Infrared multiple photon dissociation of chloromethyltrifluorosilane

Article abstract of DOI:10.1111/j.1751-1097.2009.00541.x

Infrared multiphoton absorption and dissociation of chloromethyltrifluorosilane molecules under the action of pulsed transversely excited atmospheric pressure CO₂ laser were experimentally studied. Dissociation products were analyzed. The dissociation proceeds via chlorine atom transfer from carbon to silicone. High degrees of silicon isotope separation were achieved. The presence of α-chlorine atom in a silicon organic compound brings about a significant improvement in multiple photon dissociation characteristics and an essential increase in isotopic selectivity.

Full text of DOI:10.1111/j.1751-1097.2009.00541.x

Photochemistry and Photobiology, 2009, 85: 901–908
Infrared Multiple Photon Dissociation of Chloromethyltrifluorosilane
Pavel V. Koshlyakov*1, Sergey R. Gorelik1,2,3, Evgeniy N. Chesnokov1,2, Oleg S. Aseev1,2
Asylkhan A. Rakhymzhan1,2 and Alexander K. Petrov1
,
1Institute of Chemical Kinetics and Combustion SB RAS, Institutskaia, 3, Novosibirsk, Russia
2Novosibirsk State University, Pirogova, 2, Novosibirsk, Russia
3Institute of Material of Research and Engineering, 3, Research Link, Singapore
Received 9 September 2008, accepted 13 December 2008, DOI: 10.1111 ⁄ j.1751-1097.2009.00541.x
(9,10) reported that the MPD of Si2F6 molecules induced by
CO2 laser radiation had significantly high dissociation efficiency
and high isotope selectivity under mild radiation fluences lower
than 1 J cm)2. The enrichment coefficient (selectivity) reached
30 for the 30Si isotope and 6 for the 29Si isotope.
Despite the impressive results, it is not evident that Si2F6
might be considered as the best object for silicon isotope
separation by means of IR MPD. The molecule contains two Si
atoms and the rare isotopes are contained mostly in ‘‘isotopi-
cally mixed’’ molecules such as 28SiF3-29(30)SiF3. This fact limits
the isotopic effect as during the dissociation of these ‘‘isoto-
pically mixed’’ molecules, the 28Si isotope and the rare one (29Si
or 30Si) would pass into products equally. In addition, we could
expect that the isotopic shift in the molecular vibrational
frequencies in the ‘‘isotopically mixed’’ molecules should be less
than that in the isotopically ‘‘pure’’ molecules.
Silicon isotope-selective MPD of compounds containing
single silicon atoms, such as SiH2F2 (15), SiF3C6H5 (16–18),
SiF3CH3 (19,20), SiF3CHCH2 (21) and Si(OCH3)4 (22) was
studied previously. The maximum isotope selectivities reached
for the MPD of these molecules were found to be lower than
that for Si2F6. From the analysis of spectroscopic and kinetic
characteristics of the MPD of these molecules, one can conclude
that the main reason for the higher efficiency of the Si2F6 MPD
is its low dissociation threshold (188 kJ mol)1 [10]) compared
to the molecules studied elsewhere (around 420 kJ mol)1 [14–
22]). Therefore, the further search of the compounds should be
limited to those containing only one silicon atom, and possess-
ing a relatively low dissociation threshold.
Haszeldine et al. (23–27) studied experimentally the thermal
decomposition of halogenoalkylsilanes. They showed that the
halogenoalkylsilane compounds, in which halogen is present at
a- or b-carbon relative to silicon, essentially easily underwent
the unimolecular decomposition via the intramolecular
rearrangement involving the unimolecular transfer of halogen
from carbon to silicon. It was shown that dissociation with the
b-halogen atom transfer proceeds via a four-center transition
state and leads to the formation of stable products, olefin and
halogen-substituted silane:
ABSTRACT
Infrared multiphoton absorption and dissociation of chlorometh-
yltrifluorosilane molecules under the action of pulsed transver-
sely excited atmospheric pressure CO2 laser were experimentally
studied. Dissociation products were analyzed. The dissociation
proceeds via chlorine atom transfer from carbon to silicone. High
degrees of silicon isotope separation were achieved. The presence
of a-chlorine atom in a silicon organic compound brings about a
significant improvement in multiple photon dissociation charac-
teristics and an essential increase in isotopic selectivity.
INTRODUCTION
In the past decade, interest in the applications of isotopically
pure silicon materials for semiconductor technology has been
growing worldwide, in particular, after the thermal conduc-
tivity of highly enriched 28Si had been found to be almost 60%
higher than that of naturally abundant Si at room temperature
(1). Subsequently, this result was redefined (2): It was found
that the enhancement of thermal conductivity was about 10%
at room temperature, whereas at low temperatures the thermal
conductivity of isotopically pure Si exceeds that of naturally
abundant Si by a factor of 8.
In the opinion of experts (3), the outlook for technological
applications of 28Si is quite optimistic. However, one of the
main obstacles for using isotopically pure 28Si on the techno-
logical scale is the cost of its production by means of
traditional isotope separation methods. This requires an
effective isotope separation process for Si. Laser isotope
separation involving infrared multiphoton dissociation (IR
MPD) is one of the most promising ways to satisfy the demand
because of the high selectivity of this method (4,5).
