Kinetics and mechanism of pyrolysis of methyltrichlorosilane

Article abstract of DOI:10.1021/jp0306592

The thermal decomposition of methyltrichlosilane (MTS) was studied in a flow system in the temperature range of 825-977 K and pressure range of 10-120 Torr. Yields of products were measured by gas chromatography. The rate constant, k₁, for the initiation reaction was determined from the sum of the rates of the termination reactions. The Arrhenius expression for this reaction at the high-pressure limit was obtained from a nonlinear least-squares fit to the experimental data using the Troe factorization method, k_1x = (9.6 = 2.5) × 10¹⁹ exp (-(392 ± 18) kJ mol^-1/RT) s^-1. The rate constants for hydrogen abstraction, k₂, and chlorine abstraction, k₃, from MTS by methyl radicals were also calculated on the basis of experimental measurements The Arrhenius expression for hydrogen abstraction was k₂ = (5.1 ± 04) × 10⁸ exp(-(61 ± 3) kJ mol^-1/RT) L mol^-1s^-1 and for chlorine abstraction was k₃ = (1.5 ± 0.5) × 10⁹ exp(-(72 ± 6) kJ mol^-1/RT) L mol^-1s^-1.

Full text of DOI:10.1021/jp0306592

1946
J. Phys. Chem. A 2004, 108, 1946-1952
Kinetics and Mechanism of Pyrolysis of Methyltrichlorosilane†
S. Hosein Mousavipour,* Vahid Saheb, and Shapour Ramezani
Department of Chemistry, College of Sciences, Shiraz UniVersity, Shiraz, I. R. Iran
ReceiVed: May 27, 2003; In Final Form: October 17, 2003
The thermal decomposition of methyltrichlorosilane (MTS) was studied in a flow system in the temperature
range of 825-977 K and pressure range of 10-120 Torr. Yields of products were measured by gas
chromatography. The rate constant, k1, for the initiation reaction was determined from the sum of the rates
of the termination reactions. The Arrhenius expression for this reaction at the high-pressure limit was obtained
from a nonlinear least-squares fit to the experimental data using the Troe factorization method, k1∞ ) (9.6 (
2.5) × 1019 exp (-(392 ( 18) kJ mol-1/RT) s-1. The rate constants for hydrogen abstraction, k2, and chlorine
abstraction, k3, from MTS by methyl radicals were also calculated on the basis of experimental measurements.
The Arrhenius expression for hydrogen abstraction was k2 ) (5.1 ( 0.4) × 108 exp(-(61 ( 3) kJ mol-1/RT)
L mol-1s-1 and for chlorine abstraction was k3 ) (1.5 ( 0.5) × 109 exp(-(72 ( 6) kJ mol-1/RT) L mol-1s-1.
Introduction
In 1987, Gordon and Truong6 studied the possible primary
pyrolysis processes of methylsilane, H3SiCH3, theoretically.
They suggested a chain radical mechanism for the pyrolysis of
H3SiCH3. They investigated the way of formation of methane
during the process. In 1991, Besmann and Sheldon7 worked on
the reaction of MTS and hydrogen. They suggested the
formation of SiC and HCl as the overall products of the reaction.
In 1991, Besmann and co-workers8 in vapor-phase fabrication
and properties of continuous-filament ceramic composites
studied the characteristic of SiC film produced in chemical vapor
deposition. In 1992, Besmann and co-workers9 studied the
depletion effect of silicon carbide deposition from MTS and
dependence of SiC deposition on total system pressure. In their
study, they reported the overall reaction CH3SiCl3 f SiC +
3HCl and they did not look at the mechanism of pyrolysis of
MTS.
The chemistry of chlorinated organosilanes is of intrinsic
interest because of their important role in chemical vapor
deposition processes. Gas-phase study of chemical reactions is
useful for determining the formation and consumption of
reactive intermediates important to the film-growth rate, deposit
composition, and reactor behavior in chemical vapor deposition
(CVD) processes. Silicon carbide (SiC) prepared by CVD finds
many structural and electronic applications.1 CH3SiCl3 (MTS)
is one of the precursors which could be used in CVD processes
to produce a coating film of silicon carbide SiC on the surfaces.
