil
CHe 2/00 non-aged
Aged in tube
5
Impact E50
[J]
33,18
27,12
36,65
37,21
36,06
25,94
Friction F50
[N]
303,60
218,25
262,80
234,00
243,00
281,25
Detonation velocity
[m.s-1]
7436,20
533
4.
23,34
23,15
30,42
29,98
30,20
31,66
33,66
28,11
28,31
25,81
25,63
27,31
40,49
32,14
24,19
24,02
31,66
24,19
8,55
8,98
9,57
8,74
Czech Republic
256,50
258,75
248,40
218,25
204,75
200,25
313,20
240,75
296,40
272,25
213,75
204,75
214,80
153,00
249,75
218,00
234,00
227,25
188,40
138,00
141,00
144,00
2006
[Content]
7479,50
7448,15
8285,50
8125,30
8222,55
7351,00
7278,20
7184,20
-
DISCUSSION
The results above mentioned prove that during explosive ageing process mainly friction
sensitivity was changing, for all the samples after ageing there was found a friction
sensitivity increase except Comp Che 2/00 explosive, where the sensitivity slightly
decreased. As far as the impact sensitivity either was not changed or (Comp CHc) slightly
decreased.
Impact sensitivity increase was occurred using ageing only at Comp Che 2/00, Comp
Cha explosives and Comp Che1/97 aged in a test tube. In THE, there is not so striking
difference in impact sensitivity before and after ageing process as e. g. in Comp C4, where
ageing process sensitivity has striking slumped. The reason of that is probably high volume
of deterrent in THE and high volume of highly explosive in Comp C4. The intermolecular
activity dominant influences to initiation of detonation by impact in system highly explosive
binding material and conformational stability of molecules.[7] In thermolysis and in
initiation of detonation, those nitramine-parties are primary homolysing that are, in definitive
way, participating with intermolecular potential in correspondent molecular crystals.[8] The
organic additives to individual high explosive are influencing to the sensitivity as well, and
that thanks to their specific properties as e. g. specific heat, capability to dissolve the
energetic material, capability to generate the radicals by impact.[9,10,11]
In terms of the packaging used, the least sensitivity changes were captured on the
samples aged in the Al+PE foil. Detonation velosity slightly increased using ageing of Comp
Che 2/00 while the other two explosive samples proved a slight sensitivity decrease after
ageing.
534
Czech Republic
2006
[Content]
ln
Tm
[12]
with
E 1
*
+A
R Tm
C-4
40
CHe1/97
35
THE
CHc
30
CHa
25
E (J)
CHe2/00
20
15
y = -0,0016x - 5,6325
2
R = 0,7714
10
AOa
0
-35000
-30000
-25000
-20000
-15000
-10000
Kissinger slope
Fig 1.
535
Czech Republic
2006
[Content]
400
350
F (N)
300
CHa
CHe1/97
CHe2/00
250
THE
C-4
CHc
200
y = -0,0203x - 259,21
2
R = 0,8894
150
AOa
100
-61000
-51000
-41000
-31000
-21000
-11000
-1000
Kissingerova smernica
Fig 2.
80
70
C-4-N
C-4-Al
C-4-Pe
60
Che-Al
Che-N
THE-Al
THE-Pe
D2 (m.s-2)
50
Che-Pe
THE-N
y = 0,0009x + 70,518
2
R = 0,7457
40
30
20
10
0
-23000
-22000
-21000
-20000
-19000
-18000
-17000
-16000
-15000
Kissingerova smernica
Fig 3.
536
Czech Republic
2006
[Content]
There is possible to establish from figures, that plastic and wax high explosives are not
allocating ideal linear dependency, but trend of linear dependency between impact and
friction sensitivity and results of DTA by Kissinger is evident here. The similar trend of
linear dependency of tested high explosives has expressed in relative relations between
impact sensitivity and friction sensitivity, and between detonation velocity and friction
sensitivity, the figures 4 and 5 are presented this.
300
CHe2/00
250
CHa
THE
CHe1/97
CHc
C-4
200
F (N)
y = 5,7545x + 73,602
2
R = 0,7986
150
AOa
Np10
100
NpD
50
0
0
10
15
20
25
30
35
E (J)
Fig 4.
537
Czech Republic
2006
[Content]
75
70
C-4-N
C-4-Pe
C-4-Al
D2 (m2.s-2)
65
60
Che-Al
55
Che-Pe
Che-N
THE-N
THE-Al
THE-Pe
50
y = 0,12x + 24,84
2
R = 0,8009
45
40
120
140
160
180
200
220
240
260
280
F (N)
Fig 5.
5.
CONCLUSIONS
On the basis of results of experiment, it is possible to establish, that the tested plastic and
wax phlegmatizated high explosives stayed chemical and thermally stable after artificial
ageing process, their detonation velocity as well as impact sensitivity have succumbed to just
small changes and friction sensitivity has considerably increased in majority of them after
artificial ageing process. The impact sensitivity of Comp Cha a Che has moderately
increased. In term of packing, the Al-PE foil seems the optimal for plastic high explosives as
well as for wax phlegmatized high explosives, which application with was reported the least
of negative changes of explosive sensitivity.
The thermolysis and initiation of blending high explosives detonation by impact and
friction, they are affected primarily by intermolecular interactions and inherence of organic
additives. The work has confirmed, that it is possible to apply the DTA method by Kissinger
to approximately prediction of blending high explosives sensitivity.
538
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
11
539
Czech Republic
2006
[Content]
1.
INTRODUCTION
The research and development for high explosives nowadays is dedicated to improve the
insensitivity of the materials. It is known that the sensitivity strongly depends on the surface
characteristics of the material. Defects, roughness or cracks influence the sensitivity of the
material [1], but also the density plays an important role [2]. With decreasing density the
amount of cracks or spaces increases. As these defects may be filled with air, upon impact of
a shock wave this volume element may act as a hot-spot and lead to an initiation of the explosive. Therefore, the sensitivity should increase with decreasing density. However, no
systematic experimental study of this effect exists up to now.
2.
EXPERIMENTS
The material used for these experiments was HNS-HBD (hexanitrostilbene high bulk
density). We applied different pressures while preparing the HNS-pellets and used two different pressing tools. One of them was applying the pressure only from one side, while the
second was applying pressure from both sides. The following samples were produced with
the first pressing tool, applying pressure only from one side. This pressing tool is designed in
the way that the length of the pellet is always the same. Therefore, the density of the pellet is
depending on the initial mass filled in the pressing tool. This dependency is shown in figure
1. The density is linearly increasing with increasing mass, as expected from theory.
As mentioned before, we produced the HNS-pellets with two different tools. In figure 2
the differences in the density-pressure-relation is shown. Since one tool applies the pressure
from two sides, the compression and therefore the average density of the material is much
higher than for the other case, where the pressure is applied only from the top. This effect
becomes more important with decreasing pressure applied. However, the observed density of
540
Czech Republic
2006
[Content]
the pellet produced in the one-side-tool exceeds the density of the pellet formed in the
two-side-tool at relatively high specific pressure above ~ 4 kbar. We think that this unexpected behavior might result from a change of the press. To achieve the necessary specific
pressures of 3.8 kbar for the one-side-tool we had to take another press reaching higher
total pressures. To clarify this observation further experiments have to be performed. From
theory we would expect that both pressure-density-relations would converge to one density
limit.
1,71
1,70
Density (g/cm )
1,69
1,68
1,67
1,66
1,65
130
131
132
133
13 4
135
136
137
138
Mass (mg)
Figure 1: Mass-density-relation.
1,68
Density (g/cm )
1,66
1,64
1,62
1,60
1,58
1,5
2,0
2,5
3,0
3 ,5
4,0
4,5
Pressure (kbar)
Figure 2: Higher densities were reached at same specific pressures for the tool applying the pressure
from both sides.
541
Czech Republic
2006
[Content]
The change in sensitivity of HNS-pellets with different densities was investigated by performing standard gap-tests. The density of the HNS was varied in a range that the pellet can
be expected to be insensitive. By definition insensitivity means no go-distances of 15 mm
inert barrier (water at ambient temperature in our case) for the gap-tests. The donor charge, a
HWC pellet (94.5% hexogen, 4.5% wax and 1% graphite), is initiated by a detonator and the
resulting shockwave is released into the water. Water is used to ensure that the possibly initiated detonation of the HNS is not initiated by any thermal energy output of the donor
charge, but only by the propagating shockwave. From these measurements we found that the
sensitivity of the HNS strongly depends on the density of the pellet, as expected. Figure 2
shows the results of the gap-tests performed for samples with different densities.
18
Go-values
NoGo-v alues
linear fit (Go)
linear fit (NoGo)
16
14
12
10
1 ,65
1 ,66
1,6 7
1,6 8
1,69
1,70
1,71
Density (g/cm )
Figure 3: Go- and NoGo-gap-distances for samples with different densities.
None of our HNS-pellets detonated at 15 mm inert barrier. The most insensitive samples
showed densities higher than 1.70 g cm-3. For these samples gap-distances for no go of 12
mm inert barrier were found and the most dense sample with = 1.705 g cm-3 even did not
detonate at a gap-distance of 11 mm inert barrier. We tried to fit the measured gap-distances
by a linear relation to predict the densities at which the HNS-pellets will become sensitive
(go-distances > 15 mm inert barrier). We found that the limit should be reached for pellets
with densities around 1.65 g cm-3 (s. figure 3).
542
3.
Czech Republic
2006
[Content]
RESULTS
According to our measurements it is necessary to achieve densities of 1.65 g cm-3 to receive insensitive HNS-pellets. On the other side the HNS may reach densities where no initiation is possible anymore. To ensure that the HNS-pellet is insensitive but can also be initiated with standard techniques the density of the HNS should be between 1.65 g cm-3 and
1.70 g cm-3. In this range of the density the no go-gap-distances decrease from 15 mm to
12 mm inert barrier (water at ambient temperature).
REFERENCES
[1]
[2]
L. BORNE, M. HERRMANN, C. B. SKIDMORE: Microstructure and Morphology, (Energetic Materials Particle Processing and Characterization, U. Teipel, Eds.), Pfinztal, p.333-366, 2005
A. V. D. HEIJDEN, J. TER HORST, J. KENDRICK, K.-J. KIM, H. KRBNER, F. SIMON, U. TEIPEL:
Crystallization: Product Quality of Energetic Materials, (Energetic Materials Particle Processing and Characterization, U. Teipel, Eds.), Pfinztal, p.69-76, 2005
543
Czech Republic
2006
[Content]
Abstract:
Mixtures of fluorinated graphite (CF)n with different reductants, such as NaN3, B, Ti,
ZrTi, Si, CaSi2, AlSi, TiSi, MoSi2 and CrSi2 were investigated. Reactions between these
substrates are fast and exothermic enough to proceed in a high temperature, selfsustaining regime. Heat effects accompanying the reactions were measured and the
solid reaction products were analyzed. SEM observation revealed the presence of
exfoliated graphite in each case. Nanostructures, like nanofibers, nanorods,
nanospheres and nano-sized crystals, were also present in most of the products. Phase
(XRD) and elemental composition as well as porous structure (N2 adsorption) of chosen
reaction products were also determined.
Keywords:
1.
INTRODUCTION
Pyrotechnic mixtures containing powdered metals (e.g. Mg, Al, Ti) and fluorocarbons
polymers (mainly polytetrafluoroethylene, acr. PTFE) are characterized by comparatively
high heat and temperature of combustion. Moreover their combustion products are good
emitters of infrared radiation. Therefore the compositions are commonly applied in infrared
decoy flares and in igniter devices [14].
Less recognized, though, is the fact that they should also be considered as a source of
novel carbon and ceramic nanostructures, which are created in the ultrafast chemical
reactions at very high temperatures in the combustion wave. In particular, this refers to
systems, in which silicon or its compounds or alloys are used as reductants (i.e.
defluorinating agents). Then, the combustion products contain silicon carbide monocrystals
in an unique form of nanofibers [512].
Searching for new, more reactive fluorocarbon oxidizers, E-Ch. Koch has proposed to
use strained fluorocarbons (fluorinated derivatives of tetrahedrane, prismane, cubane and
[60]-fullerane) and graphite fluoride (poly(carbon monofluoride), (CF)n) as fluorine sources.
Owing to an additional energy contained in the strained carbon skeletons, and a lower CF
bond energy in (CF)n than in PTFE, they are much better oxidizers compared to PTFE in
terms of enthalpy and temperature of combustion [13]. The application of caged fluorocarbons
is not possible at present, because some of them are too expensive and others have not been
sensitized so far, but fluorinated graphites are readily available. They can be obtained by the
direct reaction of elemental fluorine with graphite or other carbon materials [14].
Depending on the synthesis conditions, the atomic ratio F/C changes from 0.6 to 1.2.
Poly(carbon monofluroide) (CF)n is a dark-gray, crystalline substance with a density of
544
Czech Republic
2006
[Content]
2.65 g/cm3. Its thermal decomposition starts at 480 oC [13] releasing extremely reactive
carbene (:CF2) and tetrafluoroethylene (C2F4) as the main initial gaseous product of
degradation in vacuum [14].
Fluorinated graphite was first applied as an oxidizer [15] or combustion modifier [16] in
the 80s. and 90s. of the XX century, but only E.-Ch. Koch was the first to publish in 2005
complete results of a thermodynamic analysis of (CF)n defluorination with magnesium,
a mechanism of the reactions as well as radiative and thermochemical properties of a
Mg/(CF)n pyrolant containing 55% (CF)n, 40% Mg and 5% Viton [17]. In the paper, the
author also reports on detection of single walled carbon nanotubes (SWCNT) and carbon
nano carpet rolls (CNCR) in the combustion products. This discovery of great importance
was patented [18] as a new method of synthesis of carbon allotropes (fulleranes, carbon
nanotubes) and various carbon encapsulates.
In the patent the application of several dozens of metals, metalloids, their alloys and
compounds for reduction of fluorinated graphite is claimed, however only results of studies
on reactions between magnesium and (CF)n have been published so far. In the present work,
the following substances: NaN3, B, Ti, ZrTi, Si, CaSi2, AlSi, TiSi, MoSi2 and CrSi2 were
used as reductants of poly(carbon monofluoride). They were selected on the basis of our
previous experience in testing of reductant/PTFE systems [512]. Reductants containing
silicon are numerously represented because the presence of silicon in the reacting mixture
results in synthesis of silicon carbide monocrystals in the form of nanofibers. Boron,
titanium and zirconium create hard and thermally resistant carbides, and under the synthesis
conditions, the compounds may be obtained in the form of nanosized particles. Sodium azide
decomposes in the combustion wave releasing metallic sodium, which violently reacts with
fluorinated graphite producing a lot reactive carbon particles/clusters possible precursors
of carbon nanostructures.
The condensed combustion products were analyzed by X-ray diffraction (XRD) for
crystal structure and characterized using scanning electron microscopy (SEM) for
morphology. Moreover elemental composition and BET specific surface area of the
carbonaceous material obtained by carbonization of fluorinated graphite with sodium azide
were determined.
2.
EXPERIMENT
545
Czech Republic
2006
[Content]
resistance wire. Using the calorimetric bomb as a reactor, it was possible not only to collect
solid combustion products, but also to measure the heat effects accompanying the reactions.
Table 1.
Composition of the tested mixtures, assumed reactions and their standard free enthalpies
Composition
[wt. %]
NaN3 (67.8) / CF (32.2)
B (36.7) / CF (63.3)
Ti (34.0) / CF (66.0)
ZrTi (42.8) / CF (57.2)
Si (31.2) / CF (68.8)
AlSi (26.4) / CF (73.6)
CaSi2 (34.0) / CF (66.0)
TiSi (38.1) / CF (61.9)
MoSi2 (45.0) / CF (55.0)
CrSi2 (33.3) / CF (66.7)
Assumed Reactions
2NaN3 + 2CF = 2C + 2NaF + 3N2
5B + 3CF = B4C + BF3 + 2C
Ti + 3CF = TiF3 + 3C
ZrTi + 7CF = ZrF4 + TiF3 + 7C
2Si + 4CF = SiF4 + SiC + 3C
AlSi + 7CF = AlF3 + SiF4 + SiC + 6C
CaSi2 + 6CF = CaF2 + SiF4 + SiC + 5C
2TiSi + 11CF = TiF3 + 2SiF4 + TiC + 10C
MoSi2 + 6CF = MoF2 + SiC + 5C + SiF4
CrSi2 + 7CF = CrF3 + 6C + SiC + SiF4
The solid combustion products were removed from the bomb with water. The
suspension was filtered off and the deposit obtained was washed with ethanol. After drying,
the reaction products were analyzed to identify their composition (XRD, Diffractometer
Simens D5000, Cu K) and morphology (SEM). Moreover elemental analysis (PerkinElmer CHNS/O 2400) and nitrogen adsorption measurements (Micromeritics Gemini 2360)
were performed for reaction products of sodium azide with fluorinated graphite.
3.
The results of XRD analysis and the reaction heats (each of them is an average of three
experimental values, standard deviation below 20 kJ/kg) are presented in Table 2.
Table 2.
Reactant Composition
Heat of Reaction
[wt. %]
Q [kJ/kg]
NaN3 (67.8) / CF (32.2)
3780
B (36.7) / CF (63.3)
3052
Ti (34.0) / CF (66.0)
4440
ZrTi (42.8) / CF (57.2)
3780
Si (32.1) / CF (68.8)
5030
AlSi (26.6) / CF (73.8)
7160
CaSi2 (66.0) / CF (34.0)
4840
TiSi (38.1) / CF (61.9)
4580
MoSi2 (45.0) / CF (55.0)
3680
CrSi2 (66.7) / CF (33.3)
4210
* (s) strong, (m) medium, (w) weak
Phases Identified
C(s)* NaF(s)
C(s), B4C(s)
C(s), TiC(s), TiF3(s)
C(s), TiC(w), ZrF4(s), TiF3(s), Fe3Zr(m)
C(s), Si(s), SiC(s)
C(s), SiC(m), AlF3(s), Si(w), Al3.21Si0.47(m)
C(s), CaF2(s), SiC(s), Fe3C(w), Si(m)
C(s), TiC(s), SiC(w), Si(m)
C(s), Mo3C2(s) Si(w), Mo4.8Si3C0.8(w)
-
All the reactions are fast and exothermic enough to proceed in a self-sustaining
regime under the experiment conditions. The heat effect varies from c.a. 3050 kJ/kg
(B/(CF)n) to 7160 kJ/kg (AlSi/(CF)n), and this well correlates with the changes in the
standard free enthalpy of the reactions, i.e. the more negative rGo is the higher the reaction
heat is. It means that the reactions are driven rather by decreasing enthalpy than by
increasing entropy.
546
Czech Republic
2006
[Content]
Phase composition of the solid combustion products suggests that the assumed simple
reactions between (CF)n and the reductants are a fair approximation of the real course of
reactions in the tested systems. In each case, the XRD scans showed the presence of graphite
with strong, narrow peaks, so that layered structure of poly(carbon monofluoride) was not
destroyed during defluorination.
When boron, titanium or molybdenum were present in the reacting mixture, the
reaction products contained as it was expected carbides of the elements (B4C, TiC and
Mo3C2). Similarly, the presence of silicon resulted in synthesis of silicon carbide SiC. Other
silicides and mixed silicides/carbides were also created. However silicon did not react
completely in any system it was detected in each sample. The products of reactions
between ZrTi and fluorinated graphite contain zirconium(IV) and titanium(III) fluorides, TiC
and other unidentified phases with high unit cell constants (around 10 nm).
XRD analysis was combined with measurements of atom fluorescence and this
technique revealed that iron was present in each sample in spite of the fact that iron was an
impurity contained only in one reductant CaSi2. In the other samples, iron appeared as a
result of the contact of the hot reacting substances with the crucible and bomb walls made of
stainless steel.
Figures 1 and 2 show SEM micrographs of the raw combustion products.
A
Fig. 1. SEM images of the products of reactions between: NaN3/(CF)n (A), B/(CF)n (B), Ti/(CF)n
(C) and ZrTi/(CF)n (D)
When starting mixture does not contain silicon, SEM images of the combustion
products reveal exfoliated graphite (EG) as a dominating structure, Fig. 1. Apart from this,
nanocrystallites of metal fluorides and carbides (the presence of these compounds was
confirmed with XRD method) are distributed randomly within exfoliated graphite.
In order to characterize the porous structure of the exfoliated graphite, the raw products
of reactions between sodium azide and (CF)n were washed with hot water (to remove NaF),
4
547
Czech Republic
2006
[Content]
and then an isotherm (at 77 K) of nitrogen adsorption for the obtained material was recorded.
The BET specific surface area of the product was calculated to be 125 m2/g. The elemental
analysis showed that the sample contains 90.7% carbon, 3.5% fluorine and 1-1.5% of
nitrogen and hydrogen (adsorbed components of atmosphere). The remaining part of the
sample (c.a. 4%) probably includes sodium (NaF was not removed completely) and oxygen
(moisture and oxygen bound chemically on the surface).
A
Fig. 2. SEM images of the products of reactions between: Si/(CF)n (A), AlSi/(CF)n (B), CaSi2/(CF)n
(C), TiSi/(CF)n (D), MoSi2/(CF)n (E), CrSi2/(CF)n (F).
Czech Republic
2006
[Content]
reaction products in mixture of pure silicon with fluorinated graphite (Fig. 2A) is especially
unexpected. In any SEM micrographs of the products (even at low magnification), fibers
have not been detected. To find the reasons for this, more detailed study is needed, which is
in progress.
4.
CONCLUSIONS
Acknowledgment
This research was supported by the Ministry of Education and Science (Poland) through grants
3 T08D 012 28 (Warsaw University) and PBS 700 (Military University of Technology).
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
E-CH. KOCH, Review on Pyrotechnic Aerial Infrared Decoy Flares, Propellants, Explosives,
Pyrotechnics, Vol. 26, p. 3, 2001.
E-CH. KOCH, Metal/fluorocarbon Pyrolants, III, Development and Application of MTV,
Propellants, Explosives, Pyrotechnics, Vol. 27, p. 262, 2002.
E-CH. KOCH, Metal/fluorocarbon Pyrolants, IV, Thermochemical and Combustion Behavior of
MTV, Propellants, Explosives, Pyrotechnics, Vol. 27, p. 340, 2002.
T. KUWAHARA, et all, Ignition Characteristics of Ti/Teflon Pyrolants: safety igniter with
partition, Propellants, Explosives, Pyrotechnics, Vol. 30, p. 425, 2005.
A. HUCZKO, H. LANGE, G. CHOJECKI, S. CUDZILO, Y. QIU ZHU, H. W. KROTO, D. R. M.
WALTON, Synthesis of novel nanostructures by metal-polytetrafluoroethene thermolysis,
Journal of Physical Chemistry B, Vol. 107, p. 2519-2524, 2003.
S. CUDZILO, J. BLADEK, W KICINSKI, A. ARCISZEWSKA, A. HUCZKO, Synteza Materiaw
Wglowych w Fali Spalania i Ich Wykorzystanie w Analizie Chemicznej. Cz. I. Otrzymywanie i
Badania Waciwoci Fizycznych Materiaw Wglowych, Biul. WAT, Vol. 53, p. 133, 2004.
Stanislaw Cudzilo, Andrzej Huczko, Mateusz Szala, Michal Bystrzejewski
549
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
550
Czech Republic
2006
[Content]
Czech Republic
2006
[Content]
Abstract:
This paper consists of an attempt to improve the synthesis of 1,5-diaminotetrazole
(DAT) by aminating 5-aminotetrazole (5-AT) with a suitable aminating reagent.
First there are described the steps of the synthesis and the characterization of the
compounds involved in this attempt.
Keywords:
1.
Nitrophenoxyphtalimides, Nitrophenylhydroxylamines,
5-aminotetrazole, amination, characterization (Melting point, Raman,
NMR, MS, EA, X-RAY).
INTRODUCTION
1.1
Synthesis Overview
NH2
4 N
N1
N
1,5-diamino-1H-tetrazole (1,5-DAT)
A milestone in the synthesis of DAT was described by Raap[1] who reacted the sodium
salt of 5-aminotetrazole with hydroxylamine-O-sulfonic acid and obtained a mixture of
1,5-diaminotetrazole (1,5-DAT) and 2,5-diaminotetrazole (2,5-DAT) where the first
constitutes 8,5% of the reaction yield.
551
2006
H
NaOH / HOSA
- Na2SO4
N
N
[Content]
NH2
NH2
Czech Republic
NH2
N
+
H2N
5-AT
1,5-DAT (8,5%)
2,5-DAT
Previously to Raap, Gaponik & al.[2] improved the synthesis reported by Stolle & al[3].
By reacting thiosemicarbazide with lead oxide and sodium azide in a CO2 atmosphere, a
carbodiimide intermediate is formed and reacts in-situ with HN3 to lead to 1,5-DAT.
NH2
S
PbO /
NH
- PbS / H2O
H2 N
HN3
HN = C = N
NH2
NH2
NH2
Unfortunately, this reaction leads to large amounts of lead azide as the side-product
which makes this synthesis problematic for an industrial scale.
The most recent approach to this synthesis was made in our research group[4] and it
eliminates the formation of highly-explosive lead azide and produces a substancially greater
yield. After diazotation of diaminoguanodinium chloride, the reaction mixture is brought to
pH 8 to deprotonate the intermediate formed which cyclises leading to 1,5-DAT in a yield of
58%.
NH2
NH2
HN
NH
H2 N
NH2
Cl
NaNO2
HCl
N3
Na2CO3
40C
NH
H2 N
NH2
N
N
N
NH2
N
Cl
Diaminoguanidinium chloride
1,5-DAT (58%)
Nevertheless, the synthesis conditions must be perfectly controlled because the reaction
of nitrous acid with aminoguanidinium is strongly dependent of the pH value as well as the
reactants amounts, otherwise it might lead to the azide derivative which is a very explosive
by-product.
1.2
Aims
The aim of this project is to make the synthetic procedure for DAT safer and with less
steps. The idea is to find a suitable reagent which can simply aminate 5-aminotetrazole. An
efficient synthesis of two nitrophenylhydroxylamines is undergone:
552
2,4-dinitrophenylhydroxylamine (DNPH)
2,4,6-trinitrophenylhydroxylamine (TNPH).
Czech Republic
2006
[Content]
By comparing these two aminating reagents, theoratically the second one is more
advantageous because it has three nitro electron-withdrawing groups, so the amination effect
should be stronger and should lead to a higher amination rate and a higher yield.
2.
EXPERIMENTS
2.3
As a first step, the procedure requires the synthesis of the hydroxylamines precursors[5]
namely:
2,4-dinitro-O-phenyoxyphtalimide (DNPP)
2,4,6-trinitro-O-phenyoxyphtalimide (TNPP)
O
NO2
OH
O2N
(CH3)2-CO
O2N
NO2
Et3N, r.t.
O
O
O
NO2
OH
Cl
O2N
(CH3)2-CO
O2N
NO2
Et3N, 5C
2.4
NO2
NO2
O2N
NO2
NH
NO2
N2H5OH
MeOH, CH2Cl2
HCl, OC
O2N
NH2
NH
553
2.5
Czech Republic
2006
[Content]
Amination reaction
5-aminotetrazole
(5-AT)
is
espacially
soluble
in
alcohols
but
2,4-dinitrophenylhydroxylamine (DNPH) decomposes into 2,4-dinitrophenol when refluxing
in these solvents[7]. This makes the choice of the solvent relatively limited. Therefore, the
amination of 5-AT by DNPH was carried out in DMF at 60 to 70C for 50 hours and should
normally lead to these salts:
NH2
NO2
5
4
N
1N
N
2
H
H2N
N
3
NO2
5-AT
2,4-O-DNPH
DMF, 60-70C, 50 h
NH2
NH2
NO2
5
H2N
1
N
N
2
N
3
1,5 DATH
3.
NO2
H2N
+
NO2
5
4
N
1
N
N
2
N
3
2,5 DATH+
2,4-DNPO-
H
O
NO2
2,4-DNPO-
O-Nitrophenoxyphtalimide derivatives
3.1.1. 2,4-dinitro-O-phenyoxyphtalimide
DNPP is obtained in a yield of 85%, it was purified by recrystallization from EtOH and
fully characterized as follows:
Melting point (Bchi B-540): [188.0-188.8C] which agrees with the reported value of
187.5C
.
O2N
H3
c
b
554
4
6
H6
NO2
5
H5
O
Insensitive Environmentally-compatible
and Non-toxic Replacements for RDX in . . .
Czech Republic
2006
[Content]
Raman:
-NO2
DNFB
N=O
3088
836
1540
1615
1801
3090
1357
1607
1539
Int
1187 1020
837
DNPP
714
C=O
1778
3087
C=O
1714
478
1187
1017
1609
600
NHP
4000,0
3000
Fig 1.
2000
1500
Raman Shift / cm-1
1000
500
50,0
3.1.2. 2,4,6-trinitro-O-phenyoxyphtalimide
TNPP is obtained in a yield of 82%, it was purified by recrystallization from ACN and
fully characterized as follows:
Melting point (Bchi B-540): [212.3-213.7C] matching with the literature value of
213.5C
.
O2N
b
a
c
N
H3
2
O
NO2
2
O2N
3
H3
555
Czech Republic
2006
[Content]
Raman:
=C-H (disubst. arom)
3087
C=O
1714
NHP
3087
Int
1017
714
1187
1609
478
1390
1012 826
697
TNPP
1350 -NO2
N=O
1604 1555
824
3089
TNCB
4500,0
4000
Fig 2.
3000
2000
1500
Raman Shift / cm-1
1000
500
40,0
109.912(9)
Fig 3.
556
Formula unit and labeling scheme for TNPP (Diamond 3 Plot, thermal
ellipsoid represents 50% probability).
Czech Republic
Formula sum
C14H6N4O9
Formula weight
374.223 g/mol
Crystal system
triclinic
Space-group
P -1 (2)
Cell parameters
a = 8.105(1)
= 81.54(1)
Cell volume
720.3(5) 3
Calculated density
1.725 g/cm3
Rint
3.9%
R2
10.9%
3.2
O-(Nitrophenyl)-hydroxylamine derivatives
2006
b = 8.1725(9)
= 85.51(1)
[Content]
c = 12.591(1)
= 60.83(1)
3.2.1. 2,4-dinitrophenylhydroxylamine
The hydrazinolysis of DNPP or B presents some drawbacks: reprecipitation of A which
obliges to filtrate several times, this adds to the problem that B decomposes in the aqueous
phase and must be immediately extracted. In consequence, the yield of DNPH is only about
40%. DNPH was purified by recrystallisation from acetonitrile, its characterization gave the
following:
Melting point (Bchi B-540): [111.1-112.2C] which agrees with the reported value of
112C.
.
NO2
3
O2N
1
5
H5
NH2
H6
557
Czech Republic
Raman:
2006
[Content]
-NO2
1344
2,20
2,0
1316
1,8
1,6
DNPH
1,4
1370
N=O
1612
1,2
Int
1,0
0,8
1518
3261
0,4
3319
0,2
-0,02
4500,0
838
1285
N-H
0,6
1137
4000
Fig 4.
3106
3000
2000
1500
Raman Shift / cm-1
1000
500
mm
tube,
50,0
r.t.)
of
113.45(11)
Fig 5.
Formula unit and labeling scheme for DNPH (Diamond 3 Plot, thermal
ellipsoid represents 50% probability).
Formula sum
C6H5N3O5
Formula weight
199.13 g/mol
Crystal system
monoclinic
Space-group
P 21/a (14)
Cell parameters
a = 10.0146(7) b = 14.558(1)
c = 10.6425(7)
= 91.77(1)
558
Czech Republic
Cell volume
1550.9(3) 3
Calculated density
1.706 g/cm3
R1
4.74%
wR2
14.54%
GOOF
1.109
2006
[Content]
3.3
DAT Salts
3.3.1. DATH+ DNPOAfter removing DMF under reduced pressure, the residue was analysed by Raman and
we obtained the spectrum below:
-NO2
1340
Raman:
2,15
2,0
1,8
1,6
1,4
1,2
1265
Int 1,0
840
0,8
C=N- 1540
1626
0,6
0,4
0,2
N-H
3318
3213
1145
434
-0,02
4000,0
3000
Fig 6.
2000
1500
Raman Shift / cm-1
1000
500
50,0
This spectrum shows resembling reflexions compared to those of the starting materials.
The residue was also analysed by Mass Spectrometry (FAB+, xenon, 6 keV, m-NBA
matrix):
559
Czech Republic
2006
[Content]
4.
CONCLUSION
5.
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. We
are indebted to and thank Dipl.-Chem. Jrg Stierstorfer for the X-ray work. CD also wants to
thank the LMU Munich for a postdoctoral scholarship.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
560
R.
RAAP:
Amination
of
tetrazoles
with
hydroxylamine-O-sulfonic
acid:
1- and 2-aminotetrazoles, Can. J. Chem, No. 47(19), 3677-81, 1969
P. N.GAPONIK and V. P. KARAVAI: Synthesis and properties of 1,5-diaminotetrazole,
Khim. Geterotsikl. Soedin., No. (12), 1683-6, 1984
R. STOLLE, H. NETZ, O. KRAMER, S. ROTHSCHILD, E. ERBE and O. SCHICK:
Derivatives of 1-aminotetrazoles, J. Prak. Chem., No. 138, 1-17, 1933
(a) J. C. GALVEZ-RUIZ, G. HOLL, K. KARAGHIOSOFF, T. M. KLAPTKE,
K. LHNWITZ, P. MAYER, H. NTH, K. POLBORN, C. J. ROHBOGNER, M. SUTER and
J. J. WEIGAND: Derivatives of 1,5-diamino-1H-tetrazole: a new family of energetic
heterocyclic-based salts, Inorg. Chem., No. 44(12), 4237-4253, 2005
(b) J. C. GALVEZ-RUIZ, G. HOLL, K. KARAGHIOSOFF, T. M. KLAPTKE,
K. LHNWITZ, P. MAYER, H. NTH, K. POLBORN, C. J. ROHBOGNER,
M. SUTER and J. J. WEIGAND: Derivatives of 1,5-diamino-1H-tetrazole: a new family of
energetic heterocyclic-based salts, Inorg. Chem., No.44(14), 5192-5208, 2005
A. ROUGNY and M. DAUDON: Use of N-hydroxylimides for the synthesis of primary
alkoxylamines, Bulletin de la Socit Chimique de France, No. 5-6, 833-8, 1976
C. LEGAULT and A. B. CHARETTE: Highly Efficient Synthesis of
O-(2,4-Dinitrophenyl)hydroxylamine. Application to the Synthesis of Substituted
N-Benzoyliminopyridinium Ylides, J. Org. Chem., No. 68(18), 7119-22, 2003
A. S. RADHAKRISHNA, G. M. LOUDON and M. J. MILLER: Amination of ester enolates
with O-(2,4-dinitrophenyl)hydroxylamine, J. Org. Chem., No. 44(26), 4836-41, 1979
Czech Republic
2006
[Content]
Abstract
Preparation methods via nitration of isowurtzitane acyl derivatives, physicochemical
and spectral properties of a new isowurtzitane compound - 4,6,8,10,12-pentanitro-2acetyl-2,4,6,8,10,12-hexaatetracyclo [5.5.0.05,903,11] dodecane ( PNAIW), are described.
INTRODUCTION
A quest for organic compounds which, owing to their structure, crystallize with the
highest possible spatial packing have resulted in the development of synthesis of compounds
of cage structure. Instances of such compounds are: adamantants, cubans, wurtzitans and
isowurzitans. Aside from their characteristic feature mentioned, strain bonds are present in
the structure of these compound which additionally exhibit some properties characteristic of
aromaticity.
The compounds has for long been am attractive research object, as they provide a great
many new data for theory of chemistry. Moreover, they already find practical uses among
others as high energy materials. Ranking in this group are nitro derivatives of isowurtzitan,
e.g: 2,4,6,8,10,12-hexanitro-2,4,6,8,10,12,hexaatetracyclo[5.5.9.95,903,11] dodecane i.e.
HNIW. Numerous reports and patents are devoted to the preparation of this compound.
Our studies on the synthesis of this compound showed that under specific conditions of
the HNIW synthesis an intermediate product, not described yet in literature, can be obtained
in
variable
yield,
viz.
4,6,8,10,12-pentanitro-2-acetyl-2,4,6,8,10,12hexaatetracyclo[5.5.0.05,903,11]dodecane i.e. PNAIW. The synthesis conditions, along with
major properties of PNAIW and its spectral data, are reported in this communication.
561
Czech Republic
2006
[Content]
EXPERIMENTAL
Preparation of PNAIW as a reference compound
Extra
Reagent
Reaction Time
Weight of Derivative [g]
99% H2SO4
CH3COOH
20% oleum
N2O4
TADBIW
5h
1.03
Derivative
TADNOIW
5h
0.79
0.2 mol (19.6 g)
0.2 mol (12 g)
23.5 g
TAIW
1h
0.67
0.2 mol (19.6 g)
TADBIW- tetraacetyldibenzylhexaazaisowurtzitane
TADNOIW tetraacetyldinitrosohexaazaisowurtzitane
TAIW - tetracetylhexaazaisowurtzitane
562
Czech Republic
2006
[Content]
TADBIW Nitration with a Mixture of 99% Nitric Acid and 99% H2SO4 with
a 20-% Addition of NaNO2
In a reaction flask of a 50-cm3 capacity equipped with a thermometer, a mechanical
agitator, a dropping funnel and a cooling bath, 12 g nitric acid (0.2 mol) was placed. Once
the stirring was started the flask contents was cooled to 5C with the dry ice / acetone
mixture. At that temperature was added, on vigorous stirring, 1.03 g (0.02 mol) TADBIW
during 5 min. Then 19.6 g (0.2 mol) 99% H2SO4 was added dropwise. The sulfuric acid
adding was continued for 10 min while maintaining the reaction temperature below 5C.
On addition of reactants the funnel was replaced by a reflux condenser, the cooling bath
by a hemispherical heater and the reaction system was brought to a required temperature of
80C and then the reaction was carried out for 5 h. After that time the reaction mixture was
cooled to 5C (on a cooling bath) and 2.5 g NaNO2 was added. Then the temperature was
made to rise to 25C and under these conditions the reaction was carried out for 20 h.
On completion of the reaction the reaction mixture was diluted with water (with ice).
The reaction product was a light-yellow solid which was filtered off, washed with methanol
and dried in the air.
563
Czech Republic
2006
[Content]
condenser), whereupon the flask contents was diluted by pouring it over ice. The precipitate
formed was filtered off and washed with methanol.
Analysis of Product
For the determination of the content of individual products of the reaction of nitration of
isowurtzitane derivatives liquid chromatography technique was used with a Shimadzu model
10A instrument and a Supelco 300-5 C018 150 mm column and a 50:50 acetonitrile/ water
eluent system using a UV detector at a wavelength of 245 nm.
IR measurements were carried out with a Specord M-80 spectrometer. The NMR
spectrum was recorded with a Gemini 200BB apparatus and the DTA TG thermograph
was recorded with a Derivatograph-C
Results
Pentanitroacetylisowurtzitane can be obtained after Reaction 1:
W(X)2 +
HNO3 (HNO3/B)
(1)
XN
O2N
NX
NO2
N
NO2 O2N
O2N
N
NO2
HNIW
O2 N
NO2 O2N
O2 N
NO2
NO2
Mono-oxa
TEX
W(X)2
Fig. 1. Spatial structure of isowurtzitane W(X)2 and HNIW, Mono-oxa and TEX
564
Czech Republic
2006
[Content]
Substrate
Nitrating
mixture
Reaction
temperature
[oC]
HNIW
[%]
TEX
[%]
NonoOXA
[%]
PentaNitro
[%]
NO2BF4
sulfolane
25 - 60
90,3
5,2
25 80
80,7
1,4
8,4
3,9
I HNO3
H2SO4
II NaNO2
25 -80
4,9
1,0
1,2
11%
HNO3
N2O4
40
5,4
2,8
28,4
HNO3
15,8
30,8
HNO3
H2SO4
75,3
12,4
HNO3
oleum
80,8
4,8
HNO3
CH3COOH
12,9
21,5
HNO3
H2SO4
20
10,2
__
__
40,3
HNO3
5.9
__
__
75.0
I N2O4
TADBIW
TADNOIW
TAIW
II HNO3
565
Czech Republic
2006
[Content]
mixture lowers the PNAIW yield in favor of HNIW. Analogous results are obtained if a
HNO3/CH3COOH nitrating mixture is used.
The best PNAIW yield (75%) is observed for the nitration of TAIW with 99-% HNO3
alone at 0C for 1 h. Here again a raised temperature and an addition of sulfuric acid reduces
the PNAIW yield.
PANIW is a crystalline solid decomposing over the temperature range of 258260C
fig.2.
DTA
TG
50
115
180
245
310
375
566
Czech Republic
2006
[Content]
12
N+
O
O
+
10
N+
O
N+
Fig. 4.
H-NMR of PNAIW
is typical of isowurtzitanes where the signals from methylene bridge hydrogens lies in
the range of the signals of aromatic protons: of the six-ring bridges at = 8.3 singlet 4H and
the protons of the bridge that links the five-member rings at = 8.1 two singlets 1H. The
spectrum differs from the PMR spectrum of HNIW which is composed of two singlets 4H
and 2H. Singlets 4H are assigned to two dimethylene bridges at six-member ring, and singlet
2H corresponds to the dimethylene bridge that clamps the five-member rings. The presence
of these singlet is evidence for the fact that the four hydrogen atoms of the six-member ring
are in the same environment. The presence of singlet 4H and the presence of 2 singlet 1H in
the PMR spectrum of PNAIW suggests that the 4H hydrogen atoms are in the same
enivironment, whereas 1H hydrogen atoms are not equivalent.
567
Czech Republic
2006
[Content]
90
T%
60
30
0
4000cm -1
3400
2800
2200
1600
1000
400
(2)
Nitration of PNAIW with this acid in the presence of acetanhydride results in a decay of the
isowurtzitane structure to afford a new cage structure -TEX in a 21-% yield -Reaction 3
PNAIW + HNO3 + (CH3CO)2O TEX
(3)
REFERENCES
[1]
[2]
[3]
[4]
[5]
568
Nielsen A.T., Nissan R.A., Vanderah D.J., J.Org.Chem. 55, 1459-1466, 1990.
Bellamy A.J., Tetrahedron 51, 4711 4722,1995.
Tomotsu K., Masahiro T., Masanori I., EP 0 753 519 A I, 1997.
Bellamy A.J., 30th Int.Ann.Conf.ICT, 30, 109.1.-109.14, 1999.
Shu-Sen C., Wen-Ge-Q., Yong-Zhong Y., Chin.J.Chem. 18(5), 756-758, 2000.
Czech Republic
2006
[Content]
Abstract
This work concerns to summarizing of literature experimental results of initiation,
deflagration and detonation in the metal dust/air systems. Results of works of the
Institute of Chemical Physics of Academy of Sciences of Russia on initiation of
detonation of dust/air systems and works of Novosibirsk Scientific Center on ignition of
aluminum and magnesium particles in a shock tube and influence on this ignition of
organic substances are in detail considered. The main conclusions is that particles of
magnesium are ignited in a shock wave much easier, than aluminum and the most
dangerous additives to aluminum from organic liquids are nitrates of fat spirits.
Keyword: Ignition, ignition delay, initiation of detonation
1.
INTRODUCTION
2.
ICP WORKS
The central position in a row of scientific materials related to ignition and initiation of
detonation in the metal dust/oxidizing gas clouds presumably must be assigned to the work
by A.A. Borisov, B.A. Khasainov et al. carried out in the Institute of Chemical Physics of
Academy of Sciences of Russia (ICP) [1].
The shock tube made of 122 mm caliber 4.2 m long gun barrel was used for
investigation of detonation processes in aluminum/air suspensions in [1]. The flaked
aluminum of Russian production, named PAP, was taken for the experiments. Again, it is
not said what kind of aluminum, PAP-1 or PAP-2, is used, but thickness of the flakes, about
1 m, is reported.
569
Czech Republic
2006
[Content]
The powders were placed in trays inside the barrel, with 25 cm distance between them, ~
0.5 s before the experiment, the barrel was blown off by compressed air that formed the
dust/air suspension. The quantity of the powder deposited on the walls during the time of ~
0.5 s was specially estimated and appeared to be only 10-20 % of the original mass of the
powder in the trays.
Initiation of detonation was realized by a detonating cap and a sample of RDX placed at
the open end of the tube. Registration of a process inside the tube was implemented by
means of the streak-camera through the windows in the walls on the last 2 meters of the tube,
and two pressure gages located at 1.25 and 3.25 m from the open end.