This makes it relevant to study the MPD of silicon-
containing molecules in order to find the most appropriate
compounds for the laser-induced silicon isotope separation
technology.
Silicon has three stable isotopes with the natural composition
92.22% (28Si): 4.69% (29Si): 3.09% (30Si) (6). Separation of Si
isotopes by means of the infrared laser radiation has been
studied (7–14). The most impressive results were achieved in
experiments on MPD of the species Si2F6 (9–13). Kamioka et al.
*Corresponding author email: pvk@kinetics.nsc.ru (Pavel V. Koshlyakov)
ꢀ 2009 U.S. Government. Journal compilation. The American Society of Photobiology 0031-8655/09
901

902 Pavel V. Koshlyakov et al.
mental frequencies were calculated using the density functional theory
at the hybrid B3LYP level (29,30) with the 6-31G(p,d) basis set and the
unrestricted formalism for the case of radicals. The geometries of
transition states (TS) were found using the QST2 procedure (31) and
then optimized. The TS geometries were confirmed by frequency
calculations of the same level of theory, presenting one imaginary
frequency corresponding to the transition vectors pointing in the
direction of the reaction coordinate. The structures of local minima
(reactant and products) were also confirmed through frequency
calculations. This shows that all the frequencies are necessarily real
at local minima.
Similar transfer of a-halogen proceeds via a three-center
transition state and is followed by the formation of halogen-
substituted silane and carbene.
In particular, the kinetics of pyrolysis of fluoroalkylsilanes
was studied (23). The decomposition of such molecules with
either a- or b-fluorine took place at fairly low temperatures,
less than 200ꢁC, while the molecules containing no fluorine
atom in a- or b-position were stable under these conditions.
Dissociation energies obtained from the kinetics data were
within the range of 100–160 kJ mol)1 (23).
So low dissociation energies make the unimolecular decay
reactions of halogenoalkylsilanes proceeding via the halogen
atom transfer from carbon to silicon a promising way for the
isotope-selective MPD. In this study we report the experimen-
tal study of the silicon isotope selective decomposition of
SiF3CH2Cl molecules induced by pulsed transversely excited
atmospheric (TEA) pressure CO2 laser irradiation. This
molecule contains the a-chlorine atom, and the dissociation
of this molecule occurs via the chlorine atom transfer from
carbon to silicon:
RESULTS AND DISCUSSION
Infrared spectrum
The FTIR spectrum of chloromethyltrifluorosilane (1 cm)1
resolution) containing most of the fundamental vibrational
bands of the molecule is shown in Fig. 1. Two strong
absorption bands correspond to the stretching vibrational
modes of the SiF3 group of the molecule. The DFT calcula-
tions of the fundamental frequencies of the SiF3CH2Cl
molecule were performed to assign the vibrational bands in
the IR spectrum. The molecule has a plane of symmetry.
Therefore, all the vibrations are divided into two groups:
A¢—symmetrical with respect to the plane, and A¢—antisym-
metrical with respect to the plane. The calculated vibrational
frequencies, as well as those obtained from the IR spectrum,
are shown in Table 1.
SiF3CH2Cl ! :CH2 þ SiF3Cl
ðR:1Þ
The dissociation products are methylene and trifluorochlo-
rosilane. The silicon isotope composition of the latter product
varies depending on the frequency of the laser radiation.
To estimate the isotopic shifts in the IR spectrum, the
vibrational frequencies of minor Si isotopomers of the mole-
cule were calculated in this spectral region. The results are
shown in the third and fourth columns of Table 1. For the
stretching vibrations of the SiF3 group of chloromethyltri-
fluorosilane molecules, the calculated isotopic shift for the
Si isotopically substituted molecules was found to be about
7–8 cm)1 for the molecules containing 29Si and about
15–16 cm)1 for the molecules containing 30Si with respect to
28Si-containing molecules. These values are typical for the Si
isotopic shifts in the IR spectra of fluorosilanes in the spectral
region of Si–F stretching vibrations (32).
MATERIALS AND METHODS
Chemicals. SiF3CH2Cl was synthesized by the reaction of SiCl3CH2Cl
with antimony trifluoride and consequently purified using a homemade
low-temperature rectification column. Reaction yield was 40%.
SiCl3CH2Cl was received from the A. E. Favorsky Irkutsk Institute
of Chemistry SB RAS.
Radiation source. A homemade tunable pulse TEA CO2-laser was
used as a source of IR radiation. The maximum pulse energy of the
laser was about 5 J. The laser beam was collimated in front of
the input window of the cell with the aperture 1 cm in diameter.
The energy of the laser radiation pulse was varied by changing the
discharge energy and pressure in the laser cavity. To change the pulse
energy significantly, a set of parallel-sided CaF2 plates was used.
The average energy of the laser radiation was measured in front of the
cell and after the cell with an optical powermeter. The unfocused
uniform laser beam was used through all the experiments.
