Laser photodissociation and thermal pyrolysis of energetic polymers

Article abstract of DOI:10.1021/jp962936t

Decomposition reactions of glycidyl azide polymer (GAP) and poly(glycidyl nitrate) (PGN) have been investigated by pulsed infrared laser pyrolysis and ultraviolet laser photolysis of thin films at 17-77 K. Reactions are monitored by transmission FTIR spectroscopy. Photolysis of GAP at 266 nm shows that the initial reaction steps are elimination of molecular nitrogen with subsequent formation of imines. Warming leads to loss of the imines, most likely as a result of secondary polymerization reactions. Thermal decomposition of GAP by infrared laser pyrolysis reveals products similar to the UV experiments after warming. Laser pyrolysis of PGN indicated that the main steps of decomposition are elimination of NO₂ and CH₂O from the nitrate ester functional group, followed by rearrangement of the polymer backbone to form carbonyl functional groups. Higher order reactions form numerous small molecule products including NO, CO, CO₂, N₂O, H₂O, HNCO, HCN, HCNO, HONO, and CH₄. UV laser photolysis of PGN shows the major products to be CO, CO₂, N₂O, and polymeric carbonyls. A mechanism is suggested to account for the difference between products in the UV and infrared laser experiments. Quantum yields for destruction of the azide group in GAP and for destruction of the nitrate ester group in PGN are 8 × 10^-3 and 1 × 10^-3, respectively, for 266 nm excitation.

Full text of DOI:10.1021/jp962936t

2126
J. Phys. Chem. B 1997, 101, 2126-2131
Laser Photodissociation and Thermal Pyrolysis of Energetic Polymers
Ping Ling and Charles A. Wight*
Department of Chemistry, UniVersity of Utah, Salt Lake City, Utah 84112
X
ReceiVed: September 24, 1996 ; In Final Form: January 13, 1997
Decomposition reactions of glycidyl azide polymer (GAP) and poly(glycidyl nitrate) (PGN) have been
investigated by pulsed infrared laser pyrolysis and ultraviolet laser photolysis of thin films at 17-77 K.
Reactions are monitored by transmission FTIR spectroscopy. Photolysis of GAP at 266 nm shows that the
initial reaction steps are elimination of molecular nitrogen with subsequent formation of imines. Warming
leads to loss of the imines, most likely as a result of secondary polymerization reactions. Thermal
decomposition of GAP by infrared laser pyrolysis reveals products similar to the UV experiments after warming.
Laser pyrolysis of PGN indicated that the main steps of decomposition are elimination of NO
from the nitrate ester functional group, followed by rearrangement of the polymer backbone to form carbonyl
functional groups. Higher order reactions form numerous small molecule products including NO, CO, CO
O, H O, HNCO, HCN, HCNO, HONO, and CH . UV laser photolysis of PGN shows the major products
to be CO, CO , N O, and polymeric carbonyls. A mechanism is suggested to account for the difference
between products in the UV and infrared laser experiments. Quantum yields for destruction of the azide
2 2
and CH O
2
,
N
2
2
4
2
2
-
3
-3
group in GAP and for destruction of the nitrate ester group in PGN are 8 × 10 and 1 × 10 , respectively,
for 266 nm excitation.
Introduction
Brill studied the thermal decomposition kinetics of PGN and
7
found the reaction to be first-order for slow heating rates but
Glycidyl azide polymer (GAP) and poly(glycidyl nitrate)
PGN) are energetic polymers that have been proposed for use
8
second-order for rapid heating. Under rapid heating conditions,
(
the gaseous decomposition products were determined to be CO,
CO2, NO, NO2, H2CO, HCN, HONO, and HCO2H. However,
the mechanism by which these products are formed is still not
well established. In earlier work on nitrocellulose, Gelernter
as binders in solid rocket motors. The un-cross-linked oligomers
used in this study are viscous liquids at room temperature.
9
et al. had proposed that rupture of NO2 from the side chain is
the first step of thermal decomposition. However, a more recent
study suggested that the reaction begins with second-order
1
0
reaction of nitrate ester groups.