The wave velocity was measured in every run and if it were constant on the last two
meters of the tube the process was considered detonation. When it diminished the initiator
was estimated to be too small to produce the steady-state detonation, and in the subsequent
run it was increased, till the steady state process was achieved. The energy of the initiator
was calculated on the base of the theory of the point explosion, and as to the system of RDX
plus detonator taken for initiation it corresponded to about 3.3 MJ per gram of RDX.
E, MJ/m
0
100
2
200
300
400
G, g/m
Fig 1.
500
600
Results of the experiments in the tube 122 mm dia are presented in Fig. 1. The critical
energy of initiation for the flaked aluminum/air suspension (curve 1) of 330 g/m3
concentration is 3.4 MJ/m2 what corresponds to 12 g of RDX used as the initiator. For the
ultra-dispersed aluminum, UDA (~ 1 m dia) the initiation energy is much less than for the
flaked aluminum (curve 2 in Fig. 1, the minimal energy of initiation is less than 0.3 MJ/m2)
what could be explained presumably by the fact that the flakes are covered by the thin film
of aliphatic acids that form an additive obstacle (in addition to the film of aluminum oxide)
to the heating and inflammation of the flakes, in comparison with the UDA particles
deprived of the carbooxyhydrates film. The photoregistrograms of detonation process (Fig.
2) show the well formed spin-detonation process with a length of the spin-wave about 0.4 m
independently on concentration of the flaked aluminum ranged from 200 to 400 g/m3.
570
Fig 2.
3.
Czech Republic
2006
[Content]
571
Czech Republic
2006
[Content]
Fig 3.
572
Fig 4.
Czech Republic
2006
[Content]
Ignition delay of Magnesium powders. 1-d=4063 m; 2-d=6380 m; 3calculation (T=1368K, d=75 m).
Ignition of the magnesium particles is more uniform process (Fig. 4). Ignition delay as
well as in the case of aluminum does not depend on dimensions of the particles, and no
specific deviations for the overall experimental dependence is noted. The calculated lines are
deviated from the experimental points in Fig. 4 but it rather moderate difference.
Veyssiere [3] gives the method of calculation of ignition temperature developed further in
and compares the results with a formula for calculation of ignition temperature of
aluminum particles in a hot gas [8]:
[4-7]
(1)
4.
573
Czech Republic
2006
[Content]
Fig 5.
Ignition delay of droplets of aviation fuels T-15 (1) and TA (2), tridecane
(TD) (3), fuel oil, DTL (4) and droplets of nitrates: cyclohexylnitate (5),
isopropylnitrate (6), nitrates of alyphatic alcohols (7).
Together with the hydrocarbons, the data for the potential promoters cyclohexylnitate
(CHN, C6H11ONO2), isopropylnitrate (IPN, C3H7ONO2), and a mixture of the nitrates of
alyphatic alcohols (NA, CnH2n+1ONO2, at n=611).
Ignition delays for the kerosene TA and T-15 and TD are about the same:
574
ign=2.5exp(7220/T),
s, EA=60 kJ/mol
Czech Republic
2006
[Content]
ign=
0.014exp(11900/T), s, (EA=99kJ/mol),
CHN:
ign=
0.0006exp(11900/T), s, (EA=99kJ/mol),
ign=
75exp(3000/T), s, (EA=25kJ/mol).
NA:
Promoting influence of the nitrates follows from Fig. 6. One can see that all of the
promoters works in about the same manner, at a high temperature, 1500 K. At a lower
temperature difference appears. At T=1000 K only NA and IPN allow the fuel oil be ignited
at a time less 1 ms, and NA, in visible contradiction with Fig. 10, works better than the IPN
does.
Ignition delays of the fuel oil being activated by addition of NA are presented in Fig. 7.
Not only ign decreases at increasing of the NA concentration, cNA, in the solution but the
apparent energy of activation dimensions, at that
cNA=0% EA=60 kJ/mol,
cNA=5% EA=56 kJ/mol,
cNA=10% EA=46kJ/mol,
cNA=30% EA=36 kJ/mol,
obviously being directed towards the value of 25 kJ/mol which is characteristic for the
neat NA.
Mixture of 70% fuel oil and 30% NA is characterized by the Arrhenius type dependence
of the form:
ign=6exp(4330/T),
s.
Addition of the mixture to spherical aluminum powder ASD1 (99.2% Al, 0.2% Fe,
0.02% H2O, fractions of 3-5 m, 10-14 m and 14-20 m are used in the experiments).
Streak-camera photographs are presented in Fig. 7. Aluminum of 3-5 m fraction burning is
a subject of Fig. 7 experiments. Mass of the liquid in the experiments was ml=7 mg, relation
=mal/ml has been changed. Temperature was T1500 K. At =0 (no aluminum) Fig. 8a
liquid burns with quickly estinguishing flame. Addition of aluminum (mal=10 mg Fig 8b and
mal=10 mg in Fig 8c) leads to formation of voluminous flame of aluminum burning.
575
576
Czech Republic
2006
[Content]
Fig 6.
Fig 7.
Ignition delay of fuel oil with addition of NA: 1-100% NA, 2-100% fuel
oil, 3-95%fuel oil+5%NA; 4-90% fuel oil+10% NA, 5-70% fuel oil+30%
NA.
Fig 8.
Czech Republic
2006
[Content]
Fig 9.
577
Czech Republic
2006
[Content]
One would suggest, however, that this lower part of the =10 line is perhaps imaginary.
But the upper part in comparison with the smaller dependence has absolutely definite
sense: the aluminum ignition is essentially promoted by mixing it with the small quantity of
the DTL/NA solution. The small line (points 1 and 2 in Fig. 9) corresponds to the
relation:
ign=14(4330/T),
s, (EA=36 kJ/mol).
The relation is close to that for fuel oil/NA (70/10) mixture but preexponential factor is
about two times more.
ign=1.210
-3
M, kg Tetryl
The apparent activation energy for the almost dry aluminum powder is close to those for
heptane/oxygen (191 kJ/mol) [13] and methane/oxygen/nitrogen [187 kJ/mol] [13].
0.1
0.01
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
578
10
5.
Czech Republic
2006
[Content]
CONCLUSIONS
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
11
Borisov, A.A., Khasainov, B.A., Veysser, B., Saneev, E.L., Fomin, I.B., Khomik, S.V.,
Detonation of aluminum suspension in air and oxygen, Chemical physics, 1991, Vol. 10, 2,
pp.250-273.
Boiko, V.M., Lotov, V.V., Papyrin, A.N., Ignition of suspensions of metal powders in gas in
reflected shock waves, Combustion, Explosion and Shock Waves, 1989, Vol. 25, 2, pp. 6775.
Veyssiere, B., Ignition of aluminum particles in a gaseous detonation, Shock Waves,
Explosions, and Detonations, 1983, Vol. 87, 362-375.
Khasainov, B.A., Vessiere, B., Archivum Combustionis, 1987, Vol. 7, 3-4, p. 333.
Khasainov, B.A., Vessiere, B., Dynamics of explosions, AIAA Progress in Astronautics and
Aeronautics, New York, AIAA, 1988, Vol. 114, p. 284.
Veyssiere, B., Khasainov, B.A., Combustion and Flame, 1991 (in press).
Veyssiere, B., Dynamics of explosions, AIAA Progress in Astronautics and Aeronautic, NewYork, AIAA, 1986, Vol. 106, p. 522.
Ozerov, E.S., Skvortsov, I.I., Burning of magnesium particles conglomeration, Combustion,
Explosion and Shock Waves, 1971, Vol. 7, 2 pp. 227-232.
Wierzba, A., Kauffman, C.W., Nicholls, J.A., Ignition of partially shattered liquid fuel drops in
reflected shock-wave environment, Combustion Scientific Technology, 1974, Vol. 9, 5/6, pp.
233-245.
579
Czech Republic
2006
[Content]
[10] Borisov, A.A., Gel'fand, B.E., Eremenko, L.T., Features of ignition of inflammable liquids,
Report of Acad. of Science of USSR, 1979, Vol. 247, 5, pp. 1176-1179.
[11] Borisov, A.A., Gel'fand, B.E., Timofeev, E.I., et al., Self-ignition of atomized liquid fuel in
gaseous medium, Flames, Lasers, and Reactive Systems, Prog. Astronaut. Aeronaut, 1991, Vol.
132, pp. 205-219.
[12] Boiko, V.M., Lotov, V.V., Papyrin, A.N., Ignition of liquid fuel drops in shock wave,
Dynamics of Deflagration and Reactive Systems: Heterogeneous Combustion, Progress in
Astronaut. Aeronaut, 1991, Vol. 132, pp. 205-219.
[13] Boiko, V.M., Poplavsky, S.V., Ignition of aluminum powders/liquid hydrocarbon fuels
mixtures in air, Combustion, Explosion and Shock Waves, 1999, Vol. 35, 1, pp. 35-40.
[14] Atkinson, R. and Bull, D.C., Kinetic modeling of ethane/air detonability, Shock Waves,
Explosions, and Detonations, Bowen, Mabson, Oppenheim, and, Soloukhin, Editors, 1983, Vol.
87, pp.318-332.
580
12
Czech Republic
2006
[Content]
1.
INTRODUCTION
1.1
Perforating is the process of creating holes when an oil or gas well is completed, casing
and cement isolate the well bore from surrounding formations. Perforating is used to
establish fluid communication between the well bore and formation so that production
operations can proceed through perforation-hole in casing and cement.
This process can be done with use of perforating guns-devices used to perforate wells.
Sample of perforating gun is given on fig. 1. A typical gun consists of a collection of
explosive charges, a charge carrier, a detonating cord and a detonator. The chargers are load
in the charge carrier and they are connected to the detonating cord, which explosively
couples the charges. The gun is lowered into the well by wire line to the depth at which
perforations are desired.
Fig 1.
Perforating gun
581
1.2
Czech Republic
2006
[Content]
The most important component of perforating gun is the perforating charge. The charge
utilizes a secondary high explosive to propel a stream or jet of high speed metallic particles
that penetrate the casing, cement and rock formation.
Perforating charges are lined shaped charges. A cavity is shaped in the explosive and is
lined with a metallic material. Liner shaped charge have four components: the charge case,
the liner, the main explosive, and the primer.
The charge case is container that houses the other charge components. It must withstand
high temperatures and pressures. Common case materials include machined steel, cold
formed steel, die-cast aluminium and ceramics. The liner furnishes the mass needed for the
jet to penetrate casing, cement and formation. The pressures exerted on the liner when the
main explosivedetonates cause the linear to colalapse and form the jet. The shape of the
linear and its thickness and composition greatly influence the depth, diameter, and
effectivenes of the perforation. The linear is usually conical or parabolic in shape. Themain
explosive provides the energy needed to produce the jet. The mass, distribution, and
detonation velocity of the main explosive greatly affects charge performance. Granular, high
order, secondary explosives such RDX, HMX, HNS and PYX are used. The proper amount
of explosive is measured into the charge case; the lineris placed on the top of the explosive;
then pressure is applied to the case and linear to compact the explosive. The shapes of the
linear and of the case determine the distribution of the explosive. The amount of main
explosive ranges from 2 g in charges loaded in small diameter open carriers to over 30 g in
charges loaded in large diameter hollow carrers used in casing application. The primer, or
buster is composed of a small amount of explosive and its main purpose is to transfer the
shock wave from the detonating cord to main explosive. After firing, the liner collapses,
staring at the apex. The particles begin to converge along the cone axsis to form the needlike,
high speed jet. The jet shoots from the charge cone and, upon contact with the initial target,
can have pressures at the leading tip over 400 kbar. Fig. 2 shows scheme of perforator lined
charge.
Fig 2.
The perforation of the target is accomplished by the stream of metallic jet particles
radially displacing the impacted target material and thereis no fusing, glazing, or scorcing of
materials surrounding the perforation. In the rock formation, the material immediately
surrounding the perforation is crused and compacted in crused zone. Since the crused zone
Since the crused zone usually exibits reduced permeability when compared to the
582
JET Velocity and the Effects of the Shaped Charge Perforator for Use in the Oil . . .
Czech Republic
2006
[Content]
undisturbed formation, it can have significant effect upon the productivity of the well. The
jet energy is quickly expended and the penetration process ceases. The length and diameter
of the perforation depend upon the target material and upon the energy distribution in the jet.
Fig.3 shows effects of perforation in formation.
Fig 3.
2.
Testing of the perforators have been done with the aim of establishing real velocity of
detonation of the cumulative stream-jet in relation to the declared values, and determining
relation between the velocity of detonation of the raptors and penetration depths of the
stream in the environment, through the bore-hole to surrounding formations.
2.3
Technical data of lined shaped raptor, produced by Owen oil tools producer 3 3/8:
type: 3 3/8 TAG-6SPF, 60 DEGREE RAPTOR NT3
charge type: SDP 3375-411 NT 3
explosive weight (g): 25 HMX
entry hole (mm): 10,67
penetration (mm): 945
Lined shape raptor3 3/8, Owen is shown on fig.4.
Fig 4.
583
2.4
Czech Republic
2006
[Content]
Fig 5.
Concrete target
Fig 6.
Barea target
The results of the tests can serve as a basic for generalcomparasion of perforator
performance. Actual conditions at the time of perforating a well are most likely but not the
same as those existing during testing; casing, fluids, cements, formation types and pressures
are different. In order to evaluete relationship between the perforating parameter, deepth and
diameter of penetration and jet velocity of detonation, VOD of the jet was measured.
2.5
584
JET Velocity and the Effects of the Shaped Charge Perforator for Use in the Oil . . .
Czech Republic
2006
[Content]
fibre cables. In the measuring, timer calculates velocity on basis of the time and the distance
between the probes. Type of the electronic stop-timer is Explomet-FO-2000 of the
following characteristics: max measurable velocity 10.000 m/s, time interval from 0,1 do 10
s, accuracy +/- 0,1 s.
2.5.3. Measurement description
For the measurements needs, device has been constructed with possibility of setting up of
optic probes on the exactly determined distance. The device consists of: steel tube with holes
for probes, base and bearing plate for the perforators. Distance between the optic probes is
56 mm. Distance between the bottom of perforators and the first optic probe is 56 mm.
Measurement with sample no. 1 has been done in one time interval on the distance of 560
mm in order to confirm the measuring method. Measurements 2-5 are done for the purpose
of the determining velocites distribution of the stream in respect to the length of stream.
Scheme of the measurement principle is given is on the picture 7 and the photo of the
measuring equipment and the perforators on the picture 8.
Fig 7.
Measurement scheme
Fig 8.
Measurement equipment
585
Czech Republic
2006
[Content]
Interval
1
2
3
4
5
Interval
1
2
3
4
5
3.
Measurement no. 1
L(mm)
t(s)
v (m/s)
560
90,2
6208
Measurement no. 2
L (mm)
t (s)
v (m/s)
56,00
10,2
5490
56,00
9,9
5656
56,00
56,00
56,00
Measurement no. 4
L (mm)
t (s)
v (m/s)
56,00
8,2
6829
56,00
8,9
6292
56,00
7,5
7466
56,00
65,0
861
56,00
-
Measurement no. 3
L (mm)
t (s)
v (m/s)
56,00
16,1
3478
56,00
8,6
6511
56,00
9,5
5894
56,00
51,5
1087
56,00
265,7
210
Measurement no. 5
L(mm)
t (s)
v (m/s)
56,00
8,7
6436
56,00
8,9
6292
56,00
8,0
7000
56,00
10,4
5384
56,00
7,9
7088
CONCLUSION
In the measurement no. 2, 3 and 4 deviation of the jet occurred. That caused failure of
the measurement in some intervals. For the measurement no. 5, some changes in
measurement equipment have been done and that result with VOD data for all of intervals.
The results of measurements pointing on the conclusion that applied devices are
appropriate for measurement of the jet velocity. Measured velocities are in the range that
response to the manufacturer declared values (6000-7000 m/s).
Measuring equipment used in testing have properties for measuring maximum jet length
toward 560 mm. With reconstruction of the device with maximum length up to 1000 mm
(range of declared perforation length) is possible to determinate effect of shape lined chargeraptor, on the simply way without performing complicated tests.
REFERENCES
[1]
[2]
[3]
[4]
586
JET Velocity and the Effects of the Shaped Charge Perforator for Use in the Oil . . .
Czech Republic
2006
[Content]
Abstract:
In this paper, the thermochemical calculations (thermochemical properties, reaction
products nozzle, heat of explosion) and experimental study results (deflagration point,
thermal analysis and sensitivities to friction and impact, heat of combustion) of
composite solid propellants containing DADNE or NTO particles are described. The
deflagration point, the impact sensitivity and the friction sensitivity were determined by
the UN methods. Differences in properties of the propellants with DADNE or NTO are
presented.
Keywords: composite solid propellant, DADNE, NTO, thermochemical software,
sensitivity of energetic materials
1.
INTRODUCTION
587
Czech Republic
2006
[Content]
Composite solid propellants containing of HMX or RDX particles offer the advantages
of low flame temperature and low molecular mass combustion products, as well as reduced
infrared emission. The reduced infrared emission can be obtain by elimination of CO2 and
H2O from the combustion products[2].
Hence it was of importance to study the performance of propellants with other
nitrocompounds novel insensitive energetic materials i.e. 3-nitro-1,2,4-triazo-5-one (NTO)
or 1,1-diamino-2,2-dinitroethylene (DADNE, FOX-7).
In this work, results of thermochemical calculations and experimental study of
composite solid propellant (AP/PBAN-Epoxy resin) containing 20% of DADNE or NTO are
presented.
2.
PERFORMANCE CALCULATIONS
2.1.
The basic properties of ingredients selected for our studies are listed in Table 1. The
thermochemical values for the ingredients are taken from literature[4-9].
Sum Formula
Hf
[kJ/mol]
PBAN/Epoxy resin
C64.466H85.819N1.421O7.457
-263.01
1.027
-929.00
DOA
C22H42O4
-263.37
0.925
-1215.03
AP
NH4Cl4
34.04
1.950
-296.77
DADNE[4]
C2H4N4O4
-21.61
1.885
-133.89
NTO[5]
C2H2N4O3
-24.60
1.930
-59.83
Aluminum
Al
-88.95
2.702
0.00
Ferrocene
C10FeH10
-223.61
1.490
154.89
588
Czech Republic
2006
[Content]
Compounds, %
AP
PBAN/Epoxy resin
DOA
Al
Ferrocene
DADNE
NTO
thermochemical properties
density, g/cm3
mixture ratio, O/F
equivalence ratio
oxygen balance, %
temperature chamber, K
mole number, mole/kg
Ifr/Ieq, Ns/kg
Ivol, Froz/Eq., Ns/dm3
characteristic velocity (C*), m/s
thrust coefficient, CF
Propellant
1
58.0
16.2
3.6
2.0
0.2
20.0
0.0
58.0
16.2
3.6
2.0
0.2
0.0
20.0
1.642
1.381
0.337
-38.89
2537
39.31
2232/2269
3666/3726
1397
1.598
1.649
1.381
0.333
-39.49
2482
38.75
2190/2220
3611/3660
1371
1.597
78.0
16.2
3.6
2.0
0.2
0.0
0.0
1.652
3.545
0.489
-27.760
2659
39.50
2281/2330
3767/3849
1431
1.593
589
Czech Republic
2006
[Content]
25
Weight %
20
15
10
0
CO2 (g)
H2O (g)
N2 (g)
CO (g)
NH3 (g)
CH4 (g)
HCl (g)
FeCl2 (g)
Al2O3 (s)
30
Weight %
25
20
15
10
0
CO2 (g)
H2O (g)
N2 (g)
CO (g)
NH3 (g)
CH4 (g)
HCl (g)
Al2O3 (s)
FeCl2 (l)
590
Czech Republic
2006
[Content]
30
Weight %
25
20
15
10
0
CO2 (g)
H2O (g)
N2 (g)
CO (g)
NH3 (g)
CH4 (g)
HCl (g)
FeCl2 (g)
Al2O3 (s)
3.
EXPERIMANTAL RESULTS
The propellant 1 with good processing and mechanical properties was chosen as
baseline for this investigation. The purpose of examination in this study were high-energy
composite propellants containing 80% of solid components. Grain size distribution of AP,
DADNE and NTO crystals were determined by IPS-U (Infrared Particle Size) analyzer
version 8.12. The principle of the analyzer consists in measuring of changes of infrared rays
steam that diffuse on moving particles of material tested.
Bimodal AP was used as an oxidizer (small sized and large sized). Content of the small
sized (fine-grained) AP was 17.4% of the formulations. Particle diameters of the fine-grained
AP were below 50 m (Fig 4). Particle diameters of the large sized (coarse) AP were below
520 m (Fig 5). As ballistic modifiers (combustion catalyst) was used ferrocene (0.2%). As
insensitive explosives (new energetic materials) were used DADNE or NTO with content in
propellant of 20%. Particle diameters of DADNE and NTO are shown in Fig 6 and Fig. 7,
respectively. The content of a plasticizer (DOA) was 3.6%. The binder was based on PBAN
cured with epoxy resin.
Propellant formulations were mixed in a 1.5 - liter vertical mixer at 75 oC. Curing
required 48 h at 75 oC for its completion.
591
Czech Republic
2006
[Content]
4649
4346
4143
3841
3538
3335
3033
2730
2527
2225
1922
1719
1417
1114
911
69
36
20
18
16
14
12
10
8
6
4
2
0
03
Particle concentration, %
Particle diameter, m
Fig. 4. Numeric particle size concentration as a function of particle size for fine
AP.
Particle concentration, %
8
7
6
5
4
3
2
1
502515
476489
449462
423436
396409
370383
343356
316330
290303
263277
237250
210224
184197
157171
131144
104118
7891
5164
2538
011
Particle diameter, m
Fig. 5. Numeric particle size concentration as a function of particle size for coarse
AP.
592
Czech Republic
2006
[Content]
Particle concentration, %
25
20
15
10
5
8390
7583
6875
6068
5360
4553
3845
3038
2330
1523
715
07
Particle diameter, m
Fig. 6. Numeric particle size concentration as a function of particle size for
DADNE.
10
8
6
4
151158
143151
136143
128136
121128
98105
9098
8390
7583
6875
6068
5360
4553
3845
3038
2330
1523
07
715
113121
2
105113
Particle concentration, %
12
Particle diameter, m
Fig. 7. Numeric particle size concentration as a function of particle size for NTO.
3.1
Deflagration point
Deflagration points were determined by heating of the sample in a glass test tube in
Woods metal bath at a heating rate of 5 oC/min. Temperature at which the sample got
ignited decomposed, was recorded as deflagration point[9-10]. Three samples of mass 0.5 g
were tested simultaneously. The results of the deflagration point study of propellants 1 3
are presented in Table 3.
593
3.2
Czech Republic
2006
[Content]
Heat of combustion
Heat of combustion represents the caloric equivalent of the total combustion energy
of a given substance. Its value depends only on the composition of the material and not on
any other propellants, such as for example loading density. Combustion experiments are
conducted in a stationary bomb calorimeter with an automatically controlled adiabatic jacket
and without oxygen atmosphere (pressure about 400 Pa). The adiabatic bomb IKA
calorimeter system C 4000 was used to determine the heat of combustion. The calorimeter is
calibrated by burning standard propellant to determine its effective energy equivalent. The
results of the heat of combustion study of propellants 13 are presented in Table 3.
3.3
Thermal analysis
3.4
Impact sensitivity
The method (BAM Fallhammer) was used to measure the impact sensitivity[9-11] of
composite solid propellants to drop weight impact. A sample of the tested propellant was
placed in the piston device, which consists of two steel rollers and a hollow cylinder (collar).
The sample was scooped by a special 40 mm3 measuring cup. The measured sample was
placed into the open piston device. The sample response was observed. Drop weight having
a mass of 5 kg was used. The impact of the drop weight, via the piston device, onto sample
may result in either its initiation or noninitiation, depending on the sensitivity of the
propellant, the weight mass, and its drop height. The study results are shown in Table 3. The
samples were tested under the 50% probability of explosive reaction. The results obtained
according to the impact sensitivity test are shown in Fig. 9.
3.5
Friction sensitivity
The Julius Peters method (BAM Friction Apparatus) was used to measure the
friction sensitivity[9-11] of composite solid propellants. Approximately 10 mm3 of sample was
placed on the rough porcelain plate which was attached to the sliding carriage of the
apparatus. The occurrence of the initiation was observed through the sound effects
(crackling), appearance of smoke, or by the characteristic smell of the decomposition
products. The results of the friction sensitivity study of propellants 1 3 are presented in
Table 3. The samples were tested under the 50% probability of explosive reaction. The
results obtained according to the friction sensitivity test are shown in Fig. 10.
594
Czech Republic
2006
[Content]
595
Czech Republic
2006
[Content]
Propellant
Deflagration point, oC
Differential thermal analysis, oC
Tonset
Tm
Heat of combustion, J/g
Heat of explosion, J/g
Impact sensitivity, J
Friction sensitivity, N
* - calculated value[3]
1
243
240
245
2
218
220
219
3
213
213
213
245
252
5038
4934*
35
168
226
232
4207
4574*
35
144
219
228
3919
4312*
25
168
Probability of explosive
reaction [%]
100
75
50
25
0
10
20
30
40
50
propellant 2
propellant 3
150
200
250
300
350
propellant 2
propellant 3
CONCLUSIONS
On the basis of the thermochemical calculations and experimental results the following
can be concluded:
596
Czech Republic
2006
[Content]
REFERENCES
[1]
G. P. SUTTON and O. BIBLARZ, Rocket Propulsion elements, John Wiley & Sons, INC, New
York, 2001
[2] N. KUBOTA, Propellants and Explosives, Thermochemical Aspects of Combustion, WILEYVCH GmbH, Weinheim, Germany, 2002
[3] H. BATHELT, F. VOLK, M.WEINDEL, ICT-Thermodynamic Code, Windows Fronted:
Version 1.00, Fraunhofer ICT, 1999
[4] H. STMARK, A. LANGLET, H. BERGMAN, N. WINGBORG, U. WELLMAR and U.
BEMM, FOX-7 a new explosive with low sensitivity and high performance, FOA
[5] B. JANZON, H. BERGMAN, C. ELDSATER, et al., FOX-7 - A Novel, High Performance,
Low Vulnerability High Explosive For Warhead Applications, Proc. 20th International
Symposium On Ballistic, Orlando, 2002.
[6] M. W. SMITH, M.D. CLIFF, NTO Based Explosive Formulations: A Technology Review,
MDSTO-TR-0796, DSTO, 2001
[7] US Patent No 4,733,610 (1988)
[8] H. BATHELT, F. VOLK AND M. WEINDEL, The ICT-Thermodynamic-Code (ICT Code)
Version 1.0, Fraunhofer ICT, 2000
[9] R. MEYER, J. KHLER, A. Homburg, Explosives, Wiley-VCH Verlag GmbH, Weinheim,
2002
[10] T. B. FEDOROFF, Encyclopedia of Explosives and Related Items, Picating Arsenal, Dover,
NJ, USA, Vol I, p. XVI, 1960
[11] "Recommendations on the Transport of Dangerous Goods, Manual of Tests and Criteria"
(ST/SG/AC.10/11/Rev.4) United Nations, New York and Geneva, 2003
11
597
Czech Republic
2006
[Content]
INTRODUCTION
The emphasis was put on constructing software (in the MathCad pack), enabling
determination of ballistic parameters of powder, taking into account the absolute methods,
which were not included in the so far applied MIL program. The advantage of this method is
the possibility of drawing conclusions concerning the purposefulness and all the properties
of the examined powder, without the necessity of providing the standard powder. This is
the only method of examining powders without using their standard. This work presents
the computations acquired on account of the authors own software, which is then compared
to the results achieved by using the MIL program.
RESULTS
The results were worked upon on the basis of the recording course of pressure in time
function, burning of three weight amounts of examined powder in a manometric bomb. The
recording courses of pressure p = f (t ) acquired in the MIL program are presented in Fig. 1.
598
Czech Republic
2006
[Content]
And the same courses p = f (t ) from MathCad program are presented in Fig. 2.
250
250
200
150
pb 1
100
50
0.005
0.01
tb 1
0.015
0.0186
p
dp
from MIL program.
= f
dt
pm
599
Czech Republic
2006
[Content]
50
40
30
dpb1
20
10
0.2
0.4
0.6
0.8
pb 1
1
1.01
max( spb 1)
p
dp
from MathCad program.
= f
dt
pm
The characteristics gained through this study make it possible to calculate among others:
absolute force AF
absolute velocity AQ
powder force f
powder covolumen .
The exemplified results of the absolute force AF and absolute velocity AQ, achieved
in MIL and MathCad programs are as follows:
Absolute force
Absolute velocity
MIL
240,8784 MPa
6,73031GPa2/s
MathCad
241,455 MPa
7,258 GPa2/s
Powder force
Powder covolumen
0,86535
MJ
kg
1,42445
cm 3
g
arithmetic mean
0,86538
MJ
kg
1,42422
cm 3
g
MIL
600
Czech Republic
0,8653
MathCad
MJ
kg
2006
[Content]
1,4245
cm 3
g
Force and powder covolumen are calculated on the basis of Nobel-Abel Law. In order to
conduct the calculations, it is necessary to perform two burnings for two different charge
density, and proceed to drawing up the graph (pmax/)=f(pmax). The calculated points of the
specified charge densities and maximum pressures were used to calculate the force and
covolumen through linear regression.
1.25
1.2
max( pf n)
n
1.15
1.1
1.05
90
122
154
186
218
250
max( pf n)
CONCLUSIONS
1. The created software allows to use the absolute method in order to determine the
properties of the powder, which does not have its standard.
2. The results, with the accuracy of measurement at 0,25 %, are equal to the results
achieved thanks to the software used so far. The only difference of 8 % ratio in
case of absolute velocity results from application of different numeric methods
used in the calculations.
601
Czech Republic
2006
[Content]
Abstract:
One of the modern and principal trends in the development of energetic materials is the
search of the so-called green environmentally friendly primary explosives. One of the
representatives of such green explosives is hexamethylenetriperoxidediamine
(HMTD). Thin-layer composite charges based on HMTD and optically transparent
polymeric matrix (PBX) are initiated by the beam of a pulse laser of low power. This
feature makes the application of this material in the safe optical ignition charges very
promising.
Keywords:
INTRODUCTION
1.
Regular industrial primary explosives like lead azide, mercury fulminate, lead styphnate
all possess high toxicity and their application in blasting caps and pistons results in
environmental pollution with lead and mercury [1]. So the search for environmentally benign
initiation systems is one of the pressing current problems [2].
2.
DISCUSSION
2.1
Hexamethylenetriperoxidediamine, or
602
Czech Republic
2006
[Content]
O
O
O
N
O
O
Fig 1.
HMTD
CH2
N
CH2
CH2
CH2
N
CH2
+ 6 H 2O2
2 NH
+ 3 H 2O +
CH2
+
CH2-O-O-CH 2
CH2-O-O-CH 2
CH2-O-O-CH 2
N
Scheme 1
603
Czech Republic
2006
[Content]
2.2
Czech Republic
2006
[Content]
40 mJ/cm2. It is difficult to use pure HMTD in optical detonators because it possesses high
sensitivity to external stimuli and high corrosive ability. Using inert polymer matrix decides
both problems.
Optically transparent poly-2-methyl-5-vinyltetrazole with terminal
carboxylic groups (PVMT) and polymethylmethacrylate (PMMA) were used as inert
matricess in our studies.
CH2 CH
C N
N
N
N CH3
n
Fig 2. PVMT
COOCH3
|
-CH 2-C|
CH 3
Fig 3.
PMMA
We investigated thin-layer formulations of PBX based on HMTD which contained 1040% of a polymer, as light-sensitive explosives. PBX films of 1cm in diameter were used in
these experiments. 0.1 mm thick aluminimum foil sheets 20x20 mm in size were covered
with the PBX composition. Ballistic pendulum was used for measuring the momentum of
explosion products of PBX films. The laser beam was defocused in such a way that the
energy density in the center of the beam exceeded of the energy density at the edges of
samples by no more than 20%. The average energy density of the laser beam at the surface
of charges was twice as high as the threshold of initiation of PBX films. In this case planar
detonation wave of PBX films was realized. In the experiments the weights of PBX films
were varied from 35 to 110 mg/cm2. The dependence of momentum of explosion products
(J) on the weight of samples (ms) for HMTD film formulations containing 10% of PVNT
polymer is shown in Fig.4.
605
Fig 4.
Czech Republic
2006
[Content]
The experimental results were aproximated by the first degree polynomial. The vertical
pillars mark 95% probability level. Part of the charge of PBX films burned and gave no
energy impulse.
3.
CONCLUSIONS
Acknowledgements
This work was financially supported by the Ministry of Education and Science of
Russian Federation in the framework of the Intercollgiate Research and Development
Programm Development of Scientific Potential of Higher School.
REFERENCES
[1]
606
Falyakhov I. F., Gilmanov R. Z., Nesterov A.V., Fedorov O.V. and Khusainov R.M.
Technologies of preparation of enviromentally friendly primary explosives on the basis of
aromatic diazo compounds.// Proceedings of International Conference Modern problems of
technical chemistry 22-24 December 2004. Kazan. 2004. P.359-362. (in Russian).
Czech Republic
2006
[Content]
[2]
Zhoui S., Liu J. & Xie G. A Discussion on Explosive Materials and Safety // Theory and
practice of energetic materials./Ed. by Huang P., Li S. and Wang Y. Beijing/New York.
Science Press 2003. Vol. V. Part A. P. 183 185.
[3]
Bagal L.I. Chemistry and Technology of Primary Explosives. Moscow. Mashinostroenie. 1975.
455 p. (in Russian)
[4]
[5]
Matyas
R.
Chemical
decomposition
of
triacetone
triperoxide
and
hexamethylenetpiperoxidediamine//Proceed. of the VI. Seminar New trends in research of
energetic materials. Pardubice. Chech Republic. April 22-24 2003. 2003. P. 241 247
[6]
Danilov Yu.N., Ilyushin M.A. and Tselinsky I.V. Industrial explosives. Part I. Primary
explosives. Text-book. Saint-Petersburg. Synthesis. 2004. 113 p.(in Russian)
[7]
[8]
[9]
Chernay A.V., Sobolev V.V., Chernay V.A., Ilyushin M.A. and Bunchuk Yu.P. Chapter 11.
Ignition of explosives by pulse lasers.// Physics of impulse treatment of materials/ Ed. Sobolev
V.V. Dnepropetrovsk. Art-Press. 2003.p.267-314. (in Russian)
607
Czech Republic
2006
[Content]
SYNTHESIS OF 2,4-DINITRO-2,4-DIAZAPENTANE
FROM METHYLNITRAMINE AND DIHALOGENMETHANES
Abstract:
Methylnitramine reacts with dihalogenmethanes in dimethylformamide in the presence
of potassium carbonate or bicarbonate as bases to produce 2,4-dinitro-2,4diazapentane (I). The same product I is formed in the reaction of potassium salt of
methylnitramine (potassium methylnitroamide, II) with dibromomethane.
Keywords:
1.
INTRODUCTION
NO2
2 H3C N H + CH2O
H2SO4
NO2
NO2
SO3Na
2 H3C N H + CH2O
SO3Na
H3C N CH2 N CH3
SO3Na
NO2+
NO2
NO2
Na2SO4
608
NO2
2 H3C N CH2Cl +
H3C N NO2
Czech Republic
2006
NO2
NH4+
[Content]
NO2
Dichloro- and dibromomethane represent other compounds in which carbon atom has
the same oxidation state as in formaldehyde. The reaction of above mentioned
dihalogenmethanes with alkylnitramines has not been described in the literature until now.
That's why we have tried to explore the reaction of methylnitramine with dihalogenmethanes
for the synthesis of I.
2.
NO2
2 H3C N H + CH2X2
K2CO3
DMF
NO2
NO2
H3C N CH2 N CH3 +
KX
X = Cl, Br
Dichloro- and dibromomethane were chosen as halogenmethanes for the reaction with
methylnitramine. It is assumed that the reaction proceeds in three steps. The first step is the
formation of potassium methylnitramide (II). The second step is the substitution reaction of
II with dihalogenmethane affording N-(halogenmethyl)methylnitramine III. The final
product I is formed in step 3 by the reaction of the intermediate III with potassium
methylnitroamide II.
The results of experiments are summarized in Table 1. Higher yields of I are obtained
from the reactions of II with dibromomethane (examples 1-8) than with dichloromethane
(examples 9-10). The reaction of dihalogenmethane with methylnitramine takes place only in
the presence of a base. In the case when no base is used for the reaction of dibromomethane
with methylnitramine (example 1), only unreacted methylnitramine was detected. A base is
not necessary when potassium methylnitroamide II is employed (example 8). In most
examples, potassium carbonate was chosen as a base. Reaction takes place both with two
equivalents (examples 3-4) and with one equivalent of potassium carbonate to
methylnitramine (example 5). In the first case, the yields are higher. When
dihalogenmethane is used in overstoichiometric amount (examples 4 and 10), the yield of
product is higher than in the case of stoichiometric amount in corresponding experiments
(examples 2 and 9 respectively). No anticipated intermediate III was detected in the final
product. On the basis of the obtained results we suppose that the reaction of II with
2
609
Czech Republic
2006
[Content]
dibromomethane is the rate determining step of the whole reaction. The higher yields of I
were obtained when the reaction was carried at higher temperatures than at room
temperature (example 3). When potassium bicarbonate is used as a base, the reaction takes
place both with two equivalents (example 6) and with one equivalent (example 7) of
potassium bicarbonate to methylnitramine. Again, the yields are higher in the first case.
Step 1
NO2
H3C N H + K2CO3
H3C N NO2
K+
KHCO3
II
Step 2
H3C N NO2
NO2
H3C N CH2X + KX
K+ + CH2X2
II
III
Step 3
NO2
H3C N CH2X +
III
H3C N NO2
II
K+
NO2
NO2
H3C N CH2 N CH3 + KX
I
X = Cl, Br
3.
EXPERIMENTAL
610
Czech Republic
2006
[Content]
Reaction Components
Molar
Ratio
Temperature, C Yield of I,
(Time, hour)
%
CH3NHNO2
CH2Br2
no base
5.7 g
6.5 g
2
1
0
70 (3)
CH3NHNO2
CH2Br2
K2CO3
5.7 g
6.5 g
10.4 g
2
1
2
70 (3)
58
CH3NHNO2
CH2Br2
K2CO3
5.7 g
6.5 g
10.4 g
2
1
2
20 (120)
18
CH3NHNO2
CH2Br2
K2CO3
5.7 g
10.4 g
10.4 g
2
2
2
CH3NHNO2
CH2Br2
K2CO3
5.7 g
6.5 g
6.2 g
2
1
1
CH3NHNO2
CH2Br2
KHCO3
5.7 g
6.5 g
7.5 g
2
1
2
CH3NHNO2
CH2Br2
KHCO3
5.7 g
6.5 g
3.8 g
2
1
1
[CH3NNO2]- K+
CH2Br2
no base
5.7 g
6.5 g
2
1
0
CH3NHNO2
CH2Cl2
K2CO3
5.7 g
3.2
10.4 g
2
1
2
10
CH3NHNO2
CH2Cl2
K2CO3
5.7 g
3.2
10.4
2
4
2
70 (3)
62
70 (3)
47
70 (3)
48
70 (3)
27
70 (3)
47
70 (3)
70C
3 hours
15 %
611
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
612
Czech Republic
2006
[Content]
Abstract:
2,4,6-trinitrotoluene (TNT) is a widely used military explosive and a predominant
contaminant of soil and ground water. TNT has been described to be toxic to aquatic
and terrestrial organisms in the environment. TNT can be degraded, mineralized or
conjugated in the presence of bacterium or fungus. The initial products formed from
TNT biotransformation are the reduced to amino derivatives (such as 4-amino-2,6dinitrotoluene or 2-amino-4,6-dinitrotoluene). 2,4,6-triaminotoluene (TAT) is observed
under strictly anaerobic conditions. To obtain nitrocompounds from soils is necessary to
use some extraction techniques. Supercritical fluid extraction (SFE) is one of modern
extraction techniques applied for isolation of analytes from different solid samples. The
combination of SFE with gas chromatography and mass detection (GC-MS) is able to
determinate low concentrations of analyzed compounds.
Keywords:
1.
INTRODUCTION
The contamination of soils by all types of explosives is widely spread and the detection
of these compounds is a growing international problem. 2,4,6-trinitrotoluene (TNT) is one of
the most common explosives used in military munitions. These explosives have been
manufactured on a massive scale in the past. Presently soil and ground water is a worldwide
problem that started following intensive military activities (World War I and II, etc.). The
toxicity of the above energetic chemicals has been documented using different aquatic
organisms, terrestrial species such as earthworms.
TNT may be mixed with other explosives and it is constituent of many explosives. The
advantages of TNT include low cost, safety in handling, fairly high explosive power, good
chemical and thermal stability, compatibility with other explosives, a low melting point and
moderate toxicity. TNT is practically insoluble in water, but soluble in other solvents
(alcohol, acetone, etc.).
The present study describes the use of supercritical fluid extraction with gas
chromatography in combination with mass spectrometry for the isolation and determination
of TNT from soil samples. Other alternatives for the isolation analytes from solid matrix are
ultrasonic extraction and pressurised fluid extraction [1, 2].
Supercritical fluid extraction (SFE) is one of the extraction techniques usually applied
for the isolation of analytes from solid or semi-solid matrices. SFE used a supercritical fluid
as the extraction medium [3]. A supercritical fluid is a substance above its critical temperature
and pressure. Supercritical fluids can be seen as intermediates between gases and liquids.
1
Vra Jeov, Tom Fojtk, Ale Eisner, Martin Adam, Karel Ventura
613
Czech Republic
2006
[Content]
Most SFE extractions are performed using carbon dioxide (CO2) as the mobile phase. CO2
has a low supercritical temperature (31 C) and pressure (7,38 MPa). It is nontoxic and
available at high purity. Being nonpolar is a good solvent for the extraction of nonpolar and
moderately polar compounds (solvating power for polar solutes is rather poor). The
extraction efficiency of polar compounds by CO2 can be improved by the additions of small
quantities of polar organic solvents referred to as modifiers (methanol, ethanol, acetonitril,
acetone, hexane, toluene, chloroform, etc.). Other potential solvents for use in SFE (nitrous
oxide, ammonia, freons, methanol, etc.) are seldom used for reasons of safety, price,
impractical critical conditions, corrosivity, stability or purity.
Supercritical fluid extraction can be operated in two modes: static or dynamic [4]. In
static mode the supercritical fluid is held in an extraction cell for a certain amount of time
and then released to a collection device. In dynamic mode the supercritical fluid flows
continuously through the extraction cell and out into a collection device.
SFE can work in two modes: off- or on-line [4, 5]. In the on-line configuration, SFE is
coupled directly to an analytical instrument (such as gas chromatograph, supercritical fluid
chromatograph or high-performance liquid chromatograph). Off-line SFE is a stand-alone
extraction method independent of the analytical technique to be used. Off-line SFE is more
flexible and easier to perform than the on-line methods and the extract is available for
analysis by different methods. On-line coupling of supercritical fluid extraction with HPLC
is described as well for the determination of explosives in vapour phases too [6].
Two extraction techniques (ultrasonic extraction and supercritical fluid extraction) were
compared in the isolation of HMX (octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine) from
plant tissue [7]. Both extraction of plant tissues were found to be compatible in their
extraction efficiency of HMX (sonication was found to be slightly better at the higher
concentration >300 mg/kg).
This study interests in the optimalization of different parameters of SFE extraction and
GC-MS determination.
2.
EXPERIMENTAL PART
The parameters such as time, pressure, temperature and type of modifier influence the
extraction efficiency in the case of SFE. It is necessary to optimize these extraction
parameters for the higher proceeds.
The pressure and temperature of extraction were optimized in this study. The SFE
conditions for the pressure optimization were follows: 0,5 g of soil samples, extraction time
for 45 min in dynamic mode, temperature 100 C, no modifier and methanol as the
extraction solvent.
The extraction temperature was discovered at these conditions: 0,5 g of soil samples,
extraction time for 45 min in dynamic mode, pressure 25 MPa, no modifier and methanol as
the extraction solvent.
The extracts were transferred into 5-ml volumetric flask. The sample was 5 times diluted
and the internal standard (toluene, 0,02 l/ml) was added before analyzing by GC/MS.
614
Czech Republic
2006
[Content]
The diluted extracts were analysed by using gas chromatograph GC 17A coupled with
mass spectrometry detector QP 5050A (both Shimadzu). Separations were performed on a
30m x 25m i.d. capillary column coated with a 0,25 m film of polymethylsiloxane (DB-5
MS).
The column oven was isothermally maintained at 50 C for 1 minutes and than
temperature increase for 40C/min to 200C that was kept for 3 minutes. The temperature of
injector was 180C and the temperature of interface was 230C.