Experimental. All the experiments were performed under batch cell
conditions at room temperature. Samples of chloromethyltrifluorosi-
lane with natural abundance of silicon isotopes were irradiated. The
reaction cell was a Pyrex-glass cylinder with NaCl windows. The cell
length was 42 cm; the cell diameter was 3 cm. The cell content was
analyzed with a MH-1303 mass spectrometer. The cell was continually
connected to the ion source of the mass spectrometer through a glass
pinhole about 20 lm in diameter. Therefore, the cell content could be
analyzed at any moment during the laser irradiation. The concentra-
tion and isotope composition of SiF3CH2Cl were determined from the
m ⁄ e = 132–138 lines of the mass spectrum. The concentration and
isotope composition of the SiF3Cl MPD product were determined
from the m ⁄ e = 120–124 lines of the mass spectrum. The isotope
analysis of these species was complicated by the superposition of the
mass spectra of the compounds containing different chlorine and
silicon isotopes. To overcome this problem, a special software was
designed to determine the composition of 35Cl-37Cl and 28Si–29Si–30Si
isotopes taking into account all the lines in the mass spectrum. The
corresponding algorithm is described in the Appendix.
The IR spectra were measured with a Fourier transform infrared
(FTIR) spectrometer (Bruker Vector 22) with the spectral resolution of
1 cm)1
.
Figure 1. FTIR spectrum of chloromethyltrifluorosilane (1 cm)1 res-
olution), containing the major part of the fundamental vibrational
bands of the molecule. Vapor pressure SiF3CH2Cl was 3 torr, length of
the reaction cell 18 cm.
Computational details. Geometry calculations of the reactants,
products and intermediates were performed with the GAUSSIAN98
suite of programs (28). Those were fully optimized and the funda-

Photochemistry and Photobiology, 2009, 85 903
Table 1. Calculated and experimental values of vibrational frequencies
of SiF3CH2Cl isotopomers.
Calculated (B3LIP)
Scaling factor = 0.96
Type of
28Si
29Si
30Si Symmetry Experiment
vibration
1
2
3
4
5
6
7
8
9
46.8
46.8
46.8
A¢¢
A¢
A¢¢
A¢
A¢¢
A¢
A¢
A¢
A¢¢
A¢
A¢
A¢
A¢¢
A¢¢
A¢
A¢
A¢
A¢¢
–
–
–
–
–
Torsion
108.1 108.0 107.9
196.0 195.9 195.8
222.9 222.8 222.7
304.8 303.8 303.0
317.7 316.6 315.6
379.3 377.4 375.6
623.9 622.9 621.9
724.9 724.9 724.9
Si–C–Cl bending
Si–C–Cl rock
F–Si–C bending
SiF3 bending
SiF3 bending
?
–
–
668
731
768
903
989
989
1115
1201
1400
2955
–
Si–C stretching
CH2 rock
10 725.4 725.4 725.4
11 863.8 858.0 852.7
12 968.6 960.2 952.2
13 977.3 969.2 961.6
14 1095.9 1095.3 1094.7
15 1192.9 1192.5 1192.3
16 1399.0 1399.0 1399.0
17 2969.0 2969.0 2969.0
18 3026.7 3026.7 3026.7
C–Cl stretching
Si–F stretching
Si–F stretching
Si–F stretching
CH2 twist
Figure 2. Laser fluence dependence of <n>—the average number of
absorbed photons per molecule, measured at 10R (18) CO2-laser line
(977.2 cm)1) at 300 mtorr sample pressure. Solid line: polynomial
fitting of experimental points (see text for details).
CH2 wag
CH2 scissors
C–H stretching
C–H stretching
per molecule reaches 17 at 0.5 J cm)1 of the laser fluence. The
corresponding absorption cross-section, derived from the slope
of the dependence, was found to be 8 · 10)19 cm2. It is close to
the linear cross-section, which could be determined from the
IR spectrum in Fig. 1 (8.8 · 10)19 cm2).
Sample molecules were excited with the CO2-laser radiation
in the region of Si–F stretching vibrational band at 989 cm)1
.
Figure 3 presents the laser fluence dependence of IR MPD
probability b, an average probability of a single molecule to
dissociateduringthesinglelaserpulse,forSiF3CH2Clmolecules.
It was measured at 10R (18) CO2-laser line. Sample pressure was
0.1 torr. Forcomparison, thesamedataforSi2F6 molecules(34),
SiF2H2 molecules (15) and SiF3CH3 (19) are shown.
Estimation of dissociation threshold for
SiF3CH2Cl decomposition
As was mentioned above, we expected the reaction (R.1) to be
the main channel of SiF3CH2Cl decomposition, with SiF3Cl
and CH2 products. The ground state of methylene, CH2, is
triplet. Due to the spin conservation law, methylene should be
formed in the singlet excited state in reaction (R.1). The energy
difference between the fundamental triplet state and the first
excited singlet state for methylene is about 38 kJ mol)1 (33).
The B3LYP calculation of the enthalpy change in reaction
(R.1) gives: DH = 368 kJ mol)1. The activation energy is a
more important characteristic of the decomposition process.
However, we could not find the transition state (saddle point at
the potential energy surface) between the reactant SiF3CH2Cl
Dissociation probability b for SiF3CH2Cl molecules be-
comes measurable at fluences higher than 0.2 J cm)2 and
reaches several percent at the laser fluence of about 0.6 J cm)2
.
SiF3CH2Cl molecules dissociate more effectively than
difluorosilane because the latter possess the rotational bottle-
neck (15). At the same time, SiF3CH2Cl molecules dissociate
1
and the products SiF3Cl + CH2 using the QST2 procedure.
