J. Phys. Chem. 1993,97, 9392-9396
9392
Silicon-Carbon Bond Formation Kinetics: Study of the Reactions of CH3 with SiH3, Si(CH&, and Sic13 Jukka T. Niiranent and David Cutman' Department of Chemistry, Catholic University of America, Washington, D.C. 20064 Received: January 14, 1993; In Final Form: June 15, 1993 The kinetics of three Si-C bond-forming association reactions were investigated at the high-pressure limit: SiH3 C H 3 (l), Si(CH& CH3 (2), and Sic13 CH3 (3). Rate constants were measured using a heatable tubular reactor coupled to a photoionization mass spectrometer. The two radical reactants were produced simultaneously (CH3 always in great excess) using pulsed 193-nm photolysis of suitable precursors diluted in helium, and the radical decays were monitored in time-resolved experiments. The radical decay profiles were fitted to appropriate expressions to obtain the desired rate constants. Reaction 1 was studied between 301 and 526 K yielding the following Arrhenius expression for the association reaction: kl = (5.6 f 2.4) X 10-l' exp((3.0 f 1.6) kJ mol-'/RT). (All rate constants are in the units cm3 molecule-' s-l.) Reaction 2 was investigated between 306 exp((2.4 f 1.4) kJ mol-l/RT). Reaction 3 was and 526 K yielding the expression k2 = (5.2 f 2.2) X studied a t one temperature, 303 K, where k3 = (1.1 f 0.4) X Treating these association reactions as cross-combination reactions, the measured rate constants were found to be predicted with reasonable accuracy using the geometric mean rule and the rate constants of the related self-association reactions of the reactant radicals. The mechanisms of these reactions are discussed.
+
+
+
Introduction Silicon carbide (Sic), whether used as a ceramic material or in electronics applications,is generally produced using gas-phase precursors with heated substrates.'" Models of the complex gasphase and gas-surface kinetics (as well as equilibria) occurring during S i c production processes are now being developed to provide a more fundamental understanding of the chemical transformations responsible for the desired properties of the silicon carbide produced.7-11 Thermodynamic information needed in these models on participating chemical species is being provided from both experimental12.13 and computationallkl6 studies, the latter particularly using bond-additivity-corrected ab initio methods. However, the kinetic information required by these models is still very sparse, requiring extensive use of estimated rate parameters. We are investigating the chemical kinetics of substituted and unsubstituted silyl radicals (SiR,) both to broaden our understanding of the reactivity of these silicon-centered reaction intermediates and to obtain new kinetic and thermodynamic information needed to more accurately portray their role in the synthesis of silicon-containing materialssuch as S i c in gas-surface processes. To date, we have investigated the kinetics and thermochemistry of the reactions of SiH3 with HBr and HI (in part to provide a direct and accurate determination of the heat of formationof SiH3)17 and the kinetics of the reactionsof SiH3,1e Si(CH3)3,19 and Sic1319 with 02.Comparisons of the reactivities of SiR3 radicals with those of the corresponding carbon-centered radicals, CR3, were provided.17-19 We report here the results of our study of the kinetics of three direct gas-phase S i 4 bond-forming association reactions involving SiR3 radicals: SiH,
+ CH,
Si(CH,), SiCl,
-
+ CH,
+ CH,
products products
products
(1)
(2)
No kinetic studies of SiR3 + CR'3 reactions have yet been reported. Rate constants for reactions 1-3 were obtained, in two cases as a function of temperature. Details of the experimental method used and the results obtained are presented here together with a discussion of the reactivity of SiR3 radicals in association reactions.
