r e

l d
Produktions-undVerarbeitungsstufen,wobeisichdieA]rtundarev61edesimpuretespolymolphesdanstouteslesetapesdeproduc-

U]rSacheflirdieAbnahmede]rDichtewurdediethermischausgel6stetiondutraitementdumateriau.onadetemlineque1aconverslon

ErgebnisdesBinderabbausposmliertwurde.Schlagfestigkeit,Rei- Iadensiteetonasupposequelechangementdecouleuretaitle]resul-
bungundeleklrostatischeFunkenbildungwu]rdennichtvemnde]rt. tatdeladegradationduliant.LesresistancesAI'impact,aufrottement
etauxetincelles61ectrostatlquesn'ontpasetemodifiees.
8 u m m a r y
AchangemcoloranddensiC/duringprolongedheatingofafblmu-

Estane5702wasinvestigatedbyFouriertransfblminfra]red(FTm)
spectroscopicanalysis.Startingmaterials,virginandfinalcoloredfOr-

mulationwerestudied.Analysisrevealedpolymolphicimpu]mtyatall
E stagesofproductionandprocessing,thenamreandconcemrationof
whichchangedwithtreatmentofthematerial.Thermally-induced
pOIymo][phicconverslonwasdeterminedtobethecauseofthe
decreaseindensic/,whereasthecolorchangewasspeculatedtobethe

staticspalkwereunchanged.

behind the twm problems in HE fbrmulations of unwanted
crystal growth and phase conversion.
One polymomhic HE of recent interest, the polycyclic
nitramine 2,4,6,8,10,12-hexanitrohexaazaisowurtzltane
(mJIW)(1), has been the fbcus of research investigating its
polymorphs, their relative stability(2-5), and their thermal
reaction chemistry(6). The basic molecular strucmre consists
of a ]jgid carbon isowurtzitane cage modified by the substi-

group (Fig. 1). Both changes in spatial onentation of the
nitro groups about the five- and six-membered rings and
differences 1n crystal 1attice packing distinguish the four
different polymorphs that have been isolated so far. one of
the first polymorphs discovered, beta phase (orthorhombic
space group Pb2,a), is the least stable with respect to poly-
H E XA N I T R O H EXAAZA I SOW U RTZ I TA N E

O 2 N
O 2 N
- N O 2
- N O 2
- N O 2
Figul.e 1. The hexaninPohexaazaisowurtzltane, (HNIW) molecule
showing the rigid isowurtzitane cage with nill.o groups attached to the
bridging nitrogens. Changes m spatial orientation of the nin.o groups
about the cage, and differences in packing in the clystal 1attice, define
the four known polymolphs.

Itisimportantto]mowthesolubi]ityofhighexplosives
(HE)infblmulationbindersandsolventsinordertocontrol
thestabilityandsensitivityoffb][mulatedsystemsasthey
e age.Withsolubilitycomescrystalsolvationandgrowth:the
greaterthesolubility,thegreaterthechangeintheparticle
lss sizedistributionwithtime.Inaddition,fOmmlationsbased
0napolymo][phicHEcanwimessphaseconversiontome-
tastablepolymolphswhenstoredatatemperatureabove
0f thatofconversion.Temperaturecyclingexacerbatesrecrys-
tallizationof1heHEintometastablepolymorphs.Thiscon-
'versiontoahightemperaturestable,lowerdensitypoly-

metastablephasewi]1causeundesirablevolumeexpansion



I'PEPYD5,PropellantsExplosPyrotech. PEPYD5 , Propellants Explos Pyrotech. 1 9. 1 994 No 3 1 13-158 . ISSN 0721-31 15