It was used the electronic ionization in SCAN mode, the split injection 1:100 and the
helium as carrier gas.
The identification of compounds was based on the comparison their MS spectrum with
the spectrum in the library (NIST 62 and NIST 12, Shimadzu) and with MS spectrum of
standards.
3.
Since various parameters potentially affect the extraction process, the optimization of the
experimental conditions represents a critical step in the development of a SFE method. In
fact, pressure and temperature of the fluid, percentage of the modifier and the extraction
times are generally considered as the most important factors.
To evaluate the effects of pressure on the SFE efficiency of TNT, extraction were
performed at the pressures of 10 to 40 MPa by fixed temperature. As can be seen in Fig 1,
recoveries for TNT increased with increasing pressure. Because a better extraction efficiency
of TNT from soils was achieved at the extraction pressure of 25 MPa, this pressure was
chosen for latter experiments.
3
As/Aistd
0
10
15
20
25
30
35
40
Pressure [MPa]
Fig 1.
Vra Jeov, Tom Fojtk, Ale Eisner, Martin Adam, Karel Ventura
615
Czech Republic
2006
[Content]
The variation temperature during SFE affects the density of fluid, the volatile property of
the analytes, and desorption of the analytes from the matrix. At a higher temperature, the
analytes become more volatile but the supercritical CO2 density decreases.
In this study different pressures (40 100 C) were evaluated to optimize the extraction
temperature by fixed pressure (25 MPa). Fig 2 shows the effect of temperature and higher
temperatures have no significant influence on the extraction efficiency. An extraction
temperature of 50 C was chosen for later experiments because of increased amount of TNT
from soil samples.
As/Aistd
2,5
1,5
1
40
50
60
70
80
90
100
Temperature [C]
Fig 2.
4.
CONCLUSION
This study demonstrated that supercritical fluid extraction is suitable for the routine
extraction of solid samples. The optimalization of extraction parameters is necessary for the
high extraction efficiency. Extraction pressure and temperature were optimized at first. The
extraction time and the additions of modifier will be optimized in further experiments.
The SFE extracts were analysed with gas chromatography-mass spectrometry. It was
found that extraction temperature of 50 C and pressure of 25 MPa are sufficient conditions
for the extraction of 2,4,6-trinitrotoluene from soil samples.
Acknowledgement:
Experiments were performed thanks to financial support from the Ministry of Education,
Youth and Sports of the Czech Republic (Project MSM0021627502) and the Grant Agency of
Czech Republic (GAR 203/05/2106).
616
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
Vra Jeov, Tom Fojtk, Ale Eisner, Martin Adam, Karel Ventura
617
Czech Republic
2006
[Content]
Abstract:
One of major problems is the contamination of soils, ground water and air with
hazardous and toxic compounds. A large of the contamination comes from agricultural,
industrial and military activities. The possibility of microorganisms to degrade toxic
pollutants has been widely discussed.
Many synthetic organic compounds are mineralized by microbes and transformed into
CO2 and H2O. 2,4,6-trinitrotoluene (TNT) is a reactive molecule that biotransform
readily under both aerobic and anaerobic conditions to give aminodinitrotoluenes.
Various bacteria were applied into the solutions include different concentration of TNT.
The bacteria were removed with a centrifuge after defined time. The products from the
transformation of TNT were analysed with gas chromatography GC 17A with mass
spectrometry detector QP505A (both Shimadzu).
Keywords:
1.
INTRODUCITON
Czech Republic
2006
[Content]
functional group. These functional groups react under both abiotic and biological conditions
to produce azo, azoxy, hydrazo derivates [2].
In almost all studies cases, the initial products formed from TNT biotransformation are
the reduced amino derivates (such as 4-amino-2,6-dinitrotoluene, 2-amino-4,6-dinitrotoluene
or 2,6-diamino-4-nitrotoluene). 2,4,6-triaminotolune is only observed under the anaerobic
condition [2].
This study describes different types of bacterium and their ability to degrade TNT under
various experimental conditions in the culture media. The selected types of bacterium were
applied into the soils contaminated with TNT. The nitrocompounds were isolated from soils
by two extraction techniques (ultrasonic extraction or pressurised fluid extraction). The
extracts were analyzed by gas chromatography and identificated by mass spectrometry.
Ultrasonic extraction (also known as sonication - USE) uses ultrasonic vibration to
ensure contact between the sample and the solvent [3]. Sonication is relatively fast, easy, it is
possibility extraction of more samples, noisiness of the apparatus and higher consume of
solvent. After extraction, the extract is filtrated or centrifuged, and some from of cleanup is
generally needed prior to analysis.
Pressurised fluid extraction (also known as accelerated solvent extraction PFE) used
convention solvents at elevated temperatures (80-150 C) and pressures (to 10 MPa) to
enhance the extraction of organic analytes from solids [4, 5]. The solvent boiling point is
increased under high pressure, so the extraction can be conducted at higher temperatures.
PFE is easy, fast technique (15-30 min), it uses minimal amount of solvent and sample size
(1-30 g). PFE system is fully automated.
2.
EXPERIMENTAL PART
2.1
619
Czech Republic
2006
[Content]
Separations were performed on a 30m x 25m i.d. capillary column coated with a 0,25
m film of polymethylsiloxane (DB-5 MS). The volume of dosed extract was 1 l. It was
used the split injection 1:100 and the helium as carrier gas.
The column oven was isothermally maintained at 50 C for 2 minutes. Than the
temperature increase for 15 C/min to 170 C and for 5 C/min to 200 C that was kept for 2
minutes. The temperature of injector was 180 C and the temperature of interface was
230 C.
2.2
620
3.
Czech Republic
2006
[Content]
The mixture of all standards was analysed by GC/MS at first (Fig 1) and the parameters
of determination were found. The mixture of standards included the mononitrotoluenes,
dinitrotoluenes and 2,4,6-trinitrotoluene.
250000
3-MNT
4 4 0 0
4 3 0 0
200000
4 2 0 0
4 1 0 0
4 0 0 0
3 9 0 0
3 8 0 0
Intensity
150000
4-MNT
3 7 0 0
3 6 0 0
1 6 , 6
100000
2- MNT
1 6 , 7
1 6 , 8
1 6 , 9
1 7
2,3-DNT+2,4-DNT
50000
3,4-DNT
2,6-DNT
2,4,6-TNT
0
8
10
12
14
16
18
Fig 1.
The liquid samples with various types of bacteria were obtained with using GC/MS in
SIM mode. A lot of degradation products were found in the first type of samples and one of
the chromatograms is shown at the Fig 2. Mass detector reported the formations of
mononitrotoluenes or dinitrotoluenes as one of major intermediate of TNT.
The same types of compounds were observed in the case of the analysis of second group
of samples with bacteria. The incubation time of the samples was longer and that is why it
were obtained less degradation products of TNT (Fig 3).
3,4-DNT
2,6-DNT
5000
7500
2,4,6-TNT
4750
2,4,6-TNT
Intensity
Intensity
6500
2,3-DNT+2,4-DNT
5500
3,4-DNT
2,6-DNT
3-MNT
4500
2-MNT 4-MNT
4500
4250
4000
3750
3500
3500
8
10
12
14
16
18
10
Fig 2.
Fig 3.
12
14
16
18
621
3.2
Czech Republic
2006
[Content]
2,4,6-TNT
1000000
800000
Intensity
Intesity
1200000
2,3-DNT
600000
400000
3,4-DNT
2,6-DNT
200000
0
3
11
1200000
1000000
800000
600000
400000
200000
0
2,4,6-TNT
2,3-DNT
2,6-DNT
13
3,4-DNT
11
13
The transformation of TNT in soils was studied by using ultrasonic extraction during
two months. The bacteria were added into the initial samples included only different
concentration of TNT. The extracts were prepared always after one-month incubation. The
effect of TNT transformation is shown at Fig 6.
1st month
2nd month
0,035
0,03
Amount [%]
0,025
0,02
0,015
0,01
0,005
TNT
3,4-DNT
2,3+2,4-DNT
2,6-DNT
TNT
3,4-DNT
2,3+2,4-DNT
2,6-DNT
TNT
3,4-DNT
2,3+2,4-DNT
2,6-DNT
Soil sample
622
4.
Czech Republic
2006
[Content]
CONCLUSION
The aim of this study was found some type of bacterium for the transformation of 2,4,6trinitrotluene to some nitrocompounds. The time and conditions of incubation were tested
and it was discover that longer time and different types of bacteria mean less degradation
products.
Soils for optimization of extraction parameters were spiked with pure TNT 6 month
before the analysis. The TNT and its derivates were extracted from real soils included
bacterium by ultrasonic extraction or pressurised fluid extraction and the extracts were
quantified by GC/MS. The same nitroaromatic compounds were presented in all samples and
the main metabolites of TNT such as mononitrotoluenes and dinitrotoluenes were found. The
concentrations of nitrocompounds were similar and the extraction efficiencies of extraction
techniques were comparable.
Acknowledgement:
Experiments were performed thanks to financial support from the Ministry of Education,
Youth and Sports of the Czech Republic (Project MSM0021627502) and the Grant Agency of
Czech Republic (GAR 203/05/2106).
REFERENCES
[1]
[2]
[3]
[4]
[5]
623
Czech Republic
2006
[Content]
Abstract:
Investigations of TNAZ/AzDNAZ mixtures with different TNAZ contents revealed an
eutectic composition TNAZ/AzDNAZ 35/65 with melting point of 61oC. ombined
introduction of nitro- and azido groups followed by nitration of tBuDNAZ/tBuAzNAZ
mixtures in N2O5/CH3CN and NH4NO3/(CH3CO)2O systems has been investigated.
Possibility of synthesis TNAZ/AzDNAZ mixtures of various contents has been shown.
Keywords:
1.
INTRODUCTION
Recently intensive searches have been conducted of the new explosive materials
possessing high energy of explosive transformation, density and stability. The most
interesting compounds obtaining were CL-20, NTO, TNAZ, belonging to a class of cyclic
nitroamines.
1,3,3-Trinitroazetidine has been synthesized in 1984 in the USA and represents fourmembered heterocyclic nitroamine containing gem-dinitro group. Now works with TNAZ
are carried on in the USA, France, Germany, South Korea and China.
It is proposed to use TNAZ in artillery powder systems[7,10], as a component of
pyrotechnic mixtures[8,9], explosive compositions[6] and solid propellants. Also TNAZ may
be used as a softener in mixtures with CL-20 and in gunpowders containing 4% of TNAZ[7].
One of advantages of TNAZ is rather low melting point (98 - 101[1-4]) at sufficient
power (approximately as RDX) and high density (1.84 g/cm3[1,3]), that makes it attractive to
use in some special compositions.
TNAZ is capable of forming eutectic mixtures with high energetic substances, such as
HMX, 2,4-dinitroimidazole, tetryl. TNAZ/tetryl eutectic composition with the melting point
81.5 (test DSC)[5].
2.
624
OH
NaOH
O2N
+
Na
N
Czech Republic
N3
NaN3
K3[Fe(CN)]6
2006
[Content]
O2N
+
NO2
N3
N
NO2
AzDNAZ
tBuAzNAZ
-5
-10
50
Fig 1.
100
150
200
250
Wg^-1
20
10
-10
30
Fig 2.
50
70
90
110
130
150
170
190
210
230 C
625
Czech Republic
2006
[Content]
Wg^-1
-5
50
Fig 3.
100
150
200
250
105
melting, oC
95
85
75
65
55
0
20
40
60
80
100
TNAZ, %
Fig 4.
626
OH
NO2
HO
NH2 + CH2O
Czech Republic
O2 N
HCl
OH
N
OH
1
OH
NO2
O2N
N *HCl
N3
O2N
+
2) NaNO2 K3Fe(CN)6
NaN3
P(Ph)3
N *HCl
1)NaOH
4
O2N
[NO2]+
NO2
N3
O 2N
+
OH
O2 N
DEAD
NH *HCl
[Content]
OH
OH
O2 N
2006
N
NO2
N
NO2
TNAZ / AzDNAZ
tBuDNAZ / tBuAzNAZ
tBuDNAZ, %
0
0
Fig 5.
50
100
150
200
time of reacnion, min
250
300
Kinetics of nitration
tBuAzNAZ, % .
0
0
Fig 6.
200
400
600
time of reaction, min
800
1000
Kinetics of azidation
627
Czech Republic
2006
[Content]
100
contents tBuAzNAZ,%
80
60
40
20
0
0
Fig 7.
It has been shown that the 2-hours pause between addition of NaN3 and NaNO2 gives
rise to formation of a mixture of 85% tBuAzNAZ and 15% tBuDNAZ. The composition of
products was determined by HPLC method.
The earlier researches have shown that the most convenient systems for preparation
TNAZ are: HNO3/Ac2O 6/1, NH4NO3/Ac2O, and also systems containing nitric anhydride.
However, the maximum yield of TNAZ, 82%, was achieved in system NH4NO3/Ac2O
(yields were 60-80% by on different sources).
It has been shown that nitration of tBuAzNAZ occurs in this system with a small yield
(20 30%) that has been explained by low thermal stability of geminal nitroazide in acetic
anhydride. This behaviour of tBuAzNAZ can essentially change composition of the final
mixture. Therefore, it has been carried out nitration of mixtures with high contents of
tBuAzNAZ. As a result solid and liquid phases were obtained and analysed by HPLC.
Results of the analysis have shown that the solid product represents a mixture with
AzDNAZ content 92%, and the composition of the liquid phase is close to eutectic
(TNAZ/AzDNAZ 67/37).
The material balance has shown that the yield of TNAZ does not vary, and the yield of
AzDNAZ is 48%, 2 times higher than at nitration of pure tBuAzNAZ.
628
Czech Republic
2006
[Content]
EXPERIMENTAL PART.
3.
Melting points were run at 3C/min on a Boatius heating microtable. Infrared spectra
were obtained on a Thermo Nicolet 360 FT IR spectrophotometer using KBr disks.
HPLC analysis was carried out on Milichrom-4 chromatograph, column UVM
3780/12, 2.060 mm, speed of submission of eluent 200 L/min. Acetonitrile/water 1/1,
acetonitrile/water 35/65 as eluents was used. Thin-layer chromatography (TLC) was carried
out using DC-Alufolien TLC plates from Merck.
Differential Scanning Calorimetry (DSC) experiments were performed on a Mettler
Toledo DSC 822e using aluminium sample pans. The instrument was calibrated daily using
indium and aluminium standards. Traces were recorded at ramp rate of 10C/min between
25 to 200C under a flowing N2 gas atmosphere (50 ml/min). The sample weight was 1.28
mg.
3.1
4.
CONCLUSION
629
Czech Republic
2006
[Content]
REFERENCE
[1]
630
Czech Republic
2006
[Content]
SALTS OF 5-NITROTETRAZOLATE
Thomas M. Klapoetke*, Peter Mayer*, Kurt Polborn* and Jan M. Welch*
*University of Munich, Butenandtstr. 5-13 (D), D-81377 Munich, Germany
Abstract:
The previously well described sodium dihydrate (1) and silver (2) salts of 5nitrotetrazolate were prepared and characterized by Raman spectroscopy. Additionally
X-ray structure determination and measurement of the 13C and 15N solution NMR
spectra were carried out for 1. The previously incompletely described or hereto
unknown 5-nitrotetrazolate salts of 1,4-dimethyl-5-aminotetrazolium (3), guanidinium
(4), aminoguanidinium (5), diaminoguanidinium (6), ammonium (7), hydrazinium (8)
and tetraphenylphosphonium (9) by metathesis of the corresponding hydrohalide
(iodide, bromide or chloride) with silver 5-nitrotetrazolate and characterized Raman
spectroscopy and X-ray structure determination. Lastly the incompletely described
potassium salt of 5-nitrotetrazolate (10) was also synthesized and characterized by
Raman spectroscopy and X-ray structure determination. All compounds were found to
be air and moisture stable.
Keywords:
1.
INTRODUCTION
The synthesis of energetic, non-nuclear materials for possible military application has
been a long term goal in our research group.[1-3] We have recently prepared and determined
the properties and structures a variety of compounds based on and derived from 5-amino(1H)-tetrazole.[4-6] The 5-nitrotetrazolate anion has been the subject of most recent
investigations. Although the 5-nitrotetrazolate anion is well established and its metal salts
have been extensively investigated for use as, initiators, primary explosives and energetic
sensitizers,[7-10] the weakly basic 5-nitrotetrazolate anion has only recently been investigated
as a component of nitrogen rich ionic energetic materials.[11-13] Here we present our findings
in the area of nitrogen-rich ionic materials.
631
2.
Czech Republic
2006
[Content]
Synthesis
The synthesis of 1 and 2 were carried out according to literature methods.[9] Compound
10 was synthesized by appropriate modification (Substitution of potassium hydroxide for
sodium hydroxide during digestion step) of the literature method for 1. Salts 3-9 were
synthesized generally according to eq. (1).
NO2
NO2
C
N Ag
N
N
MeOH
+ [Cat][X]
+
N Cat + AgX
(1)
NH2
Cat=
NH2
C
N
N
N
(4)
(3)
H2 N
NH2
NH2
NH2
H2 N
(5)
N
H
X= Cl-
NH2
H2 N
NH2
N
H
NH4
2.2
(6)
BrI-
N
H
(7)
H3NNH2
(8)
X-ray Structures
The molecular structures of 1 and 3-10 in the crystalline state were determined by X-ray.
Generally all structures show a planar 5-nitrotetrazolate moiety as well as when possible
hydrogen bonding between cations and anions. The shortest donor-acceptor distances are
most often observed between the donor and N1 of the tetrazole ring in 5-nitrotetrazolate.
However, there are also notable hydrogen bonding interactions between the oxygen atoms of
the nitro group and donor groups in several cases and less frequently at the N2 position of
the tetrazole ring.
Salts OF 5-Nitrotetrazolate
Fig 1.
Czech Republic
2006
[Content]
Fig 2.
as for 3. Compound 4
units per unit cell. Each
, c = 6.8736(2) , a
3. The structure was
633
Czech Republic
2006
[Content]
determined from data with Rint = 0.0922 in C2/c, R1 and wR2 for all data equal to 0.0395 and
0.1039 respectively and R1 and wR2 for I>2(I) equal to 0.0370 and 0.1010 respectively.
Fig 3.
Fig 4.
634
Salts OF 5-Nitrotetrazolate
Czech Republic
2006
[Content]
Fig 5.
635
Fig 6.
Czech Republic
2006
[Content]
Fig 7.
636
Salts OF 5-Nitrotetrazolate
Fig 8.
Czech Republic
2006
[Content]
Fig 9.
2.3
7
637
Czech Republic
2006
[Content]
Salts OF 5-Nitrotetrazolate
Czech Republic
2006
[Content]
1159 (2), 1063 (30), 1041 (28), 1034 (14), 929 (5), 837 (5), 774 (3), 546 (3), 449 (3),
406 (2), 368 (2), 258 (3), 137 (2)
3.
CONCLUSIONS
639
4.
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
Czech Republic
2006
[Content]
REFERENCES
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,
pages 134-1 to 134-1.
T. M. Klaptke, B. Krumm, G. Holl, and M. Kaiser, Synthesis, Characterization and Quantum
Chemical Computations of Hexanitrohexaaza-Isowurtzitane, in: Proceedings 30th International
Annual Conference of ICT, June 30 July 3, 1999, Energetic Materials - Modeling of
Phenomena, Experimental Characterization, Environmental Engineering, Fraunhofer Institut
fr Chemische Technologie: Karlsruhe, Germany, 1999.
T. Habereder, A. Hammerl, G. Holl, T. M. Klaptke, J. Knizek, P. Mayer, and H. Nth,
Methylsubstituted Hyrdazinium Azide Derivatives, in: Proceedings 31st International Annual
Conference of ICT, June 27 30, 2000, Energetic Materials Analysis, Diagnostics and
Testing, Fraunhofer Institut fr Chemische Technologie: Karlsruhe, Germany, 2000, pages 1501 to 150-7.
T. M. Klaptke, New highly energetic materials based on hydrazine and tetrazole compounds,
ACS Spring 2003 Award Symposium, New Orleans, March 23 27, 2003.
J. Geith, T. M. Klaptke, J. Weigand, G. Holl, Calculation of the Detonation Velocities and
Detonation Pressures of Dinitrobiuret (DNB) and Diaminotetrazole Nitrate (HDAT-NO3),
Propellants Explosives and Pyrotechnics 2004, 29, 3-8.
M. v. Denffer, T. M. Klaptke, G. Kramer, G. Spie, J. M. Welch, G. Heeb, Improved
Synthesis and X-Ray Structure of 5-Aminotetrazolium Nitrate, Propellants Explosives and
Pyrotechnics 2005, 30, 191-1195.
E. v. Herz, C-nitrotetrazole Compounds, US Patent 2066954 (1932)
L. R. Bates, J. M. Jenkins, Produciton of 5-nitrotetrazole salts, US Patent 4094879 (1978).
W. H. Gilligan, M. J. Kamlet, Synthesis of Mercuric 5-nitrotetrazole, Technical Report 76-146
(1976), Navel Surface Weapons Center, White Oak Labd, Silver Spring, MD, USA.
R. J. Spear, P. P. Elischer, Studies of Stab Initiation. Sensitization of Lead Azide by Energetic
Sensitizers, Australian Journal of Chemistry 1982, 35, 1-13.
H. Xue, Y. Gao, B. Twamley, Jeanne M. Shreeve, Energetic Azolium Azolate Salts, Inorganic
Chemistry 2005, 44, 5068-5072.
H. Xue, B. Twamley, Jeanne M. Shreeve, Energetic Quaternary Salts Containing Bi(1,2,4triazoles) Inorganic Chemistry 2005, 44, 7009-7013.
T. M. Klaptke, K. Karaghiosoff, P. Mayer, A. Penger, J. M. Welch, Synthesis and
Charaterization of 1,4-dimethyl-5-aminotetrazolium 5-nitrotetrazolate, Propellants, Explosives,
Pyrotechnics, In Press.
Acknowledgment:
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.
640
Salts OF 5-Nitrotetrazolate
10
Czech Republic
2006
[Content]
Abstract:
The reaction between 5,5`-bis-(1H-tetrazolyl)amine (1) and CuX2yH2O (X = Cl, y = 2; X = ClO4, y
= 6; X = NO3, y = 2), performed in the corresponding acid and required ratio of 1:1 and 1:2
stoichiometry, afforded complexes with the formula [{CuCl2(H2bta)(H2O)}2H2O] (2),
[{CuCl(H2bta)2}Cl]
(3),
[{CuCl2(H2bta)2}2H2O]
(4),
[{Cu(H2bta)2}(ClO4)2]
(5),
[{Cu(H2bta)2}(ClO4)2H2O] (6), [{Cu(NO3)2 (H2bta)2}0.5H2O] (7). 3 crystallize in the monoclinic
space groups P-1, 3 and 4 in the monoclinic space groups P21/n and P21/c, respectively. A
description of the molecular arrangement of 3, 4 and 5 is given. The sensitivity of the investigated
compounds ranges from very insensitive (2, 4) to very (6) and extreme (5) sensitive toward impact
and friction. Except for the copper compounds 2 and 4, all investigated complexes show
photosensitive behavior and explode by laser irradiation (Laser: Nd/YAG 1064 nm).
Keywords:
1.
INTRODUCTION
Neutral as well as ionic azides can undergo cycloaddition with cyano groups yielding a
huge variety of different tetrazole derivatives [1]. An interesting precursor for cycloaddition
reactions with azide is the non-linear pseudohalide dicyanamide anion (N(CN)2-), which
exhibits a rich variety of bonding modes for coordination in for example 3d-complexes.
Dicyanamide complexes have attracted much interest in recent years in the construction of
supramolecular aggregates due to their intriguing network topologies and potential functions
as new class of materials [2].
The reaction of sodium dicyanamide with sodium azide under acid catalyzed condition
yield the corresponding 5,5`-Bis-(1H-tetrazolyl)amine as monohydrate (abbreviated as
H2bta, 1) [3]. In the continuous search of novel energetic materials with high nitrogen content
for application, for example as low-smoke producing pyrotechnic compositions, gas
generators, propellants and primers in primer charges (PC), high-energy-capacity transition
metal complexes are of special interest and 1 might play an important role in future
investigations as well applications. Still modern primary explosives may be represented by
lead azide as the main filling detonator [4], and by lead styphnate as the main filling for
primers usually associated with tetrazene as a sensitizer [5]. Recently, coordination
compounds containing near stoichiometric fuel and oxidizer fractions have evinced great
interest [6]. The extensive study on the relationship between structure of coordination
compound and explosive properties has been reported by various research groups [7,8].
Nickel, copper and cobalt complexes appear suitable for detonator applications [9]. For
1
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand
641
Czech Republic
2006
[Content]
example, nickel hydrazinium nitrate (NHN) [10] may find wide ranging applications in
conventional detonators, whereas bis-(5-nitrotetrazolato-N2)-tetraamine cobalt perchlorate
(BNCP) [11] has emerged as an energy producing component for semi-conducting bridge
(SCB) initiator applications.
Low-smoke high energy materials (HNC) are thought as suitable fuel components in
new coordination compounds used in novel PCs. Derivatives of tetrazoles seem to be the
most promising substance class since they are characterized by extreme properties among
heterocycles e.g. by the highest NH-acidity, lowest basicity and the presences of several
pyridine-like nitrogen atoms important in new ligand systems as appropriate metal binding
sites. Tetrazole containing substituents with potential coordination centres are of interest
because of their added chelating properties and it is know that tetrazoles are well known
ligands for ions of many metals [12]. With respect to new PCs, 5,5`-bis-(1H-tetrazolyl)amine
(1) was investigated to see if it could serve as new high-nitrogen content (HNC) ligand
system for new copper-based PEs [8].
The synthetic routes for 1 as its monohydrate fall into three main types: (1) reaction of
sodium dicyanamide with hydrazoic acid, in situ prepared from sodium azide and a
corresponding weak acid like trimethylammonium chloride, boric acid, ammonium chloride
and the like [13]; (2) reaction of sodium dicyanamide with sodium azide in the presence of a
catalyst like zinc chloride, bromide or perchlorate, followed by an acid work-up [14]; and (3)
reaction of 5-amino-1H-tetrazole (5-AT) with cyanogen bromide under base-catalyzed
condition followed by a subsequent cycloaddition of HN3 (Fig 1.) [15].
Fig 1.
For reaction 1 and 2 (Scheme 1), the first reaction step yields certain metal complexes
(e.g. Zn2+, Na+) which can easily transformed to 1. Treatment of 5-AT under base condition
with cyanogen bromide results in the formation of the interesting and poorly described
dianion of 5-cyanoiminotetrazoline (CIT), as sodium salt. Further treatment of the CIT salt
with hydrazoic acid yields 1 [15].
As 1 is a bidentate chelating ligand with multiproton donor sites, it is able to coordinate
to a metal with three reversible types of protonated and deprotonated modes: neutral
(H2bta), mono-deprotonated (Hbta-), and di-deprotonated (bta2-) types (Fig 2.). Moreover,
H2bta can be utilized as a new bridging ligand for controlling the molecular architectures
642
Czech Republic
2006
[Content]
which, in combination with appropriate outer- and inner-sphere ligands, allows the variation
of physicochemical and explosive properties within a wide range.
Fig 2.
The occurrence of energetic oxygen rich cations is comparable few but required to
assure an appropriate oxygen balance. Therefore, counterions such as nitrate (NO3-) or
perchlorate (ClO4-) anion are required in combination with the bta ligand system. In such
systems, of course, the bta needs to coordinate in its protonated from (H2bta). The reaction
of copper salts (e.g. Cl-, NO3-, ClO4-) in the corresponding acid and H2bta strongly depends
on the reaction conditions, like temperature, reaction time and concentration of the acid. By
variation of those parameters a serious of highly energetic bta-copper complexes were
obtained (Fig 3.). The complexes are crystalline, typically blue-colored compounds and
dissolves in water under decomposition. They were characterized by means of IR
spectroscopy, elemental analysis and X-ray structure determination.
Here we wish to report (i) the synthesis of a serious of energetic copper/H2bta salts (27), (ii) the X-ray determination of selected examples of the obtained compounds and (iii)
selected physical-chemical properties, like IR and sensitivity data [16].
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand
643
2.
Czech Republic
2006
[Content]
SYNTHESIS
Caution: H2bta in its dehydrated form shows increased friction and impact sensitivity.
Although the copper(II) chloride complexes of H2bta are kinetically stable and in most cases
are insensitive to electrostatic discharge, friction, and impact, they are nonetheless
energetic materials and appropriate safety precautions should be taken. The perchlorates
and nitrates in contrast are energetic materials and tend to explode under certain condition,
appropriate safety precautions should be taken all time, especially when this compound are
prepared on a larger scale. Laboratories and personnel should be properly grounded and
safety equipment such as Kevlar gloves, leather coat, face shield and ear plugs are
necessary when manipulating the perchlorates 5 and 6 as well as the nitrate 7. Several
severe explosions occurred during the preparation of the solvent free perchlorate 5. 5, 6 and
7 tend to explode violently when heated or manipulated with a metal spatula.
The ligand N,N-bis(1(2)H-tetrazol-5-yl)-amine (1) was synthesized by the reaction of
sodium dicyanamide with sodium azide under acid-catalysed conditions, according to the
method previously described [8]. The H2bta ligand (1) was isolated as monohydrate
(H2btaH2O). 1 is completely insoluble in aliphatic hydrocarbons, acetonitrile and
dichloromethane. It is very slightly soluble in methanol and cold water but significantly
better soluble in hot water and warm dilute acids like HCl, HNO3 and HClO4. It is worth to
note that the monohydrate of 1 can easily be dehydrated under vacuum (24h, 10-3 Torr) at
elevated temperature (120C) accompanied with an increased friction and impact sensitivity.
The result of elemental analysis indicates that the reaction between 1 and CuX2yH2O (X =
Cl, y = 2; X = ClO4, y = 6; X = NO3, y = 2), performed in the corresponding acid and
required ratio of 1:1 and 1:2 stoichiometry, afforded complexes with the formula
[{CuCl2(H2bta)(H2O)}2H2O] (2), [{CuCl(H2bta)2}Cl] (3), [{CuCl2(H2bta)2}2H2O] (4),
[{Cu(H2bta)2}(ClO4)2]
(5),
[{Cu(H2bta)2}(ClO4)2H2O]
(6),
[{Cu(NO3)2
(H2bta)2}0.5H2O] (7) (Fig 3.).
Fig 3.
The synthesis of the CuII complexes strongly depends on the volume of the solvents,
concentration of the acids, temperature, and the growing time of the crystals. Interestingly, it
is possible to obtained different complexes by changing certain reaction conditions. The 1:1
complex 2 was obtained after the addition of the copper solution to a warm (45 C) solution
644
Czech Republic
2006
[Content]
of the ligand in conc. HCl after 2 weeks. 2 is unstable during storage on air, and a few days
after the removal of the crystals from the mother liquor, 2 becomes cloudy indicating the
loss of crystal water. The solvent free complex 3 was successfully obtained when
CuCl22H2O was refluxed with the ligand in a small amount of conc. HCl for a short time. 3
crystallize already during refluxing from the solution and X-ray quality crystals can be
separated from the reaction mixture within 10 minutes. In the case of 4 we found that the
formation of 4 is favored when the reaction is performed in less concentrated solutions and
when no heat is applied while adding the copper salt to the solution of the ligand. From such
reaction mixtures 4 is obtained within a few days. 3 and 4 are air stable blue to blue-green
compounds. The reaction of Cu(ClO4)26H2O in conc. HClO4 with 2 eqivalents of the ligand
1 under harsh condition (70C) yielded in the formation of the water free complex 5.
Applying the same reaction in less concentrated perchloric acid the copper(II) perchlorate
salt 6 was obtained as monohydrate. It is important to mention that the preparation of the
salts 5 and 6 is aligned with a high risk of explosion. Never heat reaction mixtures were the
crystallization is already in progress. Especially in the case of 5, the solution with separated
crystals shows an increased shock sensitivity and tend to explode without any obvious
external impact. Special precautions have to be applied to synthesize and separate those
crystals form the reaction mixture. Crystals of 5 are very hygroscopic and unstable once
separated from the mother liquor. Crystals of 7 were obtained by the addition of a warm
copper nitrate solution to a warm H2bta solution in conc. HNO3. 7 crystallizes as beautiful
coarse plates suitable for X-ray determination within 1 day. Although we never determined
any problems with complex 7 we recommend to apply special precautions during the
synthesis and separation of those crystals.
The IR spectrum of 1 is characteristic for aminotetrazoles [17]. The changes in the bands
that are composed from combinations of the vibrations of the tetrazole ring components
indicates the formation of a coordination bond trough the nitrogen atoms of the two tetrazole
rings of 1 involved. Although those absorptions bands showing medium or weak intensities,
the shift and change in intensity can be used for the demonstration of the direct participation
of the tetrazole cycle in the complexation. The characteristic bands located at 1656 and 1556
cm-1 and assigned to the asym(CtetNCtet) and sym(CtetNCtet) stretching vibrations in 1
are shifted to lower frequencies by approximately 8 to 10 cm-1. The bands attributed to the
stretching vibrations of the N=N (1454 cm-1), C=N (1352, 1337 cm-1) and NN 1282, 1268
cm-1) bonds in the IR spectrum of the free H2bta ligand are shifted to higher wavenumbers
after complexation associated with a decrease of intensity [18]. The IR spectra of the
compounds have in common the bands for the stretching and bending modes of the NH
groups from the protons at the tetrazole ring and the sec. amino group. As it was found for
other coppertetrazole complexes the bands between 3500 and 3000 cm1 assigned to
stretching vibrations of the NH groups of 1 are shifted to higher frequencies with respect to
those of the free ligand. Furthermore it is observed that those bands in the complexes are
less well resolved compared to the free ligand 1. This can be best explained due to the
formation of intermolecular hydrogen bonds in the crystalline network. A red shift of the
stretching modes establish that the heteromolecular hydrogen-bond interaction between the
NH (donor) groups of the ligand 1 and corresponding acceptors (e.g. couteranion and
water) are strong. In all spectra the bands of the respective energetic anions (ClO4- (5, 6) and
NO3- (7)) were obvious and as these bands usually have characteristic fingerprints in IR
spectra, they could be identified easily. The presence of nonsplit, strong absorption bands at
1384 cm-1 (NO3-) [19] in 7, 1090 and 626 cm-1 (ClO4-) [20] in 5 and 6 points to a lack of strong
deviation of the trigonal planar and tetrahedral symmetry in the nitrate and perchlorate
anion, respectively [21].
5
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand
645
3.
Czech Republic
2006
[Content]
CRYSTAL STRUCTURES
3.1
Molecular structure of 3
Fig 4.
The coordination environment of the CuII ion in 3, showing the atom-labeling scheme. Displacement
ellipsoids are drawn at the 50% probability level and H atoms are shown as spheres of arbitrary radii. Selected
bond length []: CuN6 2.010(2), CuN2 1.996(2), CuN15 1.995(2), CuCl1 2.2525(7), CuN11 2.260(2);
Selected bond angle []: N2CuN15 173.21(8), N6CuCl1 158.79(7), N15CuN6 89.00(8), N6CuN2
85.63(8), N15CuCl1 88.88(6), N2CuCl1 94.62(6), N15CuN11 83.03(8), N2CuN11 101.37(8);
N6CuN11 92.95(8).
The angular structural index parameter, , of the coordination polyhedron for the CuII
ion in 3 has been calculated according to Addison et al.[22] and the distortion value for (Fig
4.; = N2CuN15, = N6CuCl1) accounts to 0.24. This indicates that the coordination
geometry around each CuII ion is a slightly distorted square pyramid and that the Cu (d9)
ions probably adopt sp2d2 hybrid orbitals to accept electrons from the ligands. The Cu(II)
being linked to three nitrogen atoms (N2, N6, N15) and one chloro ligand (Cl1) in the basal
plane, and another nitrogen atom (N11) in the apical site. The CuN bond length in the basal
plane ranges from 1.995(2) 2.010(2) , which is typical and also found in complexes with
e.g. 1,10-phenantroline as ligand system [23]. The basal CuCl1 bond distance (2.2525(7) )
is similar within experimental error, to those reported in related complexes [24]. The bond
646
Czech Republic
2006
[Content]
distance found for the apical CuN11 accounts for 2.260(2) . The mean plane in the
molecule is the six-membered H2bta chelate ring defined by atoms N2, C1, N4, C2, N6 and
Cu which is planar with negligible distortion (average r.m.s. deviation form the six-atom
plane is 0.0084 ). The plane defined by N2, N6, N15 and Cl1, deviates from planarity
(average r.m.s. deviation form the six-atom plane is 0.2463 ), since the chloro ligand lies
under this plane with a distance out of this least square plane of 0.8126 .
3.2
MOLECULAR STRUCTURE OF 4
Fig 5.
The coordination environment of the CuII ion in 4, showing the atom-labeling scheme. Displacement
ellipsoids are drawn at the 50% probability level and H atoms are shown as spheres of arbitrary radii. Selected
bond length []: CuN6 2.003(3), CuN2 2.016(3), CuCl1 2.822(6), Selected bond angle []: N2CuN6
84.61(1), N6CuCl1 87.52(3), N2CuCl1 96.59(3). [Symmetry code: (i) 1-x, -y, 2-z.].
3.3
MOLECULAR STRUCTURE OF 5
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand
647
Czech Republic
2006
[Content]
in accordance to the above discussed compound 4 also a distorted octahedral geometry and
is positioned at (0,0,0). It lies on an inversion centre, bonded to four nitrogen atoms of the
two ligands in the equatorial plane (CuN1 1.996(3); CuN6 1.974(3) ). It is bonded more
weekly to two nitrogen atoms N3 of the tetrazole ring of adjacent molecules (CuN3ii and
CuN3iii 2.943(5) ; [Symmetry code: (ii) x, -0.5+y, 0.5-z; (iii) x, 0.5-y, -0.5+z]) in the
axial position. The weak CuN3 interactions link the molecules into layers parallel to [0,1,1]
(Fig 7.), whereas parallel stacking of these layers in the x direction is achieved via hydrogen
bonds between the NH groups and the oxygen atoms of the perchlorate anion (N5HO3i
2.854(4) , N9HO1iv 2.848(4) , N9HO3iv 2.964(4) , N4HO2vi 2.958(3) ,
N4HO1iv 3.004(4) [Symmetry code: (iv) 1-x, -y, -z; (v) x, -0.5-y, 0.5+z; (vi) x, y,
1+z; (vii) x, 0.5-y, 0.5+z]). It is worth to mention that the oxygen atom O4 of the perchlorate
anion is not involved in the hydrogen bonding.
Fig 6.
The coordination environment of the CuII ion in 5, showing the atom-labeling scheme. Displacement
ellipsoids are drawn at the 50% probability level and H atoms are shown as spheres of arbitrary radii. Selected
bond length []: CuN1 1.996(3), CuN6 1.974(3), CuN3i 2.943(5), Selected bond angle []: N1CuN6
85.6(3), N1CuN6i 94.4(4), N1CuN3ii 100.3(1), N6CuN3ii 79.8(2). [Symmetry code: (i) -x, -y, -z; (ii)
x, -0.5+y, 0.5-z; (iii) x, 0.5-y, -0.5+z].
648
Czech Republic
2006
[Content]
Fig 7.
[0,1,1].
A view of the weak axial contacts in 5 (dashed lines) linking the molecules to a 2D layer parallel to
4.
SAFETY TESTING
For initial safety testing, the impact and friction sensitivity was tested according to
BAM methods with the BAM Fallhammer and BAM friction tester [26]. Due to the
extreme hygroscopicity and sensitivity of compound 5 suitable values for the friction and
impact sensitivity can not be given but we experienced that compound 5 is extremely
sensitive toward friction and impact and several explosions during preparation occurred
even without obvious external impact. In the case of the monohydrate 6 (3.4 J; 14.7 N)
values comparable to those of Pb(N3)2 (2.5 J; 10 N) indicates that this salt belongs to the
group of primary explosives. Interestingly, changing the counteranion in 6 by the nitrate
anion in 7, the impact sensitivity of 7 (3.9 J) seems to be not affected whereas the friction
sensitivity considerably decreases (196 N). The chloride 3 shows a low impact (11 J) and
very low friction (353 N) sensitivity whereas 2 and 4 are insensitive toward impact (> 40 J)
and friction (> 360 N).
Raman investigations of colored transition metal complexes fail in most cases, since the
samples tend to show fluorescence. The attempt to record Raman spectra of the compounds
3, 5, 6 and 7 resulted in the immediate explosion of the sample. In the case of 2 and 4 no
reaction was observed. This observation is quite interesting, since modern concepts of the
initiation of explosives with pulsed-laser radiation is based on the ignition at centers formed
upon hypothesis of radiation absorption by optical microheterogeneities in separate crystals
[27]
.
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand
649
5.
Czech Republic
2006
[Content]
CONCLUSION
In this contribution we presented our preliminary results on a series of copper BTA salts
(2-7). They might play an important role in the development of safe non-toxic PC
formulations, as possibly photosensitive compounds utilized in laser detonators as well as
colorants in pyrotechnic formulations. Since the starting materials are cheap and the
preparation is straight forward, 1 has a high potential as ingredient in explosive or propellant
formulations. Except for the copper compounds 2 and 4, all investigated complexes show
photosensitive behavior. They tend to explode by Laser irradiation (Laser: Nd/YAG 1064
nm), performed with different Laser power (20 to 200 mW). This observation is quite
interesting, since modern concepts of the initiation of explosives with pulsed-laser radiation
is based on the ignition at centres formed upon hypothesis of radiation absorption by optical
microheterogeneties in separate crystals. The sensitivity of the investigated compounds
ranges from very insensitive (2, 4) to very (6) and extreme (5) sensitive toward impact and
friction.
ACKNOWLEDGEMENTS
The authors are indebted to and thank Prof. Dr. P. Klfers for his generous allocation of
X-ray diffractometer time. Financial support of this work by the University of Munich
(LMU) and the Fonds der Chemischen Industrie is gratefully acknowledged. J.J.W. thanks
for a FCI Ph.D. fellowship (DO 171/46) and an Alexander von Humboldt postdoctoral
scholarship. The authors are also indebted to and like to thank Mr. Gunnar Spie for the
drop hammer and friction tests. Financial support by the WIWEB is gratefully
acknowledged.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
650
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
11
Czech Republic
2006
[Content]
Thomas Klaptke, Peter Meyer, Kurt Polborn, Jrg Stierstorfer, Jan Weigand
651
Czech Republic
2006
[Content]
Abstract:
The opportunity of deriving of effective emulsive explosives is rotined at introduction in
emulsive compositions such as a solution of saltpeter in oil of water solutions
inorganic peroxides. Responsiveness to an originating impulse and the high indexes of
serviceability (velocity of a detonation 5200-5500 m/s, critical diameter of a unclosed
charge of 80 mm) are ensured as a result of formation in volume of an emulsion of
microblisters of oxygen.
Keywords:
emulsion, explosive, sensitization , inorganic peroxides, oxygen
1.
INTRODUCTION
652
2.
Czech Republic
2006
[Content]
3.
For supply of a reliable detonation feeble fitchery grounds and decrease of a critical
diameter of a unclosed charge (for Ukrainit-HP-1 is peer 200 mm) the solution was accepted
653
Czech Republic
2006
[Content]
654
Czech Republic
2006
[Content]
Value
-0,05-0,18
2900-2960
1,25-1,30
3700-3850
0,762-0,892
80
5200-5500
As the received explosive is visible from the table has small values of design energy of
explosion. However, as have shown industrial trials, Ukrainit-HP-2 is characterized by high
indexes of serviceability at use both in fitchery grounds, and in soils of a rocky type.
As a result of a serial of industrial trials 5955 tons Ukrainit-HP-2, as a result of which
1189 thousand m3 mining and 3896 thousand m3 of rocky soils were blasted at
100 % of dampness of a mountain massif. Thus the oversize yield compounded no more than
0,01 %, and the specific consumption in rocky soils depending on a stability compounded
0,98-1,29 kg/m3.
Measuring gaseous composition of explosion products also were carried out at explosion
in a dummy underground working. Set, that the quantity of toxiferous gases in recalculation
on CO compounds 24,3 dm3/kg, that is much lower, than at explosion of regular explosives.
The high indexes of serviceability of explosive, at small values of design parameters
(energy of explosion of 2800-2900 kJ/kg) can be explained as follows. At first, during
sensitization in volume of an emulsion the microblisters of oxygen are formed, which one
accept fissile involvement in secondary oxidizing reactions in a gas phase, ensuring
practically 100 % a completeness of chemical reactions of explosive decomposition of
explosive. It results in practically complete embodying potentially of possible energy of
explosion of explosive except for thermal losses. Secondly, oxygen and the peroxides have
particular activity concerning a hydrocarbonaceous phase of an emulsion.
