- Email: [email protected]
Mass spectrometric studies of the thermal decomposition of 1,3,5,7-tetranitro-1,3,5,7-tetraazacyclooctane (HMX)
Volume 80, number 2
1 June 1981
CHEMICAL PHYSICS LETTERS
MASS SPECTROMETRIC! STUDIES OF THE THERMAL DECOMPOSITION OF 1,3,5,7-TETRANITKO-1,3,5,7-TETRAAZACYCLOOCTANE
(HMX)
Milton FAEZBER and R.D. SRJYASTAVA Space Sciences. Inc. Monrovia. Cdifornia, USA Received 17 September 1980; in fmaI form 23 February 1981
The subIimation and thermal decomposition of HMX were studied by means of Langmuir evaporation and effusion mass spectrometry in the temperature range 175-275°C. Langmuir experiments showed that the primary mechanism for thermal decomposition is ring cleavage to two equd 148 amu species. Decomposition withm the effusion cell produced numerous smaller moiecules and free radicals due to the decomposition of the 148 amu molecule.
1. Introduction Several mechanisms for the thermal decomposition of 1,3,5,7-tetranitro-1,3,5,7-tetraa2acycIooctane (HMX) have been proposed. These incIude ring cIeavage of the C-N bonds, splitting off NO, groups and atomic migration from ring attached groups [ 1.2]_ To obtain a clearer defmition of the mechanism, the sublimation and thermal decomposition of HMX in the temperature range 175-275’C were investigated mass spectrometrically. Langmuir evaporation and effusion techniques were employed in these studies, This allowed the initial gas-phase products to be observed within a few microseconds, as well as observation of further decomposition products produced within effusion cell reactions_ Several experimental precautions were taken to ensure that the gaseous products observed were from parent precursors and not from electron impact within the ionization chamber of the mass spectrometer. The identification of the products in short times has allowed us to propose a mechanism for the thermal decomposition of HMX. A similar study involving the thermal decomposition of RDX has been reported [3].
2, Experimental apparatus and procedures Sublimation and thermal decompo~~on
studies of
HMX were conducted by both the effusion-mass spectrometric and the Langmuir evaporation-mass spectrometric methods. Pressure within the effusion cells are usually several orders of magnitude higher than the surrounding vacuum, allowing the gas products to collide with each other, the cell walls, and with the condensed phase many times prior to their effusing from the cell into the mass spectrometer chamber, which may result in secondary decomposition and fragmentation. The evaporation (Langmuir) method is one in -which the material is heated within the main vacuum chamber and allowed to enter the mass spectrometer chamber without further decomposition or collision with other gaseous molecules. Details of the dual vacuum chamber-quadrupole mass spectrometer system used in these experiments have been presented previously (41. An alumina effusion cell with an elongated orifice for beam collimation positioned within 5 cm of the ionization chamber of the mass spectrometer allowed gaseous species to be measured within 10 P. Experimental procedures relating to ion intensities, mass spectrometer resolution, calibration, and other parameters have also been described previously [S] . The formation of an ion from its parent precursor requires a lower ionization energy than for an ion formed from fragmentation of a larger molecule, For example, to produce NOi from (CH&%N02 (14.6 345
CHEMICAL PHYSICS LETTERS
Volume 80, number 2
1 June 1981
56 1
Fig. 2. Langmuir evaporation of HMX at 2OOOC.
74
A,
I
0ll-l”
Fig. 1. Thermal decomposition of HbIX in an effusion cell at 229C as a function of electron impact ionizing energy.
eV) requires ~5 eV more than from NO2 -+ NO; (9.75 eV) [6] _ High electron impact energies can produce species within the mass spectrometer that are not related to the combustion processes. Fig. 1 illustrates this effect on HMX decomposition products. At an energy of 50 eV a number of peaks are seen, which disappear at 30 eV_ The fragmentation of the nitramine ring requires a higher electron impact energy than that required for the removal of a group attached to the ring by a single bond. To guard against electron impact fragmentation, the mass spectrometer was operated at arriorrizingvoltage of 14-18eV,=l-2 eV above the appearance potential, wbichin nearly all cases allows only the formation of the ion from the parent species [7-IO]. Thermal effects on the ionization processes are negligible in the temperature range of these studies.