I''
I



1y/strucmrally different from the alpha and gamma phases,

ferently into the unit cell (orthorhombic Pbca, and mono-
clinic P21/C, respectively). The alpha crystal lattice is stabi-
1ized by the addition of H20 to the unit cell, with X-ray

molecules(1a,7). The most recently discovered polymorph,
the epsilon phase (monoclinic space group P21/c), is yet
another strucmrally different fbrm of HNIW. The epsilon
phase has been shown to be the themlodynamically stable
one at ambient conditions(2,5). In addition to the three
known alpha/gamma, beta and epsilon confOmlers, a high
pressure zeta phase has been discovered arising from a
reversible transition fi.om gamma phase(3a). The polymorphs

mmphs can be easily distinguished by their FTIR Spectrum

There have been several smdies investigating the solu-
bility(1a,8) and polymo]Fhic conversion(9) of HNIW. The
solvents and binders studied have included those commonly
used in fOmmlation processing. Not included are materials
whose high viscosity defy standard solubi]ity testing meth-
ods involving solute filtration. one such viscous material
used m a number of fOnnulations is the polyester co-poly-
mer Estane [poly(urethane-ester-MDI), trade name Estane
5702, Goodrich; Fig. 2] . This polymer is known to undergo
hydrolytic degradation. This decomposition is evidenced by
reduction in intrinsic viscosity due to hydrolysis of the ester
linkage and fmgmentation into 1ow molecular weight seg-

reactivity, is also susceptible to the effects of water. Some
deterioration will invariably occur during storage, the
degree of which will depend on precautions taken to limit
exposure to atmospheric moisture. Since it is impractical to
obtain fresh binder before each formulation process, the
EStane,, that actually goes into a particular formulation
will have a somewhat different fOnnula than that shown in
Fig. 2.
ESTANE5702

Figul.e 2. The polyester co-polymer Estane, or poly(urethane-estel.-
MDI) (1rade name Estane 5702) used m the RX-39-AC fOnnulation.
The highly reactlve end-chain isocyanate group is susceptible to the
effects of water.
Thennal degradation is a second problem with Estane
fO]rmulations. Progressive yellowing and some loss of bind-
er resilience from prolonged heating have been seen in
other HE/EStane fOrmulations(11). The decomposition mech-
anism probably involves both the reaction of the isocyanate
and hydl.oxyl ends of neighboring chains and the cleavage
of these nascent segments fiFom the 1onger repeat umt. The
minimal amount of thelmal degradation seen in HE/EStane
formulations during processing has thus far not been con-
sidered detrimental to the binding properties of the Estane.






cause of the grey coloring of the PBX. There were threel smdJ'


r
ll,:







)uring water. The standard slurry coating formulation process
ven at included heatmg the HNIW/Estane mixture to 60C, above
;haped the lowest measured temperature for HNIW polymolphic
corre_ Phase transition(5). The polymo:[phic purity of the as-
period received IEllW was therefbre detemlined fbr baseline com-
ty, the parison with as-made (b]riefly heated) fblmulation and fbr-
1 ther- mulation heated for an extended period of time. Because of
d the the beaded nature of the three PBX batches, and the batch-
,dy the to-batch color valiation, pressed parts of the blended RX-

,anges ance. After prolonged heating, the appearance was that of
,mples lightgreen-dalkgreenmottling.
ipuri_ Fouriertransfbnninfmred(FTIR)spech.oscopywasthe

)losive impurity in the e-mJIW. An attempt was also made to iden-
abora_ tify the source of the green color by FTIR spectroscopy.
,1ation Spectra were taken of standard 1 % by weight KBr pellets at
ed and 4.0, 2.0, and 1.0 cm-l resolution with 10 collection scans
sd and using a benchtop Nicolet instrument (model 730). Since
39_AC Sample resolution was best matched with an insmlment res-
d (g) a 01ution of 2 cm-1 in identifying HNIW polymorphic Con-
3apons t[lminants, most of the analysis used this intemlediate set-

)drich, Component of the RX-39-AC mixture was intense, masking

lensity was analysed intact. The HNIW was therefore recovered
;e dur_ fi0m the folmulation by first dissolving the Estane in chlo-
roform,methylenechlorideorethylenedichloride,then
rinsingtheInVIWresidueinmethanol.Thesesolventswere

TheEstanebinderinsamplesthathadbeenpressedwas
velyslowtodissolve.AItematively,attemptstosmdythe
0riginofthegreencolormthebinderinvolveddissolving