As is known, the acidification of the majority of organic compounds, and including oil
of hydrocarbons takes place on the radical mechanism[5]. The introduction in an emulsion of
solutions of inorganic peroxides results in initiation of radical acidification of a
hydrocarbonaceous phase of an emulsion, raising its reactivity and sensitivity to an
originating impulse.
655
a)
Czech Republic
2006
[Content]
b)
REFERENCES:
[1]
[2]
[3]
[4]
[5]
656
I.L. Kovalenko, A.V. Kuprin//Probl. of chem. and chem. techn., 4, P. 72-76, 2004.
F.A. Baum, K.P. Stanykovich, B.I. Shechter: Physics of detonating, Moscow,1959
K.K. Shvedov, V.V. Lavrov// Chem. Physics., V.17. 3, P. 74, 1998.
Pat. UA 74500: Emulsion explosive Ukrainit-HP-2; 20040705318; 15.12.2005
A.P. Glazkova: Catalysis of combustion of explosives, Moscow, 1976
Czech Republic
2006
[Content]
Abstract
Detonation propagation in liquid mixtures of Nitromethane (NM) with Azideethanol
(AE) was investigated. It was revealed that high velocity detonation (HVD) of the
mixtures of NM contained less 70% AE propagated steady in thick-walled steel tubes.
Addition of 20% AE to NM decreased df from 13 to 9 mm (in glass tubes). Critical
pressure of initiation of detonation in this mixture was measured. Detonability of jel-like
AE was investigated too. Effective kinetic parameters of reactions in detonation wave
for NM-AE mixtures were estimated.
Keywords: detonation, Nitromethane, Azideethanol.
1.
INTRODUCTION
2.
1
EXPERIMENTAL
KozakGD@nm.ru
657
Czech Republic
2006
[Content]
Commercial Nitromethane was purified by distillation at 101 0C. Neat Azideethanol was
extracted from aqueous solution containing 30% water by means of dichloromethane.
Solutions of AE in NM were prepared by mixing of components during 10-15 minutes.
Before every experiment the density of solution was measured by means of bottle method.
The dependence of solution density vs. content of AE in mixture is shown in Fig.1.
Porous material in the form of phenol-formaldehyde microspheres (PFS) was used for
decreasing of liquid AE density. Average size of microspheres was about 0.12 mm. 5% of
colloxylin was added in liquid AE for increasing of viscosity and stabilization of emulsion.
Fig 1.
Czech Republic
2006
[Content]
develop the process propagation on streak camera record we used modifying technique,
suggested in [14,15]. The radial holes were cowered by paraffin plugs near channel of tube; the
length of plugs was the same in all holes (8 mm). Small amount (~0.02 g) of lead azide was
promptly placed into each hole on paraffin plug, and then the holes were glued up with a
scotch. Luminosity of lead azide flashes fixing in streak camera record helped to judge about
a stability of process, and to measure its velocity. As it had been shown in works [14,15], the
line connecting points-flashes of lead azide in radial holes on a streak camera record was a
straight one, if low velocity process was not damp. If a propagating process damped the
slope of line diminished to the end of a charge.
3.
EXPERIMENTAL RESULTS
Fig 2.
The detonation failure diameter (df) of NM mixture with 20% AE was measured in glass
tubes. Steady HVD propagated in tube by diameter 10.5 mm and did not propagate at d= 8,49 mm. Extinction of detonation was observed in tube by d=9.5 mm. The value df=9.3 was
determined as critical diameter of this mixture.
Critical pressure of initiation was defined for mixture of NM with 20% AE (Tabl.1).
HVD was initiated by mixture RDX/NaCl 40/60 ( =0.95 g/cm3). Corresponding value of
critical pressure of HVD was Pin = 5.6 GPa. LVD was initiated by mixture RDX/NaCl 50/50
( =0.86 g/cm3), that corresponded to Pin = 4.5 GPa. Detonation did not propagate at Pin =
1.6 Gpa.
659
Czech Republic
2006
[Content]
Character of deformation
Tube
witness-plate
85
0.55
1.6
Intact
Intact
50
0.86
4.5
Perforation
60
0.95
5.6
Intact
Destruction
in fragments
along tube
Destruction
in twain
Result
No
explosion
LVD
HVD
It was interesting to determine detonation of neat AE. To this end the method of
sensitization of detonation for low detonatability explosives by introduction of microspheres
was used [16, 17]. Experimental density dependence on PFS content in gel-like AE could be
described by equation:
=1.202-0.059 PFS,
Fig 3.
660
4.
Czech Republic
2006
[Content]
CALCULATION RESULTS
Detonation and shock parameters as well as characteristics of the reaction zone defining
the process of reinitiation of detonation wave in the zone of instability were calculated by
means of SGKR computer code developed in our Laboratory previously [13]. Computation
procedure was described in detail in our previous papers [13,18].
The input data required for the calculation were the enthalpy of formation for the
substances, experimental dependencies of density, sound speed, specific heat from initial
temperature, and shock adiabat of the substances. Physical-chemical and shock
characteristics of NM were studied well [19]. Experimental density of AE AE=1.145 g/cm3
closes to NM value NM=1.13 g/cm3. The dependence of density on AE content in mixtures
is described well by additivity law of specific volumes (Fig.1}. The same temperature
coefficient of density 0/dT0=0.0014 g/(cm3K) for NM [19], AE and mixtures was assumed.
Sound speed, specific heat of AE were not known. These parameters were calculated having
used experimental values for liquid and gaseous HN3 [20,21] and liquid ethanol. The
generalized Hugoniot of liquids U = U0+0.29+1.63u [13] was used for calculation. The same
velocity of the reaction failure wave v= 3.8 km/s for all mixtures, neat AE as well as for neat
NM [4] was accepted.
Detonation parameters were calculated for NM, AE and their mixtures. Calculated heat
of explosion (Qv,) decreased with increasing content of AE in the mixture, as expressed by
the linear equation:
Qv= 5290 8.47CAE, kJ/kg,
Fig 4.
661
Czech Republic
2006
[Content]
From the beginning characteristics of reaction zone (pressure p3, temperature T3 and
reaction time 3) near the limit of detonation propagation were calculated for neat NM and
the mixture containing 20% AE by experimental df values (Tabl.2). Parameters pcj, D, p3 for
NM and the mixture differed slightly from each other. The T3 value for NM is 50 K above
value for the mixture. Reaction time 3 at addition of 20% AE to NM reduced according to
decreasing of df.
Table 2.
The failure detonation diameter (df), detonation parameters (D, pcj) and
characteristics of reaction zone (pressure p3, temperature T3 and reaction
time 3) near the limit of detonation propagation for NM, the mixture of
NM with 20%, and AE.
CAE, %
df, mm
Exper. Calcal.
D,
m/s
pcj,
GPa
p3 ,
GPa
3,
3 ,
mcs
13
13
6334
12.8
9.2
1061
0.24
20
9.3
6334
12.9
9.3
1012
0.17
100
133
6.421
13.4
9.6
809
2.45
Earlier we established linear correlations between the pressure P3 and the detonation
pressure Pcj, and between df and 3 for liquid mixtures [18,22]. These correlations allow to
evaluate parameters p3, T3, 3 and df over a wide range of compositions, if only two df
values (for neat substances or mixtures) were determined. Since detonation pressure Pcj did
not dependence from AE content in mixture, pressure P3 was invariable too. The T3 value
was defined by shock temperature dependency T(p) at correspondent value of p3 for each
mixture (Fig.4). The addition of AE to NM decreased T3 in mixture. The calculated
dependence of T3 on the concentration of AE in solutions is expressed by the linear equation:
To predict values of df for liquid mixtures in addition to dependency T3(CAE) the kinetic
parameters of detonation reaction are necessary.
5.
DISCUSSION
662
Czech Republic
2006
[Content]
(1)
where k1 k2 are the rate constants of decomposition reaction in neat NM and AE corr.,
(3)
The dependencies of kinetic constant k1, k2, and kef vs. 1/T3 are shown in Fig.5. Constant
kef decreased from NM to AE. According to calculation the maximum value of kef occurred
at the concentration 20% of AE in mixture.
663
Fig 5.
Czech Republic
2006
[Content]
The kinetic constant kef was used for calculation of df(CAE) dependence for NM-AE
mixtures (Fig.6). Calculated df values for NM, and the mixture of NM with 20% AE are 13
and 9 mm correspondingly. The results of calculations of df are in good agreement with
experimental data. Estimation of critical diameter of AE with using eq.3 shows that df is
about 130 mm. This value is greater than equivalent critical diameter of steel thick-walled
tube deq=805 mm [18]. This significant fact explains non-detonation of AE in steel thickwalled tube.
Conversion of liquid AE into porous gel by means of introduction 5% of colloxylin and
1-4% PFS facilitated propagation both HVD and LVD. LVD was initiated at gel density
=1.1-1.2 g/cm3 and HVD was initiated at =1g/cm3.
The addition of 20% AE to NM decreased df and as well as critical pressure of HVD
initiation from 10 to 5.6 GPa. Critical pressure of LVD initiation for the mixture conversely
was higher than for NM. So introduction of AE in NM facilitates HVD initiation but
impedes propagation of LVD in liquid.
664
Fig 6.
6.
Czech Republic
2006
[Content]
CONCLUSION
Limited conditions of HVD and LVD in mixtures of NM with AE and jel-like AE were
investigated. It was revealed that NM mixtures containing 70% and less are detonable.
Addition of 20% AE to NM decreased critical diameter and initiation pressure of HVD.
Detonation reactions in the mixtures were considered in terms of A. Dremins theory and
SGKR computer code calculations. The df values of NM, AE and their mixtures were
estimated by chemical kinetic and correlative equations.
ACKNOWLEDGEMENTS
Authors are grateful to Foundation of Basic Research of RF, that supported the
investigation through grant 04-03-32713. Our post graduated students V.V. Potapov and
E.A. Likholatov took part in carrying out the experiments and calculation
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
A.N DREMIN: The failure detonation diameter of liquid explosives, Proceedings of Akad. Sci.
USSR (Doklady), v.147, p. 870-873, 1962 (Russ.)
A.N. DREMIN, V.S. TROFIMOV: Calculation of Failure Detonation Diameter of Liquid
Explosives, Problems of mechanics and theoretical physics, 1, p. 126-131, 1964 (Russ.)
J.W.ENIG AND F.W. PETRONE: The failure diameter theory of Dremin, 5th Symposium (Int.) on
Detonation, Pasadena, Ca. Aug. 18-21, p. 99-104, 1970.
C.N. TARVER, R. SHAW AND M. COWPERTHWAITE: Detonation failure diameter studies of four
liquid nitrialkanes, J.Chem.Phys. v.64, 6, p. 2666-2673, 1976.
B.N. KONDRIKOV, G.D.KOZAK, V.M. RAIKOVA, A.V. STARSHINOV: Detonation of
Nitromethane, Proceedings of Akad. Sci.USSR (Doklady), v.233, 3, p. 402-405, 1977
(Russ.)
B.N. KONDRIKOV, G.D. KOZAK, , V.M. RAIKOVA, , A.V. STARSHINOV: Homogeneous
catalysis at detonation of nitromethane, Detonation, Akad. Sci. USSR, Chernogolovka, p. 7376, 1977 (Russ.)
George Kozak, Vlada Raikova
665
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
666
Czech Republic
2006
[Content]
10
Czech Republic
2006
[Content]
11
667
Czech Republic
2006
[Content]
Abstract:
This paper reports the relationship between friction sensitiveness data of explosive
substances obtained by the BAM friction apparatus using several different velocities of
rubbing surfaces. The CO and NO analyzer was used for automatic determination of
load adequate to the first reaction (initiation) of tested substances. The little rubber
chamber for keeping decomposition gases was developed during this research work.
Keywords: Friction sensitivity, rubbing surfaces velocity, explosives, BAM friction
apparatus
1.
INTRODUCTION
2.
EXPERIMENTAL
The following explosive samples were studied:
The samples of PETN type NpS (Fig.1) and NpD (Fig.2) manufactured by Synthesia and
RDX type Hx E (Fig.3) and Hx C (Fig.4) manufactured by Dyno Nobel. The samples of
each explosive differ in particle size. The quantity of tested sample was 10 mm3.
668
Czech Republic
2006
[Content]
Measure was carried out using new developed apparatus FSKM-50/20K (Fig. 5)
manufactured by OZM Research. This instrument fully complies with UN Tests and Criteria
Manual - BAM Friction sensitivity test5. The using of intelligent compact drive with fieldbus
stepper motor is the main advantage compared to instruments driven by electromotor with
the gear box traditionally used. This solution allows to control not the speed with high
precision only, but also precise positioning of the motor to the starting positions. The
overshoots or operating cycle incompletion often occur using some older apparatuses.
The velocity, acceleration and deceleration can be changed easily with operating
software. The maximal average velocity of rubbing plates of this apparatus is closed to 10
cm.s-1 because of mechanical reasons. The minimal velocity is not limited.
669
Czech Republic
2006
[Content]
The average velocity corresponding to total time of working cycle was changed in three
levels during testing. The real average velocity of each level was verified by independent
method.
Rubbing surfaces
velocity
Time of 1 cycle
(s)
(cm.s-1)
Number of
revolutions per
minute (r.p.m)
3.33
0.600
100
4.66
0.429
140
10.00
0.200
300
3.
The change of the rubbing surface velocity influences significantly the determined
values of friction sensitiveness. The difference of 1 cm.s-1 causes the difference of 6-20 % in
determined sensitiveness. The decrease of sensitiveness with the increase of friction velocity
was observed in case of all samples. The biggest decrease was observed in case of samples
containing big crystals, as can be seen in table 1 and in figure 6. The values of friction
sensitiveness obtained at maximal friction velocity are 25-50 % smaller against the values
obtained at minimal velocity.
Velocity cm/s
3,33
4,66
62,8
59,0
47,3
15,5
24,7
59,0
53,2
41,4
17,6
29,8
RDX Hx E
Load [N]
78,5
70,6
53,2
25,3
32,2
RDX Hx C
Load [N]
78,5
62,8
39,3
39,2
49,9
670
Czech Republic
2006
[Content]
90,0
PETN NpS
PETN NpD
RDX Hx E
RDX Hx C
80,0
Load [N]
70,0
60,0
50,0
40,0
30,0
0
10
12
Velocity [cm/s]
Fig. 6: Relationship between rubbing surfaces velocity and sensitiveness by First reaction
Design of BAM Friction sensitiveness tester has one big disadvantage. The sample is
triturated only by the top of spherical part of the pin. The friction area is therefore very
small. The less sensitive samples containing big particles give very scattered results. The pin
only with problems triturates the big crystals; their moving away is more probable. The
triturating and following ignition is very irregular and comes on only in case of suitable
crystal position and/or suitable crystal shape. The probability of ignition increases with
increasing friction velocity. This rule does not show itself very much in case of PETN
samples, due to the high PETN sensitiveness (see figure 6). Both PETN curves are almost
parallel. The both RDX samples have the same level of sensitiveness at minimal friction
velocity. The first reaction friction load of RDX Hx C sample, containing big crystals,
determined at maximal velocity, falls to one half of the load determined at minimal velocity.
This sample is more sensitive than PETN samples at the friction velocity of 10 cm.s-1 and
demonstrates the sensitivity increase with increasing friction velocity in case of less sensitive
samples containing big crystals.
PETN NpD 4,66 m/s
Load [N] CO/NO [ppm]
59,0
1/7
53,2
2/8
47,3
0/2
RDX Hx E 4,66 m/s
Load [N] CO/NO [ppm]
78,5
0/5
70,6
0/9
62,8
0/4
CO/NO [ppm]
CO/NO [ppm]
CO/NO [ppm]
CO/NO [ppm]
3/9
1/7
2/7
2/8
0/7
0/5
1/7
0/5
0/0
0/1
0/0
0/1
CO/NO [ppm]
CO/NO [ppm]
CO/NO [ppm]
CO/NO [ppm]
0/7
0/11
0/7
0/8
0/8
0/8
0/8
0/9
0/1
0/2
0/3
0/2
671
Czech Republic
2006
[Content]
Some of the results closed to first reaction friction load for the samples PETN NpD and
RDX Hx E are stated in table 2 (friction velocity 4.66 cm-1). Despite the CO and NO
concentration measurement is not quantitative, it is sufficiently reproducible and it increases
the objectivity of sensitiveness determination without the respect to the experience of the
operator.
4.
CONCLUSION
The velocity of rubbing surfaces is an important factor which influences the results of
friction sensitivity testing. It was found that relationship between velocity of rubbing surface
and sensitiveness exists. The changes in sensitiveness are significant and can cause relevant
deviations of results. The big differences were observed in case of less sensitive samples
containing bigger crystals. This is one of the disadvantages of the BAM friction
sensitiveness apparatus design. The control of velocity and automated evaluation of test
results by toxic gas analyzer bring work simplification, safer and improve the reproducibility
of obtained results.
Acknowledgements
This work was created as a part of Ministry of Education, Youth and Sports f the Czeech
Republic No. MSM 0021627501.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
672
Czech Republic
2006
[Content]
673
Czech Republic
2006
[Content]
Formula
Mol. mass
Oxygen
balance
Melting
temperature, C
Heat of
explosion, J/g
TMD,
g/ccm
HMX
48N88
296
-22
278(d)*
5632
1,9
II
44N814
388
90
5138
1,96
68N1016
476
+17
178(d)
6450
1,82
674
Czech Republic
2006
[Content]
R,
r , J /g
/m in
( v ) ( %)
HMX
10
2
HMX +
1 0 %A l
10
2
HMX +
5% B
10
2
HMX +
2 0 %Zr
10
II
10
3492
(2.4)
3476
(3.9)
3600
(4.9)
3063
(11.3)
2948
(21.9)
6696
(3.0)
3109
(5.7)
4632
(45.4)
5902
(8.1)
2318
3174
3298
3036
(24.9)
2716
(16.2)
4320
(0.78)
r / t , %
4
62
62
64
54
52
119
55
82
105
41
56
59
59
53
84
d (v),
0
( %)
max (v),
0
( %)
268.8
270.0
(0.39)
(0.26)
275.0
276.0
(0.25)
(0.23)
276.7
278.0
(0.36)
(0.33)
272.9
273.6
(0.37)
(0.46)
277.0
278.1
(0.14)
(0.15)
277.7
278.4
(0.09)
(0.09)
274.5
275.4
(0.22)
(0.17)
277.0
277.7
(0.13)
(0.31)
278.0
278.8
(0.23)
(0.24)
272.6
273.7
273.8
274.9
276.9
278.7
148.9
156.1
(0.40)
(1.1)
159.8
168.7
(0.64)
(0.21)
164.5
176.3
(0.06)
(0.28)
, s
36
12
7.8
21
13
4.2
27
8.4
4.8
33
13
11
216
107
71
675
2
2
II +
10%Al
10
2
II +
5%B
10
2
II +
2 0 % Zr
10
10
2
I +
10%Al
10
2
I+
5%B
10
+20%Z
r
5
10
676
3
3098
(17.6)
2941
(4.5)
3569
(2.5)
4503
(33.1)
4436
(17.1)
5189
(5.1)
3293
(6.8)
2160
(44.7)
4384
(12.4)
4197
(1.9)
3064
(20.8)
4193
(6.6)
3823
(6.8)
4078
(10)
4539
(21)
5143
(1.6)
5037
(0.12)
6446
(13.3)
4583
4100
4869
4
60
57
69
88
86
101
64
42
85
65
48
65
59
63
70
80
78
100
71
64
75
Czech Republic
148.9
156.6
(0.98)
(0.39)
158.9
167.9
(0.16)
(0.23)
162.1
176.1
(0.13)
(1.7)
148.4
158.3
(1.9)
(1.1)
154.4
165.9
(0.86)
(0.43)
158.1
172.1
(2.4)
(2.4)
147.7
158.7
(0.072)
(1.0)
161.1
169.3
(0.11)
(0.29)
167.6
176.9
(2.1)
(0.34)
182.5
184.8
(0.11)
(0.07)
188.6
191.4
(0.24)
(0.41)
193.7
197.4
(0.12)
(0.07)
183.4
185.7
(0.16)
(0.11)
189.6
192.3
(0.04)
(0.01)
194.2
197.6
(0.09)
(0.08)
190.1
191.4
(0.048)
(0.02)
193.7
195.2
(0.022)
(0.004)
197.0
198.5
(0.18)
(0.67)
184.0
185.9
185.8
189.5
193.0
197.1
2006
[Content]
231
108
84
297
138
84
330
98
56
69
34
22
69
32
20
39
18
9.0
57
44
25
Czech Republic
2006
[Content]
677
Czech Republic
2006
[Content]
The rate of diffusion process grows sharply up at higher temperature that is realized at
decomposition of HMX. In this case the sharp increase of heat effect at high rate of heating
in a mixture of HMX with aluminum nanopowder is observed.
Thermal decomposition of substance "I" and, especially, "II" goes on much more slowly,
despite of much higher contents of oxygen in its molecule. In decomposition products of
substance II diffusion difficulties are not overcome, the heterogeneous reaction does not
go on fast. This observation proves assumption stated in work[4] that at high temperatures in
relation to particles of aluminum covered by oxide film, water steam is more active oxidizer,
than the pure oxygen.
The fact, that the main role in reaction of aluminum particles with products of
decomposition of the studied nitramines plays a diffusion factor, proves results of thermal
analysis of mixtures with the boron nanopowder.
Boron as the chemical element is closed to aluminum in respect of energy of formation
of oxide and to chemical similarity in relation to products of decomposition of nitramines.
The specific surface of a studied boron nanopowder is also close to the specific surface of
aluminum nanopowder.
It was already mentioned, essential difference of these two powders in relation to
reaction with oxygen-containing gases (H2O, CO2, CH2O) are diffusion difficulties.
As it is seen from thermal analysis results, submitted in Table 2, the sharp reduction of
diffusion barrier in the case of boron results that additional realize of decomposition heat of
HMX mixtures begins already at rate 5 K/min and grows by 20 % at the increase of rate up
to 10 K/min
In boron mixtures with substances "I" and "II" additional heat release is observed
already at the minimum rate of heating. For substance "II" the rather additional heat release
remains constant at all rates of heating, but for substance "I", as well as for HMX additional
heat release grows at the maximal rate.
The process of decomposition of substance "II" goes on at lower temperatures, than of
substance "I" and, especially, for HMX, it is less intensive, is more protracted in time. The
losses of heat are significant, that results in the decrease of temperature at a maximum rate.
This is probably due to incomplete boron reaction. This reason explains also the lowered
heat release in boron mixture with substance "II" in comparison with HMX and substance
"I".
It confirms the assumption that a major factor determining the rate of interaction of an
active additive with products of thermal decomposition the nitramines, is the diffusion
resistance of oxide film. The resistance decreases at the increase of temperature, and this
reduction depends on properties of oxide. For boron the diffusion resistance is lower, than
for aluminum even at the lower temperatures. The high specific surface of particles results in
increase of reactivity. The low reactivity of the zirconium powder is caused by strong
agglomeration of its primary particles. So, despite of the high specific surface of pores in
agglomerates, the penetration of products of decomposition inside of large agglomerates is
difficult.
It is possible to assert that from four factors influencing to the reactivity of the
inflammable solid additives to the energetic materials, namely: 1) temperature; 2) specific
surface; 3) oxygen balance; 4) diffusion resistance the first factor is the main one (the 4-th
factor is connected to the first).
678
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
Chemical encyclopedia, in five volumes, " The Soviet Encyclopedia ", ., 1988, (in Russian).
G. B. Manelis, et al. Thermal decomposition and combustion of explosives and gunpowders,
"Science", ., 1996, (in Russian).
E. Yu. Orlova et al. HMX - heat-resistant explosives. "Nedra", ., 1975, (in Russian).
V.M. Maltsev, et al. The basic characteristics of combustion. "Chemistry", ., 1977, (in
Russian).
679
Czech Republic
2006
[Content]
Abstract:
In order to obtain mixtures of low sensitivity to external stimuli, the explosives like
octogene (HMX) or hexogene (RDX) were mixed with different quantities of
nitroguanidine (NQ) or 3-nitro-1,2,4-triazol-5-one (NTO). In this paper, a method of
preparation of charges to tests is described. The measurement results of the sensitivity
to friction and shock wave generated by water layer are presented. On the basis of the
results it was found that, the addition of NQ or NTO considerably decreases friction
sensitivity of the compositions, whereas the sensitivity to shock wave is decreased in a
lesser degree.
1.
INTRODUCTION
During last years the main subject in world development of the military technique field
we could notice that so many people are interested in explosives of low sensitivity to
external stimuli in comparison with widespread like: trinitrotoluene (TNT), hexogene,
octogene, pentaerythritol trinitrate (PETN). This is the result of concerning about soldiers
security.
Unexpected explosion and unsuitable handle of explosives can sometimes lead to
accidents. The experts claim that we should reduce the number of accidents in the army by
supplying modern low sensitivity ammunition.
The following materials belong to the group of less sensitive high explosives : 1,2,4triazol -5-one (NTO), 1,3,5-triamino-2,4,6-trinitrobenzene (TATB), 2,2,4,4,6,6heksanitrostilbene (HNS), tetranitrodibenzo- 1,3a,4,6a-tetraazapentalene (TACOT), and
nitroguanidine (NQ) [1 7].
2.
For these studies, less sensitive mixtures of different chemical constitution were
selected. The mixtures included hexogene or octogene as components of high sensitivity and
NQ or NTO as components which decreased the sensitivity. In order to monitor the influence
of changes of chemical contents on explosive properties of these mixtures, we decided to
make changes in mutual percentage fraction of the particular components and finally we
received twelve mixtures. This set of compositions was put to various tests.
680
Czech Republic
2006
[Content]
(RDX+0,5%C) / NQ
70/30
50/50
30/70
(RDX+0,5%C) / NTO
(HMX+0,5%C) / NQ
70/30
50/50
30/70
70/30
50/50
30/70
(HMX+0,5%C) / NTO
70/30
50/50
30/70
The majority of the tests was realized on the samples of mixtures which was pressed by
hydraulic press to the possible high density. In order to obtain the mixtures, the components
were weighed in portions (1000 g), and mixed by shaking in foil container for 8 10
minutes. The loose mixture was divided in portions (10 - 50 g) and was pressed under
pressure 200 MPa. The diameters of dies were 20 - 40 mm. The density of obtained blocks
was determined by measurements of weight and dimensions. The obtained blocks were
glued together by nitrocellulose lacquer. The columns formed in this way were tested.
3.
The measurements of the sensitivity to friction were made according to PN-C86019:1994 standard by using Joulus-Petri apparatus. The lower insensitivity point of
mixtures was determinated. It is the maximum pressure of stamp (in Newton), which causes
minimum one positive result (explosive process) during six trials. The loose mixtures were
used to the all tests. The results of our experiments are given in table 2 and figure 1.
Table 2. Measurement results of sensitivity to friction of tested mixtures
Explosive
RDX/NTO
RDX/NQ
HMX/NTO
HMX/NQ
RDX PETN
compositions 70/30 50/50 30/70 70/30 50/50 30/70 70/30 50/50 30/70 70/30 50/50 30/70 cryst. cryst.
Friction
247
283
353
235
283
353
235
247
318
212
235
318
120
60
sensitivity
[N]
681
Czech Republic
2006
[Content]
400
350
300
250
200
283
247
353
235
283
353
235
247
318
212
235
318
150
100
120
50
60
0
RDX/NTO RDX/NTO RDX/NTO
70/30
50/50
30/70
Fig 1.
4.
RDX/NQ
70/30
RDX/NQ
50/50
RDX/NQ
30/70
HMX/NQ
50/50
HMX/NQ
30/70
RDX
PETN
Czech Republic
2006
[Content]
2
3
8
5
Fig 2.
683
Czech Republic
2006
[Content]
1
2
4
5
8
9
Fig 3.
The system for measurments of sensitivity to shock wave of explosives: 1 indicator plate, 2 - detonating cord, 3 upper PCV tube, 4 tested charge,
5 - joining tube, 6 water layer, 7 - donor charge, 8 - lower PCV tube, 9 detonator
After the initiation of detonation in tested charge (7), the shock wave permeates through
the water layer (6) and depending on wave intensity (regulated by thickness of water layer)
the charge detonates or it does not. Since the applied charges are small, only clear trace after
detonation of detonating cord (2) on indicator plate (1) confirms that the tested charge
(4)was detonated.
684
Fig 4.
2006
[Content]
Fig 5.
Czech Republic
The indicator plates and pieces of detonating cord unless tested charges
were detonated: 1 RDX/NQ 70/30-27 mm, 2 RDX/NQ 70/30-27 mm, 3
HMX/NTO 70/30-24 mm, 4 RDX/NTO 30/70-24 mm, 5 HMX/NQ
30/70-22 mm, 6 RDX/NTO 70/30 -5 mm
The appearance of indicator plates after the detonation of tested charges is presented
in figure 4, whereas the plates without detonation of the charges are shown in figure 5. The
number of millimeters which is presented in the description indicates thickness of water
layer which shock wave permeates through. The measurement results of sensitivity for
mixtures RDX/NQ, RDX/NTO, HMX/NQ, HMX/NTO are presented in table 3 and figure 6.
685
Czech Republic
2006
[Content]
Density
[g/cm3]
HMX/NTO 70/30
HMX/NTO 50/50
HMX/NTO 30/70
HMX/NQ 70/30
HMX/NQ 50/50
HMX/NQ 30/70
RDX/NTO 70/30
RDX/NTO 50/50
RDX/NTO 30/70
RDX/NQ 70/30
RDX/NQ 50/50
RDX/NQ 30/70
1,75
1,77
1,77
1,69
1,66
1,62
1,66
1,71
1,73
1,64
1,62
1,60
P [GPa]
1,2
Shock wave
pressure [GPa]
1,00
1,05
1,20
1,00
1,05
1,20
0,95
1,00
1,00
0,75
1,05
1,05
1,20
1,20
1,05
1,05
1,00
1,05
1,05
1,00
1,00
1,00
0,8
0,95
0,6
0,75
0,4
0,2
0
30/70
HMX/NTO
HMX/NQ
50/50
RDX/NTO
RDX/NQ
Fig 6.
70/30
The comparison of results reported by Timborn [8] (six columns on the left) with
chosen results which was obtained for the most sensitive tested mixtures (six columns on the
right) is presented in figure 7.
686
Czech Republic
2006
[Content]
P [GPa]
1,80
1,60
1,70
1,40
1,20
1,20
1,20
1,20
1,20
1,00
1,05
1,00
1,00
0,80
0,75
0,70
0,60
0,40
0,20
0,00
RDX/TNT 60/40
Fig 7.
5.
RDX/WOSK
95/5
PETN/WOSK
97/3
TETRYL
PETN/TNT
50/50
TNTpras.
RDX/NQ 70/30
HMX/NTO
70/30
HMX/NTO
50/50
HMX/NTO
30/70
SUMMARY
The comparison of measurement results shows that the addition of less sensitive
component causes considerable decrease of sensitivity to friction of tested mixtures. The
content of 30% NQ or NTO in mixture generates double decrease of the sensitivity. It can be
stated that, only at a 70% share of the additives these mixtures are less sensitive to friction.
A load of above 300 N is required to the initiation of these mixtures.
The studies of sensitivity to shock wave of tested mixtures were made according to
the modified gap test (Tab.3, Fig. 6-7). These tests show that the sensitivity of mixtures
with RDX is higher than the sensitivity of mixtures with HMX. In the range of changeable
quantity of the same additives, the sensitivity was changed according to expectations. Tested
mixtures have a sensitivity level such as pressed TNT, pentolite 50/50, or desensitized RDX
Only at a 70% share of low sensitive additive. At different content of the additive, the
sensitivity level correspond to the tetryl.
687
6.
Czech Republic
2006
[Content]
REFERENCES
[1]. G.I. CHIPEN, R.P. BOKALDER, V.YA. GRISHTEN, 1,2,4-Triazol-3-one and Its Nitro and
Amino Derivatives, Khimia Geterogennykh Soedinienii, 1966, 2, 1.
[2]. K-Y. LEE, M.D. COBURN, 3-Nitro-1,2,4-triazole-5-one a Less Sensitive Explosive, Los Alamos
National Laboratory Report No. LA-10303-MS, Feb 1985.
[3]. K-Y. LEE, M.D. COBURN, 3-Nitro-1,2,4-triazole-5-one a Less Sensitive Explosive, Journal of
Energetic Materials, 1987, 5, 27-33.
[4]. K-Y. LEE, M.D. COBURN, 3-Nitro-1,2,4-triazole-5-one a Less Sensitive Explosive, pat USA
4733610, 1988.
[5]. E. VON RIPPER, G. K, ber - und -Nitroguanidin, Explosivstoffe 1969, 17, 7, 145 151.
[6]. W. ENGEL, H. HEINISCH, Process for Preparation of Compact Nitroguanidine, pat. USA 4
544 769, 1985.
[7]. J.A. SANCHEZ, E.L. ROEMER, L.A. STRETZ, Spherical Nitroguanidine Process, pat. USA 4
967 000, 1990.
[8]. M. HELD, GAP TESTS, Energetic materials, Reactions of Propellants, Explosives and
Pyrotechnics, 34-th International Annual Conference of ICT, Karlsruhe, FRG, 24-27 June,
2003.
[9]. R. MEYER, J. KHLER, A. HOMBURG, Explosives, 5 rev. and extended ed., Wiley VCH
Verlag GmbH, Weincheim 2002.
688
Czech Republic
2006
[Content]
1.
INTRODUCTION
2.
1-Methyl-3,5-dinitro-1,2,4-triazole (MNTA), 3,5-dinitro-1,2,4-triazole (NTA), and N(1,2,4-triazol-3-yl)-2,4,6-trinitrophenylamine (PATO) were synthesized according to the
published methods. [8-10] The purity of compounds has been determined using melting points,
TLC, NMR, IR, and elemental analysis. All the other compounds were obtained from Sigma
and used as received.
689
Czech Republic
2006
[Content]
3.
In this work, we examined the enzymatic redox reactions of NTA, MNTA, and PATO
which may be important in their cytotoxicity: i) their single-electron reduction by
flavoenzyme electrontransferase ferredoxin:NADP+ reductase (FNR) which is frequently
used as a model system to study the free radical formation of nitroaromatic compounds and
their redox cycling ability.[4-7] The reactivity of FNR increases with the increase in electronaccepting potency of nitroaromatics (single-electron reduction potential), and is insensitive
to their structural peculiarities; and ii) their two-electron reduction by rat NAD(P)H:quinone
oxidoreductase (DT-diaphorase, NQO1), which is one of the main enzymes responsible for
the formation of toxic nitroso- and/or hydroxylamine reduction products in mammalian
cells.[4-7] We also determined the concentrations of the above compounds for 50 % survival
of FLK cells (cL50). The results obtained are presented in Table 1, together with the
previously determined enzymatic reduction rate constants and cL50 of other nitrotriazole
explosives, 5-nitro-1,2,4-triazol-3-one (NTO) and 5-nitro-1,2,4-triazol-3-amine (ANTA),
and model compounds, nitrobenzene and 2,4,6-trinitrotroluene (TNT).
According to the calculated Hf(ArNO2-.) values, the electron-accepting potency of
NTA and MNTA in vacuo is intermediate between those of TNT and nitrobenzene, whereas
the electron-accepting potency of PATO is close to TNT (Table 1). In general, the rate
constants of single-electron reduction of nitroaromatics by FNR increase with the decrease in
their Hf(ArNO2-.), although the dependence is uncertain (Table 1). It is evident that the
order of cytotoxicity of compounds, nitrobenzene ~ NTO, ANTA NTA, MNTA < PATO <
TNT almost parallels their reactivity towards FNR (Table 1), i.e, the rate of enzymatic free
radical formation. It points to the possibility of the oxidative stress-type cytotoxicity
mechanism of NTA, MNTA, and PATO. The prooxidant character of cytotoxicity is further
690
Czech Republic
2006
[Content]
evidenced by the partial protection by the low molecular weigth antioxidant N,N-diphenylp-phenylene diamine (DPPD), the iron-chelating agent desferrioxamine, the latter preventing
the Fenton reaction, and by the antioxidant enzyme catalase (Table 2). However, an inhibitor
of NQO1, dicumarol, also partly protects against the cytotoxicity (Table 2), thus pointing to
the toxic effects of the products of NQO1-catalyzed reduction.
Table 1. Enthalpies of the formation of nitroanion-radicals, rate constants of the
reduction of nitroaromatic compounds by FNR and NQO1, and their
concentrations for 50 % survival of FLK cells (cL50). aThe data taken from
our previous work;[6] bThe kcat values (s-1) are given in parentheses; ccL50
is close to the solubility limit.
Hf(ArNO2-.) (kJ/mol)
Compound
kcat/Km (M-1s-1)
cL50
b
(M)
AM1
PM3
FNR
NQO1
NTA
-280
-293
107
15 (0.03)
2700380
MNTA
-271
-245
240
40 (0.10)
2300300
PATO
-307
-318
46000
2400 (0.14)
50c
NTOa
-242
-236
55
27 (0.09)
3500c
ANTAa
-148
-159
260
29 (0.09)
3000400
-308
-316
11000
670 (1.0)
255.0
-168
-172
11
24 (0.06)
43701470
TNT
Nitrobenzene
Compound
PATO (50 M)
MNTA (2000 M)
2 M
DPPD
300 M
desferrioxamine
100 U/ml 20 M
catalase dicumarol
59.02.0
88.24.0
58.42.7
88.83.9
98.02.5
82.23.0
84.12.8
3
NTO (3000 M)
55.03.7 75.83.5 78.23.5
n.d.
68.03.8b
[6]
This work and our previous study demonstrate that the presence of triazole group in
NTO, ANTA, NTA, MNTA, and PATO does not affect their mammalian cell cytotoxicity,
691
Czech Republic
2006
[Content]
i.e., these compounds follow the same electron-accepting potency vs. cytotoxicity
relationship as polynitrobenzenes.
Acknowledgements
This work was supported in part by Lithuanian Science and Studies Foundation. We
thank Dr. M. Martinez-Julvez and Professor C. Gomez-Moreno (Zaragoza University,
Spain) for their generous gift of ferredoxin:NADP+ reductase.
REFERENCES
[1]
692
Czech Republic
2006
[Content]
OPTIMALIZATION OF CRYSTALLIZATION OF
1,1-DIAMINO-2,2-DINITROETHYLENE
Andrzej Orzechowski*, Dorota Powaa *, Bogdan Florczak*,
Andrzej Maranda**, Jerzy Nowaczewski**
Institute of Industrial Organic Chemistry, 6 Annopol St, 03-236 Warsaw, Poland
Military University of Technology, 2 Kaliskiego St, 00-908 Warsaw 49, Poland
Abstract:
In this paper, the method of synthesis and crystallizations of 1,1-diamino-2,2dinitroetylene was presented. The aim of the work was to obtain cubic crystals of
FOX-7. Several solvents for crystallization of FOX-7 were checked. The influence of
various parameters on crystallization process was investigated The mixture DMSO and
water as a solvent to crystallization was the best.. The particle size distribution and the
tap density of the obtained crystals are presented.
Keywords: crystallization, 1,1-diamino-2,2-dinitroetylene, FOX-7
1.
INTRODUCTION
Growth of terrorist threats and military technology development causes the necessity
of new insensitive explosives elaboration. Hexogen, nitropenta, octogen are commonly used
high explosives. They are characterized by high explosive energy and low sensitivity to
external stimuli. The increase of resistance to mechanical, electrical and thermal, without
worsening other parameters, is realized by the addition of such explosives like
nitroguanidine (NGU), 3-nitro-1,2,4-triazol-5-one (NTO) or 1,3,5-triamino-2,4,6trinitrobenzene (TATB) [1-3].
FOX-7 is an interesting new high explosive because of its low vulnerability [4]. In the
literature, various synthesis methods of FOX-7 were referred. The best methods of FOX-7
preparing from acetamidine are described in [5-7], and optimization of crystallization methods
of this compound are shown in [5-10].
In our laboratory, we conduct researches on crystallization of FOX-7 in order to obtain a
suitable form of this explosive for application in Plastic Bonded Explosives.
2.
EXPERIMENTAL
Properties of 1,1-diamino-2,2-dinitroetylene
FOX-7 is a yellow crystalline substance, the density is 1.885 g/cm3 [11]. The impact
sensitivity of this explosive is 24.7 Nm and the friction sensitivity is above 350 N. The
solubility of FOX-7 in popular organic solvents i.e. acetone, methyl and ethyl alcohols, ethyl
acetate, ether, carbon tetrachloride, toluene, is limited, except N,N- dimethylformamide,
dimethylsulfoxide, N-methylpyrrolidone.
2.1
Synthesis of 1,1-diamino-2,2-dinitroethylene
FOX-7 was prepared by reacting dioctyl malonate with acetamidine hydrochloride in the
solution of sodium methylate, and next nitration of obtained 2-methylpyrimidyne1
693
Czech Republic
2006
[Content]
4,6(3H,5H)-dione (Fig 1). According to [8] 1 mole of acetamidine hydrochloride was added
to the 2,5 mole of the sodium methylate solution in methanol (concentration of CH3ONa was
97 %). At room temperature 1 mole of the dioctyl malonate solution in methanol was added.
The mixture was stirred at 50 oC for 2 hours and then, the one was left for 24 hours (20 oC).
After that hydrochloric acid was added to the mixture (to pH 6). The second step was
nitration of methylpyrimidyne-4,6(3H,5H)dione. This compound was dissolved in
concentrated sulfuric acid (about 11 mole) at 20 oC, next the nitric acid was added at the
same temperature for 2 hours (H2SO4/ HNO3 3/1 molar ratio).The yield of the reaction was
74 %.
12
10
Volume [%]
0
5,7
137
270
403
535
668
801
Fig 1.
2.2
Crystallization of 1,1-diamino-2,2-dinitroethylene
Solubility [%]
80
70
DMSO
60
DMSO + H2O
DMF
50
NMP
40
30
20
10
0
20
30
40
50
60
70
80
90
Temperature [ C]
Fig 2.
694
Czech Republic
2006
[Content]
20
18
16
Volume [%]
14
12
10
8
6
4
2
0
5,7
137
270
403
535
668
801
Fig 3.
Volume [%]
0
5,7
137
270
403
535
668
801
Fig 4.
Example 2
This test was carried out in the same way just like in example 1. The difference between
these two examples was cooling rate. In this case it was about 20 C/min. The obtained
695
Czech Republic
2006
[Content]
product is presented in Fig. 5. Tap bulk density 0.63 g/cm3. We obtained lover bulk density
of FOX-7 than above.
3,5
Volume [%]
2,5
1,5
0,5
0
5,7
137
270
403
535
668
801
Fig 5.
Example 3
FOX-7 was added to a flask with DMF. The mixture was heated with mechanical
stirring (800 rpm) to 90 C. Acetone was added to the mixture. The product was filtered and
washed with water and acetone. The bulk density of FOX-7 , produced in this way, was 0.90
g/cm3. The crystals are presented in Fig. 6.
12
10
Volume [%]
0
5,7
137
270
403
535
668
801
Fig 6.
696
Czech Republic
2006
[Content]
3,5
Volume [%]
2,5
1,5
0,5
0
5,7
137
270
403
535
668
801
Fig 7.
Example 2
This test was carried out in the same way like previous (example 1). The difference
between these two examples was stirring rate. In this case it was about 400 rpm. The
obtained product is presented in Fig. 8. Tap bulk density 1.0 g/cm3.
4
3,5
Volume [%]
2,5
1,5
0,5
0
5,7
137
270
403
535
668
801
Fig 8.
Example 3
FOX-7 was added to a flask with DMSO. The mixture was heated with mechanical
stirring 850 rpm to 90 C. Then acetone was added to the mixture The product was filtered
and washed. Tap bulk density of FOX-7 was 1.04 g/cm3. The crystals are presented in Fig. 9.
9
8
7
Volume [%]
6
5
4
3
2
1
0
5,7
137
270
403
535
668
801
Fig 9.