10-ll-lO-lo atm in the temperature range 98-130°C, and Crookes and Taylor determined vapor pressure data varying from 10m7 to 10B5 atm in the temperature range 188-213°C. Our cell evaporation and decomposition pressures varied from 5 X 10m7 to 5 X lOA atm in the temperature range 175-2?5”C. Langmuir evaporation experiments of HMX indicated sublimation and thermal decomposition occurring simultaneously. An example of these results is presented in fig. 2, which shows relative concentrations of the HMX molecule at 296 amu and its decomposition product at 148 amu. No other fragments were observed from the decomposition of HMX, which was directly heated in vacuum. Effusion experiments allowed further decomposition of the gas-phase molecules as a result of numerous
b.
3. Results Sublimation of HMX occurs at temperatures as low as 175OC. Results from fairly low-temperature effusion cell measurements by Rosen and Dickinson 11 l] and by Crookes and Taylor [12] indicate that HMX evaporates without appreciable decomposition_ Rosen and Dickinson reported a vapor pressure range of 346
Fig. 3. Evaporation and thermal decomposition of HMX in an effusion cell at 229C.
reactions with the walls and with the condensed phase. Fig. 3 shows the relative intensity of X-IM%at 22S°C as well as its decomposi$on products. Relative concentrations at 148,128,120,102,74 and 56 amu are depicted in fig. 3a. The peaks appearing in the low 1846 amu range are shown in fig. 3b.
Direct evaporation in vacuum (Langmuir experiments) allows any species derived from the condensed phase to enter the mass spectrometer without prior collision with any other species or with the surface area. From these experiments the mechanism proposed for the primary mode of solid phase HMX decomposition is ring cleavage, as
HZC-N-C-~-NO2 I 1 NO2
-+ H2F-y-C-y-
z
2
ZIN*
NO2 102 amu
The 148 amu fragment may form the more stable molecule shown below:
102 amu The 102 amu fragment splits off another NO, group, yielding
HN=&N=CH2
_
56 amu
Hz ~-N--NO2 NO&-kH,
This moiecule can split to produce two H2C=N - radicals. The 148 amu fragment can also split into two equai stabIe moIecuIes,
-
148 amu The products of sublimation and decomposition within the effusion cell undergo further decomposition to form the products shown in figs. 3a and 3b. The effusion cell decomposition products shown in fig. 3 include a relatively small quantity of a molecule or radical at 222 arnu. This would indicate a decomposition mode as
N-N02--? / =3
296 amu
H2w-No2 H2wNo2 +NO,. -+ -N-CH2
N02-N-CH2
which can rearrange to form a more stable resonating molecule,
N?? 296 amu
N
or, from the bond closure molecule,
56 amu
148 amu
1
+ NO, ,
H2 I H,C=N-C-N ,
No2_Y-J?02 _ N2~-$$+*oz .
\ v=
The peaks shown in figs. 3aand 3b may result from furthe) decomposition of the major product at 148 arnu through collisions within the cell. The reaction scheme is either
148 amu
4. Discussion
NO1-
1 June 1981
CHEMICAL PHYSICS-LETTERS
Volume 80, number 2
‘2-N-No* +
H2 H, y-r;r-C-T-NO,
I
I
v=,
4=2 r
NO2
2
74 amu
.
NO2 I48 amu
H C=N_NO
N-+-N
+- 2H2C=N-NO,
2
74 amu
The peaks at 120 amu, CH2Ns04, and at 128 amu, C3H4N402, are apparently produced in the effusion cell as a result of the reaction of the gaseous products with the condensed phase and with each other. These peaks have ako been observed by Goshgarian [ 11, Stals [ 133, and Suryanarayana et al. 1141. Goshgarian [l] has‘ postulated the formation of the 128 amu peak as
222 amu 347
CHEMICAL
Volume 80, number 2
PHYSICS
LETTERS
1 June 1981
The ring migration can result in either of two bond configurations:
HNO, + O-N-O-H 128 amu
The peak at 120 amu has been postulated by Stals [13] as ring migration of the NO, group,
28 amu
120 amu 148 amu
St& also indicated that the formation of the 132 peak results from the 148 molecule losing an 0 atom,
O-Y0 H
-
However, no evidence of such migration has been found in the experiments at this laboratory, since the 249 amu peak was not observed. Also, N-N bond rupture to produce NO, species does not appear to be a likely primary decomposition mode, since the peak at 250 amu was not observed in these studies and those by others [l]. Our studies yield an activation energy of 175 kJ/ mol for decomposition to NO, groups, as shown in fig. 4. Goshgarian 113 has reported activation energies of 159 + 8 kJjmo1 from 250 to 270°C and 175 + 8 kJ/mol from 271 to 28OOC. McCuire [ 161 reported an E, of 175 kJ/moI from the decomposition of HMX
to HZC=N-NO;Z
NO2
H:!