IinsingthebinderresidueinmethanoLThecleanedEstane
wasdissolvedinchlorofbrmandappliedasasyruptoKBr
l in the platesforFTmanalysis.
Tostudytheeffectofextendedheatingaloneonthefbr-
mlllation,someoftheC.J.S.91-104batchofRX-39-AC
materialwasheatedforvaryingperiodsoftimepriorto
,estion FTIR analysis. Samples of as-made RX-39-AC fOrmulation
signa- were heated in glass and in stainless steel pans (~5 g each)
. "elevatedtemperamreinsealedovens(FonnaScientific c t l v at -
HNIW 521). Initial heating at 100oC was in a dry oven, but for
as the h6ating longer than 24 hrs a pan of water was added to
: three study the affect of humidity on themlally-mduced changes.

}rs 91_ 0v6ns once every hour for up to eight hours, and the


) were rln extended period of time, heating an additional 72 hrs at
skim- lO5oC in the presence of moisture. AII twemy samples (ten
sta'}e l
.tion in
9 1 - 1 02
2) were
/ skim-
D u r i n g
)ven at
shaped
l corre-
; period
lity, the
:d ther-

udy the
l phase
h an g e s
, amp l e s
ld puri-
lade e-
p 1 o s iv e

u 1 a t i o n
3ed and
ted and
3 9 - A C
ld (g) a
e ap o n s
ght %,
, o dri ch ,
m i n i n g
d e n s i t y
ge dur-
ll e S tl o n
s s i g n a -
d in the
l c t l v at -
H N I W
'as the

t, with



each from glass and steel pans) were checked by FTIR
analysis for HNIW polymorphic pu]rity. In addition, both an
RX-39-AC formulation made with a different 1ot of e-

weight %) were heated in a wet oven at 105C for 96 hrs
for visual observation of color change.
To study the effect of pressure coupled with heating,
samples of the C.J.S. 91-104 batch of RX-39-AC fOmmla-
tion were pressed at different temperatures using a heatable
press (BIackhawk Mfg. Co., Porto-power). Four pellets of
~500 mg each of RX-39-AC were pressed in a 0.5 inch
diameter die set. Pressing conditions dup]icated those of the
problematic green RX-39-AC fbnnulation. The sample and
die Set were preheated to allow for equilibration at the spec-
ified elevated temperamre (75oC, 85oc, 95oC, and 105oC)
prior to pressing. The samples were held under vacuum
(less than 400 Mm Hg) during three pressure cycles at
30,000 psi (208.6 MPa). The pressu]re dwells were 5 min-
utes per cycle with a relaxation period of one minute
between cycles. The pellets were allowed to cool to room
temperature for at 1east two hours before the binder was
dissolved in methylene chloride for polymorphic FTIR
analysis of the HNIW. A similar RX-39-AC pellet was also
pressed from material that had been heated 96 hrs (24 hrs at
100oC plus an additional 72 hrs at 105C).
Safetytestingincludedtheelechnstaticspalktest(1.0J,
500 Q resistance, with pigmented mylar film over sam-
ple)(14), the 50% drop-weight impact sensitivity test (Dh50,
type 12 tool, 2.5 kg weight)(15), and the friction test
(BAM)(16). Both original virgin fOrmulation and some that
had tumed green after 96 hrs heating were tested and their
results compared.
3. Results
At each step in the process of producing, heating, and
pressing the RX-39-AC formulation, some polymorphic
impurity was identified by FTIR spectroscopy in the
E-HNIW. The amount of impurity varied from on the order
of 10% to almost 100%. The concentration was seen to
change from one bead of RX-39-AC to another, and (aver-
age of Several beads) from short heating time to 1ong. The
FTIR spectra revealed the presence of small amounts of
polymomhic impurity in E-HNIW by (1) the appearance of
new, weaker spectral features, and (2) the broadening of
existing spectral feamres where there was insufficient spec-

Table 1 lists the characteristic peaks in the 1200-700 cm-1
range used to distinguish the HNIW polymorphs. The

each step, depending on the treatment of the material. The
as-received C-128 E-HNIW was found to contain primarily
beta polymorph contaminant (on the order of 10%) as
described in Ref. 8a. Figure 4 shows the 2 cm-1 resolution
mid-IR fingerprint, region FTIR spectra of C-128

e-InHW was prepared by stirring the contaminated HNIW
for 26 hrs at room temperature (R.T.) in a mixture of
xylenes plus 30% by volume acetophenone. This solvent

