- Email: [email protected]
Pulsed laser pyrolysis of dimethylnitramine and dimethylnitrosamine
Spectrochimica Acfa, Vol. 43A, Printed in Great Britain.
PULSED
Donald
No
2, pi’. 237-239,
LASER PYROLYSIS
F. McMillen,
Department
S. Esther
of Chemical
0584-85391.87 $3.00+0.00 Pergamon Journals Ltd.
1987.
OF DIMETHYLNITRAMINE
Nigenda,
Kinetics,
AND DIMETHYLNITROSAMINE
Alicia C. Gonzalez,
SRI International,
and David M. Golden
Menlo Park, CA
94025
The decomposition of dimethylnitramine (DMNA) is often invoked as a model for the thermal decomposition of the cyclic nitramines, which have been widely used for many years In spite of earlier workl-li and several recent reexaminaas explosives and propellants. tions of the thermal decomposition of DMNA, the primary step(s) and the secondary reactions we summarize laser-powered In this presentation, are not yet satisfactorily understood. The latter studies homogeneous pyrolysis (LPHP) studies carried out in this laboratory. include results from both end-product GC/MS analyses and also a real-time molecular-beammonitored
LPHP system.
ulsed-laser pyrolysis used here (GC/MS-monitored) has been The principal forn of recently described in detail. ! 9' Briefly, a CO2 laser is used to heat the substrate via an absorbing but unreactive gas (e.g., SFS): which then transfers its kinetic energy by collision (total pressure 1OD torr, collision frequency = log/s), bringing the substrate to Reaction then occurs in roughly 10 p available before cooling high temperature in 1-3 p. takes place as the acoustic expansion wave moves radially outward from the center of the Figure 1 shows a schematic of the LPHP apparatus. cell.
With the LPHP technique, surface catalyzed reactions are essentially precluded, since this time scale for heating, reaction, and cooling is far shorter than the millisecond time scale of diffusion to the walls. The major qualification to this assertion is that adsorption of IR-absorbing reactants or products on the KC1 windows could result in decomposition This is generally unlikely since only extremely strongly absorbing of this adsorbed phase. materials would be heated to a sufficiently high temperature when present at only a few monolayers thickness on the transparent, cold windows.
237
238
DONALDF.MCMILLEN~~UI. The
need
to explicitly
measure
the
temperature
corresponding
to any
particular
mea-
surement of k is eliminated by use of the "comparative rate" technique. This involves concurrent determination of the fractional decomposition of a temperature standard, a second "substrate" whose decomposition rate temperature dependence is already well known, and therefore whose fractional decomposition defines an "effective" temperature that is a convolution of temperature over time and space within the heated portion of the cell. Given that the temperature dependence of the substrate and the temperature standard are not of both materials is less than greatly different, and that the fractional decomposition about 2 % manner.gsd
the standard and the substrate "track" the temperature variations in a similar Data reduction simply involves determining the slope and intercept of a plot of log kunknown vs "9 kstandard* which correspond to the ratio of activation energies, and the difference in the log A values, respectively. As discussed in References 6 and 9, uncertainties in reaction time and in cell volumes cause no error in the derived (slope of the comparative rate plot) and only small errors in derived activation energies A-factors.
