DiscussionW. H. Duewer (Lawrence Livermore Laboratory). The stationary state assumption for various free radicals has been tested by other modelers and found to be satisfactory for several. The large discrepancies found by the authors suggest errors either in the formulation of the model or of the steady state expression. Because only a few terms were included in the steady state approximations used, I suspect the latter. If there is no error (e.g., omitted significant terms) then the significant differences between your model and other models that cause your model not to behave 2483 as a steady state species should be described. In the case of

The thermolysis of tetramethylsilane has been studied in a pulsed stirred-flow system between 840 and 1055 K, yielding si&cantly different Arrhenius parameters above and below 950 K. The high temperature results relate to a non-chain mechanism, whence D(Me3Si-Me) = 355 k 6 kJ mol-I while the low temperature results relate to a short chain sequence. A mechanism for the latteriis shown by computer-aided numerical integration to be consistent with experimental results, as is a similar chain mechanism for the thermolysis of trimethylsilane.It has become apparent in recent years that the thermolysis of several methylsilanes proceeds by short-chain reactions of moderate complexity in which short-lived " double-bonded " intermediates, Si=CH2, play a prominent ~a r t . l '~ The thermolysis of hexamethyldisilane has been studied in detail over a range of pressure and temperatureY4* and has been shown to proceed by two concurrent chain sequences which differ in relative importance as the pressure is varied. Me,Si=CH, is an intermediate in the main chain sequence at low pressure, which produces trimethylsilane as the principal product. This chain could be inhibited, enabling a kinetic measurement of D(Me,Si-SiMe,) to be made.4 An earlier investigation of the thermolysis of tetramethylsilane between 8 10 and 980 K likewise produced evidence for a chain reaction involving Me,Si=CH,, but the mechanism was not fully el~cidated.~ As has recently been pointed out,'* 2* it follows from both of these studies that results obtained some years ago ' on the kinetics of thermolysis of trimethylsilane must be reinterpreted. We describe in this paper some experiments on the thermolysis of tetramethylsilane and some computer-aided calculations on the kinetics of thermolysis of tetramethylsilane, trimethylsilane and hexamethyldisilane which enable mutually consistent detailed mechanisms to be advanced for these thermolyses and which lead to a kinetic estimate of D(Me,Si-Me). \ / EXPERIMENTAL A N D RESULTSThe thermolysis of tetramethylsilane was investigated in the same apparatus as for he~amethyldisilane,~ using the " pulsed stirred-flow " technique which has been fully described. * Analysis was by g.l.c., supplemented by mass spectrometry. The temperature range covered by the experiments was from 840 to 1055 K, with single pulses of tetramethylsilane corresponding to initial concentrations between and mol dm-3. The carrier gas was purified nitrogen,8 at above atmospheric pressure. Residence times in the reactor were varied between 13 and 120 s, partly by altering the flow rate of the carrier gas and partly by using quartz reaction vessels t Present address :

The decomposition of neopentane was studied using the very-low-pressure pyrolysis (VLPP) technique at temperatures from 1000 to 1260 K. The derived Arrhenius parameters log k(sec-') = 17.3 -80.8/0 at 1000 K are consistent with AHp(t-butyl) = 8.4 kcal/mol. Using the above A factor, data on the decomposition of tetramethyltin yield DHO(Sn(CH3)S -CH3) = 69 f 2 kcal/mol.

Low-pressure limit rate constants for the three-body reactions, (1) Cl+NO2+M→ products and (2) ClO+NO2+M→ products, have been calculated according to Troe’s method [J. Chem. Phys. 66, 4758 (1977)]. The result for the reaction of Cl+NO2+M is in excellent agreement with the experimental finding of Niki et al. [Chem. Phys. Lett. 59, 78 (1978)]. An explanation is proposed to account for apparent discrepancy between the measured rate constants for ClO+NO2+M in the forward and reverse directions. The stratospheric implications are also discussed.