The understanding of the kinetics of gas-phase pyrolysis of MTS
is of particular interest. To the best of our knowledge, the
kinetics of thermal decomposition of MTS is not studied in
detail, experimentally.
Different groups have studied the kinetics of the deposition
of some methylsilanes.2
Allendorf and Melius10 in 1993 studied the thermochemistry
of molecules in the Si-C-Cl-H system including MTS.
In 1994, Cagliostro and co-workers11 compared the pyrolysis
products of dichloro-dimethylsilane in CVD process and forma-
tion of SiC on silica in the presence of hydrogen or argon.
Burgess and Lewis3 in 1974 studied the kinetics of the
reduction of MTS by hydrogen. They reported the following
Arrhenius expression for the decomposition of MTS.
k ) 7.6 ×1014 exp(-336 kJ mol-1/RT) s-1
They reported the dissociation products of MTS in the
presence of hydrogen are likely to be SiHCl3 and CH4, although
other species, such as SiCl2, may be important.
Osterheld et al.12 in 1994 studied dissociation reaction of
MTS, theoretically. They studied different possible paths for
the decomposition of this molecule. They used RRKM theory
to calculate the rate constants for the three most probable paths
for the decomposition of MTS. According to their results, Si-C
bond cleavage to form CH3 and SiCl3 is the most likely path
for the unimolecular decomposition of CH3SiCl3. The other two
pathways for the decomposition of this compound are 1,2
elimination of HCl and C-H bound cleavage. They have
reported high-pressure limiting values of the rate constants for
each reaction in a temperature range of 800-1500 K.
In 1975, Kobayashi et al.4 worked on the formation of carbon-
excess SiC from pyrolysis of MTS. They investigated pyrolysis
of MTS in the presence of argon and hydrogen at high
temperatures. They determined the effect of pressure and
temperature on the C/Si ratio and the weight of the deposit.
They also observed that the amount of excess carbon was larger
in the presence of argon than in the presence of hydrogen. Their
study suffers the lack of mechanistic explanation for their
observation.
In 1995, Gordon and co-workers13 studied ab initio molecular
Davidson and Dean5 have studied the kinetics and mechanism
of pyrolysis of some methylchlorosilanes. They suggested that
pyrolysis of these compounds proceeded mainly by radical chain
mechanism.
+
orbital study on the unimolecular decomposition of CH3SiH2
.
In their study, they suggested a chain reaction mechanism for
the pyrolysis of methylsilane.
The purpose of the present study is to investigate the pyrolysis
of MTS at the temperature range of 825-977 K and pressure
range of 10-120 Torr.
* Address correspondence to this author.
† This article is dedicated to Professor Philip D. Pacey for his guidance
and mentorship.
10.1021/jp0306592 CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/24/2004

Pyrolysis of Methyltrichlorosilane
J. Phys. Chem. A, Vol. 108, No. 11, 2004 1947
Proposed Mechanism and Derivation of the Rate Laws.
A chain reaction mechanism for the pyrolysis of MTS is
proposed as
CH3SiCl3 f CH3 + SiCl3
CH3 + CH3SiCl3 f CH4 + CH2SiCl3
CH3 + CH3SiCl3 f CH3Cl + CH3SiCl2
CH3SiCl2 f SiCl2 + CH3
(R1)
(R2)
(R3)
(R4)
CH2SiCl3 f CH2SiCl2 + Cl
Cl + CH3SiCl3 f HCl + CH2SiCl3
Cl + CH3SiCl3 f Cl2 + CH3SiCl2
CH2SiCl3 + CH3 f C2H5SiCl2
2CH2SiCl3 f (CH2SiCl3)2
(R5)
(R6)
(R7)
(R8)
(R9)
Figure 1. Schematic diagram of the apparatus used for the study of
pyrolysis of CH3SiCl3.
CH3 + CH3 f C2H6
(R10)
(R11)
(R12)
(R13)
and k12 can be analyzed using the factorization method of Troe:
14
CH3 + SiCl3 f CH3SiCl3
2SiCl3 f (SiCl3)2
k1/k1∞ ) FLHFcSCFcWC
(4)
(5)
CH3SiCl3 f HCl + CH2SiCl2
FLH ) (k1,0 [M]/k1∞)/(1 + k1,0 [M]/k1∞)
SiCl3 + CH3SiCl3 f HSiCl3 + CH2SiCl3 (R14)
SiCl3 + CH3SiCl3 f SiCl4 + CH3SiCl2 (R15)
Here k1,∞ and k1,0 are the limiting, high-pressure, first-order and
limiting, low-pressure, second-order rate constants, respectively,
for reaction R1. The same expressions might be written for the
association reactions R8-R12, but the rate constants are second
and third order, respectively.