The extent to which the comparative rate method accommodates the temperature and reaction time variations inherent in this laser pyrolysis technique has been amply demonstrated by estimates of the errors involved and by testing the technique using a series of "unknowns" whose Arrhenius parameters have already been reported in the literature. 6,8,10 This has provided values in good agreement with the literature parameters for a series of molecular elimination reactions. The precision achievable indicates that the pulse-topulse and minute-to-minute laser variations, which existed under the conditions of this experiment, did not give rise to any distortion of the derived temperature dependence. Tests of this type make it clear that the comparative rate method, applied to pulsed laser pyrolysis, is very successful in accommodating the temperature variations inherent in the method and thereby provides an accurate measure of the temperature dependence of the substrate; that is, the physics involved in laser pyrolysis present no problem. Data from the GC/MS-monitored LPHP system give, for OMNA disappearance, an average value of log k (set-I) = 13.2 + 0.8 - 35.7 f 2.5/2.3 RT, as shown in Figure 2. This temperature dependence is considerably lower than that reported by some other workers, and at this point, neither should be considered immune to systematic error. The current results were obtained from decomposition in the presence of roughly a hundredfold excess each of cyclopentane and NO. Cycopentane is an effective scavenger for H-atoms, and NO suppresses their formation by trapping amino radicals before they can eliminate H-atoms. Thus, these results should be essentially free of reactions secondary H-atom involving either the substrate or the temperature standard. Variations in temperature
-2.0
F
-2.5
I
-2.0
I
I
-1.5
I
I
-1.0
I
I
-0.5
I
1
0
dependence aside, the central finding, common to all but the lowest pressure studies, that must be accounted for is the observation of dimethylnitrosamine (DMNO) in yields of 60 to 100%. This observation spans a wide range of temperatures, substrate pressures, reaction times, and both gas composition and pressure, and is not easy to reconcile with N-NO2 bond scission, the reaction that Initial virtually all previous efforts had led researchers to expect as the initial step. N-NO2 scission would mean DMNO would be formed largely by recombination of dimethylamino radicals with NO. The requirements that must be met among the various secondary reactions, if all the data are to be satisfactorily explained, have been quantitatively delineated with the aid of a numerical model of the reaction system. The principal requirement is
239
Pulsed laser pyrolysis of DMNA
that there be a bottleneck in the oxidation of the dimethylamino radical in the first stage of its degradation, such that dimethylamino radicals must be either completely oxidized or stability of the first oxidation This seems unlikely, given the anticipated untouched. intermediate, a nitroxyl radical. Current work in which inhibition by hydrocarbon scavenger and NO is varied, as well as work with a molecular-beam mass spectometrically-monitored version of LPHP, is expected to provide further characterization of the rapid sequence of reactions leading to the nitrosamine. This work is supported by the U.S. Army Research Office, Contract and by the U.S. Air Force Office of Scientific Research under Contract
Acknowledgement: DAAL03-86-K-0030 F49620-85-K-006.
No. No.
References 1.
2.
3.
Fluornoy,
J. M. "Thermal
1962, 36,
1106.
Korsunski,
B.
1967, 174,
1128.
McMillen,
D.
L.;
F.,
Decomposition
Dubovitskii,
Barker,
J.
F.
R.,
of Gaseous
I., and
Lewis,
K.
Dimethylnitramine,"
Sitonina,
E.,
G. V., Dokl.
Trevor,
P.
L.,
"Mechanisms of Nitramine Decomposition: Very Low-Pressure Final Dimethylnitramine," Report, SRI Project PYU-5787, 0115/117). DOE Contract No. EY-76-C-03-0115. 4.
Lloyd,
S. A., Instead,
5.
Wodtke,
A. M., private
6.
McMillen, e.,
7.
8.
9.
Pai, H.-L., 4494. (a) (b) (c)
10.
and
Golden,
Pyrolysis 18 June
Materials,
Nauk
SSSR,
D.
M.
of HMX and I979 (SAN
1985, 3,
187.
G. P., Golden,
D. M., J. Phys.
709.
A. C., Larson,
J. Phys. Chem.,
Energetic
Akad.
communication.
D. F., Lewis, K. E., Smith,
1983, 86,
Gonzalez,
M. E., and Lin, M. E., J.
J. Chem. Phys.,
1985, E,
Specht,
C. W., McMillen,
E., Berman,
Tsang, W. J., Chem. Phys., Ibid., 1964, 41, 2487. 2817. Ibid., 1967, 3,
Shaub, W. M., Bauer,
D. F., and Golden,
D. M.,
4809. M. R., and Moore,
1964, 2,
C. B., J.
Chem.
1171.