We calculated the potential surface along one specific trajec-
tory connecting the reactant with the products. For this
trajectory, the activation threshold was not higher than the
energy of the reaction products. Hence, the activation thresh-
old of the unimolecular decomposition of SiF3CH2Cl to
1
SiF3Cl + CH2 is equal to the energy change of the reaction,
.
i.e. 368 kJ mol)1
The energy change in the SiF3CH2Cl decomposition via
Si–C bond cleavage, SiF3CH2Cl fi SiF3 + CH2Cl, was
calculated to be 381 kJ mol)1
.
Infrared multiphoton absorption and dissociation
Figure 2 presents the laser fluence dependence of <n>—the
average number of absorbed photons per molecule, measured
at 10R (18) CO2-laser line (977.2 cm)1).
When the laser fluence is lower than 0.5 J cm)2, the
dependence is linear. The average number of absorbed photons
Figure 3. Laser fluence dependence of IR MPD probability for
SiF3CH2Cl molecules at 10R (18) CO2-laser line, sample pressure
0.1 torr. Same dependences for Si2F6 molecules (29), SiF2H2 molecules
(12) and SiF3CH3 (15) are shown.

904 Pavel V. Koshlyakov et al.
less effectively than hexafluorodisilane, apparently because the
dissociation energy of the latter is essentially lower—about
188 kJ mol)1 (10).
Note that the MPD of SiF3CH3, for which SiF3CH2Cl is the
chlorine-substituted derivative, could be observed only at laser
fluences higher than 0.4 J cm)2. Thus, the insertion of chlorine
atom into the a position was followed by an essential decrease
in the multiple photon dissociation threshold.
was experimentally obtained by tuning the CO2 laser
through the laser lines. Sample pressure was 0.1 torr, the
laser fluence was kept constant about 0.3 J cm)2. Under
these conditions, the dissociation probability did not exceed
0.2%.
The width of the MPD spectrum shown in Fig. 4 is about
12–15 cm)1. As shown above, the isotope shifts in the
frequency of 29,30Si–F stretching vibrations with respect to
28Si–F are about 8 cm)1 for 29Si and 16 cm)1 for 30Si
correspondingly. Thus, the width of the MPD spectrum is
comparable with the isotope shifts, or even lower. Therefore
we can expect the isotopic selectivity of SiF3CH2Cl mul-
tiphoton dissociation to be relatively high. However, the
frequencies of the silicon isotope-substituted molecules are
close to the gap between the R and P branches of the CO2-
laser, where the laser does not emit. This gas is indicated in
Fig. 4 by two solid vertical lines. This gap could make it
impossible to achieve the optimal conditions for a highly
selective Si isotope MPD of SiF3CH2Cl molecules in our
experiments.
Reaction products and secondary reactions
The mass spectrometric analysis of the reaction cell contents
after irradiation shows that the main products of dissociation
are SiF3Cl, C2H4 and C2H2. Other products, such as SiF4,
SiF3CH=CH2, SiF3CH2–CH2Cl and HCl, are present in
small amounts. The product composition confirms that the
dissociation proceeds via Cl atom transfer from carbon to
silicon:
1
SiF3CH2Cl þ nhm ! SiF3Cl þ CH2
ðR:2Þ
Figure 5 presents typical dependence of relative concentra-
tions of different silicon isotopes on irradiation time for the
sample molecule (a) and for the SiF3Cl product (b) in our
experiments. The time when the laser was turned on is
indicated with a dashed vertical line. Before that, the laser
was turned off. A decrease in the concentration during this
period is caused by gas consumption in the mass-spectrometer.
The laser was tuned to the 10P (19) laser line, the repetition
rate was 0.5 Hz. Concentrations of molecules with different Cl
isotopes were summed up.
The rate constants of the relaxation of singlet methylene in
collisions with polyatomic molecules
3
1CH2 þ M ! CH2 þ M
ðR:3Þ
are essentially high (1–2) · 10)10 cm3 s)1 (33,35). CH2 reac-
tions with hydrocarbons are very fast, with rate constants
approaching the gas kinetic collision efficiency, while the
corresponding rate constants of 3CH2 are 5–6 orders of
magnitude lower (33).
1
Therefore, the end products are formed not only in primary
dissociation reaction (R.2) but also in secondary reactions of
1CH2 formed in reaction (R.2). In particular, 1CH2 can
undergo insertion reactions (35–39). Therefore, the presence
of SiF3CH2–CH2Cl in the end products of the dissociation of
SiF3CH2Cl in our experiments could be explained by the
Assuming the first-order reaction, the following expressions
for the concentrations of different isotopomers of reagent
SiF3C2H3Cl2 and product SiF3Cl can be obtained:
n28ðtÞ
ꢀ
ꢁ
ꢀ
ꢀ
ꢁꢁ
¼ exp ꢀk28 ꢁ t p28ðtÞ ¼ n208 ꢁ 1 ꢀ exp ꢀk28 ꢁ t
ð1Þ
n208
1
insertion of CH2 into the Si–C bond of SiF3CH2Cl:
where n28(t) and p28(t) are the concentrations of the reagent
and the product molecules containing 28Si, respectively, k28 is
the effective first-order rate constant and t is irradiation time.