Recently, transient thin film laser pyrolysis techniques, which
use CO2 laser pyrolysis of target materials, following by rapid
thermal quench of reaction products, has restricted reaction time
at high temperatures to within a millisecond scale.11 Just as
shock heating is suitable for studying gaseous reactions at high
temperatures for short periods, this technique has been proved
successful in studying initial thermal reaction in the condensed
phase.
GAP and other azide-containing polymers have been the subject
1,2
of several studies over the years. Brill and co-workers found
that GAP decomposes at about 250 °C during rapid heating;
the major products are N2, CO, HCN, NH3, CH2O, CH4, C2H2,
and C2H4. Thermal decomposition at high temperature and high
3
pressure produces N2, H2, CO, C, and other minor products.
Although these studies have proved valuable for understanding
the gas phase decomposition kinetics, it has been difficult to
determine the initial steps of the decomposition mechanism
because the time scale of reactions is short and the products
are not usually detected in the condensed phase. Haas and co-
Laser photolysis is also a useful tool in studying the
decomposition mechanism of molecules. Although it is not
likely to involve electronically excited states in normal thermal
decomposition, it is suggested that electronic excited states may
4
12
workers studied the thermal decomposition of GAP by pulsed
play an important role in detonation. On the other hand, if
CO2 laser irradiation and found that both gaseous products and
smoke powder are formed. The gaseous products are mainly
N2, CO, C2H4, HCN, CH4, and C2H2. The smoke powder was
assigned to be principally imines, as indicated by a broad
internal conversion of electronically excited states to a vibra-
tionally excited ground electronic state is important, then the
photolytic decomposition product ratios should bear a strong
similarity to those resulting from thermal decomposition.
Because GAP and PGN polymers have the same backbone
structure, it is appropriate and helpful to use transient thin film
laser pyrolysis and laser photolysis techniques to study initial
reactions in both materials in one set of experiments.
-
1
infrared absorption band near 1650 cm . However, early
studies of methyl azide decomposition show that there are other
5
strong bands associated with imines that were not reported by
Haas and co-workers, casting some doubt on this assignment.
Thermal decomposition of a series of nitrate esters was
studied by Oxley and co-workers recently.6 They concluded
that the rate-determining step in nitrate ester thermolysis is
usually homolytic cleavage of the RO-NO2 bond. Chen and
Experimental Section
GAP was obtained from the 3M Company, which markets
the material as an energetic binder for propellants. The material
used in our study is the un-cross-linked polyol, which has a
linear repeating structure but is synthesized using a branched
*
Author to whom correspondence should be addressed.
Abstract published in AdVance ACS Abstracts, February 15, 1997.
X
S1089-5647(96)02936-7 CCC: $14.00 © 1997 American Chemical Society

Laser Photodissociation of Energetic Polymers
J. Phys. Chem. B, Vol. 101, No. 12, 1997 2127
initiator. PGN was supplied by Dr. Robert Wardle of Thiokol
Corp. Both GAP and PGN are viscous liquids at room
temperature.
TABLE 1: Integrated Infrared Absorption Coefficients
integrated absorption
coefficient (cm/µmol)
band position (cm-1)
molecule
Pulsed Infrared Laser Pyrolysis. The details of the
apparatus have been described elsewhere.11 Briefly, a thin film
of GAP or PGN is sandwiched between two salt windows (CsI
or CaF2) and mounted at the cold tip of an evacuated liquid
nitrogen Dewar vessel. The vacuum enclosure has two infrared
windows mounted on opposite sides for obtaining transmission
FTIR spectra and one or two quartz windows mounted on the
CO
CO
2
2344
2138
2238
1720
19.
2.3
13.
2.8
1.2
0.93
4.5
6.4
5.1
4.2
0.62
0.8
6.2
6.5
2.2
N
H
2
O
CO
2
1
495
1861
755
(NO)
2
1
2
N O
3
1861
1594
1300
two remaining sides of the cell for photolysis. The vacuum
cell is pumped to 10-4 Torr; then the cold finger is cooled to
7
84
7
7 K by liquid nitrogen. In some experiments, the sample was
(NO
2
)
2
1874
cooled to 17 K by a closed cycle helium refrigerator in a similar
vessel.