S. H., Int. J. Chem. Kinetics,
1975, I,
509.
Phys.,
1982, _' 77
PULSED
Donald
No
2, pi’. 237-239,
LASER PYROLYSIS
F. McMillen,
Department
S. Esther
of Chemical
0584-85391.87 $3.00+0.00 Pergamon Journals Ltd.
1987.
OF DIMETHYLNITRAMINE
Nigenda,
Kinetics,
AND DIMETHYLNITROSAMINE
Alicia C. Gonzalez,
SRI International,
and David M. Golden
Menlo Park, CA
94025
The decomposition of dimethylnitramine (DMNA) is often invoked as a model for the thermal decomposition of the cyclic nitramines, which have been widely used for many years In spite of earlier workl-li and several recent reexaminaas explosives and propellants. tions of the thermal decomposition of DMNA, the primary step(s) and the secondary reactions we summarize laser-powered In this presentation, are not yet satisfactorily understood. The latter studies homogeneous pyrolysis (LPHP) studies carried out in this laboratory. include results from both end-product GC/MS analyses and also a real-time molecular-beammonitored
LPHP system.
ulsed-laser pyrolysis used here (GC/MS-monitored) has been The principal forn of recently described in detail. ! 9' Briefly, a CO2 laser is used to heat the substrate via an absorbing but unreactive gas (e.g., SFS): which then transfers its kinetic energy by collision (total pressure 1OD torr, collision frequency = log/s), bringing the substrate to Reaction then occurs in roughly 10 p available before cooling high temperature in 1-3 p. takes place as the acoustic expansion wave moves radially outward from the center of the Figure 1 shows a schematic of the LPHP apparatus. cell.
With the LPHP technique, surface catalyzed reactions are essentially precluded, since this time scale for heating, reaction, and cooling is far shorter than the millisecond time scale of diffusion to the walls. The major qualification to this assertion is that adsorption of IR-absorbing reactants or products on the KC1 windows could result in decomposition This is generally unlikely since only extremely strongly absorbing of this adsorbed phase. materials would be heated to a sufficiently high temperature when present at only a few monolayers thickness on the transparent, cold windows.
237
238
DONALDF.MCMILLEN~~UI. The
need
to explicitly
measure
the
temperature
corresponding
to any
particular
mea-
surement of k is eliminated by use of the "comparative rate" technique. This involves concurrent determination of the fractional decomposition of a temperature standard, a second "substrate" whose decomposition rate temperature dependence is already well known, and therefore whose fractional decomposition defines an "effective" temperature that is a convolution of temperature over time and space within the heated portion of the cell. Given that the temperature dependence of the substrate and the temperature standard are not of both materials is less than greatly different, and that the fractional decomposition about 2 % manner.gsd
the standard and the substrate "track" the temperature variations in a similar Data reduction simply involves determining the slope and intercept of a plot of log kunknown vs "9 kstandard* which correspond to the ratio of activation energies, and the difference in the log A values, respectively. As discussed in References 6 and 9, uncertainties in reaction time and in cell volumes cause no error in the derived (slope of the comparative rate plot) and only small errors in derived activation energies A-factors.