1CH2 þSiF3CH2Cl!SiF3 ꢀCH2 ꢀCH2Cl DH¼ꢀ502kJmolꢀ1
ðR:4Þ
SiF3–CH2–CH2Cl is formed with an excess of energy,
therefore it could dissociate, producing SiF3CH=CH2:
SiF3C2H4Cl ! SiF3C2H3 þ HCl DH ¼ 82kJ molꢀ1 ðR:5Þ
All DH values were calculated using the B3LYP method.
Hydrocarbon products resulting from the dissociation of
SiF3CH2Cl are formed in reactions:
1
1CH2 þ CH2 ! C2H2 þ H2
ðR:6Þ
ðR:7Þ
3CH2 ! ðwallÞ ! 1=2 C2H4
From the above consideration, we can conclude that,
qualitatively, the end product composition confirms the
primary dissociation channel to be the dissociation of
1
SiF3CH2Cl into SiF3Cl + CH2.
Isotope-selective MPD of SiF3CH2Cl
Figure 4. Dependence of the probability of the MPD of SiF3CH2Cl
molecules on the laser radiation wavenumber–MPD spectrum
(squares). Sample pressure was 0.1 torr, the laser fluence was about
Figure 4 shows the dependence of the reaction yield on the
laser radiation wavenumber–MPD spectrum (squares). It
0.3 J cm)2
.

Photochemistry and Photobiology, 2009, 85 905
k30 n208 w30 k29 n028 w29
S30
¼
¼
ꢁ
S29
¼
¼
ꢁ
n209
w28
ð4Þ
n300
k28
w28
k28
ꢂ
where w28 ¼ dp30 dt is the rate of 30SiF3Cl formation, w28 is
the rate of 28SiF3Cl formation and w30 is the rate of 30SiF3Cl
formation.
From the initial region of growing product concentration
(Fig. 5b): w30 ⁄ w28 = 0.8.
Taking into account the fact that the silicon isotope
composition of the sample molecules before the irradiation
was equal to the natural one: g28 = 92.2%—the abundance of
28Si and g30 = 3.1%—the abundance of 30Si, we obtain for
selectivity S:
S30 ¼ g28 ꢁ w30=ðg30 ꢁ w28Þ ¼ 24
ð5Þ
One can see from Fig. 5b that the rate of the growth of the
30SiF3Cl product decreases during irradiation. Consistent with
Eq. (1), this decrease is due to the fast consumption of the
sample molecules with 30Si. According to the isotope compo-
sition analysis of the sample molecules, nearly a half of
30SiF3CH2Cl dissociated during the first 260 s of irradiation,
whereas the concentration of 28SiF3CH2Cl changed insignifi-
cantly during the same time. Therefore, the concentration of
the 28SiF3Cl molecules continues to increase at the same rate.
Figure 6 presents the spectral dependences of the selectiv-
ities for 30Si and 29Si determined using Eq. (5). The depen-
dences were obtained by tuning the laser through different
laser lines in the P- and R-branches. The initial pressure of
SiF3CH2Cl was 0.1 torr. The laser fluence was within the range
0.27–0.3 J cm)2. When the laser radiation was tuned into the
R-branch, the enrichment of products with 28Si isotope was
observed. In the P-branch of laser radiation, the products were
enriched with 30Si, 29Si isotopes. The spectral range 955–
970 cm)1 was not available for measurements, as it corre-
sponds to the gap between R- and P-branches of the laser
radiation. The maximal selectivity achieved in our experiments
was S30 = 44. This value is comparable with the values
achieved by Kamioka et al. in experiments with Si2F6 (9,10).
Figure 5. Typical dependences of relative concentrations of different
silicon isotopes on irradiation time in the sample molecule (a) and in
the SiF3Cl product (b) in our experiments.
Similar expressions could be written for the concentrations of
the molecules containing 29Si and 30Si.
The initial region of product concentration growth could be
approximated by linear function
CONCLUSION
The presence of a-chlorine atom in organosilicon compounds
brings about a significant improvement in multiple photon
p28ðtÞ ¼ n208 ꢁ k28 ꢁ t; p29ðtÞ ¼ n209 ꢁ k29 ꢁ t; p30ðtÞ ¼ n030 ꢁ k30 ꢁ t
ð2Þ
The ratio of the first-order rate constants can be used as a
measure of the isotopic selectivity:
k30
k28
k29
k28
S30
¼
S29
¼
ð3Þ
The effective rate constants can be determined either from
the decay of the reagent concentration or from the growth of
the product concentration. For 28Si and 29Si, the decay of the
reagent concentration is considerably affected by gas con-
sumption in the mass spectrometer. Therefore, the rate
constants can be determined more accurately from the growth
of the products. Using the linear approximation (Eq. 2) we
obtain:
Figure 6. Spectral dependences of the selectivities for 30Si and 29Si.

906 Pavel V. Koshlyakov et al.
dissociation characteristics, which causes an essential increase
in isotopic selectivity.
MS Excel ‘‘search of solution’’ function was used to find
numerically the minimum of S(x,y,z). As a result, Cl, Si
isotope composition in SiF3Cl was obtained. The chlorine
However, the studied molecule is not an optimal choice for
the technological applications from the following point of
view: a singlet methylene biradical is formed as a primary
product of the dissociation. It is involved in the complex
secondary reactions, including the reaction with the parent
molecules. The latter reaction is isotopically nonselective,
therefore it limits the selectivity. A desirable choice would be
the molecule which dissociates into molecular products only.