1
1
738
255
747
Pyrolysis of the sample is carried out by pulsed CO2 laser
irradiation (Pulse Systems Model LP140-G). The laser was
-1
tuned to 1039.5 cm , and the pulse length is about 35 µs. The
laser beam size can be adjusted by means of a spherical mirror
and is measured by irradiating a target of thermally sensitive
paper. The average energy per pulse is calculated from the beam
power, as measured by a Scientech Model 38-0101 disk
PGN. After about 30 min the evolved gas is admitted to an
evacuated IR gas cell for detection. In some experiments the
gas phase products were condensed onto a 17 K cold window
in a deposition chamber for IR detection in the condensed phase.
The pyrolysis temperatures were chosen as the lowest temper-
atures at which gas bubbles can be observed to evolve from
the GAP or PGN.
2
calorimeter. Typically, laser fluences in the range 0.5-2.3 J/cm
were used in these experiments.
At higher temperatures, it is easy to induce explosive thermal
decomposition. For example, when a tube containing GAP is
placed in a hot bath at 230 °C, an explosion occurs in seconds,
and all of the GAP is consumed.
Neither GAP nor PGN has any strong infrared absorption
bands within the tuning range of the pulsed CO2 laser (900-
-
1
1
100 cm ). Therefore, infrared laser pyrolysis experiments
were conducted using two different methods. In the first
method, we deposit a thin film of salt (NaBF4, which has very
strong absorption at the laser frequency) on the IR window prior
to sample preparation. The salt film is in good contact with
the GAP or PGN to ensure efficient thermal conduction during
laser pyrolysis. The second method is to use a relatively thick
sample of GAP or PGN (g15 µm). Although the energy
deposition per unit volume of sample is the same as for thin
films, the rate of conduction out of the sample is much slower,
so reactions have a greater time to take place. Usually in the
first method, a single shot is enough to reach maximum
concentration of reaction products, whereas the second method
normally requires five to seven shots at the highest available
Supplementary Experiments. GAP exhibits weak absorp-
tion around 285 nm; this transition has been assigned to the
1
4
n-π* transitions in the azide functional groups. Similarly,
PGN exhibits a weak n-π* absorption at 270 nm. The
absorption cross sections at 266 nm were measured by a
-20
2
Hewlett-Packard diode array UV-visible to be 7.5 × 10 cm
-
20
2
and 8.8 × 10
cm per functional group for GAP and PGN,
respectively.
Calibration experiments were carried out in order to determine
the relative integrated infrared absorption coefficients for CO,
CO2, N2O, NO, NO2, CH2O, and some gas mixtures. Samples
containing 1-10 µmol of the authentic compounds were
deposited onto a 2.5 cm diameter infrared window at 17 K at a
rate not exceeding 5 µmol/min. Formaldehyde was made by
thermal depolymerization of paraformaldehyde; other gases were
used as supplied from commercial sources. A good linear
relationship was observed between the integrated absorption
intensity and the amount of gas deposited. In most cases, the
band positions and widths were somewhat dependent on
temperature, but the integrated intensities were essentially
constant. The exceptions were N2O3 and (NO)2, for which some
of the integrated intensity of some bands increased as much as
2
fluence (2.3 J/cm ) to achieve similar results.
Transmission FTIR spectra were recorded before and after
pyrolysis by means of a Mattson Model Polaris FTIR spec-
trometer. Spectra are obtained by averaging 32 scans at 0.5
-
1
cm resolution.
Pulsed UV Laser Pyrolysis. The sample assembly is the
same as for laser pyrolysis. Before photolysis, the sample is
turned to face one of the quartz windows. After photolysis,
the sample is turned to face both IR windows before recording
the infrared spectrum. Photolysis is carried out by using the
fourth harmonic of a Nd:YAG laser (Continuum Model Surelite)
at 266 nm. The laser pulse duration is 7 ns, and the laser power
is typically 35 mW at 10 Hz repetition rate, as measured by the
disk calorimeter. The spatial profile of the beam is determined
by recording the laser power while a razor blade is scanned
across the beam. The data are differentiated and fit to a
Gaussian function to determine the 1/e diameter of the beam,
which was determined to be 3.9 mm. A lens is used to expand
3
0% from 17 K to 77 K. In the case of NO and NO2, the
absorption coefficients were determined by depositing these
samples as monomers in an excess of argon matrix gas.