7



Table 1. Characteristic Peaks of HNIW Polymolphs m the 1200-700

Characteristic Peaks for HNIW Polymorphs from 1200-700 [cm-1]
Ep s il o n AIpha
Beta Gamma
705.1
723.9
738.2
744.5
751.4
758.2
820.2
831.6
855.4
883.8
900.3
913.2
938.3
944.4
980.9
998.7
1 022 . 1
1 05 1 . 6
1 0 8 7 . 2
1 1 25. 1
1 139.2
1 1 8 2 . 5
1 1 9 1 .7
7 1 8.0
746.5
75 1 . 1
757.7
764.9
825.3
835.4
860.2
881.7
904.3
945.2
947.7
949. 1
951.7
953.0
989.2
990.7
1 0 5 2 . 4
1 0 7 2 . 8
1 0 7 7 . 0
1 0 7 8 3
1 0 8 0 . 3
1 0 8 2 . 0
1 0 9 4 . 9
1 1 18.3
1 1 198
1 1 2 1 . 1
1 1 68 . 1
1 1 69.6
718.9

766.8

882.9
907.7
944.9
9592
991.8
l 0 5 2 . 3
1 0 9 4 . 7
1 1 54.4
1 1 7 1 . 8
1 1 7 8 . 6
7 1 9 .5
74 1 . 0
755.8
764. 1
831.7
8344
858.3
879.2
8920
909.4
938.1
958.8
970.6
l 043 . 8
1 0 8 0 . 4
1 1 06 . 1
l 1 5 3 .0
1 1 80 .4
mixture was found in an ear]ier thelmal study(5) to increase

ly rapid recrystallization even at R.T. to the thennodynami-
ca]1y stable epsilon polymo]ph. There is modest broadening
and shoulder-smoothing (938.3 and 998.7 cm-1) of some
feamres in the as-received E-HNIW due to beta impurity,
but most pronounced are relative intensity changes in the
low frequency quartet (738.2-758.2 cm-1) and doublet
(83 1 .6- 820 .2 cm- 1) .

1200 1150 110D 1050 10DO 950 900 850 BOO 750 700

Figure 4. The fingerprmt region (2 cm-1) FTIR spectra of (a) as-
received lot C-128 E-HNIW, (b) C-128 E-HNIW cleaned of beta

by peak intensiD/ changes around in the 1ow fTequency quartet (738.2-
758.2 cm-1) and doublet (83 1 .6-820.2 cm-1).

Figure 5 shows the FTIR Speclra of as-received C-128
E-HNIW, HNIW recovered from two samples of RX-39-AC

ples no 1onger have impurity spectral features typical of
beta phase, but do have those indicative of alpha phase.
Again the quartet of 738.2-758.2 cm-1 are affected; in Figs.


feamres characterislic of alpha phase at, fOr example 835.4,
904.3, 990.7, 1094.9, and 1169.6 cm-1, vary in scale in
i]1ustration of the range in alpha phase impunty concentra-
tion in as-made RX-39-AC. The simultaneous disappeal-
ance of beta phase and appearance of alpha phase during
the fbrmulation process indicates a phase conversion in the
mWIW impurity. The reason fbr this conversion 1ies in the
fommlation procedure itself, which included jacket heating
to 60oC dming air sweep removal of the ethylene dichlolide



1200 1150 1100 105D 1000 950 900 B50 800 750 700
wavonumber[cm.5
Figm.e 5. The fingerprint region FTIR spectra of (a) as-received
C-128 e-HNIW, (b) and (c) HNIW recovered from two as-made