According to this proposed mechanism CH4, CH3Cl, SiCl2, CH2-
SiCl2, and HCl would be the major products. The above
mechanism consists of two unimolecular reactions R1 and R13.
At low conversion of MTS, no active species such as radicals
in reaction R13 are produced. Therefore, we do not expect any
contribution of reaction R13 in the other reactions in the
proposed mechanism at low conversion regime. Applying the
steady-state approximation to the radicals in this mechanism,
the rate of the initiation reaction R1 would be equal to the sum
of the rates of the termination reactions, reactions R8-R12.
This assumption would allow the calculation of the rate constant,
k1, for reaction R1 as
Experimental Section
All experiments were performed in a flow system shown in
Figure 1. The pressure in the storage bulb (10 L) and reactor
were monitored by two pressure transducer (Druck DPI 705).
The flow rate into the reactor was controlled by a needle valve.
CH3SiCl3 of 98% purity (Fluka) was used in all experiments.
During the experiments, a 40-cm-long section of the reactors
was heated by a resistive furnace. The temperature of the reactor
was controlled by a platinum/platinum-13%-rhodium thermo-
couple. The temperature was varied from 825 to 977 K. A
FORTRAN program15 was used to calculate the effective length
and temperature of the reactor according to the temperature
profile of the reactor and approximate activation energy of the
overall reaction. The average residence time, t, of molecules in
the reactor was calculated from the measured pressure, tem-
perature, and effective cross section of the reactors and from
the reactant flow rate. The conversion of reactant was between
1 and 3% to minimize the effect of possible secondary reactions.
To examine the possible surface reactions, two 80-cm-long
cylindrical quartz reactors with different surface-to-volume ratios
were used. The cross sections of the reactors were 1.26 and
0.94 cm2. The surface-to-volume ratios (s/V) for these reactors
was 17.5 and 9.48 cm-1. A six-way linear stainless steel gas
sampling valve at the exit of the reactor was used to take
samples in a 14 cm3 sample loop and a gas chromatograph
(Shimadzu GC-8A) was used to analyze the samples. The
products were separated on a 3-m silicagel column (mesh 50/
75) at different temperatures. Nitrogen was used as carrier gas.
A flame ionization detector and a thermal conductivity detector
k1 ) (R8 + R9 + R10 + R11 + R12 )/[MTS]
(1)
Here, R8-R12 are the steady-state rates of termination reactions,
respectively. All these reactions are pressure dependent and to
calculate the k1 at high-pressure limit one must calculate rates
of these termination reactions at high-pressure limit.
Applying the steady-state approximation, the ratios of rate
constants for reactions R2 and R3 to the rate constants for
reaction R10 may be calculated as
k10/k22 ) R10 [MTS]2/(R2)2
k10/k32 ) R10 [MTS]2/(R3)2
(2)
(3)
where R10, R2, and R3 are the steady-state rates of formation of
ethane, methane, and chloromethane, respectively. Reaction R10
has been studied extensively and the pressure and temperature
dependence of its rate constant are reasonably well-known. Any
pressure dependence of the rate constants k1, k8, k9, k10, k11,

1948 J. Phys. Chem. A, Vol. 108, No. 11, 2004
Mousavipour et al.
Figure 4. Chordal rate of formation of ethane at 55 Torr and 825 K
Figure 2. Chordal rate of formation of methane at 55 Torr and 825 K
in a reactor with surface-to-volume ratio of 17.5 cm-1
.
in a reactor with surface-to-volume ratio of 17.5 cm-1
.
Figure 3. Chordal rate of formation of chloromethane at 55 Torr and
825 K in a reactor with surface-to-volume ratio of 17.5 cm-1
Figure 5. Chordal rate of formation of dichlorosilane at 70 Torr and
917 K in a reactor with surface-to-volume ratio of 17.5 cm-1
.