Annealing these samples to evaporate the argon component
produced bands that have been assigned previously to the
1
5
16
dimers and (in the case of mixed NO/NO2 samples) N2O3.
The integrated absorption coefficients (in cm/ µmol) measured
in this manner are given in Table 1.
2
the beam to the desired fluence, which was typically 1 mJ/cm
Results
in this study. The bulk temperature change of the sample due
1
3
to absorption of laser light is estimated not to exceed 1 K.
UV Laser Photolysis of GAP. Transmission FTIR spectra
of thin films of GAP were obtained before and after UV laser
photolysis, as illustrated in Figure 1a,b. After photolysis, the
Ordinary Thermal Decomposition Experiments. A drop
of GAP or PGN is placed in a glass tube, and the tube is
evacuated and sealed. The end of the tube with material inside
is immersed in an oil bath at 170 °C for GAP or 160 °C for
-1
intensity of the characteristic azide band at 2100 cm is greatly
reduced compared with the other bands. New absorption bands

2128 J. Phys. Chem. B, Vol. 101, No. 12, 1997
Ling and Wight
Figure 2. Infrared spectra of GAP showing the similarity of samples
that were (a) subjected to 266 nm laser pyrolysis and warmed to room
temperature for several hours, (b) subjected to CO
7
3
2
laser pyrolysis at
7 K, and (c) a residue of a GAP film that was heated to 170 °C for
0 min.
absorption coefficient. A plot of ln(A/A0) was found to be linear
in J, and the resulting value of Φ determined from the slope is
Figure 1. Infrared spectrum of a GAP sample (a) prior to and (b)
following 266 nm laser photolysis at 77 K. Spectrum c is the same
sample after warming to room temperature for several hours.
-3
8
× 10 .
Pyrolysis of GAP. The spectra of GAP films that were
2
subjected to pulsed CO2 laser pyrolysis at 2 J/cm are illustrated
-1
-
1
in Figure 2b. A new broad band around 1660 cm and a small
appear at 3212, 1652, 1394, and 1224 cm . The intensities of
the new bands are highly correlated, suggesting that they come
from the same products or group of products. It is noteworthy
that absorption bands of GAP associated with C-O-C stretch-
-
1
band around 2032 cm are formed after pyrolysis. We found
that varying the laser fluence does not have any significant effect
on the shape or position of the new bands, but simply changes
the magnitude of the product absorptions.
-
1
ing vibrations (1082-1128 cm ) and C-H stretch vibrations
-
1
The spectra of thin film residues formed by heating GAP to
(
2750-3000 cm ) remain almost unchanged after photolysis.
170 °C for several minutes is shown in Figure 2c. The spectrum
This suggests that the backbone of the polymer is rather stable
with respect to 266 nm radiation. Photolysis for extended
periods (i.e., until all the azide groups are destroyed) revealed
a tiny infrared band due to CO, suggesting that only a tiny
portion of backbone is broken to release CO.
bears a strong resemblance to the spectrum of GAP subjected
to CO2 laser pyrolysis (Figure 2b) and to the spectrum of GAP
subjected to UV photolysis followed by warming to room
temperature for several hours (Figure 2a). The gas evolved
during the 170 °C pyrolysis period was collected and found to
be essentially infrared-inactive, except for a trace of CO.
A control experiment was carried out in which a sample of
poly(propylene glycol) polymer was photolyzed for extended
periods.
Pyrolysis and Photolysis of PGN. Infrared spectra of PGN
obtained before and after pulsed CO2 laser pyrolysis at a fluence
2
of about 2 J/cm are shown in Figure 3a,b. Reaction product
-
1
bands were observed at 1723 and 1498 cm (CH2O), 1862
-
1
(
mainly NO2 dimer), 1595 (N2O3), and 1302 cm (NO2 dimer
-
1
and N2O3). A shoulder observed at 1744 cm exists in the
reference spectra of (NO)2 and (NO2)2. This band persists to
relatively high temperatures during the warm-up cycle, sug-
gesting that it is mainly attributable to (NO2)2. In addition, small
This polymer is similar to GAP except that the azide group is
replaced by a hydrogen atom. No change in reactant bands
was observed, and no products were formed.