The silicon organic molecules with b-chlorine or fluorine
atoms should dissociate to molecular products. One could
expect that the study of such molecules will allow one to
achieve more efficient multiphoton dissociation and higher
silicon isotope selectivities.
isotope composition x = 0.231
0.005 was slightly differ-
ent from the natural one, 0.2447 (4). Silicon isotope
composition varied in a wide range depending on the laser
irradiation wavelength. The algorithm <became> nonro-
bustness appeared when the samples highly enriched with
30Si were analyzed. In these cases, the solution looked like
the samples were highly enriched with the 37Cl isotope. To
overcome this, the chlorine isotope composition was frozen
at x = 0.231. Only the silicon isotope composition was
varied.
The isotope analysis of SiF3CH2Cl was performed using
the group of mass spectrum lines m ⁄ e = (132–138). Not
only the parent ion SiF3CH2Cl+ contributes to these lines
but also fragment ions SiF3CHCl+ and SiF3CCl+, which
complicates the isotope analysis. Denoting the relative
contribution of the SiF3CHCl+ ion as p and that of
SiF3CCl+ as q, we obtain for the relative line intensities
Acknowledgement—This work was supported by the Russian Foun-
dation for Basic Research (Project No. RFBR 07-0300873-a).
APPENDIX
h132…h138
:
Isotope analysis of SiF3Cl and SiF3CH2Cl compounds
h132 ¼ qð1 ꢀ xÞð1 ꢀ y ꢀ zÞ
ðA7Þ
ðA8Þ
The mass spectrum of SiF3Cl contains a group of lines with
m ⁄ e = (120–124), which corresponds to different chlorine and
silicon isotope variations of the parent ion SiF3Cl+. Let us
denote the intensities of these lines as H120,…,H124, and their
relative intensities as
h133 ¼ pð1 ꢀ xÞð1 ꢀ y ꢀ zÞ þ qð1 ꢀ xÞy
h134 ¼ ð1 ꢀ p ꢀ qÞð1 ꢀ xÞð1 ꢀ y ꢀ zÞ þ pð1 ꢀ xÞy þ qðð1 ꢀ xÞz
Hi
hi ¼
þ xð1 ꢀ y ꢀ zÞÞ
ðA9Þ
124
P
Hj
j¼120
h135 ¼ ð1 ꢀ p ꢀ qÞð1 ꢀ xÞy þ pðð1 ꢀ xÞz þ xð1 ꢀ y ꢀ zÞÞ þ qxy
ðA10Þ
Relative intensities could be derived using different contents of
chlorine and silicon isotopes:
h120 ¼ ð1 ꢀ xÞð1 ꢀ y ꢀ zÞ
h121 ¼ ð1 ꢀ xÞy
ðA1Þ
ðA2Þ
ðA3Þ
ðA4Þ
ðA5Þ
h136 ¼ ð1 ꢀ p ꢀ qÞðð1 ꢀ xÞz þ xð1 ꢀ y ꢀ zÞÞ þ pxy þ qxz
ðA11Þ
h122 ¼ ð1 ꢀ xÞz þ xð1 ꢀ y ꢀ zÞ
h123 ¼ xy
h137 ¼ ð1 ꢀ p ꢀ qÞxy þ pxz
h138 ¼ ð1 ꢀ p ꢀ qÞxz
ðA12Þ
ðA13Þ
h124 ¼ xz
Here x is 37Cl content, y is 29Si, z is 30Si content.
Using the chi-square method, we need to find the minimum
of the function of five variables. Its expression is similar to
(A6), and is not shown here due to its complexity.
Hence, we have to determine the set of variables x, y and z
so that the intensities h120 …. h124 fit the experimental
mass spectrum in the best way. Applying the chi-square
method, we have to find a minimum of the function of three
variables:
The analysis of SiF3CH2Cl before irradiation allowed
for determination of its isotope composition and relative
probabilities of the fragment ion formation: P = 0.015 b
q = 0.006. In subsequent analysis, these two variables
were assumed to be the same. Therefore, the problem of
determining the isotope composition of SiF3CH2Cl using
the group of mass lines m ⁄ e = (132–138) was reduced to
the search of the minimum of the function of three
variables.
2
2
Sðx; y; zÞ ¼ ½h120 ꢀ ð1 ꢀ xÞð1 ꢀ y ꢀ zÞꢂ þ½h121 ꢀ ð1 ꢀ xÞyꢂ
2
þ ½h122 ꢀ ð1 ꢀ xÞz ꢀ xð1 ꢀ y ꢀ zÞꢂ
þ ðh123 ꢀ xyÞ þðh124 ꢀ xzÞ
2
2
ðA6Þ

Photochemistry and Photobiology, 2009, 85 907
Table A1. Mass spectra of SiF3Cl and SiF3CH2Cl. Each mass
spectrum was normalized to its most intensive line.
14. Polianski, M., O. V. Boyarkin, T. R. Rizzo, V. M. Apatin,
V. B. Laptev and E. A. Ryabov (2003) Infrared laser chemistry of
trichlorosilane in view of silicon isotope separation. J. Phys.