or
_
i
74 amu
v -+ O=N-N-CH,
H27-YNo2 NO,-N-CH, 148 amu
f 0 _
i32 amu
Beyer [lS] in ESR studies found considerable free radicals produced from the decomposition of HMX at 26O*C. He attributed this free radical spin resonance to the formation of HZCN. at 28 amu. Fig. 3b shows a high concentration of the 28 amu peak which, in all probability, is a mixture of the decomposition products CO, N2, and H2CN. He postuIated that the raciicai, HZCN, is derived from -+ HzC=N - + NO,
H2C=N-NO2 74 amu
E
t
N-r.02
p,A
No2
296 amu
348
,
C--H
,
to-=
+z
0
Hz= / N \ HC z
It has
28 arnu
;kh NO2 -
bonds of the highly strained ring configuration.
.
The possibility of ring migration to form HNO, has been raised by Shaw and Walker 121, 0
of sublimation obtained by Taylor and Crookes 1121 was 163 kJ/moi, also shown in fig. 4. The proposed primary themA decomposition mechanism of the cleavage of the HMX molecule into equal fragments of 148 amu is compatible with thermodynamic considerations. The very high specific heat of the solid HMX compound allows adsorption of sufficient enthalpy to cause the rupturing of the C-N
The en~~py
---e
\02-N
“* CH
Hz= f
\
\ HC 2 \
/
NO2
*
tfN02
-
y,=3
i p’oz
249 amu
Fi. 4. Temperature dependence of NO, from the thermaf decomposition of HMX in an effusion cell_
Volume 80, number 2
CHEMICAL PHYSICS LETTERS
been calculated [2] that the C.-N bond strength due to ring strains is 25 kcal/mol less than the normal C-N bond strength in aliphatic amines.
1 June 1981
t51 M. Farber and R-D. Snvastava, Combustion Flame 20 (1973) 33.
[cl H.M. Rosenstock, K. DraxI, B.W. Steiner and
J-T. Herron, J. Phys. Chem. Ref. Data 6 (1977) Suppl. 1
t71 M. Farber, R-D. Srivastava and O.M. Uy, J. Chem. Sot. Faraday Trans. I 68 (1972) 2d9.
Acknowledgement This research was supported by the Department of the Navy, Office of Naval Research, Material Sciences Division, Power Program.
Pi M. Farber and R-D. Srwastava, J. Chem. Sot. Faraday Trans. 170 (1974) 1581.
PI M. Farber and R-D. Srivastava, J. Chem. Sot. Faraday [W [ill
References [l]
B.B. Goshgarian, Final Report AFRPL-TR-78-76, October 1978, Air Force Rocket Propulsion Laboratory,
[I21 I131 1141
Edwards AFB, California. [2] R. Shaw and F.E. WaIker, J. Phys. Chem. 81 (1977) 2572.
[ 15 ]
[ 31 M. Farber and R.D. Srivastava, Chem. Phys. Letters 64 (1979) 307. [4] M. Farber, M.A. Frisch and H-C!. Ko, Trans. Faraday Sot. 65 (1969) 3202.
1161
Trans. 173 (1977) 1692. M. Farber and R.D. Snvastava, Chem Phys. Letters 51 (1977) 307. J. Rosen and C. Dlckmson, J. Chem Eng. Data 14 (1969) 120. J.W. Taylor and R.J. Crookes, J. Chem. Sot Faraday Trans. 172 (1976) 723. J. StaIs, Trans. Faraday Sot. 67 (1971) 1768. i3. Surya-yana, T. Axenrod and G-W-A. h-e, Org. hlass Spectrom. 3 (1970). R.A. Beyer, Conference on Thermal Decomposition, USAF Academy, Colorado, August 1979. R. hkGuire,Conference on Thermal Decomposition,
USAF Academy, Colorado, August 1979.