-
1
-1
bands were observed at 2341 cm (CO2); 2133 cm (CO);
2282 and 2276 cm (HNCO); 2257 and 2247 cm-1 (HOCN);
-
1
When photolyzed samples are warmed to room temperature,
additional changes occur over a time scale of several hours.
The polymer undergoes a transformation from a transparent
viscous liquid to a solid that is highly light scattering. The
infrared spectra show that the bands at 3212, 1394, and 1224
-1
-1
and 2236 cm (N2O). The product band observed at 1731 cm
belongs to a nonvolatile product because it persists after
warming to room temperature. The band observed at 1567 cm
-1
cannot be definitively assigned, though aliphatic nitro and
nitroso compounds typically have strong absorption bands near
this position.
Pyrolysis experiments carried out at 17 and 77 K exhibit
similar product absorption spectra except for smaller amounts
of CO that are observed at the higher temperature, due to the
volatility of this product.
-1
-1
cm disappear completely, and the band at 1652 cm broadens
and diminishes in intensity, as shown in Figure 1c. The
quantum yield for destruction of the azide group under 266 nm
photolysis was measured by recording the integrated absorption
-
1
intensity, A, of the azide band at 2102 cm as a function of
2
cumulative laser fluence, J (photons/cm ). For optically thin
samples, the quantum yield can be found from the relationship
Figure 3c shows the gaseous products from heated PGN at
1
60 °C in an ordinary thermal decomposition experiment. The
ln(A/A ) ) -ΦσJ
(1)
0
evolved gas was deposited onto an infrared window at 17 K.
In addition to the major products CH2O, NO, NO2, and CO2,
(which were observed in the laser pyrolysis experiment), we
where Φ is the quantum yield and σ is the measured 266 nm

Laser Photodissociation of Energetic Polymers
J. Phys. Chem. B, Vol. 101, No. 12, 1997 2129
Figure 5. Infrared spectra of a PGN sample (a) prior to and (b)
following 266 nm laser photolysis at 77 K.
assigned as NO in this spectrum rather than (NO2)2. The major
products, CO2, CO, and N2O, were produced in relative amounts
1.0/1.27/1.57, after correcting for differences in infrared absorp-
tion band intensities. An excellent correlation was also found
between the decrease in intensity of the PGN nitro group band
and the increase of the condensed phase product band at 1731
Figure 3. Infrared spectrum of a PGN sample (a) prior to and (b)
following CO
by collecting the vapor evolved from a sample of PGN heated to 160
C and condensing it onto an infrared window at 17 K in vacuum.
2
laser pyrolysis at 17 K. Spectrum c is a film produced
-
1
-1
cm . A second condensed phase product band at 1567 cm
°
-
1
was also observed. A band near 1200 cm associated with
C-O-C vibrations of the polymer backbone is essentially
unchanged during the photolysis period. The C-H stretching
-
1
bands around 2900 cm were decreased after photolysis, and
-
1
a new broad OH band around 3200-3400 cm appeared.
During warming to room temperature, the bands at 1730 and
-
1
1
567 cm decreased about 20-40%. The C-O-C band and
OH band also decreased after warming up.
The quantum yield for 266 nm laser photolysis was measured,
-
3
using the technique described above, to be 1 × 10 .
UV Laser Photolysis of Nitromethane. To aid the inter-
pretation of some of the experimental results with PGN, we
found it necessary to investigate the UV laser photolysis of
nitromethane isolated in a 17 K argon matrix at relative
concentration of 1/1200. The products were found to be CH2O,
HNO, NO, CO, CO2, HNCO, nitrosomethanol, and a small
amount of N2O, in agreement with previously published
Figure 4. (a) Partial infrared spectrum of a PGN sample that was
pyrolyzed with a CO laser and subsequently warmed to room
temperature. (b) Residue of a PGN film that subjected to ordinary
17,18
results.