Chem. A 107, 8578–8583.
m ⁄ e
SiF3Cl
SiF3CH2Cl
m ⁄ e
SiF3Cl
SiF3CH2Cl
15. Gorelik, S. R., E. N. Chesnokov, L. V. Kuibida, R. R. Akberdin
and A. K. Petrov (2004) Infrared multiphoton dissociation of
difluorosilane. Appl. Phys. B 78, 119–125.
16. Petrov, A. K., A. V. Chernyshev, E. N. Chesnokov, S. R. Gorelik,
L. V. Kuibida, K. Nomaru and H. Kuroda (2002) Separation of
silicon isotopes in IR multiphonon dissociation of trifluorophe-
nylsilane induced by a free-electron laser. Doklady Akademii Nauk
385(5), 638. [In Russian]
17. Chernyshev, A. V., K. Nomaru, A. K. Petrov, E. N. Chesnokov,
S. R. Gorelik, L. V. Kuibida, R. R. Akberdin and H. Kuroda (2003)
Isotope selective dissociation of phenyltrifluorosilane under free
electron laser irradiation. J. Phys. Chem. A 107(44), 9362–9367.
18. Baranov, V. Y., A. P. Dyad’kin and V. A. Kuz’menko (1990)
Isotopically selective dissociation of CH3SiF3 and C6H5SiF3
molecules by CO2 laser pulses. J. Quantum Electron. 20, 450. [In
Russian]
19. Koshlyakov, P. V., E. N. Chesnokov, S. R. GorelikandA. K. Petrov
(2006) Infrared multiphoton dissociation of methyltrifluorosilane.
Russ. J. Phys. Chem. B 25(5), 22.
20. Alimpiev, S. S., A. M. Velichko, S. M. Nikiforov, G. L. Odo-
bashyan, B. G. Sartakov and S. V. Sin’ko (1988) Time-of-flight
photo-ionization of products of infrared dissociation of silicon
containing molecules. Pis’ma v Zhurnal Tekhnicheskoi Fiziki 14,
1786. [In Russian]
85
86
87
98
100
4.3
28.9
1.50
0.95
1.58
70.3
4.99
5.23
0.23
0.74
0.15
2.03
15.5
1.64
5.30
0.10
0.03
0.03
26
27
28
31
32
33
34
35
47
48
49
50
51
52
61
62
63
66
67
69
79
80
81
82
83
6.46
1.61
3.68
0.08
0.21
0.48
2.42
1.09
0.78
50.7
0.51
2.27
26.4
5.87
34.8
1.88
10.6
0.14
0.11
0.09
0.28
4.90
6.20
1.54
0.63
1.12
0.97
0.28
0.27
2.03
0.15
0.55
1.19
38.6
3.09
11.8
1.15
99
100
101
102
103
104
114
115
116
117
120
121
122
123
124
132
133
134
135
136
137
138
0.60
3.87
7.00
0.40
0.48
0.12
0.22
0.69
0.73
1.37
1.26
2.90
0.17
74.1
3.38
25.7
1.08
0.54
1.30
2.98
100
7.08
35.5
2.02
1.08
0.18
0.15
0.13
1.15
1.73
21. Koshliakov, P. V., S. R. Gorelik, E. N. Chesnokov, A. V. Voro-
biev and A. K. Petrov (2006) Infrared multiphoton dissociation of
vinyltrifluorosilane. Appl. Phys. B 84, 529.
22. Laptev, V. B., L. M. Tumanova and E. A. Ryabov (1998) Selective
dissociation of Si(OCH3)4 and [(CH3)3Si]2O molecules under
action of pulsed CO2 laser radiation. High Energy Chem. 32, 108.
[In Russian]
23. Fishwick, G., R. N. Haszeldine, C. Parkinson, P. J. Robinson and
R. F. Simmons (1965) Kinetics of the thermal decomposition of
polyfluoroalkylsilicon compounds. J. Chem. Soc, Chem. Common.,
382.
24. Bevan, W. I., R. N. Haszeldine, J. Middleton and A. E. Tipping
(1970) Novel thermal rearrangements involving interchange bet-
ween groups on silicon and those on its a-carbon. J. Organomet.
Chem. 23, 17.
25. Haszeldine, R. N., P. J. Robinson and R. F. Simmons (1967)
The kinetics of the reactions of silicon compounds. Part II. The
gas-phase thermal decomposition of 3,3,3-trifluoropropyltrifluo-
rosilane. J. Chem. Soc. B, 1357.
26. Haszeldine, R. N., P. J. Robinson and R. F. Simmons (1964) The
kinetics of the reaction of silicon compounds. Part I. The gas-
phase thermal decomposition of 2,2-difluoroethyltrifluorosilane.
J. Chem. Soc. B, 1890.
27. Graham, D., R. N. Haszeldine and P. J. Robinson (1971) Kinetics
of the reactions of silicon compounds. Part VII. Unimolecular gas-
phase thermal decomposition of 2,2-difluoroethyltrimethoxysi-
lane. J. Chem. Soc. B, 611.
REFERENCES
1. Ruf, T., R. W. Henn, M. Asen-Palmer, E. Gmelin, M. Cardona,
H. J. Pohl, G. G. Devyatych and P. G. Sennikov (2000) Thermal
conductivity of isotopically enriched silicon. Sol. St. Commun.