349
1 June 1981
CHEMICAL PHYSICS LETTERS
MASS SPECTROMETRIC! STUDIES OF THE THERMAL DECOMPOSITION OF 1,3,5,7-TETRANITKO-1,3,5,7-TETRAAZACYCLOOCTANE
(HMX)
Milton FAEZBER and R.D. SRJYASTAVA Space Sciences. Inc. Monrovia. Cdifornia, USA Received 17 September 1980; in fmaI form 23 February 1981
The subIimation and thermal decomposition of HMX were studied by means of Langmuir evaporation and effusion mass spectrometry in the temperature range 175-275°C. Langmuir experiments showed that the primary mechanism for thermal decomposition is ring cleavage to two equd 148 amu species. Decomposition withm the effusion cell produced numerous smaller moiecules and free radicals due to the decomposition of the 148 amu molecule.
1. Introduction Several mechanisms for the thermal decomposition of 1,3,5,7-tetranitro-1,3,5,7-tetraa2acycIooctane (HMX) have been proposed. These incIude ring cIeavage of the C-N bonds, splitting off NO, groups and atomic migration from ring attached groups [ 1.2]_ To obtain a clearer defmition of the mechanism, the sublimation and thermal decomposition of HMX in the temperature range 175-275’C were investigated mass spectrometrically. Langmuir evaporation and effusion techniques were employed in these studies, This allowed the initial gas-phase products to be observed within a few microseconds, as well as observation of further decomposition products produced within effusion cell reactions_ Several experimental precautions were taken to ensure that the gaseous products observed were from parent precursors and not from electron impact within the ionization chamber of the mass spectrometer. The identification of the products in short times has allowed us to propose a mechanism for the thermal decomposition of HMX. A similar study involving the thermal decomposition of RDX has been reported [3].
2, Experimental apparatus and procedures Sublimation and thermal decompo~~on
studies of
HMX were conducted by both the effusion-mass spectrometric and the Langmuir evaporation-mass spectrometric methods. Pressure within the effusion cells are usually several orders of magnitude higher than the surrounding vacuum, allowing the gas products to collide with each other, the cell walls, and with the condensed phase many times prior to their effusing from the cell into the mass spectrometer chamber, which may result in secondary decomposition and fragmentation. The evaporation (Langmuir) method is one in -which the material is heated within the main vacuum chamber and allowed to enter the mass spectrometer chamber without further decomposition or collision with other gaseous molecules. Details of the dual vacuum chamber-quadrupole mass spectrometer system used in these experiments have been presented previously (41. An alumina effusion cell with an elongated orifice for beam collimation positioned within 5 cm of the ionization chamber of the mass spectrometer allowed gaseous species to be measured within 10 P. Experimental procedures relating to ion intensities, mass spectrometer resolution, calibration, and other parameters have also been described previously [S] . The formation of an ion from its parent precursor requires a lower ionization energy than for an ion formed from fragmentation of a larger molecule, For example, to produce NOi from (CH&%N02 (14.6 345
CHEMICAL PHYSICS LETTERS
Volume 80, number 2
1 June 1981
56 1
Fig. 2. Langmuir evaporation of HMX at 2OOOC.
74
A,
I
0ll-l”
Fig. 1. Thermal decomposition of HbIX in an effusion cell at 229C as a function of electron impact ionizing energy.
eV) requires ~5 eV more than from NO2 -+ NO; (9.75 eV) [6] _ High electron impact energies can produce species within the mass spectrometer that are not related to the combustion processes. Fig. 1 illustrates this effect on HMX decomposition products. At an energy of 50 eV a number of peaks are seen, which disappear at 30 eV_ The fragmentation of the nitramine ring requires a higher electron impact energy than that required for the removal of a group attached to the ring by a single bond. To guard against electron impact fragmentation, the mass spectrometer was operated at arriorrizingvoltage of 14-18eV,=l-2 eV above the appearance potential, wbichin nearly all cases allows only the formation of the ion from the parent species [7-IO]. Thermal effects on the ionization processes are negligible in the temperature range of these studies.