The most important new result in our experiment
2
is that, upon warming to 40 K, the argon matrix is evaporated
and changes in the product band intensities were recorded. The
band corresponding to HNO disappears completely, and HNCO
is diminished. At the same time, the band due to N2O is greatly
enhanced.
thermal pyrolysis at 160 °C.
see a significant enhancement in the intensity of bands attributed
to CO and N2O.
Thermally Activated and Photolytic Oxidation of Form-
aldehyde. Two experiments were carried out to determine the
thermal reaction products of CH2O with NO and NO2. Equal
amounts of CH2O and NO or NO2 were mixed in a sealed glass
tube, one end of which was immersed in oil bath at 160 °C.
After 30 min, the gaseous contents of the tube were deposited
onto an infrared window at 17 K. The principal products of
Thermal decomposition of PGN at higher temperature
produces gaseous products CH2O, NO2, NO, N2O, CO, CO2,
HCN, HONO, CH4, and H2O, which were detected both in the
gas phase and in the low-temperature condensed phase. Ex-
plosive thermal decomposition occurs in samples of PGN, but
requires somewhat higher temperatures than GAP.
In addition to the gas phase products, two bands at 1730 and
-
1
the CH O + NO reaction are N O and CO, along with smaller
1
567 cm
were observed after warming laser-pyrolyzed
2
2
amounts of CO2. The main products of the CH2O + NO2
reaction are CO and NO; moderate amounts of CO2 were also
observed, along with a small yield of N2O.
In a separate series of experiments, we prepared thin films
of formaldehyde and NO by vapor deposition of the two gases
directly onto a 17 K infrared window in vacuum. UV photolysis
at 266 nm forms CO, N2O, and CO2. Similar experiments
carried out by photolyzing films of formaldehyde and NO2 at
samples to room temperature (Figure 4a). Similar bands were
observed in the residue of PGN heated to 160 °C (Figure 4b).
UV laser photolysis of PGN at 266 nm results in the spectra
illustrated in Figure 5b. Compared with pyrolysis, the main
-
1
difference is the much smaller amounts of H2CO (1499 cm )
-
1
and NO (1862 cm ) produced in the photolysis experiments.
A thermal annealing experiment showed that the 1862 cm-
band belongs to a volatile compound and is therefore more likely
1

2130 J. Phys. Chem. B, Vol. 101, No. 12, 1997
Ling and Wight
1
7 K formed CO, NO, and CO2. After extended periods of
Subsequent reaction of CH2O with NO2 forms CO, NO, CO2,
and N2O, in agreement with the supplementary formaldehyde
oxidation experiments that were carried out in this study. It is
UV photolysis, N2O also appears as a secondary product.
Discussion
likely that H O is a reaction product in this scheme; however,
2
contamination of all samples by atmospheric water makes it
UV Photolysis and Thermal Pyrolysis of GAP. Photolytic
decomposition of GAP appears to take place in two distinct
steps. The first is elimination of molecular nitrogen and
formation of the corresponding imine,
difficult to isolate and quantify this. Other side reactions
produce minor amounts of HNCO and possibly HOCN.
-
1
We observed a product band at 1731 cm that is consistent
with formation of a polymeric carbonyl functional group, leading
us to speculate that the backbone of the polymer begins to
depolymerize according to
This reaction is analogous to the mechanism previously studied
1
9
for simple alkyl azide compounds.
By analogy to the
published spectra of methylenimine, we may readily assign the
5
Chen and Brill had previously reported a similar band, which
they attributed to formic acid decomposition product.7 How-
ever, the involatile nature of the band in our experiments leads
us to favor an assignment as a polymeric aldehyde. Some of
the polymeric radicals in (4) may react with NO or NO2 to form
nitro and nitroso functional groups along the polymer backbone,
as evidenced by the observed nonvolatile product band at 1567
cm , which is consistent with this type of assignment.