.
at 917 K and 70 Torr. Each point in Figures 2-5 is the average
of at least three measurements. Experiments were carried out
at 825 and 977 K in each of the two reactors to test for any
surface reaction. Table 1 shows the steady-state rates of
formation of different products at different pressures and
temperatures. To calculate the steady-state rate of formation of
each product at different conditions in Table 1, a figure similar
to Figures 2-5 was used. As Table 1 shows, with clean quartz
reactors at identical pressures, there was no trend evident in
the steady-state rates of formation of different products on
changing surface-to-volume ratio. The magnitude of error for
each rate in Table 1 was less than 10%. Calculated rate constants
at different pressures and temperatures are also shown in Table
1.
The rate constant for reaction R1 was calculated using the
sum of the rates of formation of termination products as in eq
1. Unfortunately, it was not possible to measure the heavy
termination products; their retention time in GC column was
longer than the retention time for MTS. A value of 6.6 × 1010
L mol-1 s-1 from ref 16 was used for the rate constant of
reaction R11. Rate constant for reaction R12 was calculated
according to eq 6.
were used for the detection of possible different products. The
spectrum of HCl at 2500-3300 cm-1 was detected qualitatively
with a spectrophotometer (Shimadzu FTIR-8300).
Procedure. MTS was degassed by condensing with liquid
nitrogen. It was then placed in the storage bulb, shown in Figure
1 by evaporation and was again degassed in liquid nitrogen.
The storage bulb was covered with dark paper to prevent
photolysis of the MTS. The vapor pressure of MTS could be
monitored by a transducer. A needle valve controlled the flow
rate of the MTS vapor into the reactor. The vapor of the MTS
passed through the furnace. The temperature profile along the
reactor was measured several times during each run to keep
the temperature along the reactor constant. The pressure in the
reactor was controlled by the pressure transducer and a needle
valve. Normally, we tried to clean the surface of the reactors
by passing air through the reactors at high temperatures.
Results
The reaction was studied in the temperature range of 825 (
5 to 977 ( 5 K and the pressure range of 10 ( 2 to 120 ( 2
Torr. Depending on the temperature, the vapor pressure of pure
MTS varied from 130 to 140 Torr. The main products observed
were CH4, CH3Cl, SiCl2, and CH2SiCl2 and the observed minor
product was C2H6. Steady-state rates of formation of the primary
products were calculated from their average rates of formation
in the plateau region. We detected HCl qualitatively by means
of an spectrometer.
R11 ) m(R10 R12)1/2
(6)
Mousavipour and Pacey15 have shown m is almost independent
to the pressure and is very close to 2.
As we will discuss later, rates of reactions R8 and R9 are
negligible compared to the rates of the other termination
reactions R10-R12. We found rate of reaction R3 the major
source for CH3SiCl2 is smaller than that for reaction R2 and
Figures 2-4 show the average rates of formation of methane,
chloromethane, and ethane, respectively, at 825 K and 55 Torr.