10-ll-lO-lo atm in the temperature range 98-130°C, and Crookes and Taylor determined vapor pressure data varying from 10m7 to 10B5 atm in the temperature range 188-213°C. Our cell evaporation and decomposition pressures varied from 5 X 10m7 to 5 X lOA atm in the temperature range 175-2?5”C. Langmuir evaporation experiments of HMX indicated sublimation and thermal decomposition occurring simultaneously. An example of these results is presented in fig. 2, which shows relative concentrations of the HMX molecule at 296 amu and its decomposition product at 148 amu. No other fragments were observed from the decomposition of HMX, which was directly heated in vacuum. Effusion experiments allowed further decomposition of the gas-phase molecules as a result of numerous
b.
3. Results Sublimation of HMX occurs at temperatures as low as 175OC. Results from fairly low-temperature effusion cell measurements by Rosen and Dickinson 11 l] and by Crookes and Taylor [12] indicate that HMX evaporates without appreciable decomposition_ Rosen and Dickinson reported a vapor pressure range of 346
Fig. 3. Evaporation and thermal decomposition of HMX in an effusion cell at 229C.
reactions with the walls and with the condensed phase. Fig. 3 shows the relative intensity of X-IM%at 22S°C as well as its decomposi$on products. Relative concentrations at 148,128,120,102,74 and 56 amu are depicted in fig. 3a. The peaks appearing in the low 1846 amu range are shown in fig. 3b.
Direct evaporation in vacuum (Langmuir experiments) allows any species derived from the condensed phase to enter the mass spectrometer without prior collision with any other species or with the surface area. From these experiments the mechanism proposed for the primary mode of solid phase HMX decomposition is ring cleavage, as
HZC-N-C-~-NO2 I 1 NO2
-+ H2F-y-C-y-
z
2
ZIN*
NO2 102 amu
The 148 amu fragment may form the more stable molecule shown below:
102 amu The 102 amu fragment splits off another NO, group, yielding
HN=&N=CH2
_
56 amu
Hz ~-N--NO2 NO&-kH,
This moiecule can split to produce two H2C=N - radicals. The 148 amu fragment can also split into two equai stabIe moIecuIes,
-
148 amu The products of sublimation and decomposition within the effusion cell undergo further decomposition to form the products shown in figs. 3a and 3b. The effusion cell decomposition products shown in fig. 3 include a relatively small quantity of a molecule or radical at 222 arnu. This would indicate a decomposition mode as
N-N02--? / =3
296 amu
H2w-No2 H2wNo2 +NO,. -+ -N-CH2
N02-N-CH2
which can rearrange to form a more stable resonating molecule,
N?? 296 amu
N
or, from the bond closure molecule,
56 amu
148 amu
1
+ NO, ,
H2 I H,C=N-C-N ,
No2_Y-J?02 _ N2~-$$+*oz .
\ v=
The peaks shown in figs. 3aand 3b may result from furthe) decomposition of the major product at 148 arnu through collisions within the cell. The reaction scheme is either
148 amu
4. Discussion
NO1-
1 June 1981
CHEMICAL PHYSICS-LETTERS
Volume 80, number 2
‘2-N-No* +
H2 H, y-r;r-C-T-NO,
I
I
v=,
4=2 r
NO2
2
74 amu
.
NO2 I48 amu
H C=N_NO
N-+-N
+- 2H2C=N-NO,
2
74 amu
The peaks at 120 amu, CH2Ns04, and at 128 amu, C3H4N402, are apparently produced in the effusion cell as a result of the reaction of the gaseous products with the condensed phase and with each other. These peaks have ako been observed by Goshgarian [ 11, Stals [ 133, and Suryanarayana et al. 1141. Goshgarian [l] has‘ postulated the formation of the 128 amu peak as
222 amu 347
CHEMICAL
Volume 80, number 2
PHYSICS
LETTERS
1 June 1981
The ring migration can result in either of two bond configurations:
HNO, + O-N-O-H 128 amu
The peak at 120 amu has been postulated by Stals [13] as ring migration of the NO, group,
28 amu
120 amu 148 amu
St& also indicated that the formation of the 132 peak results from the 148 molecule losing an 0 atom,
O-Y0 H
-
However, no evidence of such migration has been found in the experiments at this laboratory, since the 249 amu peak was not observed. Also, N-N bond rupture to produce NO, species does not appear to be a likely primary decomposition mode, since the peak at 250 amu was not observed in these studies and those by others [l]. Our studies yield an activation energy of 175 kJ/ mol for decomposition to NO, groups, as shown in fig. 4. Goshgarian 113 has reported activation energies of 159 + 8 kJjmo1 from 250 to 270°C and 175 + 8 kJ/mol from 271 to 28OOC. McCuire [ 161 reported an E, of 175 kJ/moI from the decomposition of HMX
to HZC=N-NO;Z
NO2
H:!