5
697
Czech Republic
2006
[Content]
Volume [%]
6
5
4
3
2
1
0
5,7
137
270
403
535
668
801
Analyses were carried out in open test tubes and thermocouple protected by a glass shield
was inserted into the sample. Data were evaluated using MEAVY 2.0.0.4 software of DTA
551 Ex analyser.
D TA
M ultigraf
36
Fox Pr 6 5 ( 5) 2 mg
34
32
30
28
26
Difference [C]
24
22
20
18
16
14
12
10
8
6
4
2
0
1 00
1 20
14 0
16 0
1 80
2 00
Te mp e r atur e [C]
2 20
2 40
260
2 80
Me a v y
698
Czech Republic
2006
[Content]
Yield
[%]
Bulk density
[g/cm3]
FOX-7 (crude)
70 80
0.43
0.50
50.7
0.91
1.02
50.8
0.63
0.74
72.7
0.90
0.99
40.0
0.96
1.10
40.0
1.00
1.06
58.0
1.04
1.11
84.0
0.86
0.94
48.0
0.93
1.04
53.0
0.45
0.51
In order to obtain suitable form of FOX-7 we checked some solvents. The best results
we obtained for mixture of DMSO and water. The parameters of crystallization, obtained
bulk density, and yield are presented in table 1. After these tests, coarse FOX-7 (bulk density
~1 g/cm3) was obtained. The crystals and characteristics of grain size of FOX-7 are shown in
Fig. 3-10. The best yield we obtained for FOX-7 precipitated with mixture DMSO and other
solvents like nitrobenzene.
3.
SUMMARY
699
Czech Republic
2006
[Content]
Acknowledgments
This research was supported by the State Committee of Scientific Research through
Institute of Industrial Organic Chemistry, grant 3 T09B 021 27.
References
[1]
[2]
J.P. AGRAWAL, Recent Trends In High-Energy Materials, Prog. Energy Combust. Sci., 1998.
A. BECUWE, A. DELCLOS, Low-Sensitivity Explosive Compounds for Low Vulnerability
Warheads, Propellants, Explosives, Pyrotechnic, 18, 1-10, 1993.
[3] Y. TANG, The Effect of NQ and TATB to the Insensitivity of HMX, Proc. of 33nd International
Annual Conference of ICT, Karsruhe, Germany 2002
[4] Y.N. MATYUSHIN, G.T. AFANASEV, V.P. LEBEDEV, M.N. MAHOV, V.I. PEPEKIN,
TATB and FOX-7: Thermochemistry, Performance, Detonability, Sensitivity, 34th International
Annual Conference of ICT, Karlsruhe, Germany, 24-27. 06.2003.
[5] N. Latypov, J. Bergman, A. Langlet, U. Wellmar, U. Bemm, Synthesis and reactions of , 1,1diamino-2,2-dinitroetylene, Tetrahedron 54, pp. 11525-11536, 1998.
[6] N. LATYPOV, A. LANGLET, U. WELLMAR, Chemical Compound Suitable for Use as an
Explosive, Intermediate and Method for Preparing the Compound, US Patent 6312538 B1,
Nov. 6, 2001
[7] H. OESTMARK, H. BERGMAN, U. BEMM, P. GOEDE, E. HOLMGREN, M.
JOHANSSON, A LANGLET, N.V. LATYPOV, A. PETTERSSON, M-L PETTERSSON, N.
WINGBORG, C. Vorde, H. Stenmark, L. Karisson, 2,2-Dinitroetylene-1,1-diamine (FOX-7)Properties, Analisis and Scale-up, Proc. 32nd International Annual Conference of ICT,
Karlsruhe, Germany,3-6. 06.2001.
[8] Z. CHYEK, S. CUDZIO, Jan Bdek, Sylwia Pietrzyk, Optymalization of 1,1-diamino-2,2dinitroetylene Synthesis, New Trends in Research of Energetic Materials, Pardubice, Czech
Republic, 207-212, 2005.
[9] U. TEIPEL, I. FUHR, K. HARTLIEB, A. KJELLSTROM, C. ELDSATER, Crystallization of
1,1-diamino-2,2-dinitroethylene (FOX-7), 35th International Annual Conference of ICT,
Karlsruhe, Germany, 2004.
[10] U. TEIPEL, Energetic Materials. Particle Processing and Characterization, Weinheim, 2005,
WILEY-VCH Verlag GmbH & Co. KGaA
[11] H. OESTMARK, H. BERGMAN, A. LANGLET, N. WINGBORG, U. WELLMAR, U.
BEMM, FOX-7 A New Explosives with Low Sensitivity and High Performance, Proc. 11th
Symposium (international) on Detonation, Snowmass Village, 1998, 18-23.
700
Czech Republic
2006
[Content]
Abstract:
The reference direction of the present researches was the search of structures of the
selective catalytic additives labilizing pre-flame processes of HMX transformation. At
choice of frame of the additives outwent from HMX ability to derivate molecular
[1]
complexes with a number of chemical compounds . The investigated salts and nitro-
nitroso-compounds essentially reduce tbid of HMX (on 30-80 0C). The interaction
of the investigated additives with HMX implements also in thermal
decomposition process of fuel compositions.
All researched catalytic additives are modifying agents of burning of fuel
compositions "PU+HMX". The interrelation between activity of the additive in
the thermal decomposition process of HMX and influencing this additive on
burning rate and value of -index of fuel compositions with HMX is observed.
The most effective activators of thermal decomposition of HMX and modifying
agents of burning of HMX based propellants are Cu, K-EDNA, K-NAPN and
4-NA.
Keywords:
1.
INTRODUCTION
Cyclic nitramines are perspective components of rocket propellants. They have high
energy characteristics. At addition of these compounds to propellant the combustion
temperature and single impulse is considerably increased. Typical representative of
nitramines class is HMX, which widely applied in rocket propellants of different types. But
the application of formulas with nitramines is complicated by high dependence of burning
rate of such propellants on pressure that upsets jet engine work stability and reduces
efficiency of rocket.
Is established, that HMX behaves in K-phase (condensate phase) of combustion as the
inert component, it decreases burning surface temperature and development of heat in
reactionary layer of K-phase. Disadvantage of HMX is as major ignition delay time [2]. The
particles of HMX have not time to ignite in K-phase, are carried out in gas phase, where
burn down with major development of heat. The entrance of heat from the flame zone does
not compensate energy consumption conditioned by transformation of HMX in K-phase. In
outcome the burning rate of propellant is reduced and the -index in the burning rate law is
701
Czech Republic
2006
[Content]
2.
EXPERIMENTS
2.1
702
Formula
Melting
point tm,C
Decomposition
point td,C
K2[Cu(CH2NNO2)2]H2O
225
u(H2O)[O2NNC(O)OCH3]2
100
u(NH3)[O2NNC(O)OCH3]2
200
NCCH2CH2NHNO2
49-50
190
NCCH2CH2NNO2K
51-54
180
[NCCH2CH2NNO2]2Cu
160
Czech Republic
2006
[Content]
Formula
NO
1-Nitroso-3,5,7-trinitro-1,3,5,7tetraazacyclooctane (1-NO)
N
O2N N
N NO2
N
NO2
NO
N
O2N N
N NO2
N
NO
1,4,5,9-Tetranitroso-1,4,5,9-tetraazabicyclo
[4,4,0]decaline (4-NA)
NO
NO
NO
NO
In work was used high-dispersive HMX with a specific surface area equal 256 m2/g. The
size of 1-NO and 2-NO particles has compounded 100 to 200 microns, 4-NA - 4 microns.
For manufacturing model fuel compositions fissile was selected binding presenting by itself
polyurethane caoutchouck plasticized by blasting oil (PU). Model fuel contained 60 % mass.
crystalline explosives with a relation HMX: nitrosamines equal 3:1. The content of the
catalytic agent did not exceed 3 % mass. from mass of a sample. As a starting composition
the intermixture containing 60 % of HMX and 40 % of PU was used.
2.2
The thermal decomposition of individual explosives, their intermixtures and model fuel
compositions on their basis was explored on derivatograph Paulik - Paulik - Erdey,
atmosphere - air, heat rate - 10 0C / min., interval of temperatures - 20-6000C.
2.3
The performances of burning of fuel compositions were determined with a chimney type
strand burner (CTSB) which was pressurized with nitrogen at pressure range 2,0 to 10,0
MPa. Not hardened samples placed in fluorine-plastic glasses were investigated. During
operation pressure in the device and intensity of a flame glow were automatically fixed. The
burning time was determined on inflections on the curve "pressure - time". Burning rate
depending on the pressures were approximated in logarithmic coordinates by the equation
U=. The statistical processing of results was conducted under the reference computer
programs. The precision of definition of burning rate were about 3 %, -exponent were 5 %
703
Czech Republic
2006
[Content]
at confidence probability 0,97. The ignition of each sample was conducted by an electrically
heated nichrom wire which was attached on the top of sample.
3.
1
2
3
4
5
6
7
Composition
ts,
HMX
270
HMX+Cu,K-EDNA
240
HMX+Cu-MNU
265
HMX+Cu,NH3-MNU 265
HMX+NAPN
150
HMX+K-NAPN
190
HMX+Cu-NAPN
260
1 stage
tf,
350
280
280
320
210
210
300
, %
99
87
39
99
6
27
93
ts,
300
380
380
260
260
380
2 stage
tf, , %
540
89
430 100
530 100
310
94
280
81
480
95
ts,
340
3 stage
tf, , %
380
90
The note:
ts start temperature of stage;
tf final temperature of stage;
transmutation degree, determined on the equation:
= (ms-mi)/ ms100 %
where: ms start mass of a sample, g.
mi mass of the heels in the end of i-th stage of decomposition, g.
At presence K-NAPN during decomposition of HMX it is possible to bleed two stages
(figure 2):
704
Czech Republic
2006
[Content]
W, %
100
80
exo
endo
60
40
20
0
0
100
200
300
400
500
o
600 t, C
1
8
2
18
5
35
7
38
9
50
15
51
The intermixture of PU with HMX is decomposed in 3 stages which are flowing past in
temperature bands 160-2000, 260-2700 and 360-5000. All stages are exothermic (figure
2).
W, %
100
80
exo
endo
60
40
20
0
0
100
200
300
400
500
600 t,
705
Czech Republic
2006
[Content]
1
2
3
4
5
6
7
8
Composition
PU
PU+HMX
PU+HMX+Cu,-EDNA
PU+HMX+Cu-MNU
PU+HMX+Cu,NH3-MNU
PU+HMX+NAPN
PU+HMX+K-NAPN
PU+HMX+Cu-NAPN
t s,
170
160
150
165
160
140
170
180
1 stage
tf ,
260
230
215
230
190
210
210
210
,
%
80
29
37
32
30
31
32
26
t s,
300
260
230
240
220
250
220
240
2 stage
tf ,
420
320
275
280
260
300
290
290
,
%
92
91
91
90
90
91
99
89
t s,
320
330
390
370
390
360
3 stage
tf ,
,
%
510 100
510 98
510 100
520 96
510 100
510 93
(1)
(2)
(3)
706
(4)
Czech Republic
2006
[Content]
As in explored temperature range the potassium nitrite is inconvertible and is solid, the
reaction (4) is not implemented at decomposition of potassium salt of NAPNA, and the
fractional oxidizing of propionitrilimine implements under the plan:
NCCH2CH=NH+NO2N2+N2O+CO+CO2+H2O.
(5)
In HMX always there is a reactionary centre with a deficit of electrons and on it there
can be a nucleophilic attack by anion, that results in acceleration of HMX decomposition.
In case of K-NAPN and Cu-NAPN salts, explored in our work, nucleophilic particles are
propionitrilimine and its oligomer having a lone electronic pair on amines nitrogen.
On the basis of results obtained by us, it is possible to suspect that the reason of
activation of HMX decomposition in a case K-NAPN is the chemical interaction between
HMX and yields of decay of the additive, that confirms by dependence of HMX lowtemperature transmutation degree on K-NAPN content. Thus, at use K-NAPN, the
considerable importance in activation of HMX have the ionic processes.
The additives from a class of nitro-nitrosamines have exhibited high activity in thermal
decomposition and burning processes of HMX contained propellants.The results of research
have shown, that unlike HMX, for which one temperature of beginning of intensive
decomposition (tbid) is 270 0C, nitrosamines are decomposed at lower temperatures: to 1NHMX - at 230 0C, 2-NHMX 220 0C, 4-NA - at 200 0C. Thus, with increment of amount of
nitroso-groups, tbid of nitrosamine is reduced. The intermixtures of HMX with nitrosamines
are decomposed by one peak in temperature range, typical for decomposition of nitrosamine.
The degree of transmutation thus makes 92 % for intermixtures of HMX with its nitrosoderivatives and 74 % for an intermixture of HMX with 4-NA (Table 6).
Table 6. Parameters of thermal decomposition of nitro-nitrosamines and
propellants on their basis
Composition
HMX
1-NO
HMX +1-NO
PU
PU+ HMX
PU+ HMX +1- NO
2-NO
HMX +2-NO
PU+ HMX +2- NO
4-
HMX +4-
PU+ HMX +4-
t s,
270
180
240
170
160
160
160
220
150
200
190
180
1 stage
tf ,
350
210
320
260
230
200
210
340
200
270
270
205
, %
99
6
92
80
29
33
3
93
31
74
74
63
t s,
230
320
300
260
240
220
340
220
370
510
210
2 stage
tf ,
310
520
420
320
320
330
510
320
560
560
340
, %
92
94
92
91
84
90
95
91
97
89
85
t s,
320
330
320
440
3 stage
tf ,
590
470
535
620
, %
100
96
98
98
707
Czech Republic
2006
[Content]
Hence, the yield of decay of nitrosamines, interacting with HMX, produces its
destruction at lower temperatures. The number of fissile centres generated during
decomposition of nitrosamines, apparently, proportional to amount of nitroso-groups.
Therefore, it is possible to suspect, that the fissile centres are formed during break of N-Nconnections in nitrosamines.
The interaction of nitrosamines with HMX occurs also in thermal decomposition process
of fuel compositions. In fuel compositions the intermixtures of HMX with nitrosamines are
decomposed by one peak in temperature range, typical for decomposition of respective
nitrosamine (Fig.3).
W, %
100
exo
endo
80
60
40
20
0
0
100
200
300
400
500
600
t, oC
Thus, hence the analysis of the obtained experimental data it is possible to make
following deductions:
-the investigated nitrosamines reduce tbid of HMX owing to interaction of yields of their
decay with HMX;
-the value of a decrease tbid of HMX is proportional to amount of nitro-groups in a
molecule of the additive;
-the interaction of yields of decay of nitrosamines with HMX implements both at
decomposition of intermixtures of explosives, and in thermal decomposition process of
model fuel compositions.
708
3.2
Czech Republic
2006
[Content]
The investigated additives practically do not influence quantity of burning rate and index of clean binding. In intermixtures PU with HMX the catalytic activity displays Cu, KEDNA and K- NAPN. At their insertion the burning rate at 4,0 MPa is incremented in 1,2
times, -index is descended from 0,88 (original compositions) to 0,75 in case Cu, K-EDNA
and to 0,68 in case K-NAPN. Results of definition of burning rate and -exponent are listed
in Table 7.
Table 7. Basic performances of burning of fuel compositions with the additives of
nitramines salts.
Composition
PU
PU +HMX
PU +Cu,-EDNA
PU +HMX+Cu,K-EDNA
PU +Cu,NH3-MNU
PU +HMX+Cu,NH3-MNU
PU +Cu-NAPN
PU + HMX +Cu- NAPN
PU +Cu-MNU
PU + HMX +Cu-MNU
PU + NAPN
PU + HMX + NAPN
PU +- NAPN
PU + HMX +- NAPN
, mm/s
MPa
3,45
1,30
3,00
1,95
2,97
1,36
3,06
1,17
3,17
1,07
3,03
1,26
2,82
2,03
0,60
0,88
0,72
0,75
0,62
0,81
0,64
0,93
0,56
0,99
0,64
0,84
0,70
0,68
U, mm/s, at
P, MPa
4
10
7,9
13,7
4,4
10,0
8,2
15,8
5,5
10,9
7,0
12,4
4,2
8,8
7,4
13,4
4,3
10,0
6,8
11,4
4,2
10,4
7,3
13,1
4,1
8,8
7,5
14,3
5,2
9,7
On the basis of high catalytic activity K-NAPN and Cu, K-EDNA at thermal
decomposition of HMX and its intermixtures with PU, it was possible to expect more
essential influencing of these salts on the performances of propellants burning. But the
activity of the additives is reduced by effect of a number of the factors:
-short time of contact of HMX particles and additives;
-presence of layers of binding between particles;
-inadhesion binding to HMX;
Ability of HMX fractionally dissolve in PU results in formation on a surface of its
particles of liquid layer, that sharply reduces an possibility of adhesion interaction between
polymer and HMX.
The intensive gas-making during a burnup of PU results in an offset of HMX particles to
a gas phase. Therefore the residence time of HMX in reaction layer of K-phase and,
accordingly, time of its contact with yields of decomposition of the catalytic additive is
reduced. And so, the problem of activation of HMX transmutation in K-phase of burning
needs to be solved complexly, together with modification combustible - binding, in order to
ensure sufficient adhesion to filling agent.
9
709
Czech Republic
2006
[Content]
4.
A,
mm/s MPa
1,30
1,51
1,74
2,53
1,63
1,69
1,91
2,40
2,37
2,38
2,58
0,88
0,75
0,81
0,50
0,70
0,74
0,71
0,57
0,60
0,49
0,45
U, mm/s at
P, MPa
4
10
4,4
10,0
4,3
8,5
4,5
9,5
5,1
8,1
4,3
8,2
4,7
9,3
5,1
9,8
5,3
9,0
5,5
9,5
4,7
7,3
5,2
7,1
CONCLUSIONS
Czech Republic
2006
[Content]
propellants are Cu, K-EDNA, K-NAPN and 4-NA. The insertion of soot boosts catalytic
effect during burning. Combination in fuel HMX with nitramines and combined additive KNAPN+C allows to decrease -index on 30-50 % as contrasted to by HMX based propellant.
5.
REFERNCES
[1]
[2]
A.A. Zenin, V.M. Puchkov, S.V. Finjakov HMX combustion wave structures at a
different temperature and pressure//FGB Journal, Vol. 34, 2, p. 59-66, 1988.
[3]
[4]
[5]
[6] A.A. Zenin About one model of reactionary bed of a condensed phase of ballistite
gun powder//tesis. Moscow: IHF AN USSR, Vol. 213, p. 1357-1360, 1973.
11
711
Czech Republic
2006
[Content]
Abstract:
Molecular simulations methods were used for calculation and possible prediction of
crystal structures of energetic materials. Calculated results for selected energetic
material (RDX, TNA) were compared with real crystal structures. Mainly these items
were compared: arrangement of individual molecules, space group and energy
minimum for experimental and calculated models. Calculations were done in different
types of empirical force fields, in different modeling environment (Cerius2 modules
crystal builder, off setup, minimizer, crystal packer, Material Studio module builders,
discover, charge modification), under different conditions (1) with and without respect
to conformation of real crystal structure, (2) with the same space group for calculated
and experimental sample or with neglecting of space group symmetry, (3) with different
arrangement of individual molecules in the crystal structure etc. Crystal structures of
calculated models with minimum energy were compared with published experimental
crystal structures. Calculated and experimental crystal structures show significant
similarities with respect to arrangement of individual molecules and also with respect to
energy. This way of prediction of energetic materials show that on the base of general
information the crystal structure can be partially calculated by molecular simulations
method.
Keywords:
energetic materials, RDX, TNA, molecular simulations, classical
molecular dynamics simulations, hydrogen bond
1.
INTRODUCTION
This work is derived from a series of our previous articles describing and analyzing
energetic materials by molecular mechanics and classical molecular dynamics simulations
under different conditions [1]. The most important conclusion from these results for this paper
is that influence of non bond interactions is significant for structure parameters and crystal
packing [2-3]. Calculated non bond parameters include van der Waals, Coulomb and
Hydrogen bond terms. For example, hydrogen bond interactions play important role for
measurement of physical quantity sensitivity [4]. On the base of knowledge of
intermolecular properties the crystal structures of cyclotrimethylene-trinitramine, known as
RDX [5] and crystal structure of 1-amino-2,4,6-trinitrobenzene, known as TNA [6] were
calculated by molecular simulations methods in Cerius2 and Material studio modeling
environments [7]. Resulted minimized models were compared with experimental data. The
design of new and proposed energetic compounds by different computational characteristics
was described by P. Politzer et al [8]. Computational methods like molecular dynamics
712
Czech Republic
2006
[Content]
simulation [9, 10], Monte Carlo methods [11], Ab initio study [12] are widely used to describe
decomposition processes of energetic materials. Attention was also focused for important
role of weak interaction in crystal engineering [13, 14]. Crystal structures of energetic materials
are described by intermolecular potentials which very well characterize the real structure [15,
16]
. Moreover, influence of edge dislocation during hydrostatic compression on the electronic
structure can be also calculated by the ab ititio calculation [17]. These several examples show
increasing influence of computational chemistry for description and characterization of
energetic materials and their properties and bring promises for design of new energetic
materials with desirable properties.
2.
MOLECULAR SIMULATIONS
The experimentally described crystal structures of RDX and TNA were built on the
base of crystallographic data [5, 6]. The total crystal energy of these experimental structures
was calculated in individual empirical force field (COMPASS [18, 19], PCFF [20, 21], CVFF [22],
UFF [23], Tripos [24], Dreiding[25]) to choose the most convenient force field for following
calculation. After the first total energy calculation of RDX and TNA, structures were
minimized and total crystal energy of minimized structures was established. RDX was tested
and calculated in Material Studio modeling environment in discover module. TNA was
calculated in Cerius2 in Crystal Packer module, which allows predicting crystal structure.
After force field selection follow these general steps: (i) the space group of real RDX
and TNA was changed to P1 and minimized. (ii) Individual molecules of RDX and TNA
were randomly arranged in the crystal cell with fixed parameters related to real crystal cell.
A series of model with different initial position of molecules were minimized and on the
base of minimum crystal energy were selected and compared with previously minimized real
crystal structure. Models with minimum energy were compared with minimized real
structures.
After first energy minimization the conformational search was done by classical
molecular dynamics under following conditions: quench type dynamics - the model is after
100 steps of dynamics minimized. These minimized models are saved into quench trajectory
file and dynamics continue from the point before minimization. Dynamics were performed
in NVT ensemble, the number of atoms, volume and temperature are constant. Temperature
was 300 K, dynamics time step was 1fs. The maximum length of calculation in ps was 80 ps.
Bonding tolerances were 1.150 and molecular dynamics were calculated with ideal
periodical crystal structure without any defects. Charges were calculated by Charge
equilibration method (Qeq) [26].
3.
Empirical force fields used in this calculation do not describe the real crystal structure
precisely, so we can see difference of total crystal energy of real sample and minimized
sample, see table 1. The lower difference was calculated for COMPASS force field.
COMPASS force field was used for subsequent calculation of RDX (see Fig. 1) in Material
Studio. In table 2 we can see comparison of force field for experimental and minimized
crystal structure of TNA (see Fig. 2). Dreiding force field was selected for subsequent
calculations although it is only moderately accurate for geometries, conformational energies,
intermolecular binding energies, and crystal packing.
2
713
Czech Republic
2006
Fig 1.
Fig 2.
[Content]
COMPASS
PCFF
CVFF
experimental
-748
-480
-327
minimized
-841
-803
-721
Czech Republic
2006
TNA
Dreiding
Universal
Tripos
experimental
-160
-151
-145
minimized
-167
-159
-186
[Content]
Crystal structure of RDX and TNA was changed to P1 space group. After minimization with
fix space groups parameters we obtain the same value of total crystal energy. Consequently
individual molecules of RDX and TNA were randomly placed in the space group with fixed
parameters and minimized. Results with minimum total energy for the best models are in the
table 3. Initial random arrangement of RDX and TNA molecules after minimization was
more organized than in initial phase. Calculated crystal structure of RDX - test1 is shown in
the Fig. 3 and for TNA - test1 in the Fig. 4.
Table 3. Total crystal energy [kcal/mol] of minimized RDX and TNA crystal
structures for series
test1
test2
test3
RDX
(COMPASS)
-804,65
-804,61
-803,01
TNA
(Dreiding)
-159,3
-140,0
-134,4
Fig 3.
715
Fig 4.
4.
Czech Republic
2006
[Content]
CONCLUSIONS
Molecular mechanics and classical dynamics simulations in this work were used for
showing several possible steps on the way of design new energetic materials on the base of
their general properties derived from non bond interactions. Calculated models can be
compared with real experimental structures in case, that this real crystal structures is
optimized in the same force field. The force field describing one crystal structure precisely is
not usually suitable for different type of energetic material. On the other hand force field
with very general parameters can describe a series of energetic material but with high error
in comparison with real crystal structure. The aim of this work was to find some consensus
among these two point of view for describing energetic materials by force field calculations.
REFERENCES
[1]
[2]
[3]
[4]
[5]
716
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
Czech Republic
2006
[Content]
717
Czech Republic
2006
[Content]
[25] S. L. MAYO, B. D. OLAFSON, W. A. GODDARD III: DREIDING: a generic force field for
molecular simulations, J. Phys. Chem., 94, p. 8897-8909, 1990.
[26] A. K. RAPP AND W. A. GODDARD III: Charge Equilibration for Molecular Dynamics
Simulations, J. Phys. Chem., 95, p. 3358-3363, 1991.
Acknowledgment:
M.P. and P.K. thanks to financial support from the Ministry of Education, Youth and
Sports of the Czech Republic (Project MSM0021620835).
P.V. thanks to financial support from the Ministry of Education, Youth and Sports of the
Czech Republic (Project MSM0021627501).
718
Czech Republic
2006
[Content]
Abstract:
Several substances used in present as deterrents for coating of propellants were
compared in the sense of efficiency of utilisation, thickness and look of surface layer,
heat of explosion, gun chamber maximal pressure and bullet velocity.
Keywords:
1.
INTRODUCTION
At firing is generally required to obtain the highest bullet velocity and the lowest
maximal gun chamber pressure and temperature for the minimal abrasion and moderate
construction demands. The processes in gun barrel may be affected by several ways - one of
them is coating of propellant grain. This depends on the choice of suitable deterrent and on
the conditions and way of treatment (e. g. temperature, solvent system, time and mode of
deterrent adding). The aim of this contribution is to compare coating of propellants by
several substances - both substances used for this purpose for a longer time and substances
potentially applicable in future. For the possibility of simple comparing were the processes
made in the same way.
2.
EXPERIMENTAL
2.1
Sample preparation
The coating process was performed under stirring in water with ratio of uncoated grain
and water 1 : 4.2 at the temperature 80 C with adding of deterrent in six portions at time
intervals of 20 minutes with following 3-hour idle period [1].
The two series of samples were prepared from the two types of uncoated grain with
different grain size. For the possibility of comparing every set of samples includes one
standard sample coated with dibutylphthalate.
The set of samples 10, 11, 12 and 14/05 was prepared from uncoated grain with bigger
grain size, as deterrent were used DBP (dibutylphthalate), iNBP (isononylbenzylphthalate),
BuNENA (N-n-butyl-N-(2-nitroxy-ethyl)nitramine) [2,3] and DNDA57 (mixture of linear
nitramines 5, 6 and 7 - similar to DNDA7) [2,3] in amounts 5 % to uncoated grain weight.
The set of samples 21, 23, 24, 25, 26 and 27/05 and 02/06 was prepared from uncoated
grain with smaller grain size, as deterrent were used DBP, DOP (dioctylphthalate), DNT
(2,4-dinitrotoluene), DNDA7 (3,5-dinitro-3,5-diazaheptane) [2,3], CI (Centralite I), CII
(Centralite II) and DOTP (dioctylterephthalate) in amounts 6.7 % to uncoated grain weight.
Pavel Prchal
719
2.2
Czech Republic
2006
[Content]
Gun firing
All samples were fired in gun 5.56 NATO with bullet weight 4 g - every sample in five
shots. The chamber pressure was detected by piezo-sensor, bullet velocity was measured in
distance 9,5 m from muzzle.
3.
RESULTS
10/05-DBP
surface
layer
thickness
heat of
explosion
[%]
[
m]
[kJ.kg-1]
[MPa]
[m.s-1]
96
42
3750
188.6
723.6
gun
chamber
bullet
velocity
pressure
11/05-iNBP
4330
429.2
900.2
12/05-BuNENA
93
4220
370.5
865.6
14/05-DNDA57
85
4305
413.0
900.0
21/05-DBP
92
54
3928
247.3
769.7
23/05-DOP
4543
434.6
867.9
24/05-DNT
97
209
4368
412.2
849.0
25/05-DNDA7
95
214
4464
408.1
848.5
26/05-CI
97
44
3809
235.8
759.0
27/05-CII
89
199
3741
294.4
761.5
02/06-DOTP
17
4349
432.4
855.2
Fig 1. 10/05-DBP
720
efficiency of
utilisation
Fig 2. 11/05-iNBP
Czech Republic
2006
Fig 3. 12/05-BuNENA
Fig 4. 14/05-DNDA57
Fig 5. 21/05-DBP
Fig 6. 23/05-DOP
Fig 7. 24/05-DNT
Fig 8. 25/05-DNDA7
Fig 9. 26/05-CI
Pavel Prchal
[Content]
721
Czech Republic
2006
[Content]
4.
CONCLUSIONS
Classical deterrents (DBP, DNT, CI a CII) show good efficiency of utilisation between
90 and 100 %, surface layer thickness for CI is similar to standard with DBP, in the case of
DNT and CII was selected time of deterrent coating too long (diffusion practically through
the whole grain).
New compounds (BuNENA, DNDA7 a DNDA57) show good efficiency of utilisation
(about 90 %), in the case of surface layer thickness probably the diffusion through the whole
grain occurred - for BuNENA and DNDA57 is surface layer not visible, for DNDA7 is
diffusion through the whole grain visible.
The other tested phthalates with heavier molecules (iNBP, DOP a DOTP) have no
visible surface layer, analyses of amounts of phthalates in samples showed efficiency of
utilisation nearby zero for iNBP and DOP and the low 17 % efficiency for DOTP, i. e. they
are not suitable for coating in wet diffusion regime.
The values of heat of explosion, chamber pressure and bullet velocity are corresponding
to deterrent amount in sample, to depth of deterrent diffusion into grain and to
energetic/nonenergetic character of molecule of individual deterrents.
REFERENCES
[1]
[2]
[3]
722
Czech Republic
2006
[Content]
Abstract:
This paper describes the methods of analysing nitroglycerin (NG) - ethylene glycol
dinitrate (EGDN) mixture present in wastewaters. In practice, for application reasons
the mixture of NG and EGDN is produced by nitration of glycerine - ethylene glycol
mixture. These substances have excellent explosive properties, they are, however, toxic
for living organisms. Techniques of chemical wastewaters treatment are based on
various chemical or physical-chemical operations. The aim is to clean or effectively preclean these waters.
At present situation, the utilization of chemical reactions seems to be the most suitable
method of treatment of wastewaters from nitrate esters production. Three methods were
used for nitrate esters degradation - reduction of nitrocompounds with fine-grained
iron, continual and discontinual, and reduction with lime milk (calcium hydroxide water
solution). Nitrate esters were separated, their quantity was determined by means of
high-pressure liquid chromatography with ultraviolet detection (HPLC/UV) with
internal standard used.
Keywords:
1.
INTRODUCTION
723
Czech Republic
2006
[Content]
2.
EXPERIMENTAL
2.1
2.2
Czech Republic
2006
[Content]
used, containing 60.5 % methanol and 39.5 % redistilled water, at flow rate of 0.6 ml/min
and absorbance was observed at 230 nm. Internal standard method was used[2].
2.3
2Cr3+ +
7 H2O
2Cr3+ + 6Fe3+
+ 7 H2O
Maximum permitted CODCr value in the wastewaters entering the pre-treatment system
on wastewater treatment plant inlet is 2700 mg/l.
2.3.2. Biochemical Oxygen Demand BOD5 - Dilution and Inoculation Method
of Biochemical Oxygen Demand Determination with Nitrification
Suppression
(BOD5) - biochemical 5 days oxygen demand: mass concentration of dissolved oxygen,
consumed under defined conditions (5 days incubation at dark at 20C) by biochemical
oxidation of organic, possibly also inorganic compounds in water. As organic compounds
represent one of the major water pollutants, BOD5 is one of the key water pollution
indicators.
This method can be applied for all waters with BOD5 value ranging from 3 mg/l to
6000 mg/l. The results obtained are a combination of biochemical and chemical reactions.
This is why they do not represent as unambiguous and precise result as it is with some
simple and well-defined chemical process. Nevertheless it is important parameter, which
the water quality can be concluded from.
Principle of the method: BOD5 is determined as a difference between the concentrations
of dissolved oxygen in a wastewater sample prior and after biochemical oxidation.
Determined loss in dissolved oxygen (i.e. BOD5 value) is proportional to the content of
biochemically degradable compounds in water.
Procedure: The wastewater sample is diluted with various volumes of diluting water
with sufficient concentration of dissolved oxygen under addition of aerobic organisms
inoculum. Nitrification is inhibited by addition of allylthiourea. Incubation proceeds in dark,
in fully filled and closed vial at 20C. The quantity of dissolved oxygen is determined by
means of oxi-electrode prior and after incubation. The mass of oxygen consumed in 1 litre of
water is calculated.
725
Czech Republic
2006
[Content]
3.
NG [g/l]
0.510
0.310
EGDN [g/l]
3.810
3.289
CODCr [mg/l]
2175
1773
BOD5[mg/l]
--27
The experiments were carried out under various reduction times at ambient temperature
(Table 2). Agitation rate did not affect the values measured. The results are summarized in
Figure 1.
concentration [g/l]
4
3
NG
2
EGDN
1
0
0
50
100
150
200
tim e [m in]
Fig 1.
726
NG [g/l]
0.510
0.340
0.220
0.178
0.000
0.000
0.000
EGDN [g/l]
3.810
2.731
2.717
2.458
1.568
1.055
1.090
Czech Republic
2006
[Content]
Utilization of iron filings for nitrate esters degradation confirmed the previous finding,
that removing of NG from a wastewater requires reaction time 1.5 - 2 hours. NG is degraded
completely, while starting EGDN content decreases to 1/3 only within the same time.
3.2
Further, the possibility was tested of reduction of nitrate esters, present in wastewater B
from NG and EGDN production, with iron filings in continual regime in flow-through
column (Table 3). The experiments were carried out at ambient temperature. Again,
the decrease in nitrate esters content achieved depends on reaction time, i.e. on the time of
delay of the water in the column (Figure 2).
concentration [g/l]
4
3
NG
2
EGDN
1
0
0
50
100
150
200
tim e [m in]
Fig 2.
NG [g/l]
0.310
0.262
0.186
0.000
EGDN [g/l]
3.289
3.287
2.569
0.221
The best result so far was achieved in decrease in EGDN content to 7 % of its starting
content, with NG being completely removed.
As to operation, this arrangement of the future production equipment would be less
demanding than discontinual process.
3.3
The possibilities were tested of discontinual reduction of nitrate esters in the sample of
model water using lime milk. NG content in this model water was 0.635 g/l and EGDN
content was 3.551 g/l.
727
Czech Republic
2006
[Content]
concentration [g/l]
4
3
NG
EGDN
1
0
0
100
200
300
tim e [m in]
Fig 3.
NG [g/l]
0.635
0.443
0.313
0.124
0.026
0.002
0.003
0.000
0.000
EGDN [g/l]
3.551
2.980
2.874
2.870
2.686
2.658
2.603
2.607
2.605
Utilization of this method for nitrate esters degradation confirmed the previous finding,
that removing of NG from a wastewater requires reaction time 2 hours. NG is degraded
completely, while starting EGDN content decreases to 2/3 only within the same time and its
degradation rate depends on the presence of NG. After degradation of NG, EGDN content
remains unchanged even after 7 days (10080 minutes). This method is demanding for
preparation of water samples prior to analysis as well.
4.
CONCLUSION
728
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
J. YINON, S. ZITRIN: Modern Methods and Applications in Analysis of Explosives, John Wiley
& Sons, Chichester 1993
J. SKLDAL, V. JEOV, A. EISNER, M. KOUBA, K. VENTURA: Determination of Nitrate Esters
Content in Water Samples, Proc. of the 8th Seminar New Trends in Research of Energetic
Materials, Univ. of Pardubice, April 2005
729
Czech Republic
2006
[Content]
Abstract:
This paper presents the results of testing of burning process of rocket solid propellants
when low pressure closed vessel or rocket motor are used for measurements.
The dependence is measured of burning rate of solid propellant on gas pressure in
combustion chamber.
Keywords:
1.
INTRODUCTION
Stable burning of the solid propellant used is an essential condition of proper function of
a rocket motor. The most important parameter indicating the stability of burning of solid
propellant is the dependence of burning rate on gas pressure in combustion chamber. This
paper presents the results of testing of burning process of rocket solid propellants when low
pressure closed vessel (LPCV) or rocket motor (RM) are used for measurements.
2.
THEORETICAL
The values of gas pressure in combustion chamber, burning temperature and gas flow
rate are important parameters to express the function of solid propellant rocket motor. These
values define the rate of burning, which can be expressed as follows [1,2]
u = u0 .g ( p ). f (TPH ).h( w) ,
(1)
(2)
Determination of burning rate - gas pressure dependence u(p) is usually carried out using
Crawford bomb. It is based on measurement of burning time of solid propellant grain of
defined dimension at defined pressures. Certain disadvantage is the necessity to use
complicated equipment, special modified solid propellant grain and time consumption. For
730
The Use of Low Pressure Closed Vessel and Rocket Motor for Measurements of . . .
Czech Republic
2006
[Content]
these reasons other methods were studied. As another possibility proved to be utilization of
LPCV or RM.
In this paper burning rate - gas pressure dependencies are compared, measured with
i)
ii)
The main advantages of measuring with LPCV are: simple equipment, utilization of
standard solid propellant grains, time savings. The main advantages of measuring with
rocket motor are: complexity of measurements (i.e. possibility of measuring any parameter),
utilization of standard test procedures.
(i) Low Pressure Closed Vessel (LPCV)
Closed vessel (CV) is a standard apparatus for determination of gun propellant burning
rate. LPCV is a low-pressure variant (pmax << 100 MPa) of closed vessel, serving for
observation of rocket solid propellant. Evaluation of burning rate involves processing of
measured pressure-time curves.
Gas pressure in CV as well as in LPCV defined as function of proportionate quantity of
burned propellant 0.1 can be expressed as [4]
p( ) =
rTv
(W ( 1 ) )
(3)
u ( p) =
e0
dp
. ,
k B ( ) p max dt
(4)
where e0 is unit burning thickness of solid propellant grain, pmax is maximal pressure in CV
and coefficient kB() is equal to
k B ( ) =
(W )(W )
[W ( 1 ) ]2
(5)
As in LPCV /W0.03 kg/dm3 (see below) and therefore kB() 1, from Eq. (4) it can
be written
u ( p)
e0
dp
. ,
p max p z dt
(6)
Petr Stojan
731
Czech Republic
2006
[Content]
u th ( pth ) =
e0
,
th t z
(7)
where e0 is unit burning thickness of propellant grain, time tz (start of burning) and time th
(end of burning) are shown in Fig. 1, and average pressure pth in time interval from tz to th is
defined as
pth =
1
th t z
th
p(t )dt .
(8)
tz
tz
Fig 1.
th
Pressure p - time t dependence in combustion chamber of rocket motor definition of time tz and th.
Equation (7) is used for determination of one point of burning rate-gas pressure
dependence (see Eq. (2)).
3.
EXPERIMENTAL
The Use of Low Pressure Closed Vessel and Rocket Motor for Measurements of . . .
Czech Republic
2006
[Content]
p [MPa]
tested in CV. Propellant grain is safely ignited by 5 g black powder and ignition pressure
pz = 1.8 MPa. Pressure p - time t curves for the tested propellants are shown in Fig. 2.
Burning times are approx. 0.25 s and the values of maximal pressures amount to 50 MPa.
Note that these values are very close to values achieved in rocket motor (see below).
50
Propellant B
45
Propellant C
40
Propellant A
35
30
25
20
15
10
5
0
0
Fig 2.
50
100
150
200
250
t [ms]
300
p [MPa]
Gas pressure around the tested propellant grains in combustion chamber of the test
rocket motor are adjusted by critical diameter of output nozzle. Typical pressure p - time t
graphs for different nozzles for propellant B grains are shown in Fig. 3.
7,8 mm
14
7,4 mm
12
7,0 mm
16
6,7 mm
10
8
6
4
2
0
0
Fig 3.
Petr Stojan
100
200
300
400
t [ms]
500
Pressure p - time t curves measured in RM for four various nozzles propellant B grains
733
4.
Czech Republic
2006
[Content]
RESULTS
From the data of experiments in the LPCV (see Fig. 2) burning rate - pressure
dependencies were calculated by means of Eq. (6). Calculated curves for propellants A, B
and C are shown in Fig. 4.
u [m/s]
0,016
0,011
Propellant B
Propellant C
Propellant A
0,006
4
Fig 4.
10
12
14
16
p [MPa]
18
From the pressure p - time t curve measured in RM, burning rate was calculated (Eq. (7))
for corresponding average pressure calculated by (Eq. (8)). Calculated values for propellants
A, B and C are shown in Fig. 5.
u [m/s]
0,016
0,011
Propellant C
Propellant A
Propellant B
0,006
4
Fig 5.
734
10
p12
[MPa]
14
Burning rate u th [m/s] - presure pth [MPa] values calculated from results
of testing of propellants A, B and C in RM.
The Use of Low Pressure Closed Vessel and Rocket Motor for Measurements of . . .
Czech Republic
2006
[Content]
Differences between the values of burning rates achieved in LPCV (see Fig. 4) and
in RM (see Fig. 5) at the same pressure levels are about 10%. There are several possible
reasons for the differences. One of them could be that real pressure course in LPCV is
influenced by thermal losses. Therefore values of maximal pressure and pressure rise rate are
lower. Calculation of burning rate gas pressure dependencies is influenced too. Another
reason could be acceleration of burning due to gas flow (i.e. erosive burning) in RM and
inaccurate definition of start-end burning time points (see Fig. 1).
5.
CONCLUSION
This paper presents the results of testing of burning process of rocket solid propellants
when low pressure closed vessel or rocket motor are used for measurements.
The dependence is measured of burning rate of various solid propellants on gas pressure
in combustion chamber. Differences between the values of burning rates achieved in LPCV
and in RM at the same pressure levels are about 10%.
REFERENCES
[1]
[2]
[3]
[4]
Petr Stojan
735
Czech Republic
2006
[Content]
Ji TOLLER*
* kpt. Ing. Ji TOLLER, Ph.D., University of Defence, K 203, Kounicva 65, 612 00 Brno, Czech
Republic
Abstract:
The article presents the results of experimental tests carried out on anti-spalling layer
made of steel fibre reinforced concrete, intended for protective structures. In particular,
the effects of plastic explosive PlHx 30 on slabs made of plain concrete or steel fibre
reinforced concrete are shown.
Keywords:
1.
INTRODUCTION
736
Czech Republic
2006
[Content]
examination of the dynamic load, the tests of fiber concrete slabs have been carried out with
the PIHx 30 plastic explosive.
The search for new materials used for building of protective constructions is a result of
development of new weapons that have been used against resistive targets. This is why it is
necessary to look for the new materials for new and existing structures as well and adapt
such structures to the new requirements.
2.
Some test slabs were manufactured of conventional concrete and of fiber concrete, too.
To set the guaranteed concrete strength, the test objects for strength tests were being made
simultaneously such as 150 mm edge test cubes and for tensile strength tests, the 150 150
600 mm test squared logs. The test body dimensions were selected with respect to the used
DRAMIX RC-65/50-BN fiber.
The dimension of test slabs was determined with regard to material properties of fiber
concrete and slab manipulation while being tested. The tested slabs dimension was 1000
1000 150 mm and weighted approximately 370 kg.
2.1
The quality of used concrete and fiber concrete was stated prior to dynamic loading by
plastic explosive by impulse ultrasound method. The reason to determine quality of material
and changes to come after the slab has been loaded by plastic explosive lead to selection of
above mentioned method. This very method was also employed to learn about dynamic
modulus of elasticity of concrete and fiber concrete used. Measurement was taken twice
once just before the initial test of loading by blast and for the second time after the intrinsic
blast test. The test was carried out and evaluated according to standard CSN 73 1371.
Fig 1.
2
Ji toller
Czech Republic
2006
[Content]
Considering the results of ultra sound measuring and the result of variation factor (for
concrete slabs v=1.9%, fiber concrete slabs v=1.87%) we can conclude the conventional
concrete slabs as well as the fiber concrete slabs were made of high quality material. In
respect to the variation factor we can say that despite the fact they were manufactured out of
lab they were of laboratory quality (up to v=2.5%).
Having the results taken after the loading of slabs by PIHx 30 plastic explosive blast, we
could assess what way the specific charge had influenced the existing slab. It was stated
there were great differences between the conventional and fiber concrete slabs.
2.2
Assessment of spalling
After preparation of slabs the loading was carried out by plastic explosive blasting.
Firstly, the conventional concrete slabs were loaded. After being assessed, the charges for
fiber slabs were set. Graphic assessment of relation between the spalling volume and the
used explosive is mentioned in Fig.2.