The extent to which the comparative rate method accommodates the temperature and reaction time variations inherent in this laser pyrolysis technique has been amply demonstrated by estimates of the errors involved and by testing the technique using a series of "unknowns" whose Arrhenius parameters have already been reported in the literature. 6,8,10 This has provided values in good agreement with the literature parameters for a series of molecular elimination reactions. The precision achievable indicates that the pulse-topulse and minute-to-minute laser variations, which existed under the conditions of this experiment, did not give rise to any distortion of the derived temperature dependence. Tests of this type make it clear that the comparative rate method, applied to pulsed laser pyrolysis, is very successful in accommodating the temperature variations inherent in the method and thereby provides an accurate measure of the temperature dependence of the substrate; that is, the physics involved in laser pyrolysis present no problem. Data from the GC/MS-monitored LPHP system give, for OMNA disappearance, an average value of log k (set-I) = 13.2 + 0.8 - 35.7 f 2.5/2.3 RT, as shown in Figure 2. This temperature dependence is considerably lower than that reported by some other workers, and at this point, neither should be considered immune to systematic error. The current results were obtained from decomposition in the presence of roughly a hundredfold excess each of cyclopentane and NO. Cycopentane is an effective scavenger for H-atoms, and NO suppresses their formation by trapping amino radicals before they can eliminate H-atoms. Thus, these results should be essentially free of reactions secondary H-atom involving either the substrate or the temperature standard. Variations in temperature
-2.0
F
-2.5
I
-2.0
I
I
-1.5
I
I
-1.0
I
I
-0.5
I
1
0
dependence aside, the central finding, common to all but the lowest pressure studies, that must be accounted for is the observation of dimethylnitrosamine (DMNO) in yields of 60 to 100%. This observation spans a wide range of temperatures, substrate pressures, reaction times, and both gas composition and pressure, and is not easy to reconcile with N-NO2 bond scission, the reaction that Initial virtually all previous efforts had led researchers to expect as the initial step. N-NO2 scission would mean DMNO would be formed largely by recombination of dimethylamino radicals with NO. The requirements that must be met among the various secondary reactions, if all the data are to be satisfactorily explained, have been quantitatively delineated with the aid of a numerical model of the reaction system. The principal requirement is
239
Pulsed laser pyrolysis of DMNA
that there be a bottleneck in the oxidation of the dimethylamino radical in the first stage of its degradation, such that dimethylamino radicals must be either completely oxidized or stability of the first oxidation This seems unlikely, given the anticipated untouched. intermediate, a nitroxyl radical. Current work in which inhibition by hydrocarbon scavenger and NO is varied, as well as work with a molecular-beam mass spectometrically-monitored version of LPHP, is expected to provide further characterization of the rapid sequence of reactions leading to the nitrosamine. This work is supported by the U.S. Army Research Office, Contract and by the U.S. Air Force Office of Scientific Research under Contract
Acknowledgement: DAAL03-86-K-0030 F49620-85-K-006.
No. No.
References 1.
2.
3.
Fluornoy,
J. M. "Thermal
1962, 36,
1106.
Korsunski,
B.
1967, 174,
1128.
McMillen,
D.
L.;
F.,
Decomposition
Dubovitskii,
Barker,
J.
F.
R.,
of Gaseous
I., and
Lewis,
K.
Dimethylnitramine,"
Sitonina,
E.,
G. V., Dokl.
Trevor,
P.
L.,
"Mechanisms of Nitramine Decomposition: Very Low-Pressure Final Dimethylnitramine," Report, SRI Project PYU-5787, 0115/117). DOE Contract No. EY-76-C-03-0115. 4.
Lloyd,
S. A., Instead,
5.
Wodtke,
A. M., private
6.
McMillen, e.,
7.
8.
9.
Pai, H.-L., 4494. (a) (b) (c)
10.
and
Golden,
Pyrolysis 18 June
Materials,
Nauk
SSSR,
D.
M.
of HMX and I979 (SAN
1985, 3,
187.
G. P., Golden,
D. M., J. Phys.
709.
A. C., Larson,
J. Phys. Chem.,
Energetic
Akad.
communication.
D. F., Lewis, K. E., Smith,
1983, 86,
Gonzalez,
M. E., and Lin, M. E., J.
J. Chem. Phys.,
1985, E,
Specht,
C. W., McMillen,
E., Berman,
Tsang, W. J., Chem. Phys., Ibid., 1964, 41, 2487. 2817. Ibid., 1967, 3,
Shaub, W. M., Bauer,
D. F., and Golden,
D. M.,
4809. M. R., and Moore,
1964, 2,
C. B., J.
Chem.
1171.
S. H., Int. J. Chem. Kinetics,
1975, I,
509.
Phys.,
1982, _' 77