In the supplementary formaldehyde oxidation experiments,
the amount of CO2 produced was always less than CO. The
thermal pyrolysis experiment on PGN also bears this out, but
the pulsed infrared laser pyrolysis experiment produced greatly
enhanced amounts of CO2. It is possible that the higher
temperature environment associated with the laser pyrolysis
experiment causes the branching ratio for the NO2 + CH2O
reaction to favor the CO2 product channel. The other obvious
difference between the thermal and laser pyrolysis experiments
is the formation of N2O in the former. This product is presumed
to arise from CH2O + NO2 reactions (as shown by the
supplementary formaldehyde oxidation experiments), but the
short time scale of the laser pyrolysis experiment (∼1 ms)
precludes the formation of significant amounts of N2O because
products from thermolysis of at least two different functional
groups are required to form it.
UV Laser Photolysis of PGN. The most obvious difference
between the pyrolysis and photolysis experiments on PGN is
that the former produces large quantities of NO2, whereas the
latter does not. Instead, we observed production of small
amounts of NO and a relatively large amount of N2O.
Therefore, it is likely that in this case the photolytic reaction
occurs by a different mechanism than the pyrolysis reaction.
One mechanism consistent with our observations involves initial
elimination of CH2O and NO (or HNO) with formation of the
corresponding polymeric carbonyl compound:
,8
-
1
1
652 cm product band observed in our GAP experiments to
the CdN stretching vibration of the imine product. Similarly,
-1
the 3212 cm band is assigned to the N-H stretching vibration,
-
1
and the bands at 1394 and 1224 cm belong to NH and/or CH
deformation modes.
The second stage of reaction occurs after warming to room
temperature. The imine bands disappear, and the sample
becomes highly light scattering. This behavior is consistent with
cross-linking of the imine functional groups to form a higher
molecular weight polymer. The cross-linking reaction most
likely involves reaction of two imine groups with each other,
rather than imine with azide. This is supported by the
observation that during the room-temperature reaction there is
little change in the intensity of the remaining azide band at 2103
-
1
-
1
cm .
It is important to note that essentially all of the chemistry
that occurs in the GAP reactions involves only the azide
functional groups. Infrared bands associated with vibrational
modes of the polymer backbone are nearly unaffected by the
photolysis reaction. Moreover, the unfunctionalized poly-
(propylene glycol) polymer was completely unaffected by UV
laser photolysis at 266 nm.
The pyrolysis experiments (infrared laser and ordinary thermal
pyrolysis) give results that are substantially similar to the UV
photolysis experiments, except that the intermediate imine bands
are not observed. The evolved gas from the thermal pyrolysis
experiment is essentially infrared-inactive, consistent with
elimination of N2. Figure 2 demonstrates the similarity of the
product spectra obtained in the two pyrolysis experiments and
the UV photolysis experiment after warming. The similarity
-
1
of the broad band near 1660 cm in all three experiments is
particularly significant.
Thermal Pyrolysis of PGN. Interpretation of the PGN
pyrolysis experiments is straightforward. Like GAP, the
reaction of PGN is centered at the nitrate ester functional groups.
The observed products in these experiments are NO2 and CH2O,
leading us to conclude that the initial step in the photolytic
reaction is
Again, reaction of the polymeric radicals in (5b) with NO or
-1
NO2 can easily afford the observed product band at 1567 cm ,
which is associated with nonvolatile nitro or nitroso products.
The N2O, CO, and CO2 products are presumed to arise from
reaction of NO with CH2O, consistent with our supplementary
formaldehyde oxidation experiments. Of these, the N2O product
clearly arises from reactions involving at least two of the nitrate
ester functional groups, e.g.
This is consistent with the conclusions of Oxley et al., who
studied the thermal pyrolysis of a series of nitrate ester
compounds.6 According to them, the elimination of NO2 and
CH2O is not concerted, although we have been unable to cleanly
separate the two sequential reaction steps in our experiments.

Laser Photodissociation of Energetic Polymers
HNO f N O + H O
J. Phys. Chem. B, Vol. 101, No. 12, 1997 2131
2
(6)
Acknowledgment. This work was supported by the U.S.
Air Force Office of Scientific Research under Grant No.
F-049620-94-1-0125. We thank Dr. Rob Hunter (3M Company)
and Dr. Robert Wardle (Thiokol Corporation) for generously
supplying samples of GAP and PGN, respectively.