Experimental Section The experimental facility, which consists of a heatable 2.2cm-i.d. tubular Pyrex reactor coupled to a photoionization mass spectrometer, has been described.*O In order to reduce the heterogeneous loss of SiR3 radicals wall coatingswere used, either poly(tetrafluoroethylene)2* or Halocarbon Wax.22 Radicals are produced nearly uniformly along the reactor using pulsed unfocused excimer laser photolysis of suitableprecursor molecules. Diffusion in this narrow low-pressure reactor produces homogeneous conditions very near the start of the reaction. Gas is continuously sampled from a small conical orifice in the wall of the reactor and is formed into a beam by a conical skimmer before passing through the ion source of the mass spectrometer. Ion signals of reactants and/or products are monitored in timeresolved experiments. In the current experiments, CH3 and SiRp radicals were produced simultaneouslyin the tubular reactor using pulsed (-3 Hz) 193-nmphotolysis of two suitable precursor molecules which were highly diluted in helium carrier gas. Both CH3 and SiR3 decay profiles were monitored. The gas flow rate through the reactor, =3 m s-1, was adequate to completely replace the photolyzed gases between laser pulses. CH3 was produced from acetone:
-
SiH,
+I
(5)
-
C,H,
+ SiC1,
(7)
193 nm
CH,COCH3 2CH3+C0 (4) The different silyl radicals were obtained using the following sources: 193 nm
SiH,I
(3)
Present address: Department of Physical Chemistry, University of Helsinki, Helsinki, Finland.
0022-3654/93/2097-9392$04.00/0
193 nm
C,H,SiCl,
0 1993 American Chemical Society
For each of these precursors,there are additional photolysis routes at 193 nm. In the case of reaction 4, the route indicated is dominant. Pilling et al. have reported that reaction 4 accounts for 95% of the photodecomposition of acetone at 193 nm1.23 There are estimates or measurements of the importance of reaction 5 (10%) in the photolysis of SiH3I at 193 nmZ4 and reaction 6 (71 f 7%) in the photolysis of Si2(CH3)6 at 206 nm.26 Our observations indicate that reaction 7 is a major photolysis route at 193 nm. In our experiments, initial conditions (in particular, the two precursor concentrations and the laser fluence) were selected (1) to produce CH3 radicals in a concentration range in which selfCH3) could easily be monitored and combination (CH3 characterized(( 1-10) X 1012radicalscm-3)and (2) to haveinitial CH3 concentrations in large excess over the total combined concentration of all the remaining radicals in the system. The initial concentration of the radical pool from the photolysis of the silicon-radical precursor (estimated from its observed extent of photodecomposition) was typically a factor of 20-100 less than that of [CH3]0. Under these conditions, the self-combination of methyl radicals was essentially unperturbed by the presence of the other radicals in the system. [SiR& was kept low so that self-combination (SiR3 SiR3) was a negligible loss process for this radical. However, the cross-recombination reaction was always important (reaction 1, 2, or 3). Heterogeneous loss of both SiR3 and CH3 occurred and was a major sink for the SiR3 radicals. The kinetic mechanism for the important loss processes of CH3 and SiR3 in these experiments is as follows:
+
+
+ -
CH3 CH,
C,H,
(8)
heterogeneous loss
SiR, SiR,
+ CH,
-
CH,
product
(9) (10)
heterogeneous loss
(11)
For this mechanism and for the initial conditionsdescribed above, the temporal behavior of the SiR3+ ion signal, SI(SiR3+) (which is proportional to the SiR3 concentration), is provided by eq I: SI(SiR3+)
-
So(SiR3+) k9 exp(-k,,r)[(lk~[CH,I,(l
- exp(-kgt))
+ kg)
In our experiments, the SiR3+profile is fitted to eq I to obtain the cross-combination rate constant, klo. No knowledge of the initial concentrationof SiR3is required to obtain this rate constant. Previously, Slagle et al?' as well as Hanning-Lee and Pilling2s have taken advantage of such initial conditions to determine rate constants of the 0 C3H5 and H C3H5 cross-combination reactions, respectively. In these studies, decay profiles of 0 (or H ) and C3H5 were monitored. No knowledge of [0]0 (or [H]o) was required to obtain the desired rate constants from the observations since [0]0 (or [Hlo)