9000
8000
90
7000
6000
5000
concrete
4000
200
70
SFRC
3000
2000
50
1000
0
0
20
40
50
60
120
90100
110 130
150
70
80 100 120 140 160 180 200 220
Explosive PlHx 30 [g]
Fig 2.
Distribution of fibers on the spalling surface is somehow equal. The number of fibers
matches the size of spalling and indicates the distribution of fibers in the test body. When
conventional slabs were loaded by 50g of plastic explosive, the spalling did not occur. Only
some fissures with little fragments appeared on the slab surface. The spalling occurred after
the second test was performed. With increased load the volume and depth of spalling rose.
Looking at the graph of relation between spalling volume and used explosive (Fig.2) you
will find the spalling had risen up to the value of 200g after the destruction of slab took place
two times consecutively.
It also happened that the spalling did not occur in fiber concrete slabs while being loaded
by 50g and next by 70g of explosive. In the first case only micro fissures appeared and in the
738
Czech Republic
2006
[Content]
second case the fissures with little fragments showed up. Further, while being loaded the
spalling volume had increased rather slowly as seen in Fig.2
Fig 3.
3.
Fibre concrete slab being initiated by plastic explosives (in the center you
can see creation of spalling. The picture was taken by high-speed camera)
CONCLUSION
Fiber concrete is one of concrete materials with fiber reinforced structure. It is a concrete
that varies from others by structure reinforced by scattered steel fibers. This fact while
being reinforced by certain minimal amount of fibers is a ground for important change of
properties that have been marked as basic drawbacks of conventional concrete. Such
reinforcement results in more strength and resistance against dynamic load.
The results revealed the fiber concrete is not suitable for anti-spalling layer. When
loaded by blast strength there is considerable tension on inner face that the fiber concrete is
not capable to carry, thus resulting in spalling.
The right way to find new materials used for anti-spalling layer could be the sandwich
structure. It also applies to armored vehicles where basically the same effect at protective
constructions takes place while being hit by high explosive ammunition.
Well, it is necessary to conclude that fiber concrete probably will not replace the
armoring that has been used for reinforcing the protective constructions. However, being
compared with conventional concrete the fiber concrete has displayed up to three times
higher endurance. Such a fact predetermines afore mentioned material to be used at
protective constructions. More resistant structure (main supporting structure) will be created
by incorporation of reinforced concrete and fiber concrete. This way created structure should
Ji toller
739
Czech Republic
2006
[Content]
prove higher endurance against penetration of bombs and projectiles designed to destroy
resistant targets.
740
Fig 4.
Fig 5.
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
Ji toller
741
Czech Republic
2006
[Content]
Abstract:
This paper presents experimental results on detonation transmission through sharp
interface between stoichiometric hydrogen-oxygen mixture and less reactive
stoichiometric hydrogen-air mixture. Experiments were conducted in detonation tube of
square cross-section 110110 mm and 8 m long. Detonation wave was monitored by
pressure transducers and smoked foils inserted into the tube. The effect of initial
pressure on detonation transmission was studied.
Keywords: detonation, transmission, hydrogen, gradient
1.
INTRODUCTION
The gradient transmission problem is that of a detonation moving from one mixture
into another, either through discontinuous change in mixture composition or a continuous
mixture composition variation. A detonation propagating from a more to less sensitive
mixture can continuously adjust and travel as a steady CJ wave corresponding to the local
conditions, fail and then re- initiate, or fail completely.
Paterson[1] performed an analysis of a detonation impinging on a contact surface
between a combustible and inert mixture using a simple model. He took the gas state behind
the incident wave as the CJ conditions of the detonation wave (i.e. infinite kinetic rates).
Paterson also noted that this assumption of a continuous region behind the wave at CJ
conditions is not realistic (neglecting the following Taylor expansion wave), but expected
this to be valid for calculating the initial transmitted shock strength. The analysis was
extended to combustible mixtures on both sides of the contact surface by introducing an ad
hoc one parameter characterizing the degree of reaction in the receptor mixture.
Thomas[2] used Patersons non-reacting receptor analysis and found excellent
agreement with the initial strength of the transmitted shock when compared experiments in
which a detonation impinged on an inert contact surface. The experimental observations
were of an initially decaying transmitted shock which leveled off a short time after the
contact surface interaction. This decay to a plateau value was accurately captured by Thomas
with computations taking the Taylor expansion following the detonation wave into account.
Experiments in which the contact surface was allowed time to diffuse demonstrated the
detonation wave velocity and cellular structure rapidly conforming to the local mixture
conditions.
Bjerketvedt[3], Bull[4], Engebretsen[5], Thomas[2] and Teodorczyk [6,7] have studied the
transmission of detonations across an air gap between combustible mixtures, often
determining critical air gap lengths and time/distance for re- initiation once the transmitted
shock entered the second combustible mixture. Thomas showed that gradual gradients
742
Czech Republic
2006
[Content]
significantly assist the transmission process, whereas sharp gradients tend to cause
immediate decoupling of the shock and reaction zone.
Engebretsen[5], Strehlow[8] and Benoan[9] have conducted experiments of detonation
transmission through combustible mixture gradients, focusing on identifying conditions
under which the detonation could continue to propagate and investigating the effect of
mixture perturbation on the cellular structure.
This paper reports recent experimental study on detonation transmission through
sharp change in mixture composition from stoichiometric hydrogen-oxygen mixture into
stoichiometric hydrogen-air.
2.
EXPERIMENTAL
The detonation tube shown in Figs.1 and 2 was used in the experiments. It consists of
a 1 m long booster as well as 2 m and 4 m long square cross-section channels with internal
dimensions 110110 mm. The booster is filled with the oxy-acetylene stoichiometric
mixture at 0.5 bar, which ignited, by a 1 J electric spark rapidly detonates initiating in turn
detonation in the first 2 m long channel filled with the stoichiometric hydrogen-oxygen
mixture in all experiments.
The second 4 m long channel was separated by plastic membrane and filled with
weaker stoichiometric hydrogen-air mixture. The experiments were conducted for three
values of initial pressure: 0.1, 0.2 and 0.3 bar. The pressure was equal in both channels
before ignition. Detonation velocity in the first cha nnel was calculated from the signals of
two pressure transducers C1 and C2 and the behavior of combustion wave (fast deflagration
or detonation) in the second channel was monitored by four pressure transducers (C3, C4, C5
and C6) and photodiode. In additio n the structure of detonation wave was studied with the
use of smoked metal foils inserted along the bottom wall of second channel.
Fig 1.
743
Fig 2.
3.
Czech Republic
2006
[Content]
RESULTS
Table 1 presents the summary of the experimental results obtained for hydrogen-air
mixture in the second channel. Three experiments for each initial pressure are listed in the
table. From the velocities calculated from pressure signals it is evident that detonation wave
successfully transmits from stronger hydrogen-oxygen mixture to weaker hydrogen-air and
quickly adjusts its velocity to steady CJ value. The larger velocities between transducers C3
and C4 as well as pressure values in Fig. 3 show that the wave is at the beginning slightly
overdriven.
Table 1. Summary of experimental results
Channel A
Chanel B
H2 -O2
H2 -air
Initial
pressure
[bar]
CJ
detonation
velocity
[m/s]
Velocity
between C1
and C2
[m/s]
CJ
detonation
velocity
[m/s]
0.1
2716.40
2870.26
1918.30
1914.24
1879.70
0.1
2716.40
2912.06
1918.30
1965.41
1912.78
0.1
2716.40
2604.17
1918.30
1884.66
1876.88
0.2
2754.00
2597.40
1935.30
2017.76
1935.73
0.2
2754.00
2564.10
1935.30
1987.28
1947.80
No
744
Velocity
Velocity
between C3 between C5
and C4
and C6
[m/s]
[m/s]
Czech Republic
2006
0.2
2754.00
2568.05
1935.30
1933.49
1930.50
0.3
2776.20
2639.92
1944.90
2024.29
1934.98
0.3
2776.20
2653.93
1944.90
2099.08
1960.02
0.3
2776.20
2660.99
1944.90
2055.92
1987,28
[Content]
745
Czech Republic
2006
[Content]
Fig.5. Cell sizes for transmitted detonation in hydrogen-air compared with cell size
reported in literature for steady wave
Figure 4 presents detonation cells recorded on smoked metal foil for transmitted
detonation in hydrogen-air mixture at 0.2 bar. This cells are shown statistically in Fig.5
together with average value (blue line) and steady detonation value taken from literature [].
The fact that all the cells are smaller than for the steady case also indicates that detonation
wave is slightly overdriven at that phase of propagation. Figure 6 presents cell sizes of the
transmitted detonation wave for all three pressures compared with steady values.
The important question for the problem of detonation transmission from one mixture to
another is: does the wave transmit smoothly with continuously growing cells or there is
some transition distance for re- initiation after failure of regular wave cellular structure. The
smoked foils revealed that after transmission cells start to appear after some distance (Table
2).
Table 2. Transition distance for detonation cell appearance on smoked foils
Initial pressure [bar]
0.1
0.2
500
0.3
230
For lowest initial pressure no detonation cells were observed on smoked foils which
suggests that transition to detonation occurred further away from diaphragm. The length of
smoke foil was only 600 mm. This conclusion is also supported by pressure profile recorded
by transducer C3 (Fig.6) which shows first strong shock wave followed by detonation wave.
746
Czech Republic
2006
[Content]
4.
CONCLUSIONS
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
Patterson S.: Contact transmission of detonation, Proc.of the Combustion Institute, vol.4, 1953,
pp.468-471
Thomas G.O., Sutton P., Edwards D.H.: The behavior of detonation waves at concentration
gradients, Combustion and Flame, vol.84, 1991, pp.312-322
Bjerketvedt D., Sonju O.K., Moen I.O.: The influence of experimental condition on the
reinitiation of detonation across an inert region, Dynamics of Explosions, Prog.Astro.Aeron.,
vol.106, 1986, pp.109-130
Bull D.C., Elsworth J.E., McLeod M.A., Hughes D.: Initiation of unconfined gas detonations
in hydrocarbon-air mixtures by a sympathetic mechanism, Prog.Astro.Aeron., vol.75, 1981,
pp.61-72
Engebretsen T., Bjerketvedt D., Sonju O.K.: Propagation of gaseous detonations through
regions of low reactivity, Dynamics Aspects of Detonations. Prog.Astro.Aeron., vol.153, 1993,
pp.324-346
Teodorczyk A., Benoan F.: Interaction of Detonation with Inert gas Zone, Shock Waves 6, no
4, 1996, pp.211-223
Teodorczyk A., Thomas G.O., Ward S.M.: Transmission of a Detonation Across an Air Gap,
Proc. of the 20th Symposium on Shock Waves, 1996, pp.1095-1100
Strehlow R.A., Adamczyk A.A., Stiles R.J.: Transient Studies of Detonation Waves,
Astronautica Acta, 17, 1972, pp.509-527
Benoan F., Teodorczyk A.: Experimental Investigation of Transmission of DetonationThrough
Composition Gradients, Archivum Combustionis, 13, 1993, pp.21-32
747
Czech Republic
2006
[Content]
Abstract:
Two methods of determination of the detonation energy of explosives are discussed in
the paper. The first one is a calorimetric method, in the second one the results of a socalled cylinder test are used. A calorimetric system for measurement of the heat of
detonation of explosive charges is applied in which compressed inert gas is used to stop
secondary reactions in the detonation products. The basis of the second method is the
result of a cylinder test. In this test a copper tube is accelerated by the detonation
products of explosive. The expansion of the tube is recorded. From the cylinder test
results the kinetic energy of the tube and that of detonation products can be calculated.
The kinetic energies for the tested and standard explosives enable us to determine the
detonation energy. The method is verified by the use of chosen cylinder test results.
Keywords:
1.
INTRODUCTION
Czech Republic
2006
[Content]
detonation products were proposed in Refs. [5-6]. In this paper the improved method of
determination of the detonation energy from the cylinder test results is presented.
2.
Figures 1 and 2 show the calorimetric system used for measurement of the heat of
detonation of explosive charges. The spherical steel bomb having an internal volume of
5.6 dm3 was placed in the steel calorimeter vessel. The vessel contained about 26.5 dm3 of
water. The water temperature was measured by a thermometer sensor and recorded by a
memory system of the control unit. The calorimeter was surrounded by a constanttemperature jacket.
The heat capacity (thermal equivalent) of the calorimeter was determined by burning
about 5 g of benzoic acid in oxygen at a pressure of 2 MPa. The thermal equivalent, taken as
the average of five tests, was 154 470 J/oC.
The heat of detonation was measured in the bomb filled with compressed nitrogen,
which replaced Ornellas solid confinement of a charge. The charge of explosive was
suspended in the centre of the bomb. Weighed samples of explosives were pressed in
matrices of an inner diameter of 30 mm. Each charge has a fuse cavity 8 mm in diameter and
15 mm long. The heat of explosion of an electrical fuse was estimated in Ref.[3] and it was
11720 J.
At first step, the influence of the initial pressure of gas filling the calorimetric bomb on
the heat effect of detonation was determined. TNT charges of 20 g mass were detonated in
nitrogen under pressure from 0.1 to 2.2 MPa. Measured heat effects are presented in Fig. 3.
Above the value of initial pressure of 1 MPa the heat changes in a interval 4530 210 J/g (
5%). This value is close to the detonation heat measured in Ref.[1] for heavily confined TNT
charges (average value of 4575 J/g). On that account, the heat of detonation of other
explosives was measured in a bomb filled with nitrogen at a pressure of 2 MPa.
749
Czech Republic
2006
[Content]
TNT of 1.58 g/cm3, RDXph (RDX/(CH2)n 94/6) of 1.64 g/cm3, HMXph (HMX/(CH2)n
96/4) of 1.74 density, FOX-7 of 1.78, and NTO of 1.77 g/cm3 were tested. The results
obtained for different mass of charges are presented in Table 1.
5500
5000
4500
Qd [J/g]
4000
3500
TNT
3000
2500
experiment
regresion
2000
1500
1000
0.0
0.5
1.0
1.5
2.0
2.5
p [MPa]
Figure 3. Dependence of the detonation heat of TNT on initial pressure of nitrogen filling
a calorimetric bomb
Table 1. Detonation heat of high explosives for different charge mass
TNT
RDXph
HMXph
FOX-7
NTO
m [g]
Qd [J/g]
m [g]
Qd [J/g]
m [g]
Qd [J/g]
m [g]
Qd [J/g]
m [g]
Qd [J/g]
15.018
19.980
20.073
20.095
25.073
4351
15.076
5341
15.110
5906
19,912
4837
19,994
3070
4316
20.073
5624
15.024
5912
19,894
4930
19,908
3227
4501
25.097
5557
19.989
5802
20,037
4814
19,447
3298
24.983
5839
4515
4512
(Qd)aver = 4439
(Qd)aver = 5508
(Qd)aver = 5865
(Qd)aver = 4860
(Qd)aver = 3199
750
3.
Czech Republic
2006
[Content]
In the cylinder expansion test, a metal tube is driven by the detonation products of a
cylindrical charge of explosive. The expansion of the tube is monitored by precise methods
of measurements of the quickly changing processes, such as streak photography, laser
interferometry, X-ray photography. The cylinder test is mainly used in the methods in which
parameters of the Jones, Wilkins, Lee (JWL) equation of state of the detonation products of
explosives are determined. However, the cylinder test can be also applied for determining
some energetic characteristics of explosives, including the acceleration ability and
detonation energy of the explosive [6].
The result of the cylinder test is the experimental dependence of the radius of external
surface of the tube on the time or the photograph of the expanding tube. From the latter, the
dependence of the position of external surface on the axis co-ordinate can be obtained. On
assuming that the motion of the detonation products and the tube material is stationary, the
time-dependence of the external radius of tube can be calculated. At the end, the velocity of
the tube is determined [6].
During the process of tube acceleration, the work done by the detonation products is
converted to the kinetic energy of the tube, detonation products and air surrounding the tube,
and to the internal energy of tube material and air. If the kinetic energy of air and the heat
losses in the tube material and air are neglected, then the law of energy conservation can be
written in the form
e0 = ekL ( v ) + ekDP ( v ) + pi ( v ) dv
(2.1)
where e0 denotes the detonation energy, ekL and ekDP are the kinetic energy of the tube and
detonation products, respectively. The quantity pi(v) is the pressure on the isentrope starting
from the CJ point, and v is the specific volume of expanding detonation products. The
integral in Eq. (2.1) corresponds to the internal energy remained in the detonation products.
All quantities in Eq. (2.1) are related to unit mass of the explosive.
The sum of kinetic energies of the copper tube and the detonation products per unit mass
of the explosive is called the Gurney energy, and for cylindrical tube it can be expressed by
the following relation
1 u2
EG ekL + ekDP = + L
2 2
(2.2)
where is the ratio of the tube and explosive masses per unit area.
If we assume that the integral in (2.1), i.e., the internal energy of the detonation
products, is proportional to the Gurney energy in the identical way for all explosives, the
following relations are true for the given and standard explosives
e0 = A ( ekL + ekDP )
(2.3)
st
st
e0st = A ( ekL
+ ekDP
)
(2.4)
751
Czech Republic
2006
[Content]
Taking into account the dependence (2.2), the following equation can be obtained from
(2.3) and (2.4)
1
+
e0
2
=
e0st st + 1
2
uL
st
uL
(2.5)
where uL and uLst are velocities of the tubes driven by the detonation products of the given
and standard explosives, respectively.
If the detonation energy is related to unit volume of explosive, then Eq. (2.5) has the
form
1
+ 0
e0
2
=
e0st st 1 st
+ 0
2
uL
st
uL
(2.6)
where 0 and 0st denote the density of the given and standard explosives, respectively.
From Eqs. (2.5) or (2.6) the detonation energy of the given explosive can be estimated if
the detonation energy of standard explosive and the results of cylinder test for both
explosives are known.
To verify the method of determining the detonation energy, the experimental results of
the cylinder test given in Ref.[7] were used. The velocity of the tube was calculated from the
function approximating the time-dependence of radius of the external surface of the tube.
Dependence of the square of velocity on the reciprocal volume of the detonation products is
presented in Fig. 4. The values of uL2 for v0/v = 0.1 were taken into account. Pentryt (PETN)
was chosen as a standard explosive. It was assumed that the detonation energy of PETN was
equal to the energy E0 from the JWL equation of state[7]. The detonation energies of chosen
explosives from Ref.[7] calculated by using relation (2.6) are presented in Table 2.
From Table 2 it follows that, for high explosives, differences between the detonation
energies calculated by the method proposed and those given in literature do not exceed 12 %.
The highest one is for nitromethane. The detonation energy of TNT estimated from the
cylinder test data is 3865 J/g and that given in Ref.[7] is 4295 kJ/kg. Also the values of the
calorimetric heat of detonation of TNT given in literature differ significantly from each other
and, in reality, they depend on the measurements conditions.
The method proposed cannot be applied for nonideal explosives. From Table 2 it follows
that the discrepancy Rw exceeds considerably 10 % for these explosives (last four explosives
in Table 2). Such a high disagreement may be caused by the fact that heat releasing in
nonideal explosives takes place also behind the detonation zone. Probably, this energy is
ineffectively transferred to the tube during later stage of the acceleration.
752
Czech Republic
2006
[Content]
3.7
3.6
3.5
LX-09
uL2 [m2/s2]
3.4
LX-10
3.3
3.2
3.1
PETN
LX-07
3.0
LX-14
LX-04
2.9
2.8
PBX-9011
2.7
2.6
0.00
0.05
0.10
0.15
0.20
0.25
v0/v
Figure 4. Dependence of the square of velocity of the copper tube on reciprocal volume
of the detonation products
Table 2. Comparison of the detonation energy calculated from the cylinder test data,
E0cal, with the detonation energy given in literature, E0 [7],
Rw = [(E0cal-E0)/E0] 100 %
Explosive
PETN
Comp. B
Cyclotol
E0
[GPa]
10.1
8.5
E0cal
Rw
[GPa]
-
[%]
-
8.59
+1.1
9.2
9.28
+0.8
HMX
10.5
11.26
+7.2
LX-04
9.5
10.37
+1.7
LX-07
9.6
9.98
+4.0
LX-09
10.5
10.70
+2.8
LX-10
10.4
10.57
+1.6
LX-14
10.2
10.37
+1.7
PBX-9011
8.9
9.27
+4.1
PBX-9004
10.2
10.41
+2.1
TNT
7.0
6.30
-9.9
Nitromethane
5.1
4.50
-11.7
Aquanal
5.5
2.51
-54.4
Pourvex
4.5
3.82
-15.1
ANFO
2.9
2.21
-23.6
Unigel
5.1
3.91
-23.4
753
Czech Republic
2006
[Content]
To record the copper tube driven by the detonation products of explosives tested in
calorimetric experiments, Scandiflash X-ray set was used. The cooper tube filled with the
explosive tested was 300 mm in length. Its internal diameter and wall thickness were 25 mm
and 2.5 mm, respectively.
An exemplary radiograph of the copper tube driven by detonation products is shown in
Fig. 5. From the photograph, dependence of the external surface radius of the tube on the
axial co-ordinate was determined using graphical computer programs. The velocity of the
tube was determined by using the method described in Ref. [6].
Figure 5.
As a standard explosive, phegmatised RDX was chosen. Its heat of detonation (Qdet =
5508 J/g) was determined in a calorimetric bomb filled with compressed nitrogen. The heat
of detonation of RDXph was assumed to be close to the detonation energy (e0) Ref.[6] .
Results of the calculation of the detonation energy of TNT, HMXph, FOX-7 and NTO are
4190, 5980, 5060 and 3989, respectively. NTO, differences between the detonation energies
calculated from the cylinder test results and the calorimetric heats of detonation for these
explosives do not exceed 6 %. But NTO is extremely insensitive high explosives and the
detonation of charges of about 20 g mass in the calorimetric bomb is probably incomplete.
4.
SUMMARY
The calorimetric set-up for measurement of the heat of detonation of explosive was
tested in which inert gas filled a steel bomb. It was proved that compressed nitrogen stopped
secondary reactions in the detonation products and it behaved like a solid envelope confining
the explosive charge. The improved method of determination of the detonation energy from
the cylinder test results was also presented. The method was verified and used for
determination of the detonation energy of explosives tested in the calorimetric system. Quite
good agreement between the results of both methods was observed.
754
Czech Republic
2006
[Content]
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
755
Czech Republic
2006
[Content]
Abstract
1.
INTRODUCTION
The flegmatised powders are the powders, usually of simple form, treated with
substances which decrease burning rate. As a result, in case of this powders the ballistic
characteristics - the powder force f , the powder gases covolume , the burning rate
characteristic u1 - vary stratum by stratum due to the penetration depth of flegmatising
substance in the thickness of powder grain, fact which determines that the forming process
of gases to be distinct in every stratum. Consequently, the fundamental problem solving of
interior ballistics for this powders which is based on burning geometrical law, proposed by
French ballistic researcher Vieille, can to lead at significant errors in the value evaluation of
main ballistic magnitudes: the powder gases pressure p ; the projectile velocity v ; the
displacement of projectile inside of the barrel l . In order to avoid this errors in the value
evaluation of main ballistic magnitudes in case of using of flegmatised powders it is
modified the analytical relations for this magnitudes, which was obtained in l med method,
taking into account known simplified hypotheses [1, 2, 4]. In this case, besides these simplified
hypotheses, are accepted following hypotheses:
a. All powder elements have same nonhomogenity of physical-chemical composition of
strata. Every stratum has its individual characteristics: thickness 2e1.i ; powder force fi ;
burning rate characteristic u1.i ; gases covolume i ;
b. Every stratum has its personal composition of burning produces, which does not mix
during of the powder burning or adiabatic detente;
756
Czech Republic
2006
[Content]
c. First period of firing phenomenon is divided on phases, in every phase burns a stratum
which has its individual characteristics; in the preliminary period and in the first phase of I
period the first stratum burns.
The respective relations are established for the powders with simple form, which was
flegmatised.
The notations from this paper are those notations that are used usually in interior
ballistics [1, 2, 4, 5, 6, 7].
2.
The used relations in the preliminary period are usually relations from specialty
literature [1, 2, 4] .
The relation gives the fraction of burned powder at the end of preliminary period
0 =
1 1
f1
p0 pa
+ 1 1
(1)
The relative surface of burned powder is calculated with the aim of relation
0 = 1 + 4 0 .
(2)
z0 = (
(3)
t 0 = 2,303
I k .1(11 )
f1
lg p 0 .
(4)
The fictitious length of free volume of loading chamber at the end of preliminary period
is calculated with the aim of relation
l 0 = l 0 [1 (1 1 ) 0 ] .
3.
(5)
3.1
THE I PHASE
The relations for the calculation of main ballistic magnitudes in the I phase of I period
are those known relations from specialty literature [1, 2, 3, 4].
The projectile velocity is calculated with the aim of relation
v=
2
sI k .1
q
x,
(6)
757
Czech Republic
2006
[Content]
in which:
e
1.1
e1.1
e1
; z 0 = e0 ;
1
e
= 0 + k1 x + x 2 ,
(7)
where
k1 = (1 + 2z 0 ) .
The projectile displacement inside of barrel in this phase it is expressed using the
relation
l = l med ( Z x B1 1) ,
B
(8)
in which:
B=
; B1 =
B
2
; Z x = 1
) (1 )
b +1
2b
= k1 x ; 1 = 1+2b ; 2 = 12b ; b = 1 + 4 ; =
B
l0 =
W0
s
; =
W0
b 1
2b
B1
k12
0 ;
Relation gives the pressure of powder gases inside of barrel in this phase
p=
(9)
in which
l = l 0 [1 (1 1 ) ] .
The position of maximum pressure and its value are calculated through successive
approximations using the relation
xm =
k1
.
B( + 1)
2
pmax
1
1 + (1 )
f1
(10)
If xm [0, x1.k ] , then the maximum pressure is achieved in this phase, in contrary case
it will achieve in following phases of I period.
758
Czech Republic
2006
[Content]
THE i PHASE
3.2
At the end of phase I can be calculated the values v1.k , p1.k , l 1.k and 1.k for the
independently variable x = x1.k which represents the border between the first phase and the
second phase. The borders between the following phases are noted with
x2.k , x3.k ,, x i.k ,, xn.k , which will represent the values x at the end of every phase.
If these values are expressed with the aid of relative thickness of strata, then it is
obtained
x i .k =
(11)
v = sq
pdt = sq ( I I 0 ) ,
where:
I=
I0 =
t0
In the case of simple flegmatised powders, taking into account the supplementary
hypotheses, the I impulse can be expressed in following way:
I = I k .1
e1.1
e1
+ I k .2
e1.2
e1
+ ... + I k .i
ei
.
e1
I 0 = I k .1 z 0 .
Taking into account both relations of the impulses I and I 0 , the projectile velocity in
i phase is written
vi = sq I med .i + sq I k .i ( xi xi 1.k ) ,
(12)
where
759
Czech Republic
sp(l + l i ) = f med . i f i + f i i
2006
qvi2
2
[Content]
(13)
where
f med . i = f1 1. k + f 2 ( 2. k 1. k ) + ... + f i (1 i 1. k ) ;
1.k , 2.k ,..., i1.k - the fraction of burned powder at the end of first phase, second
phase,, (i 1) th phase.
The differential equation of projectile displacement inside of barrel has the following
aspect:
dli
l +l i
qvi d vi
f med . i f i + f i i
qv 2 .
i
2
Taking into account the relations of vi and i . versus xi , the differential equation of
projectile displacement inside of barrel becomes
dl i
l + li
( A2.i + Bi xi )dxi
,
Ai + A1.i x i B1.i xi2
(14)
where:
Bi =
s 2 I k .2i
;
fi q
B1.i =
Bi
2
2 2
I med .i
fi q ;
F1.i =
f med .i
fi
2;
1;
I k .i I med .i
fi q ;
Taking into account that l = l med .i , by integration of the relation (14) for l i , since
Bi
(15)
where:
Zx = (x
xi xi .1
i 1.k xi .1
760
(x
xi xi .2
i 1.k xi .2
2
Bi
bi A1.i
i i 1.k
];
2
xi.1 =
A1.i (1+bi )
;
2 B1.i
xi.2 =
A1.i (1bi )
;
2 B1.i
Czech Republic
bi = 1 +
4 B1.i Ai
A12.i
2006
[Content]
The gases pressure results from the fundamental equation of interior ballistics, which is
written in the form
(16)
in which
xi.m =
A1.i A 2.i +
fi
pi. max ( i 1 )
]
Bi ( + 1) 211[1 +
fi
which is solved by successive approximation, taking for pi. max in first approximation a
close value of expected maximum pressure.
If xi.m [ xi 1,k , xi.k ] , then the maximum pressure is achieved in this phase, in
contrary case it will achieve in following phases of I period.
4.
The ballistic magnitudes p and v , versus l , in second period of firing phenomenon. are
calculated with the aid of the relations that are presented in [1, 2, 4].
The values of main ballistic magnitudes at the beginning of this period will be the same
values at the end of first period, namely: = k = 1; v = vk ; p = p k ; l = l k ;
l = l 1 = l 0 (1 ) ; f = f med .n .
761
5.
Czech Republic
2006
[Content]
Can be establish the time of projectile motion inside of barrel following integration of
relation
dt = 1 dl ,
v
(18)
l that is not zero, but is very small and that corresponds at the time t .
If it is admitted the hypothesis that on the space l the projectile velocity varies linear,
then the time t is calculated with the aim of relation
t =
l = 2l .
vmed
v
(19)
n
l
l
t = l 1 dl = ( 1 + 1 ) i + 1 i .
l v
v vi + 1
2
i =1 i
(20)
In this way it is obtained the dependence t = t (l) . On the bases of this dependence and
the established dependencies v = v(l) and p = p(l) , can be draw the curves of variation of
gases pressure and projectile velocity versus time.
6.
In the Fig.1 and Fig.2 it is presented the variation of gases pressure versus displacement
and time, obtained with the aid of the interior ballistics software that was elaborated on the
bases of proposed new algorithm, as well as usually algorithm, for an extant gun.
From the analysis of the data that are presented in the diagrams from Fig.1 and Fig.2,
can be concluded that there are some diferences of the fundamental problem solving of
interior ballistics in the case of flegmatised powders with the aid of both algorithms.
From the comparing of the experimental data with the theoretical values obtained in the
case of the proposed new algorithm had resulted that the difference between the muzzle
velocities values is approximately 0,3% and between the maximum pressures values is
approximately 5%.
762
Czech Republic
2006
[Content]
250000
200000
150000
Serie1
100000
Serie2
50000
0
0
10
20
30
Serie1
100000
Serie2
50000
0
0,000
0,005
0,010
REFERENCES
[1]
[2]
[3]
I.D. SATNIK: Modifitirovannoi method reseniia osnovoi zadaci vnutrennei ballistiki dlia
zariadov iz porohov s peremennami termodinamiceskimi parametrami, Sbornik trudov
4-oi mejdunarodnoi conferentii, Kiev, 2000, pg. 206-221.
[4]
T. VASILE: Internal ballistics of guns, vol.1, II, Military Academy Publishing House,
Bucharest, 1993, 1996.
T. VASILE: Contributions to the fundamental problem solving of interior ballistics,
PROCEEDINGS of the II-nd International Conference Artillery Barrel Systems, Ammunition,
Means of Artillery Reconnaissance and Fire Control, October, 27-29, 1998, Kiev, pg.440-445.
T. VASILE: Studies and researches regarding the fundamental problem solving of interior
ballistics for weapon with grooved barrel, PROCEEDINGS of III-rd International Armament
Conference on Scientific Aspect of Armament Technology, October, 11-13, 2000, Waplewo,
pg.157-165.
T. VASILE, C. BARBU and D. SAFTA: The Fundamental Problem Solving of Interior
Ballistics Using the Phisical Law of Powder Burning, Academic and Applied Research in
Military Science Journal, Vol. 3, ISSN 1588-8789, 2004, Budapest, pg. 407-414.
[5]
[6]
[7]
763
Czech Republic
2006
[Content]
INTRODUCTION
The paper presents the numerical calculations of the physical chemical effects
observed during the gas generators work. Gas generator system works to provide the
constant gas emission at the accurate operation parameters that allows the proper operation
within the power transmission systems work. Until now, the applied calculation methods
have not inquired into the structure of local pressure, temperature and density changes of
gases. Thus, the numerical calculations the authors have undertaken to make are to enhance
the system.
RESULTS
The following simplifications in the calculations have been applied:
-
Flat frontage of burning has been accepted. The existence of the expanded fuel
form in the initial step of burning has not been taken into consideration.
The initial number of molecules has been 3.5 million. However, the given
number has risen up to 10.2 million after the ending of the calculations.
Molecules have been emitted randomly on the surface of burning while the
speeds have been consistent with Maxwell distribution (that is precisely beyond
the speeds backward to the propellant).
Even though the making of calculations lasted a long time (14 days non-stop), slight
percentage of the whole process has been made. The real time of gas generators work has
been from 12 to 14 seconds, and the presented results have described the first 5 s. Despite
the applied simplifications, the character of the effects and local changes of gradients have
been consistent with the physical nature of the effect. The authors present the initial results
and perform further calculations.
764
200.00
Czech Republic
2006
[Content]
200.00
150.00
150.00
100.00
100.00
50.00
50.00
Fig. 1a and 1b present the changes of pressure in the gas generators burning chamber.
The greater amount of generated molecules and the increase of local pressure have become
visible after the 40 steps of calculations.
200.00
200.00
150.00
150.00
100.00
100.00
50.00
50.00
765
Czech Republic
2006
[Content]
Fig. 2a and 2b present the changes of gas densities that flow in the gas generator. The
applied units have been the relative values. At the initial stage of burning, the density has
been comparable to that on the front of burning. The density has been getting its maximum
values in the contraction section (about 4.3 to 6 of initial value), which is consistent with
physical predictions.
temp. otoczenia)
12.00
200.00
Temperatura (krotno
150.00
100.00
8.00
4.00
50.00
0.00
0.00
50.00
100.00
150.00
200.00
250.00
Przekrj komory
3.00
Zmiana g sto ci
2.00
1.00
0.00
0.00
50.00
100.00
150.00
200.00
250.00
766
Czech Republic
2006
[Content]
CONCLUSIONS
1. The losses of energy connected with the flows are insignificant at such selection
of engine dimensions, and their occurrences are related with walls cooling.
2. Too intensive walls cooling answers for the decrease in temperature.
3. There is the subsonic flow in the chamber area whereas the outflow from the
contraction that is at the end of the sleeve reveals the supersonic character.
4. For the sake of the linear pressure maintenance in a time function (in the
chamber and the contraction), the management of the whole engine will depend
on the speed of gas emission and not on the speed of flow, which means that the
form of a jet has a decisive role in the process.
767
768
100
unit
(16)
25
25
12.5
(20)
9 mm
125 unit
mm)
28.5
0.8 mm
mmm
2.5 unit
Czech Republic
2006
80.5
Burning front
Czech Republic
2006
[Content]
Abstract:
Using Ab Initio Molecular Dynamics (AIMD) method and Density Functional Theory
(DFT), the mechanisms of thermal decompositions of five kinds of hydronitrogen
compounds were studied, which are N2H2, N2H4, N3H, N4H2, N4H4. For the ab initio
molecular dynamics studies on the decomposition of the five hydronitrogen compounds,
dissociation trajectories are simulated at different temperatures using the VASP
package. Based on these results, further calculations of minimum points and transition
state structures are performed using Gaussian 03 at B3LYP/6-31G(d,p) level, whereas
CCSD(T)=full/6-311+G(3df,2p) are used for the high accuracy single point calculation.
For N2H4, the main decomposition pathway is N-H bond rupture which is further
verified by rate constant calculation. For N2H4, N3H, N4H2 and N4H4, they decompose
mainly through different N-N bond dissociation. According to the result by B3LYP, the
thermal stability increases by N4H2 < N4H4 < N2H2 < N2H4 < N3H. Nevertheless, at
CCSD(T) level, the order is N4H2 < N4H4 < N3H < N2H2 < N2H4. Both the two
theoretical levels show that the energy barrier of N4H2 dissociation is too low for using
as structure unit of energetic materials. Our work provided some useful information for
further studying the mechanisms of thermal decompositions of various complex high
nitrogen energetic materials theoretically and experimentally..
Keywords:
1.
INTRODUCTION
In the past 20 years, the compound containing N atoms has been extensively studied as
possible candidate for high energy density materials (HEDMs), due to the dissociation of the
compound containing N atoms with the release of large amount of energy through the
slightest input of external energy, for example pressure, impact, or heat, which can be used
as propellants and explosives[1-12]. However, as a kind of good candidate for HEDMs, it is
necessary to learn the thermal decomposition mechanism in detail. So, thermolysis behavior
of the compound containing N becomes the focus of recent experimental and theoretical
research [13]. There have been many theoretical studies at various computational levels on the
compound containing nitrogen atoms [14-16]. In all kinds of the compounds containing N
atoms, the hydronitrogen compound received considerable attentions [17,18]. Industrial interest
in organic compounds containing N atoms began with the use of azides for the synthesis of
heterocycles such as triazoles and tetrazoles as well as with them as blowing agents and as
1. E-mail:syjfree@sina.com
1
769
Czech Republic
2006
[Content]
functional groups in pharmaceuticals [19]. In this work, some simply hydronitrogen molecules
are selected as the model molecules (see Fig.1). Their thermal decomposition behavior are
studied theoretically. Ab inito molecular dynamics (AIMD) was used to discuss the
decomposition process.The doublet potential energy surfaces had been obtained and the rate
constants were calculated to confirm the mechanism of thermal decomposition. Results of
pyrolysis are compared for the small model compounds which contain N-N, N=N, and azide
group.
N2H2
N2H4
Fig 1.
2.
N3H
N4H2
N4H4
COMPUTATIONAL DETAILS
The principles of ab initio MD (AIMD) method have been described in [20-24]. The
Vienna Ab Initio Simulation Package (VASP) program was used for the thermal
decomposition trajectory studies since it uses a rather traditional self-consistency scheme to
evaluate the instantaneous electronic groundstate at each MD step. The molecule being
simulated was put in a cubic box with a length of 8 to imitate gas phase condition. The
PerdewWang [25] gradient correction was added to the exchange-correlation functional [26].
For the core region, the optimized Vanderbilt ultrasoft pseudopotentials [27] supplied with the
VASP package [28, 29] were used for N and H atoms. For the trajectory study, a preliminary
equilibrating simulation was first performed with micro canonical ensemble, starting with
the ground-state equilibrium geometry. The time-step used was 0.5 fs, and the length of the
simulation was 5 ps. One hundred geometrical configurations were selected randomly from
the equilibrium results as the starting structures for the trajectory study. The resulted system
was then used as the initial state for the simulations run at higher temprature with canonical
ensemble. Each trajectory lasted 3 ps with the time-step of 0.5 fs (6000 steps). A trajectory
was terminated after 6000 steps if no reaction was observed [30].
The local energy minimum structures predicted in the trajectory study and the
corresponding transitional state structures have been fully optimized by the analytic gradient
techniques using the density functional theory with Becke s three parameter (B3) [31]
exchange functional along with the LeeYangParr (LYP) nonlocal-correlation functional
[32, 33]
(B3LYP). The standard basis set 6-31G(d,p) was used in conjunction with the DFT
method. Vibrational frequencies were also calculated at the B3LYP/6-31G(d,p) level of
theory to take account of the zerro point energy and identify the transition structures and the
local minimum structures. The coupled cluster method CCSD(T)=full/6-311+G(3df,2p))was
used for the single-point energy calculations in order to ensure the reliable energy
predictions. The GAUSSIAN 03 program package [34] was used in the calculations. For the
decomposition pathways with similar energetic barriers, the rate constants were investigated
with Vklab program, where the required energy and potential surface were abtained at
B3LYP/6-31G(d,p) level.For direct dissociation, the VT ensemble was adopted, and for
reaction through TS, the translation state theory (TST) was used.
3.
770
Czech Republic
2006
[Content]
The selected model molecules are N2H2, N2H4, N3H, N4H2, N4H4 (Fig 1). The details of
thermal decomposition pathways of each compound are discussed below.
3.1
N 2H 2
The preliminary equilibrating simulation was performed at 6000K, and the trajectory
study was performed at 6500K. For each 100 trajectories, 71 decomposed at 6500K. Out of
the 71 reactions, 69 followed channel 1: N-H cleavage; 2 followed channel 2: hydrogen
migration. The two different reaction channels observed in the MD trajectories are listed in
Table 1.
Table 1. The decomposition paths of N2H2 observed in the 100 trajectories at
6500K
Path
1
Reaction Observed
N
3
Total
N
N
H H
Freqs
69
2
N
No reaction
100
29
The geometrical parameters of the various transition states and local minima are
depicted in Fig.2 and their energy properties are summarized in Table 2.
1.040
1.021
1.246
1.016
N1.040
N2 H2
1.436
1.016
N 1.021
N 1.385
N 1.385 H N1.217 N 1.050
1.013 N 1.249
N 1.013
H
1.05
1.020
H
H
M1
1.018
1.018 H
1.045
N
H
M4
NH3
M5
1.033
1.030
N 1.453 N 1.044
H
0.743
1.106
1.044
H H
N2
M3
M2
1.184 1.184
N4H2
1.021
H 1.033
N4H4
1.018
N3H
N2H4
1.020
1.060 N 1.180
1.182
H2
N 1.422 H
N 1.014
1.020
M6
M7
H
N1.052 H
2.113
1.017
N1.017H
1.017
M8
771
H
1.111
N
1.122
1.550
H
1.062
H
TS1
1.341
1.280
Czech Republic
1.457 H 1.079 H
N 1.012
N
1.013
1.033 1.655
TS3
Fig 2.
1.154 N 1.606
H
N 1.022
1.025
TS4
TS5
[Content]
TS2
2006
3.105
1.0490 H
1.017
N1.017
H
1.017
TS6
Bond lengths (in ) of the minimum structures the transition states
Table 2. The energy and the zero-point correction of the minimum structures
and the transition states (in hartree)
N2H2
N2H4
N3H
N4H2
N4H4
M1
M2
M3
N2
NH3
M4
M5
M6
M7
M8
H
H2
TS1
TS2
TS3
TS4
TS5
TS6
EZP
B3LYP/6-31G(d,p) CCSD(T)=full/6-311+G(3df,2p)
0.028265
-110.643011
-110.522547
0.053304
-111.868658
-111.742141
0.021417
-164.785840
-164.609540
0.037854
-220.081037
-219.847144
0.063210
-221.335428
-221.089118
0.026569
-110.607411
-110.482821
-55.878981
-55.814154
0.018971
0.051281
-111.790888
-111.670299
0.005599
-109.524129
-109.419147
0.034441
-56.557769
-56.494011
0.007485
-55.137670
-55.090147
0.009601
-164.138543
-163.959949
0.013261
-110.032726
-109.910763
0.032552
-165.370432
-165.185219
0.044807
-111.705640
-111.591533
0.00000
-0.500273
-0.499810
-1.178539
-1.170823
0.010174
0.006593
-110.013307
-109.895261
0.020697
-110.522739
-110.400769
-111.761264
-111.634122
0.046768
0.030419
-165.368462
-165.184272
-111.747464
-111.609185
0.041514
0.044419
-111.705531
-111.591074
In Fig.3, the calculated energy profiles at B3LYP/6-31G(d,p) and CCSD(T)=full/6311+G(3df,2p) are presented. In channel 1, the energy barriers for the first N-H bond
breaking step are 249.45 kJ/mol and 254.59 kJ/mol respectively. The products, consisting of
N2 and H2, lie significantly below N2H2. The hydrogen migration pathway is shown in
channel 2. A relatively large energy amount of 295.90 KJ/mol and 299.86 kJ/mol is required
772
Czech Republic
2006
[Content]
to transfer H from one nitrogen to another. Accordingly, the direct N-H bond cleavage is the
main decomposition pathway, which is consistent with the result predicted by the ab initio
MD simulation.
Channel 1: The cleavage of the N-H bond
N
N
N
33.48
H 23.19
31.74
64.82
249.45
254.59
N
H
206.89
200.01
295.90
299.86
H N
440.61
422.78
N H
H N
N
Fig 3.
3.2
NH
H
N N
H
N H
N 2H 4
The preliminary equilibrating simulation was performed at 4000K, and the trajectory
study was performed at 5500K. For each 100 trajectories, 69 decomposed at 5500K. Out of
the 69 reactions, 52 followed channel 1: N-N bond cleavage; 9 followed channel 2: hydrogen
migration; 8 follow channel 3: H2 elimination. In channel 2, N2H4 is easy to tautomerize to
NHNH3, but it can usually recover to N2H4 form. Three different reaction channels observed
in the MD trajectories are listed in Table 3.