115, 243.
2. Kremer, R. K., K. Graf, M. Cardona, G. G. Devyatykh,
A. V. Gusev, A. M. Gibin, A. V. Inyushkin, A. N. Taldenkov and
H.-J. Pohl (2000) Thermal conductivity of isotopically enriched
28Si: Revisited. Sol. St. Commun. 131, 499.
3. Business Wire. FindArticles.com (2004) Isotopically Pure Silicon-
28 Recognized by International Semiconductor Industry. Avail-
able at: http://findarticles.com/p/articles/mi_m0EIN/is_2004_
March_9/ai_114045413. Accessed on 9 March 2004.
4. Bagratashvili, V. N., V. S. Letokhov, A. A. Makarov and
E. A. Ryabov (1985) Multiple Photon Infared Laser Photophysics
and Photochemistry. Harwood Academic Publ., New York.
5. Letokhov, V. S. (1989) Laser Spectroscopy of Highly Vibrationally
Excited Molecules. Adam-Hilger, Bristol.
6. Lide, D. R. (ed.) (1998) CRC Handbook of Chemistry and Physics,
79th edn, CRC Press, Boca Raton, FL.
7. Lyman, J. L. and S. D. Rockwood (1976) Enrichment of boron,
carbon, and silicon isotopes by multiple-photon absorption of
10.6 lm laser radiation. J. Appl. Phys. 47, 595–601.
8. Serdyuk, N. K., E. N. Chesnokov and V. N. Panfilov (1981)
Vibrational excitation effects on silane reactivity in its reaction
with bromine atoms. React. Kinet. Catal. Lett. 17, 19.
9. Kamioka, M., S. Arai, Y. Ishikawa, S. Isomura and N. Takamiya
(1985) 29Si and 30Si enrichment by IR MPD of Si2F6. Chem. Phys.
Lett. 119, 357–360.
10. Kamioka, M., Y. Ishikawa, H. Kaetsu, S. Isomura and S. Arai
(1986) Isotope-selective infrared multiple photon decomposition
of hexafluorodisilane. J. Phys. Chem. 90, 5727–5730.
11. Suzuki, H., H. Araki and T. Noda (1997) Enrichment of silicon
isotopes by infrared laser irradiation. J. Jpn. Inst. Met. 61, 145.
12. Lyman, J. L., B. E. Newman, T. Noda and H. Suzuki (1999)
Enrichment of silicon isotopes with infrared free-electron laser
radiation. J. Phys. Chem. A 103, 4227–4232.
28. Frisch, M. J., G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Mont-
gomery Jr, R. E. Stratmann, J. C. Burant, S. Dapprich,
J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci,
C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson,
P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck,
K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz,
A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong,
J. L. Andres, M. Head-Gordon, E. S. Replogle and J. A. Pople
(1998) Revision A.6–A.11. Gaussian, Inc., Pittsburgh, PA.
29. Becke, A. D. (1993) Density-functional thermochemistry. III. The
role of exact exchange. J. Chem. Phys. 98, 5648.
30. Lee, C., W. Yang and R. G. Parr (1988) Development of the
Colle-Salvetti correlation-energy formula into a functional of the
electron density. Phys. Rev. B 37, 785.
13. Arai, S., H. Kaetsu and S. Isomura (1991) Practical separation
of silicon isotopes by IRMPD of Si2F6. Appl. Phys. B 53, 199–202.

908 Pavel V. Koshlyakov et al.
31. Curtiss, L. A., K. Raghavachari, G. W. Trucks and J. A. Pople
(1991) Gaussian-2 theory for molecular energies of first- and sec-
ond-row compounds. J. Chem. Phys. 94, 7221.
32. Chesnokov, E. N., P. V. Koshlyakov and S. R. Gorelik (2005)
Absorption bands in IR spectrum of methyltrifluorosilane.
Calculation and comparison with experiment. Optika i Spektro-
skopiya 99(6), 923. [In Russian]
35. Kirmse, W. (1971) Carbene Chemistry, 2nd edn. Academic, New
York.
36. Roth, J. J. (1971) Photolysis of diazomethane in tetrachlorome-
thane. Abstraction reaction of singlet methylene. Am. Chem. Soc.
93, 1527.
37. March, J. (1986) Advanced Organic Chemistry, 3rd edn. Wiley,
New York.
33. Hancock, G. and M. R. Heal (1992) Temperature dependences of
methylene (ꢃa 1A1) removal rates by argon, nitric oxide, hydro-
gen, and ketene in the range 295–859 K. J. Phys. Chem. 96, 10316.
34. Okamura, H., V. Tosa, T. Ishii and K. Takeuchi (1996) Collisional
effects in the IR multiphoton absorption and dissociation of Si2F6.
J. Photochem. Photobiol. A 95, 203.
38. Krzyzanowski, S. and R. J. Cventanovic (1967) Relative rates of
reaction of olefins with the ground and the first excited electronic
states of methylene. Can. J. Chem. 45, 665.
39. Bradley, J. N. and A. Ledwith (1963) The reactions of carbenes
and diazoalkanes with ketones. Part I. The reaction of carbene
with acetone. J Chem. Soc., 3480.