alpha not beta impurity in (b) and (c) can be seen m the appearance of
alpha peaks at 835.4, 904.3, 990.6, 1094.9, and 1 169.6 [cm-1].
The FTIR spectra shown in Fig. 6 are those taken of
HNIW recovered from RX-39-AC heated in stainless steel
pans in a sealed oven fbr periods of one to 96 hrs. Included
for comparison are the FTIR spectra of HNIW recovered
fTom as-made RX-39-AC and pure gamma phase InJIW,
The amount of polymolphic impurity does not increase
monotonically from one scan to the next, i.e. with heating
tlme. This non-monotomc growth is probably due to sam-
pling error, since there is varying impurlty concentration
between individual beads of formulation. There is neverthe-
less a trend with longer heating time of decreasing intensity
of alpha features together with increasing intensity of
gamma peaks. Most pronounced is the rising prominence of
the gamma peaks at 755.8-764.1, 938.1-970.6, 1106.1, and
1153.0 [cm-1]. Spectra corresponding to longer heating
tlmes show greater conversion of InnW to gamma phase

r "
( L )
/ (
( K )
/ (
( J )
A

( H )
/
( G )

( F )
/
( E ]
'
( D )
/
( C )
'
( D )
/ '
( A )
4

1 2UO
F i g U
R X _ :
hrs a
1 0 5 c
g f 1 m ]
eP s 1
in g .

of S
l o n E
ed j
w h i 1
p m I
on 1
c l e a




lF ;
P E


ial of
l h a s e .

E x tr a
l 3 5 . 4,
llle in
8 n t r a =
' ; : :
ln the
l atin g

l l \ e d

7 o 0
b e i v e d
m a d e
lce of
1ce of
n of
steel
l u d e d
N/ er e d
NI W
re a s e

( D)
/
(C)
J
(0)

( A)

1200 1150 11@0 105o 10CO 950 9Oo 88o 800 75o 700
wavenumber[cm-11
Figul.e 6. The finge]print FTm spec of HNIW from (a) as-made
RX-39AC; (b) through (j) firom RX -39-AC samples heated 1-8, 24
hrs at 100oc; 0f) from RX-39-AVC heated a total of 96 hrs at 100oC-
105oC; and (l) pure gamma phase. There is a trend in increasing
8(lmma impuno/ concemratlon with 1onger heating time.
:atlng 6psilon mHW converted to gamma phase by 96 hrs heat-
sam- ing. Material heated in both glass and stainless steel pans
ation show 1he same trend in polymolphic conversion. Both Sets


Ce of which had been heated 24 hrs were as green as the initial,
, and problematic RX-39-AC fOmmlation. The affect of humidity
ating 0n the samples during additional 72 hrs heating was not

11 the conversion is modified by exposure to moismre, the rate of


PYD5 . Propellants Explos. Pyrotech. 19.


l2DO 1150 11Do 1050 1000 950 900 850 800 750 700
wavenumber[cm-1]

RX-39-AC; (b) through (e) from RX-39-AC samples heated at 75oc,
85oC, 95oc and 105 C; (f) ffom RX=39-AC heated a total of 96 hrs al
100.c-105oC; and (g) pure gamma phase. There is no hend in HNIW
polymmphic conversion.
EStane bmder degradation could be affected. AIthough both

were obviously greatest for material heated 96 hrs, neither
observable could be considered extreme compared to ther-

Figure7showstheFIn%spectraofRX-39-ACfOmmla-
tion samples that were heated (75oC, 85C, 95oc, and
105oC) du]ring cycle-pressing. For comparison, also shown

ed (24 hrs at 100oC) RX-39-AC formulation and pure
gamma phase HNIW. Unlike Fig. 6, there is no clear-cut

to the next, i.e. with pressing temperature. The only pro-
nounced change from alpha to gamma HNIW impurity is
seen m Fig. 7f, which is identical to Fig. 6k. This conver-
sion is pnmarily due to extended heating at elevated tem-
peramre, not elevated pressure.
The final materials tested included the heated and
pressed green RX-39-AC fOmmlation in question, an RX-
39-AC fonnulation made from a different 1ot of E-InJIW, a

%), and Estane recovered from heated RX-39-AC (rinsed
clean of HNIW). The Spectra of mnW retrieved fTom the
problematic gleen RX-39-AC material is shown in Fig. 8,
along with that of IEJIW from as-made RX-39-AC, pure
gamma phase ImJIW, and HNIW from RX-39-AC that had
1994 NO 3 113-158 . ISSN O721-3115