Table 3. The decomposition paths of N2H4 observed in the 100 trajectories at
5500K
Path
1
Reaction Observed
H
H
N
N
H
N
H
H
H
N
H
N
N
H
N + NH3
H
H N
H
Freqs
52
H
N
4
Total
H
H
H
H
H
N
H H
N
H
No reaction
31
100
The three main reaction pathways and their energy barriers are shown in Fig.4. The
energy barrier of N-N bond split is 250.3 kJ/mol at B3LYP/6-31G(d,p) level and 258.54
kJ/mol at CCSD(T)=full/6-311+G(3df,2p) level in channel 1, which is similar to the energy
barrier of N-N bond split in N2H4. The slightly higher energy is required for proton transfer
5
773
Czech Republic
2006
[Content]
in channel 2 (264.81 kJ/mol at DFT level and 266.44 kJ/mol at CCSD(T) level). However,
the decomposition of the intermediate need much higher energy than that of back to the
reactant. The previously studies [35] carried using MP2 and QCISD(T) show that the
intermediate NHNH3 can exit stably in term of thermodynamics ,which is in agreement with
our result from quantum chemistry calculation, but inconsistent with the conclusion by
AIMD. In order to confirm the dominant decomposition mechanism, we further performed
rate constant calculation for channel 1 and channel 2 with Vklab program.The rate constants
obtained are shown in Table 4.We found the rate constant of N-N cleavage is at least 600
times larger than that of hydrogen migration. Moreover,the reverse rate constant of hydrogen
migration is more than 106 times larger than the forward one.Consequently, channel 2 is not
easy to take place, as observed in the trajectories simulation. In channel 3, the energy barrier
for H2 elimination is the highest (287.24 kJ/mol at B3LYP level and 318.12 kJ/mol at
CCSD(T) level), so it is also a minor way, as predicted by ab initio MD simulation.
Channel 1: The cleavage of the N-H bond
H
H
N
250.30
258.54
206.09
189.99
65.93
83.13
H
H
N
264.81
266.44
N
H
NH3 N
-0.74
0.19
19.22
11.80
H N
H
H
N
N
H
Channel 3: H2 elimination
H
H N
H
N
H
113.57
129.25
287.24
318.12
H
H
N
H
H
N
Fig 4.
774
N
H
Czech Republic
2006
[Content]
Table 4. The rate constants of N-N cleavage (channel 1) and hydrogen migration
(channel 2)
T(K)
k(s-1)
H migrationforwardTST H migrationreverseTST
1.157E-19
1.058E+06
5.280E-16
7.511E+06
4.532E-13
3.657E+07
1.152E-10
1.350E+08
1.173E-08
4.047E+08
5.901E-07
1.031E+09
1.706E-05
2.312E+09
3.162E-04
4.673E+09
4.083E-03
8.680E+09
3.913E-02
1.503E+10
2.925E-01
2.454E+10
1.773E+00
3.813E+10
8.987E+00
5.676E+10
1.490E+02
1.132E+11
1.553E+03
2.019E+11
1.133E+04
3.302E+11
6.235E+04
5.042E+11
2.740E+05
7.284E+11
N-N cleavageVT
5.465E-13
6.934E-10
2.135E-07
2.339E-05
1.176E-03
3.247E-02
5.583E-01
6.570E+00
5.679E+01
3.804E+02
2.061E+03
9.331E+03
3.627E+04
3.767E+05
2.632E+06
1.356E+07
5.492E+07
1.836E+08
400
450
500
550
600
650
700
750
800
850
900
950
1000
1100
1200
1300
1400
1500
3.3
N 3H
The preliminary equilibrating simulation was performed at 4500K, and the trajectory
study was performed at 7000K. For each 100 trajectories, 51 decomposed at 7000K. Out of
the 51 reactions, 45 followed channel 1: N2-N3 bond cleavage, 6 follow channel 2: N2-H1
bond cleavage. Two different reaction channels observed in the MD trajectories are listed in
Table 5.
Table 5. The decomposition paths of N3H observed in the 100 trajectories at 7000K
Path
1
2
Reaction Observed
4
N
4
N
3
N
1
Freqs
45
N
1
3
Total
No reaction
49
100
The geometry parameters of the reactant and the intermediates are depicted in Fig.2 and
their energy properties are summarized in Table 2. The two main reaction pathways and their
energy barriers are shown in Fig.5. the energy barrier for N2-N3 bond breaking is 303.79
kJ/mol at B3LYP level and 241.32 kJ/mol at CCSD(T) level. In channel 2, the cleavage of
775
Czech Republic
2006
[Content]
N2-H1 need higher energy (354.99 kJ/mol and 362.23 kJ/mol respectively), so it is minor in
the decomposition of N3H, as predicted by using VASP.
Channel 1: The cleavage of the N2-N3 bond. Channel 2: The cleavage of the N-H bond
N
N N N H
303.79
241.32
354.99
362.23
Fig 5.
3.4
N 4H 2
The preliminary equilibrating simulation was performed at 1500K, and the trajectory
study was performed at 1800K. For each 100 trajectories, 93 decomposed at 1800K. The 93
reactions followed the same way at first step: N2-N3 bond cleavage (Table 6). From the
trajectory simulation, we found N2-N3 is very easy to break.
Table 6. The decomposition paths of N4H2 observed in the 100 trajectories at
1800K
Path
1
Reaction Observed
4
N
N
N
N
N
H
N N H H
N N
Freqs
57
28
N N
N
H
N
H
4
Total
N
H
N
N
No reaction
100
H
N
H
N
N
H
776
Czech Republic
H
N
N
2006
[Content]
11.17
37.51
Fig 6.
3.5
N 4H 4
The preliminary equilibrating simulation was performed at 3000K, and the trajectory
study was performed at 3800K. For each 100 trajectories, 98 decomposed at 3800K. The 98
reactions followed the same way at first step: N2-N3 bond cleavage (Table 7).
Table 7. The decomposition paths of N4H4 observed in the 100 trajectories at
3800K
Path
1
Reaction Observed
H
N
N
N
N
N
N
H
N
1 H
1 H
N
3
N
N
H
Freqs
96
N
H
4
Total
H
H
N N
H
N
N
H
No reaction
N
H
2
100
We also further examined the first step of decomposition with ab initio MO method. The
geometry parameters of reactant and the intermediates are depicted in fig.2. Their energy
properties are summarized in Table 6. The energy barrier for N2-N3 breaking is
195.15kJ/mol at B3LYP/6-31g(d,p) level and 204.94kJ/mol at CCSD(T)/6-311+G(3df,2p)
level) (see Fig.7).
H
N
N
195.15
204.94
H
N
H
Fig 7.
N
N
H
N
H
777
4.
Czech Republic
2006
[Content]
CONCLUSION
The combined ab initio MD and ab intio MO study has revealed the dissociation
mechanisms for some simple hydronitrogen compounds. For N2H2, the main decomposition
pathway is N-H bond rupture, which need 249.45 kJ/mol in energy at B3LYP/6-31G(d,p)
level and 254.59 kJ/mol at CCSD(T)/6-311+G(3df,2p) level. For N2H4, N3H, N4H2 and
N4H4, they decompose mainly through different N-N bond breaking. The energy barriers are
250.30 kJ/mol, 303.79 kJ/mol, 11.17 kJ/mol and 195.15 kJ/mol respectively at B3LYP/631G(d,p) level and 258.54 kJ/mol, 241.32 kJ/mol, 37.51 kJ/mol and 204.94 kJ/mol at
CCSD(T)/6-311+G(3df,2p) level. According to the result by B3LYP, the thermal stability
increases by N4H2 < N4H4 < N2H2 < N2H4 < N3H. Nevertheless, at CCSD(T) level, the order
is N4H2 < N4H4 < N3H < N2H2 < N2H4. Both the two theoretical levels show that the energy
barrier of N4H2 dissociation is too low for using as structure unit of energetic materials. Our
further work is to add N4H2, N4H4, N2H4, N2H2 to heterocyclic nitrogen compounds as
bridges or substitute groups to construct some potential high nitrogen energetic materials
with good properties.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
778
Van der putten NHeijdentijk D and Schenk H. Tetrazole, Cryst. Struct. Commun. ,1974, 3:
321
Lesnikovich A I,Levchik S V.Balabanovich A I,Ivashkevich O A,Gaponik P N. The thermal
decomposition of tetrazoles, Thermochimica Acta,1992,200:427.
Thomas M.Klapke, Peter Mayer, Axel Schulz and Jan J. Weigand, 1,4-Bis-[1-Methyltetrazol5-yl]-1,4-Dimethyl-2-Tetrazene: A Stable, Highly Energetic Hexamer of Diazomethane
(CH2N2)6, Propellants, Explosives, Pyrotechnics 29 (2004), No.6:325-332
Wong M W. Chem. Phys.Lett.,1996,256:391.
Yoshiaki A,Yoshio O,Masamitsu T.Proc. Int.Pyrotech.Semin.,1996,22nd :13-22
Anton Hammerl, T.M.Klaptke,G.Holl , Peter Mayer, and Jan J. Weigand, Synthesis, Structure,
Molecular Orbital Calculations andDecomposition Mechanism for Tetrazolylazide CHN7, its
Phenyl Derivative PhCN7 and Tetrazolylpentazole CHN9, Propellants, Explosives,
Pyrotechnics 30 (2005), No. 1
Kathleen B , Isabella L K. The crystal and molecular structure of 5-aminotetrazole
monohydrate, [ J ] . Acta Crystallography ,1967 ,22 : 308.
Mary M W ,William S M , Ronald A H. The heats of combustion of substituted triazoles and
related high nitrogen compounds, [ J ] . Journal of Physics & Chemstry ,1957 ,61 : 261.
Leonard E B. Electro impact2induced fragmentation of 5-aminotetrazole, [J ] . Journal of
Heterocycle Chemistry , 1970 ,7 (6) : 1223.
Levchik S V , Ivashkevich O A ,Balabanovich A I ,et al. Thermal decomposition of 5aminotetrazoles, [J ] . Thermochimica Acta ,1992 ,207 : 115.
My-Hang V. Huynh, Michael A. Hiskey, Ernest L. Hartline, Polyazido High-Nitrogen
Compounds: Hydrazo and Azo-1,3,5-triazine, Angew. Chem. Int. Ed. 2004, 43, 4924 4928,
D.E. Chavez, M.A. Hiskey, J. Energetic Mater. 17 (1999) 357.
Hiskey M, Chavez D, Naud D. Progress in high-nitrogen chemistry in explosives ,propellants
and pyrotechnics, [A] . Proceedings of 27th International Pyrotechnics Seminar [C] ,July16-21
,USA ,Colorado ,2000.3-14.
Y. Oyumi,T . B. Brill, Thermal Decomposition of Energetic Materials 4. High-Rate,In Situ
Thermolysis of the Four,Six, and Eight Membered,Oxygen-Rich, Gem-Dinitroalkyl Cyclic
Nitramines,TN AZ, DNNC, and HNDZ, Combust. Flame 1985, 62,225.
Czech Republic
2006
[Content]
11
779
Czech Republic
2006
[Content]
Abstract:
Design of the HE warhead is process confronted with series of contradictory
requirements. Influence of warhead design on lethal efficiency is very complex. Lethal
efficiency of HE warhead depends on the form and dimension of the warhead, quantity
and type of explosive, warhead case material, warhead case thickness, fuse type,
explosive train etc. At warheads natural fragmentation, fragments geometry, their mass
and spatial distribution are functions of designed shape of the warhead case (shell),
mechanical properties of case material (tensile strength and yield strength) and
performances of explosive (physical and energetic). It is essential to have a capability to
make warhead performance predictions in the earliest phases of ammunition or
warheads preliminary design.
Keywords:
1.
INTRODUCTION
780
Czech Republic
2006
[Content]
steel warhead cases, initial elastic-plastic expansion of the case occurs when it is extended
from the original volume to about 1,44 times. When the current case volume being risen to
about 2,56 to 3,24 times of the initial warhead volume, the detonation products are released
through cracks and subsequently an expanding detonation products cloud is developed
beyond the fractured warhead case [9].
Fig 1.
Natural fragments spatial distribution (including their geometrical shapes and masses) is
a complex function depending on internal and external geometry of the warhead case
surface, warhead case mechanical properties (tensile strength and yield strength) and
energetic characteristics of the explosive.
Geometrical shape of natural fragments, their mass and spatial distribution are functions
of geometrical forms of internal and external warhead case surfaces, mechanical properties
of warhead case material (tensile strength and yield strength) and performances of explosive
(physical and energetic).
It is essential to have a capability to make warhead performance prediction in the earliest
phases of HE ammunition or warheads preliminary design. Ability for warhead
performances prediction depends on comprehensive data base of warheads natural
fragmentation features, including data on fragment numbers, initial fragment velocities,
warhead case material performances, fragment shape features and spatial fragment
distributions, etc.
Measurements of warhead performances require very complex measuring equipment and
measuring process itself is expensive as well.
Capability of warheads performances and efficiency assessment is based on complexity
of our database of natural fragmentation parameters which should encompass data on
number, mass, initial velocity, fragments shape factor and spatial distribution, material
characteristics of warhead case and explosive charge, etc.
Researches performed in USA were aimed on development of simulation methods for
prediction of fragmentation characteristics of HE ammunition [5]. The recent attainment is
development of CALE computer simulation program, which is able to simulate
fragmentation performances of two and three - dimensional axial symmetric warheads and
HE ammunition.
2.
781
2.1
Czech Republic
2006
[Content]
Pit test
In Pit test, warhead is detonated in closed space (pitfall), filled with sand (Fig.2.). After
the fragmentation of warhead, fragments are obtained from the sand. Mass and shape of
fragments are determined, and fragments are classified by their mass groups.
Experimental researches performed by
authors were undertaken in order to estimate all
relevant performances of warhead natural
fragmentation (number, mass and fragments
shape) when the material of the warhead case is
changed (three types of steel with different
mechanical properties). Number, mass and
fragments shape of each mass group are
determined using the Pit test, and relative
estimation of spatial fragments efficiency was
done by Arena test.
Prediction of fragments mass distribution is
usually performed by application of Mott
formula, or Held formula. Each of mentioned
formulas has certain limitations. These empirical
formulas are based on experimental data gained
from many fragmentations in Pit and Arena
facilities.
Fig 2.
)]
N (m ) = M 0 2 M k2 e
m 0 , 5 M k
(1)
where: N(m) is fragments number with a mass greater than m, M0 is the total fragments mass
and Mk is parameter which characterizes fragments mass distribution. Parameter Mk is a
function of warhead case thickness ti, internal diameter of the warhead di and explosive
charge. In the parameter Mk, constant B depends on the explosive charge and casing material.
1
M k = B t i 6 d i3 (1 + t i d i )
(2)
For a mild steel case, the constant B obviously decreases with increasing of detonation
pressure (yielding smaller fragments); but it also decreases with increasing case hardness [8].
2.1.2 Held formula
For prediction of mass distribution of natural fragmented warhead, Held formula with
two parameters and the total mass M0 or the best fit of total mass M0Best as inputs, gives an
excellent description of the experimentally found mass distributions of a natural fragmented
warhead [6]. An improved fit to natural-fragmentation data can be obtained using equation:
M (n ) = M 0 1 e Bn
782
(3)
Czech Republic
2006
[Content]
where; B and are both empirically determined constants, with B = const. d / t and of
order 10-2 and of order 2/3. In Held equation M0 is the total mass of all fragments, M(n) and
n are cumulative fragments mass and cumulative fragments number beginning with the
heaviest fragment. Held found that it was necessary to discard a few of the heaviest
fragments in order to obtain a curve fit to data over the rest of the range. The constants B and
are determined from above equation by mathematical transformation:
[M 0 M (n )] M 0 = e Bn
(4)
ln[(M 0 M (n )) M 0 ] = B n
(5)
(6)
From the fragment mass distribution (FMD) log-log diagram, constant B and exponent
with correlation coefficient r2 are obtained. If in the log-log diagram, the straight line does
not fit the measuring data very well, given total mass M0 is not an optimum mass for such
fragments mass distribution. Now, an optimum mass (or best mass) M0Best is calculated:
M 0 Best = M (n ) 1 e Bn
(7)
The constants B and are determined based on original mass of projectile metal body.
The new constants BB and B are determined with the total mass M0Best
M (n ) = M 0 Best 1 e BB n
(8)
2.2
Arena tests
Spatial fragments distributions were determined through Arena tests. At the arena
fragmentation test, fragments concentration density per steradian is measured when
projectile or warhead are put in horizontal position at a certain distance from the ground
(Fig.3).
Fig 3. Arena fragmentation test (measurement of the fragments concentration density per
steradian)
4
783
Czech Republic
2006
[Content]
After the warhead detonation, number of hits and perforations at wooden sectors are
counted and then hits number and perforations number per m2 are calculated for each sector.
From such obtained data, dependency of fragments concentration density per m2 as a
function of sector distance is established. Further, a characteristic distance, at which
fragments concentration density of 1 perforation per m2 is obtained, is determined from the
above function. This distance is called warhead efficiency radius, which means that warhead
with greater efficiency radius has corresponding greater lethal zone.
When it is necessary to compare lethal efficiencies for similar warheads, then a
projectile or a warhead is put on the ground in the vertical nose-down position (Fig.4).
Fig 4. Arena fragmentation test (measurement the fragments concentration density per m2)
(9)
where Npen is number of perforations at arena sector made of wooden panels, A is the sectors
area and R is radial distance between the warhead explosion center and a certain sector.
Spatial distribution and density of effective fragments is determined experimentally,
with fragmentation of HE projectiles in Arena facility.
Arena consists of k semicircular
sectors, divided on nk segments, where
increment of polar angle (k) is constant.
As a result of conducted experiments,
number of penetrations is determined in
every segment, and for every sector of
Arena. Based on given results it is possible
to determine total number of fragments in
j-th polar zone and for every distance
(radius of sector). After the experiment,
number of fragments that penetrated the
target is obtained (nj, for j = 1-22). As the
surface of every segment (target) is equal,
we can determine the number of fragments
penetrating particular polar zone, with
following equation:
784
N jk = n jk
Czech Republic
2006
S jk
[Content]
(10)
Sk
where:
Njk - Scaled fragment density or total number of fragments in j-th polar zone of k-th sector
njk - Number of fragments (penetrating the targets) in j-th segment of k-th sector
Sk
Sjk - Total surface of j-th polar zone in k-th sector, and it is determined by formula
If we introduce spatial angle (steradian) of j-th polar zone:
j = 2 (cos j cos j +1 )
(11)
2.3
(12)
Fragment velocities
Fragment velocities are measured using different techniques as electronic, optical or xray. If these techniques are not available, initial velocity of fragments released from
explosion of a warhead is approximated by the Gurney formula. The simplest expression of
the Gurney formula for symmetrical configurations is:
vGurney = 2 E 1 / (0,5 + M / C )
(13)
where 2 E is Gurney constant, M-metal mass of warhead case and C - mass of explosive
charge [7].
The Gurney constant can be approximated by the simple expression 2 E = 0,338 D ,
where D is the detonation velocity (depending on explosive type and its density).
3.
EXPERIMENTAL RESEARCH
3.1
785
Czech Republic
2006
[Content]
0,565. Dimensionless thickness of the warhead shell W2 was t/d=0,0629, and ratio of metal
warhead shell and explosive charge mass was CTNT/m = 0,653 [12].
Tested warheads cases were made of following steels 45Cr2 (ratio of tensile strength and
yield strength Rm/Rv=1,09), C70D (Rm/Rv=1,59) and steel AB or 9180VP (Rm/Rv=1,41)
marked according to Army standard, former SFRJ.
Four warheads fragmentation tests were performed for each casing material of the
warhead and obtained data were processed according to Mott methodology (Fig.8).
3500
Warhead W1
t/d=0,0687
C/m = 0,586
Explosive TNT
3000
2500
3500
2000
Steel 45Cr2
1500
Steel C70D
1000
500
0
Warhead W2
t/d=0,063
C/m = 0,653
Explosive TNT
3000
2500
2000
Steel 45Cr2
1500
Steel C70D
1000
500
0
2
3
4
5
6
Mean mass of fragments group - m aver , g
2
3
4
5
6
Mean mass of fragments group - m aver , g
Fig 8. Dependency of fragments number variation greater than m from the mean fragments
mass for warheads W1 and W2
Comparing the two warheads W1 and W2, greatest number of fragment has warhead
W2, made of steel C70D. Warhead W1 with case made of the steel 9180 has greatest
fragments number and minimal fragments number with case stell 45Cr2. Warhead W2 has
minimal number of fragments for case made of steel 45Cr2. Between steels C70D and 9180
there are slightly differences in number of fragments for warhead W2.
Authors noted a strong dependency of fragments number and their mass from the ratio
Rm/Rv. Steels with higher ratios Rm/Rv generated considerable higher fragments number
(steels C70D and 9180). However, it was not possible to conclude from Fig.8 which material
would be the most favorable. Because of that, authors presented their results in a different
way (fig. 9). The mean fragments mass of each particular group was taken as a variable.
Relationship between fragments mass groups number, whose mean masses are less than the
mean mass of the total fragments number, or between fragments mass of each group and the
total warhead fragments mass were established [12].
From above (fig. 9) diagrams, comparing number of fragments and mass participation it
can be more precisely done by optimization of warhead type as function of warhead body
material. In warhead W1 steel 9180 gives greater number of fragments and greater mass
participation comparing to steel C70D for fragments with mean mass smaller than 5 gr,
while for larger fragments mass more favorable steel is C70D. For warhead W2, for
fragments mass smaller than 4 gr most favorable is steel C70D, while for larger fragments
mass (greater from 4gr) most favorable is steel 9180. As a conclusion, steel C70D has better
lethality performances in warheads which are used mainly against infantry, while steel 9180
would have better penetration capability against light armored vehicles.
786
1
0.8
0.6
0.6
0.4
Warhead W1
t/d=0,0687
C/m = 0,586
Explosive TNT
0.2
0
0
Steel
Steel
Steel
Steel
Steel
Steel
9180 - Ni/N
45Cr2 - Ni/N
C70D - Ni/N
9180 - Mi/M
45Cr2 - Mi/M
C70D - Mi/M
2
4
6
8
Mean mass of fragments group - maver, g
0.4
0.2
0
10
[Content]
1
0.8
Ni /N for m i <maver
0.8
2006
Mi /M for m i <maver
Ni /N for m i <maver
Czech Republic
0.8
0.6
Mi /M for m i <maver
0.6
Steel
Steel
Steel
Steel
Steel
Steel
0.4
0.2
0
0
9180 - Ni/N
0.4
45Cr2 - Ni/N
Warhead W2
C70D - Ni/N
t/d=0,063
9180 - Mi/M 0.2
C/m = 0,653
45Cr2 - Mi/M
C70D - Mi/M
Explosive TNT
0
2
4
6
8
10
Mean mass of fragments group - m aver, g
Fig 9. Fragments number or mass participation as a function of mean fragments mass for
warheads W1 and W2
Warheads with case made of the steel with larger ratio of Rm/Rv generates greater
number of fragments but with less mean mass and greater fragments mass participation.
Authors explained this phenomenon as a consequence of warhead case ability to expand into
more considerable volume with thinner case walls under detonation products pressure before
fragmentation (warhead body mass conservation condition). Experimental researches have
shown that initial volume of a warhead can be increased several times before fragmentation
of warhead case material has been occurred (Fig. 1). Authors also found the relationship
ti/t0=(Vi/V0)-0,5 between relative volume rise and relative wall thickness during the warhead
case expansion [12]. At 4 times increase of warhead volume, new corresponding warhead case
thickness decreases at 0,52 of initial case thickness. When the ratio of Rm/Rv rises, the ratio
Vi/V0 increases as well and ti/t0 decreases, what results in greater fragments number but with
less mass.
Second analysis was analysis of natural fragmentation experimental data for projectiles
155 mm HE M107 (steel AISI 1045) and Bofors 155 mm HE M54 (High fragmentation
steel), made in Bofors facilities. Projectiles were geometrically identical, with different types
of projectile body material [2].
Before the tests, the ordinary bursting charges in the shells were removed, and the shells
were filled with cast hexotol corresponding to "Composition B". The composition of this
hexotol was RDX 59.5%, TNT 39.5% and wax 1%. The weight of the explosive was approx.
7 kg. The average weights of the complete shells (without fuzes used for the tests were:
Bofors 155 mm HE shell M54 42.4 kg, US 155 mm HE shell M107 42.6 kg [2].
The shells were exploded statically, in target arrangements A and B of different designs.
The shells were placed in horizontal and vertical positions.
The target arrangement type A (shell in horizontal position) was built up in the form of
an enclosure. Three sides consisted of 2 mm mild steel plate, size 1 x 6 m. The fourth side
was made of 7 mm steel plate, brand SIS 1411 (s = 250 MPa, B = 420 MPa) size 1 x 6 m.
The distance at right angles from the pole on which the shell was placed to the sides was 3
m. The target arrangement type B (shell in vertical position) was made of steel plates with
the dimensions 1 x 2 m, with thicknesses of 2 mm and 7 mm, respectively, and with the
same mechanical properties as those used in target arrangement type A. The plates were
placed on edge to a height of 3 m and a width of 2 m. For fragments efficiency assessment
on unprotected personnel, 19mm chipboards with the dimensions 1,22 x 2,5m was used, at
distances of 6m and 12 m [2].
The targets were placed so that the bottom edges of the lower parts were on the same
horizontal plane. The height at which the shell was placed, in an upright position, was fixed
at 1,50 m, and the line of sight to each target was drawn in [2].
8
787
Czech Republic
2006
[Content]
In Bofors study, the general impression obtained from these comparative static bursting
tests was that the Bofors shell gave considerably better fragmentation. Bofors 155 mm HE
shell M54, gave twice as many perforations as the U.S. 155 mm HE shell M 107. The
appearance of the holes also differed. The 155 mm HE shell M107, gave large, longish
perforations, while the Bofors 155 mm HE shell M54, produced holes of more regular sizes.
This was entirely due to the properties of the materials in the two types of shells.
Authors have determined, using the earlier explained method of fragments spatial
distribution, following:
Target arrangement type A: In first polar zone (-45 to +45) there is no significant
difference in penetration density through targets (thickness 2mm). In polar zone of 45 to
135 (for 2mm target thickness) penetration density is 35% greater, and in polar zone of
225 to 315 (for 7mm target thickness), penetration density is 100% greater. Further more,
in polar zone of 135 to 225 there is 92 % greater penetration density for Bofors 155mm
M54 projectile. General conclusion is that Bofors warhead has about 35% better efficiency
against infantry, and about 100% better efficiency against light armored and mechanized
vehicles.
By analysis of experimental results of projectiles natural fragmentation in B arena
(influence of lateral fragments), on different distances (3 m, 6 m and 12 m) and against
targets of different thicknesses (2mm and 7mm steel plate) authors concluded that efficiency
of Bofors 155mm M54 projectile against armored targets is 1,8 to 2,05 times better at
distance from 3m to 12m, respectively, while efficiency of Bofors 155mm M54 projectile
against infantry is from 2,1 to 1,42 times better at distances 6m and 12m, respectively.
As a general conclusion for both tests, Bofors 155mm M54 projectile with high
fragmentation steel has significantly greater number of fragments, compared to US projectile
155mm M107. Bofors 155mm M54 projectiles efficiency is significantly better against light
armored and mechanized vehicles and that penetration capability (min 80%) doesnt
significantly change with increasing the distance, comparing to US projectile 155mm M107.
Against infantry, that lethal efficiency is minimum 35% greater, comparing to 155mm M107
and what is important to mention, it decreases with increasing the distance. That penetration
capability of fragments doesnt considerably change with increasing the distance, even
though projectile 155mm M54 has considerably greater number of fragments, and smaller
mean mass of fragments, with similar velocity of fragments; this points to fact that important
role plays shape of fragments, during process of penetration through target, specific energy
of fragment.
3.2
Czech Republic
2006
[Content]
warhead W2 was t/d = 0,0629. Four fragmentation tests were carried out for each type of
explosive charge.
For warheads W1 and W2 fragments were classified by mass groups and experimental
data were processed using Held methodology (Fig.10).
6000
Cumulative fragments mass, g
6000
5000
4000
Warhead W1
Steel C70D
t/d = 0,0707
3000
2000
1000
5000
4000
Warhead W2
Steel C70D
t/d = 0,0629
3000
2000
1000
1000
2000
3000
4000
Cumulative fragments number
5000
1000
2000
3000
4000
Cumulative fragments number
5000
Fig 10. Cumulative fragments mass dependency on cumulative fragments number for
warheads W1 and W2 [13]
Warheads W1 and W2 with the case made of steel C70D and Comp B produces larger
number of fragments compared to the warheads with TNT explosive charge. At the same
time, in these two warheads with Comp B greater loss of fragment mass has been noticed
(metal part of warhead body has been converted to very small fragments), for W1 11,7 %
(Comp B) and 8,7 % (TNT), comparing to W2 10,8% (Comp B) and 6,1% (TNT).
However, it is not possible to conclude from Fig.10 which warhead has better
fragmentation performances. Because of that, authors presented their results in a different
way (Fig. 11).
0.8
0.6
0.4
0.2
0
0
2
4
6
8
Mean mass of fragments group - m aver , g
0.6
0.4
0.2
0
10
0.8
Ni /N for m i <m aver
Warhead W1
Steel C70D
t/d = 0,0707
0.8
Ni /N for m i <maver
0.8
Warhead W2
Steel C70D
t/d = 0,0629
0.6
0.4
0.2
0
0
2
4
6
8
Mean mass of fragments group - m aver , g
0.6
0.4
Mi /M for m i <maver
0.2
0
10
Fig 11. Fragments number contribution, or fragment mass contribution as a function of the
mean fragments mass for warheads W1 and W2 [13]
The mean mass of particular groups was taken as a variable. Correlation between
fragments number of mass groups, whose mean masses are less than mean fragments mass
related to total fragments number, or between fragments mass of these mass groups and total
fragments mass of the warhead were established. The warheads W1 and W2 with explosive
charges of Comp. B generate more fragments but with less mean mass, compared to
warheads W1 and W2 filled with TNT. Warhead W1, with Comp. B has in fragment groups,
with mean mass smaller than 10gr, dominantly larger number of fragments and larger mass
participation of these fragments, while in warhead W2 these differences arent significantly
greater.
Authors explained such results as consequence of more intensive influence of detonation
products generated by more powerful explosive. When more powerful explosive is used,
10
789
Czech Republic
2006
[Content]
initial warhead volume is increased more intensive and warhead case thickness is decreased
(condition of mass conservation).
Since the sophisticated measuring equipment for fragment velocity measurement was
not available to authors, the Gurneys formula was used in order to predict initial fragments
velocities [7]. Initial fragment velocity variation for warheads filled with TNT and
Composition B explosives charge and for particular warhead segments was determined using
methodology described in the Crulls report [3] (fig. 12 and fig. 13).
2000
2000
1800
1800
From these diagrams of velocity of fragments vs. relative distance of warhead case
segments it can be seen that mean fragment velocity is about 300 m/s larger in warheads
filled with Comp. B, fragments with less mass will have less energy and radius of lethality.
1600
1400
Warhead W1
Explosive Comp B, C/m = 0,587
Explosive TNT, C/m = 0,565
1200
1000
800
600
400
200
0
1600
1400
Warhead W2
Explosive Comp B, C/m = 0,667
1200
1000
800
600
400
200
0
0.2
0.4
0.6
0.8
Relative distance of warhead case segment
0.2
0.4
0.6
0.8
Relative distance of warhead case segment
2000
2000
1800
1800
Fig 12. Variation of fragment velocity as a function of explosive type for warheads W1 and
W2 [13]
1600
1400
1200
1000
800
600
400
200
0
0
0.2
0.4
0.6
Relative position
0.8
1600
1400
1200
1000
800
600
400
200
0
0
0.2
0.4
0.6
0.8
Relative position
Fig 13. Variation of fragment velocity as a function of explosive type for warheads 128 mm
M87 and 152 mm M84 with explosive Comp B and TNT
Fragments concentration density per steradian as a function of polar zone for warheads
128 mm M87 with explosive Comp B and TNT is presented in Fig. 14 [11]. Lethality zone for
warheads 152 mm M84 with explosive Comp B and TNT is presented in Fig. 15.
790
11
1200
1000
Explosive TNT,
C/m = 0,558
600
[Content]
Warhead 152 mm M84
Steel 45Cr2
t/d = 0,117
Explosive Comp B,
C/m = 0,299
Explosive TNT,
C/m = 0,282
Explosive Comp
B, C/m = 0,598
800
2006
Density, frags/m 2
1400
Czech Republic
400
5
4
3
2
1
200
0
0
20
40
60
80
100
120
140
160
180
15
20
25
30
35
40
Diagram on fig. 15 shows that projectile 155mm M84 filled with Comp. B has 33%
larger lethality area.
3.3
Experimental researches carried out by the authors were aimed on natural fragmentation
performances estimation (number, mass and fragment form) of the warhead when the
warhead or projectile design shape is changed [14]. Spatial fragments distributions were
determined through arena tests.
Experimental fragmentation tests in order to determine the influence of projectile design
on fragmentation characteristics were carried out with 11 types of HE warheads 105 mm M1
[1]
, 120 mm W2, 120 mm M62P3, 122 mm OF-462, 122 mm M76, 128 mm M63, 128 mm
M87 [11], 152 mm OF-540, 152 mm M84, 155 mm M107 [1] and 155 mm ERFB.
Fragments concentration density per steradian (spatial fragments density) for each polar
sector of the arena:
12
791
2000
1800
1600
2000
1400
1200
1000
800
600
400
200
0
0
20
40
60
Czech Republic
80
100
120
Polar Zone, deg
140
160
[Content]
1800
1600
1400
1200
1000
800
600
400
200
0
0
180
2000
2006
20
40
60
80
100
120
140
160
180
1800
1600
1400
1200
1000
800
600
400
200
0
0
20
40
60
80
100
120
Polar Zone, deg
140
160
180
Fig 17. Fragments concentration density per steradian as a function of polar zone
18
16
Density, frags/m 2
Density, frags/m 2
20
18
14
12
10
8
6
4
2
10
15
20
25
30
35
40
10
15
18
Density, frags/m2
16
14
12
10
8
6
4
20
25
30
35
40
18
16
Density, frags/m2
20
20
14
12
10
8
6
4
2
0
10
15
20
25
30
Radius of lethality zones, m
35
40
10
15
20
25
30
Radius of lethality zones, m
35
40
792
13
14
12
[Content]
155 mm M107, Explosive TNT,
C/m=0,221, t/d = 0,138
155 mm ERFB, Explosive TNT,
C/m = 0,249, t/d = 0,128
18
16
Density, frags/m 2
Density, frags/m 2
16
2006
20
18
Czech Republic
10
8
6
4
2
14
12
10
8
6
4
2
0
10
15
20
25
30
35
40
10
15
20
25
30
35
40
Comparative analysis of lethality radius for projectiles with the same caliber and
different design confirm observations mentioned above. Projectile 120 mm W2 has
considerably greater ratio C/m and smaller ratio of t/d comparing to projectile 120mm
M62P3, so the lethality zone of this warhead is 2,8 time greater. Projectile 122 mm M76
(Comp. B) has 44% larger lethality area compared to projectile 122 mm OF-462 (TNT),
because it has greater detonation velocity, greater C/m ratio, greater slenderness of ogive and
smaller ratio of t/d. Warhead 128 mm M87 (TNT) has 50% larger lethality area or 80%
larger lethality area when it is filled with RDX compared to projectile 128 mm M63 (TNT).
2000
1800
1200
1400
1000
800
600
400
200
1200
1000
800
600
400
0
0.2
0.4
0.6
Relative position
0.8
2000
1800
1800
2000
1600
1400
1200
1000
800
600
400
200
0
0
0.2
0.4
0.6
Relative position
0.8
800
600
200
0
1800
1200
1000
800
Warhead 152 mm M84,
Explosive TNT, C/m = 0,282
Warhead 152 mm OF-540,
Explosive TNT, C/m = 0,289
200
0
0
0.2
0.4
0.6
Relative position
400
2000
1400
1000
1600
0.8
1200
1800
400
0.4
0.6
Relative position
1400
2000
600
0.2
1600
1400
200
1600
0.2
0.4
0.6
Relative position
0.8
1600
1400
1200
1000
800
600
400
200
0
0.8
0.2
0.4
0.6
Relative position
0.8
14
793
Czech Republic
2006
[Content]
Warhead 128 mm M87 has greater C/m ratio and slightly smaller ratio of t/d and greater
slenderness of ogive. Comparing two projectiles 152 mm OF-540 and 152 mm M84 is
interesting because of the influence of hollow base at the projectile (M84) bottom on
fragmentation efficiency. This projectile has greater slenderness of ogive and similar ratios
C/m and t/d, but the influence of hollow base reduces his lethality capabilities.
Projectile 155 mm ERFB has greater ratio C/m and smaller ratio t/d and significantly
greater slenderness of ogive comparing to projectile 155 mm M107, so his lethality zone is
85% greater.
Significant influence on projectile efficiency is made by fragment velocity distribution
along projectile symmetry axis. It is especially important for projectile ogive and his
cylindrical part, because the fragments from these parts have highest influence on spatial
distribution of lateral fragment spread.
Maximum fragment velocities are achieved at warheads having greater explosive charge
mass and metal case mass ratio C/m, which are appeared within central fragments polar zone
(fig. 20).
4.
CONCLUSION
Influence of warhead design on lethal efficiency is very complex. Very important role
play design factors as external and internal geometrical paths, relative warhead case
thickness, explosive charge mass and warhead metal case ratio, type of explosive, etc.
Until recently, any prediction method, which is capable to quantify particular influence
of above parameters on lethal efficiency radius, has not been developed yet.
It was not possible to clearly define the influence of explosive type on natural
fragmentation performances by using Held method.
Mott formula cannot provide a clear influence of warhead metal case variation on
natural fragmentation performances.
Authors proposed a new approach to presentation of experimental results, which enables
selection of an optimal warhead explosive charge. Authors gave also different approach to
presentation of fragmentation experimental data, which enables selection of an optimal
warhead case material [12, 13].
Authors also established relationship between ti/t0 and Vi/V0 and connected it with the
ratio Rm/Rv in order to find how above parameters affect fragment number and their mass.
Warhead with steel case, which has higher ratio Rm/Rv generates simultaneously greater
fragments number but with less mean fragments mass and with greater fragments mass
participation. Authors also concluded that it was necessary a further development of
empirical relationship between Rm/Rv, and Vi/V0 [12].
In attempt to introspect complexity of influence of design, material of body, types of
explosive and methods of conducting the experimental testing, authors made comprehensive
analysis of eleven types of projectiles and warheads. Authors observed significant influence
of ratio C/m (mass of explosive to mass of metal body ratio), ratio t/d (equivalent thickness
of projectile body wall to caliber ratio), projectile ogive slenderness, fragment velocity
distributions along projectile symmetry axis and geometrical shapes of fragments.
Influence of geometrical shapes of fragments is not investigated in detail, and in
following period numerical simulation (drag, center of pressure, center of mass, etc.) of
behavior of fragment should be conducted during his flight through air and his behavior
794
15
Czech Republic
2006
[Content]
during the penetration through target, because of his influence on penetrating capabilities.
Analysis of geometrical shape is conducted only for warhead 128 mm M87 [11], but it is
insufficient for making scientific conclusions. Our activities are directed to creating expert
base for natural fragmentation processes.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
16
ANON: US Army Test and Evaluation Command Test Operations Procedure Fragment
Penetration Test of Armor, Report number TOP 2-2-722, ADA 125824., 1983
BOFORS: Fragmentation effect of 155m HE shells, Results of static test for comparison between
US 155mm HE shell M107 and Bofors 155mm HE shell M54
M. M. CRULL: Prediction of Primary Fragmentation Characteristics of Cased Munitions,
UXO/Countermine Forum 1998, The Joint UXO Coordination Office, 1998
O. GARLAND: The use of Physics-Based Models to Prediction Fragmentation of Ordnance,
Naval Air Warfare Center Aircraft Division, OMB N0. 0704-0188., 1997
V. GOLD, E. BAKER, NG K. and J. HIRLINGER: A Method for Prediction Fragmentation
Characteristics of Natural and Preformed Explosive Fragmentation Munitions, ARWEC-TR01007, US Army Armament Research, Development and Engineering Center, 2001
M. HELD: Fragmentation Warhead, Tactical Missile Warhead, Edited by Carleone J., Progress
in Astronautics and Aeronautics, Volume 155, AIAA, Washington, 1993.
R. KARPP and W. PREDEBON: Calculation of Fragment Velocity from Natural Fragmenting
Munitions, BRL Memorandum Report N0. 2509, USA Ballistic Research Laboratories,
Aberdeen Proving Ground, Maryland, 1975
R. LLOYD: Conventional Warhead System Physics and Engineering Design, Progress in
Astronautics and Aeronautics, Volume 179, AIAA, (1999).
J. PEARSON: A Fragmentation Model Applied to Shear-Control Warheads, Naval Weapons
Center, 1991
A. VICTOR: Warhead performance Calculation for Threat Hazard Assessment, Department of
Defense Explosive Safety Seminar, Las Vegas, (1996).
B. ZECEVIC and D. EROR: Terminal Ballistic of 128 mm M87 Rocket, FTV-270, Sarajevo,
(1986).
B. ZECEVIC, J. TERZIC and A. CATOVIC: Influence of Warhead Case Material on Natural
Fragmentation Performances, Annuals of DAAAM for 2004 & Proceedings of the 15th
International DAAAM Symposium, Viena, 2004
B. ZECEVIC, J. TERZIC and A. CATOVIC: Experimental Research on Influence of Explosive
Charge to Natural Fragment Size Distribution, Annuals of DAAAM for 2004 & Proceedings of
the 15th International DAAAM Symposium, Viena, 2004
B. ZECEVIC, J. TERZIC and A. CATOVIC: Influence of Warhead Design on Natural
Fragmentation Performances, Annuals of DAAAM for 2004 & Proceedings of the 15th
International DAAAM Symposium, Viena, 2004
795
Czech Republic
2006
[Content]
Abstract
An ESZ instrument of a new, relatively simple construction has been applied to
determination of electric spark sensitivity (EES) of 16 nitramines. Results obtained are
compared with those from measurements by means of older RDAD instrument. Attention was
focused both on the relationships between the EES values from both the instruments and
theoretical B3LYP/6-311+G(d,p) N-NO2 bond dissociation energies, B3LYP/6-31G(d,p)
Mulliken net charges of the nitro group, heats of fusion and 15N NMR chemical shifts of the
nitrogen atoms of the most reactive nitro groups, respectively. It is stated a larger difference
between results of both instruments. It has been found that these relationships for ESZ KTTV
results are strongly affected by molecular structure factors. It is also pointed out that the
dislocations in the crystals should have some relation to electric spark sensitivity.
Keywords: bond dissociation energy, electric spark, sensitivity, nitramines
1.
INTRODUCTION
Recent studies on sensitivity of energetic materials to the electric spark [1-23] have
shown that this sensitivity depends not only on the chemical entity of the material given
[1,4,6,8,9,12,14-16], its granulometry [1-3,12,16] and grain shape [3], mechanical properties
[5,6], temperature [1,13], moisture content [2,4], its thermal [2,6,7,12,17,19] and impact
[7,16,20] reactivities but also on the configuration of electrodes, general structure of the
circuit [2,4] and finally on the discharge time [26]. Consequently, it has not yet been possible
to devise an international standard test specifying this stability [9,10,18]. Irrespective of the
above-mentioned spectrum of effects influencing the said sensitivity mentioned above, a
linear relationship between the tests results obtained with various apparatuses can be found
[9,10,11].
Our Institute in cooperation with the R&D Department of Austin Detonator Ltd.,
(Vsetn, Czech Republic) took part in constructing of an instrument for determination of
electric spark sensitivity [8-10]. With the help of this instrument, marked as RDAD,
measurements were carried out on a large set of polynitro compounds of secondary
explosive type [8-10,14,19-23]. From the results obtained it was possible to make
conclusions on the chemical mechanisms of initiation of these compounds [14,19-23]. The
outputs from the RDAD instrument correlate satisfactory with those obtained by other
methods of determination of the electric spark sensitivity in some recognized laboratories
[10]. However, this instrument is not suitable for determination of the said sensitivity of
primers and pyrotechnics. For this reason, we have developed with a financial support from
796
Czech Republic
2006
[Content]
Czech Ministry of Industry and Commerce [24] a new instrument [25-27], marked as ESZ
KTTV [26]. This instrument is suitable for measurements of energetic materials within the
whole scope of technically attractive sensitivities.
The present paper shows difference between outputs of both instruments mentioned
above from the point of view not only numerical values but also molecular structure. These
differences are demonstrated on some nitramines, theirs molecular structure are relatively
simple polynitro compounds, which mechanism of primary homolysis of N-NO2 bonds is
well understood [28-31].