Figure 5 shows the average rate of formation of dichlorosilane

Pyrolysis of Methyltrichlorosilane
J. Phys. Chem. A, Vol. 108, No. 11, 2004 1949
TABLE 1: Steady-State Rates and Rate Constants from Experiments at Different Pressures and Temperatures in Two
Reactorsa
2
2
R2 × 107
R3 × 107
Rd × 107
R4 × 107
R10 × 108
k2 /k10
L
k3 /k10 L
P Torr
mol L-1 s-1
mol L-1 s-1
mol L-1 s-1
mol L-1 s-1
mol L-1 s-1
mol-1 s-1
mol-1 s-1
k2/k3
k1 × 104 s-1
T ) 977 K, S/V ) 17.5 cm-1
10
20
30
40
50
60
70
80
90
1.3
1.7
5.0
4.4
7.5
11
13
14
13
13
0.7
0.5
2.8
4.5
7.5
8.5
9.5
13
0.61
4.3
0.8
2.0
3.1
129
22
26
25
20
17
7
40
23
27
15
11
15
3
1.8
1.0
1.0
1.3
1.3
1.1
1.5
1.3
1.6
0.6
1.2
1.5
2.0
2.2
2.0
2.7
2.9
5.1
18
30
5.0
6.4
6.7
5.2
8.9
8.0
9.0
14
10
12
25
100
120
5
5
3
2
22
T ) 977 K, S/V ) 9.48 cm-1
20
40
70
2.7
5.4
13.2
1.0
5.1
12
1.0
2.7
4.5
68
25
29
9
22
24
2.7
1.1
1.1
1.2
1.6
1.6
T ) 917 K, S/V ) 17.5 cm-1
10
20
30
50
70
0.2
0.4
0.7
1.4
2.6
5.9
9.0
9.4
14
0.3
0.5
0.7
1.0
1.6
4.1
6.7
7.5
10
1.4
0.3
0.5
0.9
1.5
4.4
2.6
2.2
1.7
1.5
2.2
3.6
3.6
4.4
9.8
4.1
2.2
1.0
0.6
1.1
2.0
1.8
2.3
0.7
0.8
1.0
1.3
1.6
1.4
1.3
1.4
1.4
0.7
0.6
0.7
0.7
1.0
1.6
1.7
1.4
1.9
5.9
7.1
9.0
15
5.4
3.0
6.3
7.3
6.6
90
100
110
120
10
T ) 882 K, S/V ) 17.5 cm-1
25
35
50
70
95
0.2
0.2
0.8
1.0
1.4
0.1
0.1
0.4
0.7
1.2
0.2
0.3
0.6
0.9
1.1
1.0
0.3
1.3
0.7
0.6
0.2
0.1
0.3
0.3
0.4
2.2
1.7
2.1
1.5
1.2
0.18
0.19
0.26
0.28
0.25
T ) 825 K, S/V ) 17.5 cm-1
15
25
55
75
95
0.07
0.1
0.7
0.8
1.2
0.03
0.05
0.3
0.5
0.7
0.1
0.2
0.5
0.6
0.9
0.6
0.2
0.6
0.5
0.5
0.1
0.1
0.2
0.2
0.2
2.4
1.4
1.7
1.6
1.6
0.14
0.16
0.19
0.16
0.19
T ) 825 K, S/V ) 9.48 cm-1
55
75
0.6
0.7
0.2
0.3
0.4
0.6
0.8
0.4
0.1
0.1
2.8
2.0
0.14
0.16
T ) 909 K, S/V ) 17.5 cm-1
30
40
0.3
0.5
0.3
0.5
0.3
0.4
1.1
1.3
1.1
1.3
1.0
1.0
0.23
0.23
T ) 891 K, S/V ) 17.5 cm-1
50
70
0.6
1.2
0.5
1.0
0.1
0.3
4.4
3.0
3.1
2.1
1.2
1.2
0.05
0.10
T ) 831 K, S/V ) 9.5 cm-1
25
55
75
0.1
0.2
0.4
0.1
0.1
0.2
0.1
0.3
0.4
0.5
0.1
0.2
0.5
0.03
0.05
1.0
1.8
2.2
0.07
0.11
0.11
a R2, R3, Rd, R4, and R10 are the steady-state rates of formation of methane, chloromethane, dichloromethylensilane, dichlorosilane, and ethane,
respectively. S/V is the reactor surface-to-volume ratio.
also the possibility of reaction R5 could diminish the concentra-
tion of CH2SiCl3. Values of k1 calculated at different pressures
and temperatures are listed in Table 1 and shown in Figure 6.
The Troe expression, eq 4, was fitted to the experimental data
by nonlinear least squares to estimate the low- and high-pressure
limiting values of k1. The results showed that k1∞ was not
sensitive to FCWC and that changing the collision efficiency from
unity to 0.1 changed k1∞ by only 10%. Values of the Troe input
parameters Sk, BT, and FCSC for reaction R1 were calculated at
different temperatures. Values of Sk varied from 10.227 to
11.285, BT varied from 0.69 to 0.74, and FCSC varied from 0.107
to 0.074 as temperature increased from 825 to 977 K. An
Arrhenius plot for k1∞ is shown in Figure 7. The Arrhenius
expression for this rate constant from the present work was
found by least-squares method to be
k1∞ ) (9.6 ( 2.5) × 1019 exp(-(392 (
18) kJ mol-1/RT) s-1
In the present work, we also calculated the values of k2 and
k3, the rate constants for hydrogen abstraction or chlorine
abstraction from MTS by methyl radicals, respectively. Values

1950 J. Phys. Chem. A, Vol. 108, No. 11, 2004
Mousavipour et al.