2
2
In our PGN experiments we did not detect HNO directly.
However, the UV photolysis experiments that we carried out
on matrix-isolated nitromethane provide indirect evidence for
the participation of this reaction in our study. When isolated
in an argon matrix, we observe the formation of HNO as a
photolysis product. However, upon warming the argon vapor-
izes and allows HNO molecules to react with each other,
forming N2O.
Finally, we note that the C-O-C backbone vibrational modes
of the PGN polymer are not strongly affected by the UV
photolysis reaction. This suggests that depolymerization of the
C-O bond in the polymer backbone is evidently not extensive.
References and Notes
(
1) Chen, J. K.; Brill, T. B. Combust. Flame 1991, 87, 157.
(2) Oyumi Y.; Brill, T. B. Combust. Flame 1987, 65, 127.
(3) Kubota, N.; Sonobe, T. Propellants Explos. Pyrotech. 1988, 13,
1
72.
(
(
(
4) Hass, Y.; Eliahu, Y. B.; Welner, S. Combust. Flame 1994, 96, 212.
5) Milligan, D. E. J. Chem. Phys. 1961, 35, 1491.
6) Hiskey, M. A.; Brower, K. R.; Oxley, J. C. J. Phys. Chem. 1991,
95, 3955.
(
7) Chen, J. K.; Brill, T. B. Thermochim. Acta 1991, 181, 71.
Conclusions
(8) Chen, J. K.; Brill, T. B. Combust. Flame 1991, 85, 479.
(9) Gelernter, G.; Browning, L. G.; Harris, S. R.; Mason, C. M. J.
UV laser photolysis and thermal pyrolysis of GAP appear to
follow the same general mechanism. This begins with elimina-
tion of molecular nitrogen from the azide functional groups and
rearrangement to a polymeric imine. Subsequent cross-linking
of the imine groups results in a higher molecular weight
polymer. For PGN, the UV laser photolysis proceeds by
elimination of NO and CH2O, followed by reactions of these
species to generate several other reaction products. In contrast,
pyrolysis begins with elimination of NO2 and CH2O, which is
followed by similar side reactions that oxidize the formaldehyde
to CO and CO2. In both compounds, the reactions begin with
and are centered around the azide or nitrate ester functional
groups. Aside from formation of polymeric carbonyl com-
pounds and small amounts of nonvolatile nitro and nitroso
polymers, the backbone of the polymer appears to be largely
inert to application of moderate heat or light stimuli.
Phys. Chem. 1956, 60, 1260.
(10) Leider, H. R.; Seaton, D. L. Nitrate Ester Decomposition and
Degradation of Molecular Weight in Nitrocellulose from Thermal Decom-
position of PBX-9404 Below 100C; Nat. Tech. Info. Serv.: Springfield,
VA, 1979.
(
11) Botcher, T. R.; Wight, C. A. J. Phys. Chem. 1993, 97, 9149.
Botcher, T. R.; Wight, C. A. J. Phys. Chem. 1994, 98, 5441.
12) Odiot, S. In Chemistry and Physics of Energetic Materials; Bulusu,
S. N., Ed.; Kluwer Academic Publishers: The Nertherlands, 1990; p 79.
(
(
(
13) Benderskii, A. V.; Wight, C. A. J. Chem. Phys. 1994, 101, 292.
14) Rao, C. N. R. Ultra-Violet and Visible Spectroscopy; The Whitefriars
Press Ltd.: London, 1975; p 30.
(15) Fateley, W. G.; Bent, H. A.; Crawford, B., Jr. J. Chem. Phys. 1959,
1, 204.
3
(
(
(
(
16) Hisatsune, I. C.; Devlin, J. P. Spectrochim. Acta 1961, 17, 218.
17) Brown, H. W.; Pimentel, G. C. J. Chem. Phys. 1958, 29, 883.
18) Jacox, M. E.; Rook, F. L. J. Phys. Chem. 1982, 86, 2899.
19) Wentrup, C. In Azides and Nitrenes; Scriven, E. F. V., Ed.;
Academic Press. Inc.: New York, 1984; p 395.