2.
EXPERIMENTAL
2.1.
(1)
Beside the working condenser and test chamber with sample, the series resonance
circuit consists of the loss resistor of circuit (whose value varies from 0.27 to 10.0 ) and
total parasitic inductance (ranging from 1 to 1.6 H). In the course of oscillation discharge,
the increase in output in the space between the electrodes of testing spark gap reaches the
values of dP/dt 5.1018 W/s. The period of oscillations (from 0.5 to 20 s) as well as the
parameters of the instrument mentioned above depend considerably on the capacity value
(C), testing voltage (U), and the sum of Joule losses in the test substance.
2.2.
The schemes of its wiring and its spark gap are presented in papers [8-10]. The
bottom electrode is a steel base in whose centre there is attached a screw as a leak electrode.
Attached to this screw is a plastic cylinder with a cut for fastening a piece of flexible tubing
of 5 mm height and 5 mm diameter serving as a container of the sample tested. The top
electrode is a steel cylinder of conical shape with attached resistance wire protected by a
plastic distance stop, which presses an Umafol lid to the flexible tubing. The capacity of the
capacitors was chosen so as to allow measurements in the voltage interval from 8 to 14 kV.
If the initiation was successful, the next measurement was carried out with a voltage lowered
by 0.2 kV; if it was unsuccessful, the voltage was increased by the same value. We
considered an initiation to be successful if the sample disappeared or the flexible tubing was
torn. Altogether 25 measurements were carried out with each substance and the results were
treated in the sense of eq. 1 by means of the up and down method.
2.3.
797
Czech Republic
2006
[Content]
mounted on the lower cylindrical metal electrode. The upper electrode is equipped with
manual vertically adjustable positioner. The micro container itself is placed in a separate test
box with ventilation. The time behavior of voltage and current at the spark gap is registered
with a scope and than evaluated using a microcomputer to give the effective energy
transmitted to the sample.
Table 1 Summary of the studied nitramines, their code designation and sensitivities to
electric spark EES (RDAD and ESZ KTTV instruments)
EES required for 50 % initiation
probability obtained
RDAD
ESZ KTTV
Nitramines
No.
a
b
Chemical name
Code
designation
(J)
Ref.
(mJ)
N-Methyl-N,2,4,6-tetranitroaniline
TETRYL
5.49
23
574.2 7.6
2,5-Dinitro-2,5-diazahexane
DMEDNA
8.24
23
392.051.7
3,5-Dinitro-3,5-diazaheptane
DNDAH
12.49
23
225.011.3
2,4,6-Trinitro-2,4,6-triazaheptane
ORDX
8.08
23
243.314.4
1,9-Diacetoxy-2,4,6,8-tetranitro2,4,6,8-tetraazanonane
AcAn
13.93
23
206.714.4
1,3-Dinitroimidazolidine
CPX
9.68
23
403.314.4
1,3-Dinitro-1,3-diazetidine
TETROGEN
6.25
23
494.0a
1,4-Dinitropiperazine
DNDC
15.97
23
190.0 8.3
1,3,5-Trinitro-1,3,5-triazinane
RDX
2.49
23
216.415.6
10
1,3,5-Trinitro-1,3,5-triazepane
HOMO
3.70
23
431.319.3
11
1,3,5,7-Tetranitro-1,3,5,7-tetrazocane
HMX
2.89
23
236.410.1
12
1,3,5,7,9-Pentanitro-1,3,5,7,9pentazecane
DECAGEN
2.96
23
276.5a
13
trans-1,4,5,8-Tetranitrodecahydropyrazino[2,3-b]pyrazine
TNAD
5.43
23
520.053.3
14
1,3,3-Trinitroazetidine
TNAZ
8.76
20
78.37.2
15
2,4,6,8,10,12-Hexanitro-2,4,6,8,10,12hexaazaisowurtzitane
HNIW
4.70
20
462.041.4
16
4,10-Dinitro-2,6,8,12-tetraoxa-4,10diazaisowurtzitane
TEX
285.525.3
the value predicted by means of Fig. 1 (here line A for DECAGEN and line B for TETROGEN);
the method of measurement (i.e., according to the section 2.2.) failed in this case;
798
2.4.
Czech Republic
2006
[Content]
A survey of the tested nitramines is given in Table 1. All the nitramines were purified
by crystallization in motion, and often by precipitation crystallization (with a good stirring)
from organic solvents. The crystallization was carried out in such a way as to obtain the
product appearing as individual in thin layer chromatograph (Silufol, hexane-acetone eluent
and developing the chromatograms by 2 % diphenylamine in ethanol with UV detection).
However, the granulometry of the substances investigated was not specified.
2.5.
3.
DISCUSSION
The analysis of relationship between the EES values obtained by measuring on RDAD
and ESZ KTTV instruments revealed that the relationship is not unequivocal as is illustrated
in Figure 1. Similar trends in EES values obtained from the two instruments are exhibited by
class A of high-energetic cyclic nitramines. On the other hand, class C of linear nitramines
and particularly class B of cyclic nitramines show opposite trends in EES values. DMEDNA
data correlate with the last class probably due to a fact that its molecular skeleton can be
found as a part of skeletons of remaining members of this class. Surprisingly TETRYL data
correlate with data of the A class and TNAZ is unexpectedly more sensitive in measurements
by means of ESZ KTTV instrument.
Large difference between outputs of measurements on DRAD instrument on the one
hand, and those from ESZ KTTV instrument, on the other (both the values see Table 1),
should be due to mutual difference in construction namely spark gaps of each instrument;
scheme of this gap in the first instrument is presented in Ref. 10 and in the second one in
Ref. 26. Large energetic losses in the spark gap of the RDAD instrument might manifest
oneself as a strong thermal effect of the relevant electric discharge, i.e., outputs of this
measurement connected with thermolysis of the given samples. This statement should be
valid for all equipments whose outputs are in good unambiguous correlation with those of
DRAD instrument (see in Ref. 10).
It is know that initiation reactivity of secondary nitramines is connected with N-NO2
bond homolysis (see Ref. 15 and reference therein). It is logical that the weakest bond will
be broken as the first. Therefore, energy of electric spark, EES, should correlate with the
4
799
Czech Republic
2006
[Content]
lowest values of the bond dissociation energy DH(N-N). The verification can be found in
Figs. 2 and 3. While the relationships for outputs from RDAD instrument have a logical
regularity (Fig. 2), the opposed direction of these relationships is beyond expectation typical
for ESZ KTTV measurements (Fig. 3). Some authors are inclined to accept the idea of
conversion of the spark energy into a thermal effect, i.e., thermolytic mechanism of the
initiation by electric spark [3,6,15,17,19]. However, Rat et al. [5] considers this initiation to
be a multidisciplinary problem. Results of this paper give these authors [5] the truth. Namely
facts in Fig. 2 signalize the multidisciplinarity and that the thermal component of electric
discharge not participate in the ESZ KTTV measurements; the reversed relationships in Fig.
2 still wait for an interpretation (the most probably by means of a method of quantum
chemistry simulation).
Table 2: Calculated N-NO2 bond dissociation energies, DH(N-N), Mulliken net charges of
the nitro group, Qnitro, heat of fussion, Hm,tr, and 15N NMR chemical shifts, N, of
NO2 nitrogen atoms
Code
designation
TETRYL
DMEDNA
ORDX
CPX
TETROGEN
DNDC
RDX
HOMO
HMX
DECAGEN
TNAD
TNAZ
HNIW
TEX
DH(NN)
(kJ mol1
)
130.95
176.55
194.86
186.50
186.83
163.96
168.65
188.72
154.84
161.78
177.70
177.16
179.05
168.09
167.16
167.29
140.58
176.05
154.36
175.24
177.06
151.12
169.46
186.79
Qnitro
Heat of
fussionc
Hm,tr
Ref
.
(e)
(kJ mol
1
)
-0.135261
-0.201818
-0.176718
-0.178246
-0.198959
-0.167031
-0.154150
-0.165965
-0.127376
-0.133985
-0.142745
-0.107995
-0.120368
-0.127184
-0.143869
-0.128259
-0.118018
-0.093269
-0.156531
-0.144025
-0.152032
-0.080958
-0.074584
-0.111044
25.850
29.409
34
34
34.031
34
22.570
26.320
33.823
33.012
27.738
15
position
in
molecule
(ppm)
Ref.
N(CH3)
2,52,64-
-23.1
-27.8
-28.5
-32.0
37
37
34
34
34
34
34
1,31,31,41,3,51,53-
-31.2
-27.8
-26.2
-32.9
-33.0
-34.4
37
37
37
37
32.100
34
1,3,5,7-
-34.7
37
34.930
34
1,3,5,7,9-
-33.2
37
46.400
34
30.310
43.030
36
35
35
20.5
-40.3
-43.4
-33.4
38
36.100
12,6,8,124,104,10-
37
37
37
37
800
Czech Republic
2006
[Content]
700
TETRYL
600
TNAD
TNAD
500
HOMO
CPX
HNIW
400
DMEDNA
Ay = 104.57x - 31.935
y = -30.283x + 674.01
2
R = 0.9704
R = 0.8966
300
200
TEX
HMX
DNDC
ORDX
RDX
DNDAH
y = -5.7614x + 291.26
2
R = 0.9209
AcAn
100
TNAZ
0
0
10
12
14
16
18
Fig. 1 Relationship between the outputs from ESZ KTTV and RDAD instruments.
3.5
y = 0.0526x - 7.1785
2
R = 0.9888
DNDC
TEX
y = 0.0506x - 6.0647
2
R = 0.9908
2.5
CPX
TNAZ
DIGEN
2
y = -0.0006x + 2.2133
2
R = 0.8621
TETRYL
TNAD
HNIW
1.5
ORDX
DMEDNA
TETROGEN
HOMO
1
DECAGEN
HMX
RDX
0.5
0
135
145
155
165
175
185
195
205
-1
Fig. 2 Relationship between EES values from RDAD instrument and DH(N-N) energy.
Mulliken net charges of the total nitro group, Qnitro, were used in the study of impact
sensitivities of nitro compounds [33]. Application of this approach to electric spark
sensitivity gave in this study correlations which are present in Figs. 4 and 5. The highest
Qnitro values (i. e., the lowest absolute values) of each nitramine were used in both cases. The
outputs of the RDAD instrument are relatively scattered even though some relationships are
6
801
Czech Republic
2006
[Content]
intimated here (Fig. 4). As a large part of discharge is enacted in the air, also other factors
should have an influence on the nitramines initiation (they might be partly in floatation
during the discharge). The EES values from ESZ KTTV measurements, on the other hand,
gave the relationships (Fig. 5) which are strongly determined by molecular structure (namely
for compounds along the lines with positive slope).
7
y = -0,039x + 12,857
2
R = 0,9502
y = -0,0286x + 10,618
2
R = 0,9645
6,5
TETRYL
TETROGEN
TNAD
HOMO
HNIW
DMEDNA
CPX
DECAGEN
TEX
ORDX
5,5
HMX
RDX
DNDC
5
y = -0,0356x + 10,721
2
R = 0,9549
4,5
TNAZ
4
140
150
160
170
180
190
200
-1
Fig. 3 Relationship between EES values from ESZ KTTV instrument and DH(N-N) energy.
18
DNDC
16
14
TEX
12
10
CPX
CPX
DMEDNA
y = -174,93x - 20,522
2
R = 0,9053
TNAZ
ORDX
TETROGEN
y = -57,627x + 0,0293
2
R = 0,9999
TETRYL
TNAD
HNIW
HOMO
RDX
y = 13,854x + 4,2944
2
R = 0,8982
0
-0,22
RDX
HMX
DECAGEN
DIGEN
-0,2
-0,18
-0,16
-0,14
-0,12
-0,1
-0,08
-0,06
-0,04
-0,02
Fig. 4 Relationship between EES values from RDAD instrument and net electronic charge of
the nitro group Qnitro.
802
Czech Republic
2006
[Content]
700
y = -5423,9x - 326,54
2
R = 0,9713
600
TETRYL
TNAD
TETROGEN
500
y = 463,44x + 485,54
2
R = 0,9476
HNIW
HOMO
DMEDNA
400
HOMO
CPX
DMEDNA
300
y = -6147,7x - 846,97
2
R = 0,9922
TEX
HMX
ORDX
HMX
DNDC
200
y = 1095,8x + 365,34
2
R = 0,8942
DECAGEN
RDX
DNDC
100
TNAZ
0
-0,22
-0,2
-0,18
-0,16
-0,14
-0,12
-0,1
-0,08
-0,06
Fig. 5 Relationship between EES values from ESZ KTTV instrument and net electronic
charge of the nitro group Qnitro.
18
16
y = 0,7199x - 12,942
2
R = 0,997
DNDC
14
TEX
TEX
12
10
CPX
DMEDNA
DIGEN
TNAZ
y = -0,0281x + 9,0479
2
R = 0,9877
DMEDNA
TETRYL
ORDX
y = 0,2185x - 4,7043
2
R = 0,9999
TETROGEN
HNIW
TNAD
HOMO
HMX
y = -0,2156x + 9,7002
2
R = 0,9721
HNIW
RDX
DECAGEN
RDX
y = -1,2206x + 57,207
2
R = 0,9999
0
20
25
30
35
40
45
50
55
-1
Fig. 6 Relationship between EES values from RDAD instrument and heats of fusion.
803
Czech Republic
2006
[Content]
700
y = -45,4x + 1712,7
2
R = 0,9775
y = 22,887x - 533,86
2
R = 0,9951
600
TETRYL
TNAD
TETROGEN
500
HOMO
CPX
400
HNIW
HOMO
y = 2,6726x + 349,05
2
R = 0,938
HNIW
DMEDNA
CPX
300
DECAGEN
200
TEX
ORDX
HMX
RDX
RDX
DNDC
100
TNAZ
0
20
25
30
35
-1
40
45
50
Figure 7: Relationship between EES values from ESZ KTTV instrument and heats of fusion.
Figs. 6 and 7 document an interrelationship of the EES values and intermolecular
interactions which are represented here by heats of fusion, Hm,tr, of the studied nitramines.
This approach was successfully applied in the study of influence of intermolecular
interaction on the impact sensitivity of polynitro compounds [15,39-41]. The heat of fusion
is defined as a sum of all heats of the polymorphous transitions and of the heat of melting
[34] and represents the work needed for formation of defects in crystal lattice. Fig. 6,
analogous to Fig. 4, documents again complexity of the initiation in the spark gap of the
RDAD instrument.
Fig. 7 reminds of analogous relationships between impact sensitivity (taken as the
first reaction) and the Hm,tr, values [15,39,41] of nitramines. However, these relationships
strongly differentiate cyclic and linear nitramines [15,39,41] while according to the Fig. 7
this differentiation is not evident. It might mean that a role namely of plastic deformations of
crystal and perhaps of dislocations in energetic crystals doesn't need to have such
signification in the initiation by electric spark as in the case of impact or shock. But the
dislocations should be a reason of existence of the relationships in Fig. 7. That means that
results from measurements by means of SEZ KTTV instrument should be more depended on
the shape and size of crystals as in case of RDAD. Satisfactory correlation of the data from
the first mentioned instrument with characteristics of nitramine molecular structure might be
caused by wide spectrum of their granulometry (given by method of samples crystallization)
which can leads to the averaged EES values.
804
Czech Republic
2006
[Content]
20
18
DPT
DNDC
16
AcAn
Spark Energy / J
14
OCPX
y = 1,1439x + 45,214
2
R = 0,9794
12
y = 8,0187x + 281,68
2
R = 0,9896
CPX
10
ORDX
TETROGEN
DIGEN
TNAZ
OHMX
HNIW
y = 0,4859x + 19,278
2
R = 0,9576
DECAGEN
HMX
RDX
2
0
-43
-38
-33
-28
-23
-18
15
Fig. 8 Relationship between EES values from RDAD instrument and 15N NMR chemical
shifts of nitrogen atoms of the primarily split off nitro groups (taken from [15,41].
700
y = 41.832x + 1662.5
2
R = 0.8606
600
TETRYL
y = -8.379x + 152.35
2
R = 0.9731
HNIW
500
TETROGEN
HOMO
CPX
400
DMEDNA
CPX
TEX
300
DECAGEN DECAGEN
ORDX
HMX
HMX
RDX
RDX
200
DNDC
y = -6.2253x + 30.052
2
R = 0.6326
100
TNAZ
0
-45
-40
-35
-30
15
-25
-20
-15
Figure 9: Relationship between EES values from ESZ KTTV instrument and 15N NMR
chemical shifts of nitrogen atoms of the primarily split off nitro groups.
10
805
Czech Republic
2006
[Content]
Completely mutually different are data in the Figs. 8 and 9 which represent
relationships between 15N NMR chemical shifts of nitrogen atom in the most reactive nitro
groups (i e., primarily split off nitro groups) and the EES values. Fig. 8 somewhat reminds the
relationship between these shifts and square of detonation velocities of nitramines [15,41].
On the other hand, relationships in Fig. 9 go near to those in Fig. 5. It seems the outputs from
ESZ KTTV instrument have more close relation to molecular structure as those from RDAD
instrument.
4.
CONCLUSIONS
From the facts given above it possible to come to the conclusion that the mechanisms
of spark energy transfer into the reaction centre of molecule should be different in
measurement on the two alternative instruments (RDAD and ESZ KTTV). The thermal
component of discharge in the RDAD instrument may be one of the reasons of this
difference. This assumption is also supported by the differences (in order of magnitude)
between the EES values obtained from the individual instruments (see Table 1). Further, the
considerable electric energy losses in the area between the upper electrode and the sample
surface in the RDAD system measurements contribute to the difference mentioned above
most likely a large part of discharge is enacted in the air.
Data from ESZ KTTV instrument have close relation to molecular structure of
measured polynitro compounds [16]. In line with knowledge about initiation reactivity of
polynitro compounds [15,41] the EES values correlate with such characteristics of molecular
structure which correspond to the primarily leaving nitro group in the nitramine molecule.
Reversed relationship between the EES values and dissociation energies DH(N-N) (Fig. 3)
waits for a detailed, preferably quantum chemical interpretation so far. The EES values also
correlate with heats of fusion of the studied nitramines, probably due to influence of
dislocations in their crystals on electric spark sensitivity. This assumption can have some
relation to significant influence of the shapes and size of the crystals on electric spark
sensitivity of the energetic compounds in general.
Acknowledgements
The work was carried partly out under the auspices of the Ministry of Education,
Youth & Sports of the Czech Republic as a part of its research project No. MSM
0021627501 and partly under support from the Ministry of Industry & Commerce of the
Czech Republic within the framework of research project No. FC-M2/05.
5.
REFERENCES
[1]
[2]
M. ROUX, M. AUZANNEAU AND C. BRASSY, Electric Spark and ESD Sensitivity of Reactive
Solids. Part. I: Experimental Results and Reflection Factors for Sensitivity Test
Optimization, Propellants Explos. Pyrotech. 18, 317-324 (1993).
[3]
M. AUZANNEAU, M. ROUX, Electric Spark and ESD Sensitivity of Reactive Solids. Part. II:
Energy Transfer Mechanism and Comprehensive Study on E50, Propellants Explos.
Pyrotech. 20, 99-101 (1995).
806
Czech Republic
2006
[Content]
[4]
[5]
R. RAT, M. ROUX, J. P. CHAUMAT, Etudes theoriques sur les phenomenes dinitiation des
matieres explosibles, in M. ROUX (Ed.) [Proc.], Recueil des communications Journees
detudes sur la sensibilit des composants et des substances nergt. lelectricite statique,
Aussois, Mai 1996, pp. 147-168.
[6]
J. P. CHAUMAT, Linitiation est elle thermique? in M. ROUX (Ed.) [Proc.], Recueil des
communications Journees detudes sur la sensibilit des composants et des substances
nergt. lelectricite statique, Aussois, Mai 1996, pp. 173-179.
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
12
807
Czech Republic
2006
[Content]
[21]
V. ZEMAN, S. ZEMAN, Relationship Between the Electric Spark Sensitivity and Detonation
Velocities of Some Polynitro Compounds,[Proc.] 28th Int. Annual Conference of ICT,
Karlsruhe, June 1997, pp. 67/1-67/10.
[22]
[23]
[24]
Special Scientific and Development Activity in the Field of Energetic Materials, Project No.
FC-M2/05, Ministry of Industry and Commerce, Prague, 2000-2002, solver Univ. Pardubice.
[25]
[26]
[27]
[34]
S. ZEMAN, Some Predictions in the Field of the Physical Thermal Stability of Nitramines,
Thermochim. Acta 302, 11-16 (1997).
[35]
[36]
[37]
[38]
Z. JALOV, S. ZEMAN, M. SUESKA, P. VVRA, K. DUDEK, M. RAJI, 1,3,3Trinitroazetidine (TNAZ). Part I: Syntheses and Properties, J. Energet. Mater. 19, 219-239
(2001).
[39]
808
Czech Republic
2006
[Content]
[40]
[41]
14
809
Czech Republic
2006
[Content]
Abstract
Attention was focused on the influence of grain size upon the electric spark sensitivity of 1,3,5trinitro-1,3,5-triazinane (hexogen) and 1,3,5,7-tetranitro-1,3,5,7-tetrazocane (octogen) and on the
dependence of this sensitivity upon the content of hard and fine admixtures to these compounds. It
was found that decreasing of the sensitivity corresponds with the increase of the grain size of
corresponding crystals; decreasing of a number of the intergrain points (hot spots) in the volume
unit of these nitramines can be reason in this case. It was suggested that dislocations in the crystals
should play some role here. Electric spark sensitivity of both the compounds also will decrease if
hard and fine particles (hot spots) are added. Also this fact should correspond with a number of
intergrain points of the nitramines particles in the corresponding mixtures. It is stated that an
impact component absents in relevant electric discharge.
Keywords: electric spark, grain size, hexogen, octogen, hot spots, sensitivity
1.
INTRODUCTION
Some molecular structural aspects of the spark sensitivity of nitramines were dealt
with in the first part [1] of this series. An assumption was pronounced that significant
influence of shape [2] and size [2-5] of crystals on the electric spark sensitivity might
connect with dislocations in these crystals [1]. The said factors influence a mechanism of
energy transfer of the primary impulse into the reaction centre of the given molecule. At
present there is no unified model of the said mechanism [6].
A probable influence of the above-.mentioned dislocations might have been also
connected with impact component of the electric discharge. However, Stengach confirmed
[7] the absence of this component in his experiment with electric spark sensitivity of lead(II)
azide. This paper deals with an examination ofthe presence or absence of the impact effect in
the electric spark sensitivity measurements. Hexogen (1,3,5-trinitro-1,3,5-triazinane) and
octogen (1,3,5,7-tetranitro-1,3,5,7-tetrazocane) are used as the studied samples in which hot
spots are artificially introduced. Attention is paid also to the influence of the crystal size of
both the nitramines on their electric spark sensitivity.
810
2.
EXPERIMENTAL
2.1.
Apparatus
Czech Republic
2006
[Content]
Nitramines
The following samples were tested: hexogen; produced by Chemko Strske in seven
granulometric qualities according to MIL standard R-398C (1977)} and octogen of foreign
provenance in five granulometric qualities according to MIL standard H-45444B (1978)}.
The individual classes of these nitamines are characterized in the Table 1 by their average
grain size (the mean value of limit magnitudes in the given class). This Table also presents
the results of measurements. Also a sample of Dyno Nobels hexogen RDX-RS (grain sizes
from 50 up to 630 m) was used.
Table 1:
RDX
Mean value of
grain size in m
600
450
350
250
150
75
30
2.3.
EES
in mJ
1170
1060
900
535
215
137
55
HMX
Mean value of
grain size in m
1000
400
150
40
27
EES
in mJ
875
495
242
152
88
Hot spots
In the nature of the hot spots are meaning fine and hard particles, here crushed
glass and synthetic diamonds dust of Russian provenance [12]. For evaluation of the
influence of the crushed glass on the electric spark sensitivity, the lowest-grain hexogen (see
Table 1) was mixed with this additive (with grain size below 20 m). Analogue influence of
diamonds dust, whose particle size (its agglomerates) does not exceed 5 m, was evaluated
by means of hexogen RDX-RS. Mixtures containing various amounts of these hard particles
were then measured: composition of the mixtures and the results of measurements are
presented in Table 2.
811
Table 2:
Czech Republic
2006
[Content]
Crushed glass
content in %
wt.
0
5
10
15
EES
in mJ
55
120
180
200
3.
DISCUSSION
3.1.
Diamonds dust
content in %
wt.
0
5
10
15
20
EES
in mJ
291
430
560
706
830
The relationship between octogens granulometry and its sensitivity to electric spark
has been dealt with recently by Roux et. al. [2, 4]. Our results obtained by measurement on
the ESZ KTTV instrument are given in Fig. 1. In contrast to the results obtained by Roux et.
al. [2, 4], our dependences have opposite trend. The difference between our and Roux et. al.
finding might be due to mutual different construction of the apparatus circuits and spark
gaps in their and our cases. However, we suppose that our finding confirms the statement by
Auzanneau et. al. [4]: hot spots develop at the thinnest part of the solid, i.e. at intergrain
points. The number of these points in volume unit decreases with the increasing grain size,
and that increases the corresponding EES values. However, on the basis of relationships
between the EES values and heats of fusion of nitramines [1] we should suppose that also
dislocations in the crystals have some effect on the initiation of these compounds by spark.
3.4.
Stengach studied the electric spark sensitivity of lead(II) azide and found out that an
addition of the hard inert admixtures to this azide (up to 10% by wt.) has practically no effect
on its electric spark sensitivity [7] (while the impact sensitivity is increased). We have
chosen a similar strategy in the measurements. The sample measured was hexogen of mean
grain size equal to 30 m with an admixture of crushed glass and hexogen RDX-RS with
admixture of diamonds dust. The results are given in Fig. 2, and are in accordance with
those given by Stengach [7], stating the absence of effect of impact component in the electric
discharge used. The trend shown in Fig. 2 can be interpreted by the introduced glass or
diamond particles separating the hexogen grains from each other. Thereby the number of the
intergrain contact points of hexogen or octogen grains in volume unit is decreased, and
hence the electric spark sensitivity of the respective mixture is decreased, too. This statement
again stands in accordance with the above-mentioned view by Auzanneau et. al. [4].
812
Czech Republic
2006
[Content]
1400
1200
RDX
Spark energy / mJ
1000
HMX
800
600
400
200
0
0
200
400
600
800
1000
1200
Grain size / m
Figure 1: Relationship between the grain size and electric spark sensitivity of hexogen
(RDX) and octogen (HMX).
900
diamond
800
700
600
500
400
300
crushed glass
200
100
0
0
10
15
20
25
Figure 2: Electric spark sensitivity of mixtures of the hexogen with crushed glass or
diamonds dust.
813
4.
Czech Republic
2006
[Content]
CONCLUSION
Decreasing of the electric spark sensitivity (i. e. increasing of the EES values), when
grains size of nitramines crystals increase, confirms the idea by Auzanneau et. al. [4] about
the mechanism of spark energy transfer into the powdered reactive solid, i. e. in this case to
decrease the number of intergrain points in volume unit. However, also dislocations in the
crystals should have some effect on this type of energetic materials initiation [1].
Artificially introduction fine and hard particles in the crystalline nitramines decreases
the electric spark sensitivity of the resulting mixtures, i. e. the particles behave as
flegmatizing additive. This fact also agrees with the idea by Auzanneau et. al. [4] and
confirms the Stengachs stating [7] about the absence of the effect of the impact component
in the electric discharge at determination of the electric spark sensitivity.
Acknowledgements
The work was carried partly out under the auspices of the Ministry of Education,
Youth & Sports of the Czech Republic as a part of its Research project No. MSM
0021627501 and partly under support from the Ministry of Industry and Commerce of the
Czech Republic within the framework of research project No. FC-M2/05.
5.
REFERENCES
[1]
[2]
M. AUZANNEAU, M. ROUX, Electric Spark and ESD Sensitivity of Reactive Solids. Part. II:
Energy Transfer Mechanism And Comprehensive Study on E50, Propellants, Explos.,
Pyrotech. 20, 99-101 (1995).
[3]
[4]
M. AUZANNEAU, M. ROUX, Electric Spark and ESD Sensitivity of Reactive Solids. Part. II:
Eergy Transfer Mechanisms and Comprehensive Study on E50, Propellants, Explos.,
Pyrotech. 20, 96-101(1995).
[5]
[6]
R. RAT, M. ROUX, J. P. CHAUMAT, Etudes theoriques sur les phenomenes dinitiation des
matieres explosibles, in M. ROUX (Ed.), Recueil des communications Journees detudes sur
la sensibilit des composants et des substances nergt. lelectricite statique. Aussois, Mai
1996, pp. 147-168.
[7]
[8]
814
Czech Republic
2006
[Content]
[9]
[10]
[11]
Special Scientific and Development Activity in the Field of Energetic Materials, Project No.
FC-M2/05, Ministry of Industry and Commerce, Prague, 2000-2002, solved by University of
Pardubice.
[12]
815
Czech Republic
2006
[Content]
1.
INTRODUCTION
In recent years there has been a significant effort to develop new candidates which might
lead to improved explosives in terms of performance, safety and other favorable properties.
FOX-7 is an energetic material with high performance approaching that of RDX, but with
low sensitivity comparable to that of TATB. For this reason FOX-7 is an explosive receiving
a good deal of attentions. It has a density magnitude of 1.878g/cm3. Very much has been
studied concerning FOX-7. We were aware that the synthesis had been reported [1] and the
crystal structure had been determined [2, 3]. The activation energy values scatter in papers of
different authors from 50 to 90kcal/mol [4, 5] at temperature range 200-250.
An understanding of the decomposition mechanism, however, also plays a key role in
the evaluation of an explosive. In this investigation of the decomposition mechanism of
FOX-7, we have focused on discovering the most possible initial decomposition pathway of
FOX-7. We have used the ab initio molecular dynamics simulation to investigate the
possible means of decomposition, then to demonstrate them by means of energy barrier and
rate constant calculations.
2.
COMPUTATIONAL DETAILS
syjfree@sina.com or syjfree@sohu.com
816
Czech Republic
2006
[Content]
3.
H2N
C
H2N
H2N
NO2
C
NO2
H2N
O
N
O
pathway 2:
817
H2N
NO2
H2N
H2N
C
O
ON
C
H2N
H2N
Czech Republic
2006
H2N
C
H2N
NO2
[Content]
O
N
O
C
NO2
O
C
NO2
C
H2N
pathway 3:
C
H2N
H N
NO2
H2N
N
C
C
H2N
NO2
H O
H N
C
NO2
H2N
N
C
C
NO2
H N
C
H2N
H O
NO2
N O
Fig 1. A scheme of pathways for the main three initial decompositions of FOX-7
3.2
Quantum mechanical calculations were performed for the three pathways respectively.
The structures and optimized geometrical parameters of all species involved were obtained,
as well as their corresponding absolute energies.
3.2.1. C-NO2 dissociation.
We have calculated the reactant, intermediates and transition state of C-NO2
dissociation reaction. The scheme of the pathway is shown in figure 2. The total energy
and zero-point correction energy were tabulated in table 1.The main geometry
parameters are shown on the molecules. The energetic profile of the pathway can be seen
from figure 3.
3.06
NO2
H2N
1.43
1.34
1.43 C
C
1.34
H2N 1.43
NO
H2N
1.36
NO2
1.39
C
C 1.38
1.37
H2N
P1
NO2
H N
C
H2N
H 0.97
O
1.46
1.29N O
C 1.23
1.45
TS1
P2
H N
1.29
1.38C
H2N
1.45
H
O 0.97
1.43
1.29 N
O
1.24
C
M1
R
P1
P2
TS1
M1
point Energies(Ha)
-598.232973
-393.062668
-205.63380
-393.031713
-393.034178
Czech Republic
2006
[Content]
correction(Ha)
0.092451
0.077013
0.008826
0.074121
0.074791
P1+P2
TS1
70.13
1.97
21.24
R
M1
P1
Fig 3. The energetic profile of FOX-7 decomposition via C-NO2 dissociation reaction
pathway. The units are kcal/mol. The energy barriers are uncorrected for ZPVE.
The nitro dissociation from FOX-7 molecule has no transition state which has been
verified by scanning the potential energy surface of C-NO2 bond. It is can be seen from fig
4.
total energy vs C-NO2 bond length
-598.12
-598.14 0
-598.16
total energy
-598.18
-598.2
-598.22
-598.24
-598.26
-598.28
-598.3
-598.32
-598.34
C-NO2 bond length
Fig4. Potential energy surface of C-NO2 (the units are Ha and for energy and bond
length respectively)
The hydrogen migration from nitro to the amino for product P1 may take place through
transition state TS1 and then produce M1. The transition state TS1 has an imaginary
frequency of -307.5. IRC analysis verified it is the transition structure connecting P1 and
M1.The energy barrier for P1 is not high so the product may easily react after nitro
cleavage.
3.2.2. Nitro-to-nitrite rearrangement
We have calculated the reactant, intermediates and transition state of rearrangement
reaction. The scheme of the pathway is shown in figure 5. The total energy and zeropoint correction energy were tabulated in table 2.The main geometry parameters are
4
819
Czech Republic
2006
[Content]
shown on the molecules. The energetic profile of the pathway can be seen from figure 6.
1.16O
N
1.34
O
O 1.23
H2N
N
1.80
C
C 1.53
1.40
1.42 NO
H2N
2
NO2
H2N
C
H2N
NO2
1.58
H2N
C
H2N
M1
TS1
O
1.46 1.25
C 1.46
H N
N 1.22
O1.26 O
H
H2N
H2N
1.50
O
1.23
M2
H2N
C
1.23
H N
1.33 N O
1.25 H
O
1.21
NO
1.36O
C
1.39 1.41
NO2
H N
1.23
N O
H O 1.37
1.03
M4
TS2
M3
H2N
O
C
H N
M5
H O
N O
M6
TS1
M5+M6
62.66
64.94
TS2
M2+M3
1.91
M1
37.99
1.40
4.64
M4
M2
Czech Republic
2006
[Content]
has an imaginary frequency of -819.1. The TS2 has an imaginary -1091.2. Vibration and
IRC analysis conformed they are the true transition states.
3.2.3. Hydrogen migration from amino to nitro
We have calculated the reactant, intermediates and transition state of rearrangement
reaction. The scheme of the pathway is shown in figure 7. The total energy and zeropoint correction energy were tabulated in table 3.The main geometry parameters are
shown on the molecules. The energetic profile of the pathway can be seen from figure 8.
H2N
C
NO2
H N
NO2
H2N
0.98
1.22
H2N
2.79 H 0.97
O1.43
O
1.33 N
1.50
H N
1.45
NO2
H2N
TS1
O1.40
O
1.32 N1.23
C 1.44
NO2
M1
H N
C
H2N
NO2
H O
M2
N O
M3
R
TS1
M1
M2
M3
M2+M3
58.32
TS1
2.24
30.18
M1
Fig 8. The energetic profile of FOX-7 decomposition via hydrogen translation from amino
to nitro pathway. The units are kcal/mol. The energy barriers are uncorrected for ZPVE.
The structures and optimized geometrical parameters of all the species involved were
obtained. The TS1 has an imaginary frequency of -365.5.
3.3
6
821
Czech Republic
2006
[Content]
In the molecular dynamics simulations, we have found that C-NO2 bond dissociation is
the main initial decomposition pathway. But the quantum calculation results are not
consistent with it, but C-NO2 bond dissociation energy is only about 7kcal/mol higher than
the energy barrier of nitro-to-nitrite rearrangement reaction. So we calculated these reaction
rate constants to confirm the initial decomposition pathway. The rate constants of three
decomposition pathways are shown in table 4.
Table 4 The reaction rate constants of FOX-7
temperatures
rearrangement
C-NO2
T(K)
cleavage
forward
backward
400
1.388E-18
3.474E-20 4.770E-23
450
9.623E-15
1.770E-16 3.672E-19
500
1.158E-11
1.654E-13 4.733E-16
550
3.877E-09
4.492E-11 1.663E-13
600
4.956E-07
4.825E-09 2.203E-11
650
3.002E-05
2.536E-07 1.379E-09
700
1.006E-03
7.595E-06 4.787E-08
backward
4.587E+11
6.435E+11
8.452E+11
1.058E+12
1.276E+12
1.497E+12
1.718E+12
750
800
850
900
950
1000
2.091E-02
2.940E-01
2.994E+00
2.330E+01
1.448E+02
7.441E+02
1.450E-04
1.919E-03
1.878E-02
1.428E-01
8.787E-01
4.512E+00
1.037E-06
1.531E-05
1.648E-04
1.363E-03
9.034E-03
4.958E-02
2.755E+05
9.708E+05
2.951E+06
7.934E+06
1.923E+07
4.267E+07
1.936E+12
2.151E+12
2.360E+12
2.564E+12
2.762E+12
2.953E+12
1100
1200
1300
1400
1500
1.241E+04
1.282E+05
9.208E+05
4.986E+06
2.153E+07
7.631E+01
8.074E+02
5.953E+03
3.303E+04
1.460E+05
9.402E-01
1.093E+01
8.725E+01
5.180E+02
2.427E+03
1.692E+08
5.337E+08
1.411E+09
3.249E+09
6.695E+09
3.317E+12
3.655E+12
3.968E+12
4.259E+12
4.529E+12
From table 4, we can find that the rate constants of C-NO2 cleavage are approximately
100 times to that of the arrangement reaction at the same temperature. The rates are so high
for hydrogen migration but backward rate is much higher and the total energy barrier from R
to M1 and M2 is over 80kcal/mol, so this pathway is not the dominant initial decomposition
pathway.
4.
CONCLUSION
In this paper the calculations made and the conclusions drawn are based on ab initio
molecular dynamics, DFT methodology and chemical kinetics. Using AIMD simulation, we
observed C-NO2 bond dissociation is the dominant trajectories. Quantum mechanical
calculations were performed for the three main initial decomposition pathways. The energy
barrier of hydrogen migration from amino to nitro is the lowest, but its intermediate has high
energy and its consequent reaction requires comparatively high energy. The BDE of C-NO2
is about 7kcal/mol than the energy barrier of nitro-to-nitrite rearrangement reaction at
B3LYP/6-31G(d, p) level without zero-point correction. The chemical kinetics calculations
indicate that the rate constant of C-NO2 cleavage is higher than that of rearrangement
822
Czech Republic
2006
[Content]
reaction, and the rate constant of hydrogen migration is very high but that of backward
reaction is much higher. So the results observed and calculated can make out that C-NO2
bond cleavage is the most important initial decomposition pathway. Its bond dissociation
energy is 70.13kcal/mol uncorrected for ZPVE at B3LYP/6-31G (d, p) level.
REFERENCES
[1] Nikolai V LatypovJan Bergman. Synthesis and reactions of 1,1-diamino-2,2-dinitroethylene.
Tetrahedron1998(54)11525-11536.
[2] Bemm, U., and stmark, H. (1998), .1,1-Diamino-2,2-dinitroethylene: a novel energetic material
with infinite layers in two dimensions,. Acta Crystallographica C54, 1997-1999.
[3].Gilardi, R Crystallographic data (excluding structure factors) for the DADNE structure have been
deposited with the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2
1EZ,U.K.(deposition number CCDC127539).
[4] stmark H.,Langlet A.,Bergman H.,Wingborg,N.,Welmar U. and Bermm
U.(1998),FOX-7-A new explosive with low sensitivity and high performance
Proc.Eleventh(Int.)Det.Symp. Preprints1821.
[5] V.P.Sinditskii,A.L.Levshenkov,V.Y.Egorshev,v.V.Serushkin
Study on Combustion and Thermal Decomposition of 1,1-Diamino-2,2-dinitroethylene(FOX7)//8th International Seminar EUROPYRO2003,Saint Molo,France,2003.
[6] Car, R.; Parrinello, M. Phys. Rev. Lett., 1985, 55, 2471.
[7] Payne, M. C.; Teter, M. P.; Allan, D. C.; Arias, T. A.; Joannapoulos, J. D. Rev. Mod. Phys.,
1992, 64, 1045.
[8] Kresse, G.; Hanfner, J. Phys. Rev. B, 1993, 47, 558.
[9] Kresse, G.; Hafner, J. Phys. Rev. B, 1994, 49, 14251.
[10] Kresse, G.; Furthmller, J. Comput. Mat. Sci., 1996, 6,15.
[11] Kresse, G.; Furthmller, J. Phys. Rev. B, 1996, 54, 11169.
[12] Vanderbilt, D. Phys. Rev. B, 1990, 41, 7892.
[13]
[14]
[15]
[16]
[17]
Gaussian 98, Revision A.3, M. J. Frisch etal., Gaussian, Inc., Pittsburgh PA, 1998.
Becke, A. D. J. Chem. Phys.1993,98,5648.
Lee, C.; Yang, W. ; Parr, R. G. Phys. Rev. B1988, 37, 785.
Zhang S, Truong T N. VKLab. Version 1.0 ed.; University of Utah, 2001.
Politzer, P., Concha, M. Grice, M., Murray, J., Lane, P., and Habibollazadeh, D.(1998)
.Computational investigation of the structures and relative stabilities of amino/nitro derivatives
of ethylene,. Journal of Molecular Structure (THEOCHEM) 452, 75-83.
[18] Sta Gindulyte and Lou Massa. Proposed Mechanism of 1,1 Diamino Dinitroethylene
Decomposition A Density Functional Theory Study. J. Phys. Chem. A 1999,103,11045-11051.
823
Abstract
Czech Republic
2006
[Content]
E.N. Egorychev, V.N. Lashkov, A.V. Strikanov, A.A. Tikhonova, V.I. Rybakov
RFNC-VNIIEF, Sarov, Russia
Abstract:
Ammonium nitrate (AN) is the basic component of explosive compositions (EC) in
explosion welding technology. During mix preparation of the explosive composition,
ammonium nitrate is reduced to fine particles in ball mill and then mixed with
trinitrotoluene to obtain a composition with 5%-mass content of trinitrotoluene.
Actually, we are carring out studies to obtain EC without trinitrotoluene to utilize its in
explosion welding technology.
The main requirements to such EC are:
Detonation initiation by standard detonator
Detonation velocity at charge porosity of 50% is not more than 1500 m/s
Critical thickness is not more than 20mm, slight dependence of detonation velocity on
thickness
We propose the following approach of obtaining satisfactory values in the factors:
Grinding should be more fine, with control of the degree of grinding;
Additives of dispersed metal powder are to be used.
Crystalline structure activation by polymorfic transformation.
In this report we present the results of thermal analysis of commercially available AN
and finely dispersed AN. We determined the parameters of polymorfic transformations
(direct and inverse). We also determined effect of aluminum and titanium powder
additives on thermal decomposition process of ammonium nitrate. It was shown that
inverse polymorphic transitions are running without a delay which means, that the
phase structure of trial sample is recovered after cooling.
Partical size of metal additives affects considerabled on the process of thermal
decomposition of ammonium nitrate. For example, coarse titanium powder that has
particle size more than 20 micrometer decreases heat effect of AN thermal
decomposition while dispersed titanium powder (particle size less than 5 micrometer)
increases it.
825
Czech Republic
2006
[Content]
Author Index
Adam Martin 613
Adamk Vladislav 451
Alekseev Alexander 293
Aleshkina Ekaterina 464
Aleynikov Aleksey 304
An-Min Tan 356
Annikov Vladimir 464
Anusevicius Zilvinas 285,
689
Apolenis Alexei 464
Arinina Svetlana 473
Armstrong K. 42
Astachov Alexander 480
Astratev Alexandr 377
Backofen Joseph 76
Bachurina Irina 143
Bajpai Rakesh 251
Bakhmatova Eugeniya 90
Baka Sanko 214
Barbu Cristian 756
Bazrafshan Esmaeil 393
Bellamy Anthony 491
Bladek Jan 500
Blagojevic Djordje 513
Boniuk Henryka 279
Bozic Vladica 513
Bratowski M. 742
Buczkowski Daniel 507
Buka Eduard 480
Burov Yurii 98, 526
Bystrzejewski Michal 544
Cartwright Michael 107
Catovic Alan 780
Cenas Narimantas 285, 689
Clment Dominik 540
Cudzilo Stanislaw 500, 544
Cumming Adam 11
Czerski Helen 118
Dabkowski A. 742
Darwich Chaza 551
Dashko Dmitry 377
Dimitrov A. 416
Dobrilovi Mario 581
Duda Maciej 561
Dudek Kamil 608
Egorova Irina 569
Egorshev Valerii 314
826
723
Czech Republic
329
Tasev M. 416
Teipel Ulrich 135
Teodorczyk Andrzej
2006
[Content]
742
827
UP 06-19
ISBN 80-7194-849-7