Figure 6. Dependence of k1 on MTS pressure: a, 825 K; b, 882 K; c,
917 K; d, 977 K. Points, experimental data; solid lines, nonlinear least-
squares fit of eq 4 to the data with âC ) 1.0.
Figure 8. Dependence of k2 on temperature: b, experiment; solid line,
linear least-squares fit.
Figure 7. Dependence of the high-pressure limit of k1 on tempera-
ture: b, experiment; solid line, linear least-squares fit.
Figure 9. Dependence of k3 on temperature: b, experiment; solid line,
linear least-squares fit.
of the ratios of (k2)2/k10 and (k3)2/k10 at different pressures and
temperatures are listed in Table 1. As one expected, k10 is
pressure dependent while k2 or k3 should be independent to the
pressure. As shown in Table 1, these ratios are almost
independent to the pressure specially, at higher pressures.
Therefore, we found it was not necessary to extrapolate these
ratios to find their values at infinite pressure.
The Arrhenius expression for the rate constant of reaction
R3 from the present work was found by least squares to be
k3 ) (1.5 ( 0.5) × 109 ×
exp(-(72 ( 6) kJ mol-1/RT) L mol-1 s-1
Discussion
It was not possible to separately determine k2, k3, and k10
from the present experiments, so a value of k10,∞ was taken from
the literature as17
Thermal decomposition of MTS was studied in the temper-
ature range of 825-977 K and pressure range of 10-120 Torr.
As indicated in the suggested mechanism in the present study,
the major products are CH4, CH3Cl, SiCl2, CH2SiCl2, and HCl.
As shown in the suggested mechanism, CH2SiCl2 could be
produced in several steps during the pyrolysis of MTS.
Therefore, it was difficult to investigate the role of reaction R13,
1-2 elimination of HCl, for the production of CH2SiCl2.
As one expected, reaction R1 is pressure dependent. Values
of k1 at different conditions were calculated from the sum of
the rates of the termination reactions R10-R12. We were
interested to study the pyrolysis of MTS at low conversion
regime, typically between 1 and 3%. Therefore, it was difficult
to measure the products with masses heavier than the MTS mass.
We calculated the rate of reaction R11 from the results reported
by Niiranen and Gutman16 and the rate of reaction R12 from
k10, ∞ (L mol-1 s-1) ) 9.05 × 1013 (T/K)-1.18
exp(-2.74 (kJ mol-1)/RT)
The values of k2 and k3 in Table 1 were calculated by
substituting these values of k10,∞ into the values of (k2)2/k10 and
(k3)2/k10 ratios. Arrhenius plots for k2 and k3 are shown in
Figures8 and 9, respectively. According to the our results,
Arrhenius expression for reaction R2 was
k2 ) (5.1 ( 0.4) × 108 ×
exp(-(61 ( 3) kJ mol-1/RT) L mol-1 s-1

Pyrolysis of Methyltrichlorosilane
J. Phys. Chem. A, Vol. 108, No. 11, 2004 1951
geometric mean rule, eq 6. We assumed the contribution of
termination reactions R8 and R9 should be small enough to
ignore their effect in calculating k1, especially at higher
temperatures, when reaction R5 becomes more important. At
the B3LYP/6-311++G(d,p) level of theory, we found a barrier
height equal to 295 kJ mol-1 for the decomposition of CH2-
SiCl3 to CH2SiCl2 and Cl, R5.
As shown in Figure 6, at 825 K the value of k1 reached its
high-pressure limit when the pressure was about 100 Torr, while
at higher temperatures the dependency of k1 to the pressure was
stronger.
According to our suggested mechanism, the concentration of
SiCl2 should be almost equal to the sum of the concentrations
of methane and chloromethane. The effect of presence of SiCl2
on the rate of deposition of SiC is considered by different
researchers. Sachdev and Scheid19 have studied the formation
of SiC by rf-plasma CVD. They suggested the formation of
silicon halides results in a loss of silicon needed for the correct
stoichiometry of the SiC deposits. Thus, the deposits become
carbon-rich because of different reaction kinetics of the silicon-
and carbon-containing species. Kaneko and co-workers20 have
studied the growth process of SiC from MTS by rf-plasma. From
the compositional analysis of the film, they suggested that the
competitive decomposition of SiCl3 and CH3 fragments is rate-
determining in an electron collision process. They believe the
existence of SiCl2 instead of SiCl3 supports their conclusion.