or
_
i
74 amu
v -+ O=N-N-CH,
H27-YNo2 NO,-N-CH, 148 amu
f 0 _
i32 amu
Beyer [lS] in ESR studies found considerable free radicals produced from the decomposition of HMX at 26O*C. He attributed this free radical spin resonance to the formation of HZCN. at 28 amu. Fig. 3b shows a high concentration of the 28 amu peak which, in all probability, is a mixture of the decomposition products CO, N2, and H2CN. He postuIated that the raciicai, HZCN, is derived from -+ HzC=N - + NO,
H2C=N-NO2 74 amu
E
t
N-r.02
p,A
No2
296 amu
348
,
C--H
,
to-=
+z
0
Hz= / N \ HC z
It has
28 arnu
;kh NO2 -
bonds of the highly strained ring configuration.
.
The possibility of ring migration to form HNO, has been raised by Shaw and Walker 121, 0
of sublimation obtained by Taylor and Crookes 1121 was 163 kJ/moi, also shown in fig. 4. The proposed primary themA decomposition mechanism of the cleavage of the HMX molecule into equal fragments of 148 amu is compatible with thermodynamic considerations. The very high specific heat of the solid HMX compound allows adsorption of sufficient enthalpy to cause the rupturing of the C-N
The en~~py
---e
\02-N
“* CH
Hz= f
\
\ HC 2 \
/
NO2
*
tfN02
-
y,=3
i p’oz
249 amu
Fi. 4. Temperature dependence of NO, from the thermaf decomposition of HMX in an effusion cell_
Volume 80, number 2
CHEMICAL PHYSICS LETTERS
been calculated [2] that the C.-N bond strength due to ring strains is 25 kcal/mol less than the normal C-N bond strength in aliphatic amines.
1 June 1981
t51 M. Farber and R-D. Snvastava, Combustion Flame 20 (1973) 33.
[cl H.M. Rosenstock, K. DraxI, B.W. Steiner and
J-T. Herron, J. Phys. Chem. Ref. Data 6 (1977) Suppl. 1
t71 M. Farber, R-D. Srivastava and O.M. Uy, J. Chem. Sot. Faraday Trans. I 68 (1972) 2d9.
Acknowledgement This research was supported by the Department of the Navy, Office of Naval Research, Material Sciences Division, Power Program.
Pi M. Farber and R-D. Srwastava, J. Chem. Sot. Faraday Trans. 170 (1974) 1581.
PI M. Farber and R-D. Srivastava, J. Chem. Sot. Faraday [W [ill
References [l]
B.B. Goshgarian, Final Report AFRPL-TR-78-76, October 1978, Air Force Rocket Propulsion Laboratory,
[I21 I131 1141
Edwards AFB, California. [2] R. Shaw and F.E. WaIker, J. Phys. Chem. 81 (1977) 2572.
[ 15 ]
[ 31 M. Farber and R.D. Srivastava, Chem. Phys. Letters 64 (1979) 307. [4] M. Farber, M.A. Frisch and H-C!. Ko, Trans. Faraday Sot. 65 (1969) 3202.
1161
Trans. 173 (1977) 1692. M. Farber and R.D. Snvastava, Chem Phys. Letters 51 (1977) 307. J. Rosen and C. Dlckmson, J. Chem Eng. Data 14 (1969) 120. J.W. Taylor and R.J. Crookes, J. Chem. Sot Faraday Trans. 172 (1976) 723. J. StaIs, Trans. Faraday Sot. 67 (1971) 1768. i3. Surya-yana, T. Axenrod and G-W-A. h-e, Org. hlass Spectrom. 3 (1970). R.A. Beyer, Conference on Thermal Decomposition, USAF Academy, Colorado, August 1979. R. hkGuire,Conference on Thermal Decomposition,
USAF Academy, Colorado, August 1979.
349