( A )


1200 1150 110D 1050 1000 950 900 850 8DO 75D 700

Figure 8. The finge]rprint FTIR spechPa of HNIW ffom (a) as-made

problematic geen RX-39-AC fblmulation; and (d) pure gamma phase.
The presence of gamma impurity in (C) is shown by new peaks at
755.8-764. 1 . 938. 1-970.7, 1 106. 1, and 1 153.0 [cm-1].
been heated at 100oC for 24 hrs. The amount of gamma
phase impurity in the problematic RX-39-AC is 1ess than
that seen in the material heated 24 hrs, as typified by the
relative intensity of the same gamma peaks noted in Fig. 6.
Since the gamma phase has the 1owest density of the four
polymorphs, the change in sample density from one pressed
part to the next is apparently due to growth of this phase
during heating prior to pressing.

e-HNIW did not change color appreciably with heating
96 hrs at 105C (wet oven), other than slight yellowing that
can probably be attributed to themlal EStane degradation.
The new 1ot of HNIW, unlike the original batch, did not
have activated charcoal. In addition, the non-Estane fbrmu-
lation, also without activated charcoal impurity, did
not change color during 96 hrs heating at 105oC in a wet
oven.
Severalattemptsweremadetoidentifythesourceofthe
green color in heated RX-39-AC fOrmulation. The HNIW
was dissolved in acetone and filtered, and the binder resi-
due rinsed in methanoL The mJIW/acetone solution was
only faintly yellow. The cleaned binder residue was quite
rubbery, and when extracted fiPom formulation that had

was dissolved in chlorofonn and applied as a Syrup to KBr
plates for FTIR analysis. Virgin Estane itself and Estane
from both heated and as-made formulation was studied.
There was no identifiable difference seen between the mid-

tration of species that caused the strong coloration in the
heated binder was low, and/or the species exhibited a mid-
IR FTIR Spectrum not substantially different from virgin
Estane.
Resultsfromthesafetytestsshowed1ittledifference
between as-made and heated (96 hrs heating at 105C,
green) RX-39-AC fbnnulations. The results are shown in
Table 2. Two values are 1isted fbr the drop-weight impact

Tab]e 2. Safety Tests of the RX-39-AC Fommlations
Safeb/ Test
nnpact Drop-weight
Test Dh50(a)
EIectrostatic Spark
T e s t
Friction Test
B AM ( b )
RX-39-AC
as-made heated24hrs(green)
36.7, 39.7 cm
0/10
(no go)
1/10
at 1 1 .6 kg

0/10
(no go)
1/10
at 1 1 .2 kg
(a)For comparison, LX-14-9 (HMX/Estane 5702-FI, 95.5/4.5
weight%)hasa50%drop-weightheightDh50=50.0cm.
(b) The as-received e-HNIW had one reaction in ten at 7.2 kg.
test(15) fOr as-made RX-39-AC, both of which are close to
the value measured fbr the heated sample. The approach=

as-made samples gave no reaction m ten mes. The weight
required to give one reaction in ten for friction sensitiv.
ity(16) essentially did not change wilh heating. The amount
of polymo]T)hic change in the InJIW or chemical change m
the EStane binder was therefore msufficient to affect notice~
ably the sensitivity of the formulation with respect to
impact, Spark and fiiction.
4. Discussion

change. The beta phase impurity initiaUy present in the as.
received E-mUW converted, all or in part, to alpha phas61
during the 60oC solvent stripping stage of formulation Thil
conversion is consistent with a previously measured con-
version(5) of beta to alpha at 37oC. Heating of the as-made
RX-39-AC at 100oC before pressing allowed not only th6
alpha phase, but also some fraction of the bulk epsilo]n
phase HNIW, to convert to the 1owest density gamma
phase. These conversions are consistent with phase transi-
tlons seen m earlier work, the E 'Y conversion(5) occuning
at 64 1oC. Near complete conversion of the InJIW to
gamma was seen to occur after 96 hrs heating at 105oc
The most probable reason for the decrease in sample den.
sity seen in the initial, problematic RX-39-AC formulation