They detected SiCl2 by IR adsorption. They suggested that the
precursor of SiC is SiCl2. Josiek and Langlais21 have studied
the residence-time-dependent kinetics of CVD growth of SiC
in the MTS/H2 system. They suggested the simplest precursors
SiCl3 and CH3 which are directly formed by MTS decomposition
and adsorb on surfaces site cannot fulfill their experimental
results. They suggested SiCl2, C2H2, or C2H4 are the dominant
reactive intermediate source species in the intermediate. SiCl3
adsorbed more strongly on surface site than SiCl2. They
concluded, at low residence time and temperature, the gas-phase
concentration of SiCl2 should be lower than that of SiCl3. SiCl2
must adsorb on two surface sites to play a more important role
for the growth kinetics than SiCl3. At higher residence times, it
In the present work, we were able to calculate the ratios of
rate constants (k2)2/k10 and (k3)2/k10. As shown in Table 1, these
ratios are independent to the pressure at higher pressures at
different temperatures. We were not able to calculate the value
of k10 from our experimental results. We used the expression
given by Wagner and Wardlaw17 for the recombination of
methyl radical to calculate the values of k2 and k3 from our
results. As one expected, rate of hydrogen abstraction is higher
than the rate of chlorine abstraction from MTS by methyl
radicals. As shown in Table 1, the ratio of k2/k3 is close to 1.1
at higher temperatures and decreases to a value about 1.6 at
lower temperatures. This should be due to the stronger Si---Cl
bond relative to C---H bond. We found a value of 61 ( 3 kJ
mol-1 for the activation energy of reaction R2. Niiranen and
Gutman16 reported a value of 48.1 kJ mol-1 for the activation
energy of reaction R2 in the temperature range of 378-478 K.
As one expected, hydrogen abstraction reaction R2 should have
a curved Arrhenius plot because of tunneling effect. Therefore,
it is reasonable that this reaction has higher activation energy
at higher temperatures.
22
23
is possible for SiCl2 to react with H2 or CH4 in gas phase
to produce SiH2Cl2 or CH3SiHCl2, respectively.
From the previous paragraph, it is difficult to suggest a
reliable mechanism for the formation of SiC film on the surfaces.
According to the suggested mechanism in the present study, it
is expected that a large amount of SiCl2 produces during the
initial steps of the thermal decomposition of MTS. SiCl2
produces in a loop that consists of reactions R2, R3, R4, R7,
and R15. If we accept a key role for SiCl2 in the formation of
SiC film, therefore, further studies are necessary to investigate
this role of SiCl2. According to our suggested mechanism, we
do not expect SiCl3 produces in a large amount during the initial
steps in the pyrolysis of MTS.
Osterheld and co-workers12 have suggested three different
unimolecular decomposition paths for MTS as R1, R13, and
C-H bond cleavage R16.
CH3SiCl3 f CH2SiCl3 + H
(R16)
In their study, they reported the rate of reaction R16 is almost
100 times lower than the rate of reaction R1 and activation
energy for R16 is higher than that for reaction R1 by 26 kJ
mol-1 and has a lower preexponential factor. They reported their
results as the following Arrhenius expressions in the temperature
range of 800-1000 K.
Acknowledgment. The authors are pleased to acknowledge
financial support from the Research Council of Shiraz Univer-
sity. S.H.M. wishes to thank Professor M. Rashidi for valuable
discussion.
k1 ) 2.0 × 1018 exp(-407.1 kJ mol-1/RT) s-1
k13 ) 1.3 × 1015 exp(-399.2 kJ mol-1/RT) s-1
k16 ) 5.0 × 1017 exp(-433.0 kJ mol-1/RT) s-1
References and Notes
(1) De Jong, F.; Meyyappan, M. Diamond Relat. Mater. 1996, 5, 141.
