No abstract

The homogeneous gas-phase decomposition kinetics of silane has been investigated using the single-pulse shock tube comparative rate technique (T = 1035-1184°K, P~d = 4000 Torr).The initial reaction of the decomposition SiH4 ! + SiHz + H2 is a unimolecular process in its pressure fall-off regime with experimental Arrhenius parameters of logkl (sec-I) = 13.33 f 0.28-52,700 f 1400/2.303RT. The decomposition has also been studied at lower temperatures by conventional methods. The results confirm the total pressure effect, indicate a small but not negligible extent of induced reaction, and show that the decomposition is first order in silane at constant total pressures. RRKM-pressure fall-off calculations for four different transition-state models are reported, and good agreement with all the data is obtained with a model whose high-pressure parameters are logAl (sec-') = 15.5, El(,) = 56.9 kcal, and AE;;, = 55.9 kcal. The mechanism of the decomposition is discussed, and it is concluded that hydrogen atoms are not involved. It is further suggested that silylene in the pure silane pyrolysis ultimately reacts with itself to give hydrogen: 2SiH2 -(SiZHd)* -(SiH3SiH)* -SizHz + Hz. The mechanism of H -D exchange absorbed in the pyrolysis of SiD4-hydrocarbon systems is also discussed.In a prior paper [l], we reported preliminary single-pulse shock tube kinetic results on the silane decomposition. We showed that the initial reaction of the decomposition is molecular Hz elimination [reaction (l)]+ (1)Production of hydrogen atoms in the overall reaction was suggested by large yields of HD found in the pyrolysis of SiD4 in the presence of excess toluene (see Table I). Silylene decomposition [reaction (2)] was postulated to be the source of the D atoms [l]. However, we now believe that this conclusion n + The simple bond rupture process, SiH4 -SiH3 + H, was eliminated as a possible initiation reaction because its high activation energy (93 kcal) would require a chain process with chain lengths in excess of lo6 in order to match observed reaction rates. Such long chains under shock conditions are clearly impossible. Thus one calculates that on average fewer than 50 collisions between silane (0.01%) and product molecules occur in a typical 200 Fsec shock period.

Part I : Kinetic d a t a for t h e s t a t i c system s i l a n e pyrolysis ( f r o m 640-703 K , 60-400 torr) are presented. For conversion from 3-304. first-order kinetics a r e obtained, with silane loss rates equal to half the hydrogen formation rates. At conversions greater than 4 0 4 , rate inhibition attributable to the back reaction of hydrogen with silylene occurs. Overall reaction rates are not surface sensitive. but disilane and trisilane yield maxima under some conditions are. A nonchain mechanism capable of describing quantitatively all stages of the silane pyrolysis is proposed. Post 1.05; initiation is both homogeneous (gas phase) and heterogeneous (on the walls), and reaction intermediates are silylenes and disilenes. Free radicals are not involved a t any stage of the reaction. Rate data a t high conversions and with added hydrogen provide kinetics for the addition of silylene to hydrogen [reaction (-1iI relative to its addition to silane [reaction (2)l: k , / k , = 1 0~0 6 5 x e 3 2 0 " c~i ' H T . With E2 = 1300 cal, this gives a highpressure activation energy for silylene insertion into hydrogen of E -, = 8200 cal.Part 11: An analysis is made of each rate constant of the silane mechanism and the modeling results are compared with experimental results. Agreement is excellent. It is concluded that the dominant sink reaction for silylene intermediates is 1,2-H, elitnination from disilane (followed by SiYHl polymerization and wall deposition). The model is in accord with slow isomerization between disilene and silylsilylene and near exclusive 1,2-H2 elimination from S i L H G .It is also concluded t h a t disilene is about 10 kcal/mol more stable than silylsilylene and that the activation energy for isomerization of silylsilylene to disilene is greater than 26 kcal/;nol.

No abstract

No abstract