Solventswithcarbonylgroupshavebeenshowntohave
high solubility towards HNIW(8,9). As a nitrate ester with
high carbonyl content, Estane would be expected to hav6
moderate HNIW solubility. The relatively 1ow carbonyl
content of the EVA binder used in the non-Estane PBX
suggests a lower HNIW solubility for this material. The
role of solvation in polymo]Thic phase transitions is to
1ower the bamer to conversion, thus increasing the rate of
conversion(8a). Therefbre, over the relatively short heatin8
period used in this work (days instead of months/years) (l
greater degree of conversion would be expected to be seel]
in the Estane PBX. It is not clear whether a 1ower solubility
binder (e.g. EVA) could nevertheless still undergo signifi.
cant E 'Y conversion during very 1ong telm storage at tem
peramres above 64oc. Without prolonged sample heating in


s o 1 ub i
w h e t h
m u 1 a t i
Co:
olive (

the ac
1 m p u r
b in d e l
of the
r e 1 e a S ,
mg. V
the EE
tion r,
1oss il
also il

r e V e a
identi
trOSCC
natlor
5. CO:
Pr(
C o n S I (
~ 1 0 0
in col
g at e d
an al y
m u 1 a 1
at all
c o nc E
rial. ']
1mp u]
durin
ing t(
100o(
1mpu
versi(
whic]
attrib
tum i
polyr
dle d(
AC v
bindE
charc
ven I
the al
duce
vity 1
ged.
6. Rf
(1) (a
H

T 7


g r e e n )

5. 5 /4 . 5
)Se tO
) a c h -
l and
e i gh t
s i t i v -
1 o u n t
ge in
) t i c e -
Ct tO
P u n t y
1 P h a s e
he as-
P h a s e
. This

. m a d e
ly the
) s i 1 o n
1 m In a
r a n s i -
m n j n g
[W to
0 5 o C .
} den-
1 ati o n
l have
. with
have
b o n y 1
PBX
. The
is to
ate of
# a t i n g
al.s) a
seen
b i 1 i t y
g n i f i -

ng in
an envlronmental chamber, and/or knowledge of HNIW
solubility in such binders, it is not possible to predict

m u 1 a t i o n .
Color change in the RX-39-AC from greyish-white to
01ive drab green during prolonged heating is not due to the
ImlW phase conversions, but most 1ikely originates with
the activated charcoal impurity. The role played by this
impurity could 1ie in sulface catalyzed decomposition of the
binder. Altematively, themlally-driven reactive degradation
0f the Estane binder could originate with impurities that are
released fi?om the activated charcoal duling prolonged heat-
ing. Washing of the fomlulation to isolate the InJIW and
the Estane components for analysis showed intense colora-
tion remaining in the stiffer, rubbery Estane residue. The

also indicative of long chain fi.agmentation. However, mid-
m FTIR spectra of dissolved, strongly colored binder
revealed no new spectral features that could be used to
identify the source of coloration. UItraviolet olr far-IR spec-
troscopic analysis might be more usefUl in such a determi-
n a t i o n .
5. Conclusion
Pmlonged heating of a PBX fOnnulation (RX-39-AC)


m color and sample density(12). These changes were investi-
gated by Fourier transfbml infmred (FTIR) Spectroscopic
analysis of starting mate]rials, as-made and final colored for-

at all stages of production and processing, the nature and

rial. The small amount (on the order of 10%) of beta phase
impurity m the as~received HNIW converted to alpha phase
during the fommlation process itself, which involved heat-
mg to 60oC. Extended heating of RX-39-AC samples at

impurity to low density gamma, but also (after 96 hrs) con-
ve]rsion of the bulk epsilon phase to gamma. The speed with
which this phase conversion went to completion can be
attributed to the high HNIW solubility of Estane, which in
tum is due to its high carbonyl content. Thennally-induced
poIymorphic conversion was detennined to be the cause of
the decrease in sample density. The color change of RX-39-
AC was not identified, but is speculated to be the result of
binde]r degradation arising fi?om the presence of activated
ch@rcoal impulity. Either su]face catalysis or themlally-dri-
ve]1 reactions of the Estane with impurities released fiFom
the activated charcoal during prolonged heating, could pro-
duce colored, short chain decomposition products. Sensiti-
vjty to impact, fiiction and electrostatic spa]k was unchan-
8 e d ,
6, Refbrences
(l) (a) A.T. Nielsen, M.L. Chan, K.J. Kraeutle, C.K. Lowe-Ma, R.A.
Hollins, M.P. Nadler, R.A. Nissan, W.P. Nonis, D.J. vanderah,