It was difficult to verify the role of reactions R13 and R16
in production of CH2SiCl2 or CH2SiCl3 because we expected
other reactions such as reactions R5 (directly) and R6 (indirectly)
to produce the same products. As listed in Table 1, the average
rate of formation of CH2SiCl2 is smaller than the rate of
formation of methane. Therefore, we can conclude from these
results that the role of reactions R13 and R16 could not be very
important in the pyrolysis of MTS.
(2) Davidson, I. M. T.; Ring, M. A. J. Chem. Soc., Faraday Trans. 1
1980, 76, 1520. Neudorfl, P. S.; Lown, E. M.; Sararik, I.; Jodhan, A.;
Strausz, O. P. J. Am. Chem. Soc. 1987, 109, 5780. Lee, M.-S.; Bent, S. F.
J. Phys. Chem. B 1997, 101, 9195.
(3) Burgess, J. N.; Lewis, T. J. Chem. Ind. 1974, 76.
(4) Kobayashi, F.; Ikawa, K; Iwamoto, K. J. Cryst. Growth 1975, 28,
395.
(5) Davidson, I. M.; Dean, C. E. Organometallics 1987, 6, 966.
(6) Gordon, M. S.; Truong, T. N. Chem. Phys. Lett. 1987, 142, 110.
(7) Besmann T. M.; Sheldon B. W. J. Am. Ceram. Soc. 1991, 74, 3046.
According to ref 10, ∆fH° for MTS, CH3, SiCl3, CH2SiCl3,
H, and CH2SiCl2 are reported as -576 ( 4, 146 ( 5, -318 (
7, -367 ( 6, 217 ( 0.0, -143 ( 5 kJ mol-1 at 298 K,
respectively, and for HCl -92.31 ( 0.10 kJ at 298 K from ref
18. According to these data, ∆H0298 for reactions R1, R13, and
(8) Besmann, T. M.; Sheldon, B. W; Lowden, R. A.; Stinton, D. P.
Science 1991, 253, 1104.
(9) Besmann, T. M.; Sheldon, B. W.; Moss, T. S., III; Kaster, M. D.
J. Am. Ceram. Soc. 1992, 75, 2899.
(10) Allendorf, M. D.; Melius, C. F. J. Phys. Chem. 1993, 97, 720.
(11) Cagliostro, D. E.; Riccitiello, S. R.; Ren, J.; Zaghi, F. J. Am. Ceram.
Soc. 1994, 77, 2721.
R16 equaled 404 ( 16, 341 ( 9, and 426 ( 10 kJ mol-1
respectively.
,

1952 J. Phys. Chem. A, Vol. 108, No. 11, 2004
Mousavipour et al.
(12) Osterheld, T. H.; Allendorf, M. D.; Melius, C. F. J. Phys. Chem.
1994, 98, 6995.
(18) Cox, J. D.; Wagman, D. D.; Medvedev, V. A. CODATA Key Values
for Thermodynamics; Hemisphere: New York, 1989.
(13) Gordon, M. S.; Pederson, L. A.; Bakhtiar, R.; Jacobson, D. B. J.
Phys. Chem. 1995, 99, 148.
(14) Troe, J. J. Phys. Chem. 1979, 83, 114.
(15) Mousavipour, S. H.; Pacey, D. P. J. Phys. Chem. 1996, 100, 3573.
(16) Niiranen, J. T.; Gutman, D. J. Phys. Chem. 1993, 97, 9392.
(17) Wagner, A. F.; Wardlaw, D. M. J. Phys. Chem. 1988, 92, 2462.
Slagle I. R.; Gutman, D.; Davies, J. W.; Pilling, M. J. J. Phys. Chem. 1988,
92, 2455.
(19) Sachdev, H.; Scheid, P. Diamond Relat. Mater. 2001, 10, 1160.
(20) Kaneko, T.; Miyakawa, N.; Sone, H.; Yamazakia, H. Surf. Coat.
Technol. 2001, 142, 360.
(21) Josiek, A.; Langlais, F. J. Cryst. Growth 1996, 160, 253.
(22) Wittbrodt, J. M.; Schlegel, H. B. Chem. Phys. Lett. 1997. 265, 527.
(23) Ring, A.; O’Neal, H. E.; Walker, K. L. Int. J. Chem. Kinet. 1998,
30, 89.