ReportNWCTP7020(Nov.1989),Navalweaponscenter,
ChinaLake,CA;(b)D.J.vanderah,M.P.Nadler,R.Y.Yee,and
A.T.Nielsen,ReportNWCTP7058(Sept.1990),NavalWeapons
Center,ChinaLake,CA.
(2) P. Politzer, Report ONR NOOOI4-85-K-0217 (May 1991), Depalt-
mentofchemisdy,Universic/ofNewOrleans,NewOrleans,LA.
(3) (a) T.P. Russell, P.J. Miller, G.J. Piemlarini, and S. BIock, High


Pielmarini,ands.BIock,Pressure/TemperamrePhaseDiagram
ofHexanitrohexaazaisowurtzitane,,J.P/iy,5'.C/he/".97,1993
(1993).

malPropertiesofHexanitrohexaazalsowurtzitane,',mPJ.oc.IQ90
JAMV[AFP/.Op.Me"zg,Anaheim,CA,(Oct.1990),p.204.
(5) M.F. Foltz, C.L. Coon, F. Garcia, and A.L. Nichols m, Thelmal
StabiliD/ofthePolymolphsofHexanitrohexaazaisowurtzitane,

(6) (a) K.J. Kraeutle, The Themlal Decomposition of a New Propel-

Pub.No.498,NASA-MarshalSpaceFIightcenter,Huntsville,
Oct.1988;(b)R.J.Doyle,Jr.,TheGas-phaseDissociationofa
NewPolyazapolycyclicNitramine:Hexanitrohexaazalsowurtzi-

Bri]l,ThennalDecompositionofEnergeticMate]rials.53.Kine-
tlcsandMechanismofThelmolysisofHexaninohexaazaisowurt-

guez,R.A.Fifer,K.L.McNesby,J.B.Morris,M.A.Schroeder,



ingGround,MD.
(7) R. Gilardi, private commumcatlon.
(8) (a) E. von Holtz, D. Omellas, M.F. Foltz, and J.E. CIalkson, The
Solubilio/ofE-mllwinSelectedMaterials,submittedtoPropel-
Iants,Explosives,Pyrotechnics;(b)N.C.Johnson,Y.Tran,L.


NASAKennedySpacecenter,April1992.
(9) C. Coon, prlvate communication.
(10)(a) D.W. Brown, R.E. Lowry, and L.E. Smith, Kinetics of

I3,248(1980);(b)J.W.S.HearleandM.MiraftabTheFIex
FatigueofPolyamideandPolyesterFibres,,'J.Mm,gcj.26,2861
(1991).
( 1 1 )C. Walkup and G. Moody, Lawl.ence Livelmore National Labora-

(12)Physics hltemational Company, San Leandro, CA.
(13)Gen Co]poration, AerOjet Propulsion Division, Sacramento, CA.



tinnyArsenal,DOver,NewJersey.


ReportUCRL-52997Change2(Jan.1985),LawrenceLivelmore
NationalLaboratory,Livermore,CA.
(16)R. Meyer, EXp/oJjvey, 3rd ed., VCH Publishers, New York,
1987,pp.147-9.

The author would 1ike to gratefUlly acknowledge Chris Gatrousis
and Tom Sugihara for their support, and Rich whipple, Rich Simpson,
Marc Hoffinan, Mike MuIphy, and Randy Simpson fO]r their technical
asslstance and advice.
Work performed under the auspices of the U.S. Department of
Energy by the Lawrence Livelmore National Laboratoly under con-
hact No. W-7405-ENG-48.
(Received November 16, 1992; Ms 62/92)

W PYD5 . Propellants Explos. Pyrotech 19 1994 No 3 1 13-158 . lSSN O721 31 15
, F
~