Laser-induced decomposition of silacyclobutane: Extensive H(Si)/H(C) scrambling via 1,2-H shift in silene and radical reactions

Article abstract of DOI:10.1016/S0022-328X(98)00764-5

The mechanism of the silacyclobutane decomposition has been further refined through a study of the laser induced decomposition of 1,1-dideuterio-1-silacyclobutane. It is concluded that the identified volatile and solid products and the hydrogen and deuterium content in them are in accord with 1,2-H(D)-shift in intermediate silene and with radical reactions.

Full text of DOI:10.1016/S0022-328X(98)00764-5

Journal of Organometallic Chemistry 566 (1998) 263–270
Laser-induced decomposition of silacyclobutane: extensive H(Si)/H(C)
scrambling via 1,2-H shift in silene and radical reactions
Lavrenti Khachatryan a, Elvira A. Volnina b, Radek Fajgar a, Josef Pola a,*
a Institute of Chemical Process Fundamentals, Academy of Sciences of the Czech Republic, Roz6ojo6a Str-135, 165 02 Prague 6, Czech Republic
b Institute of Petrochemical Synthesis, Russian Academy of Sciences, 117912 Moscow, Russia
Received 8 May 1998
Abstract
The mechanism of the silacyclobutane decomposition has been further refined through a study of the laser induced
decomposition of 1,1-dideuterio-1-silacyclobutane. It is concluded that the identified volatile and solid products and the hydrogen
and deuterium content in them are in accord with 1,2-H(D)-shift in intermediate silene and with radical reactions. © 1998 Elsevier
Science S.A. All rights reserved.
Keywords: Silacyclobutane; Laser-induced decomposition; Silene; H(Si)/H(C)scrambling
The predominant initial products in the hot sub-
strate-assisted decomposition in a cold wall low-pres-
sure reactor (above 1000°C) are [7] dihydrogen and
ethene in the ratio H2:C2H4 ca. 2, the latter being in
accord with the silene extrusion, while the former re-
maining unexplained, though its formation together
with 1-silacyclobutylidene is in accord with the known
ease of 1,1-H2 elimination from mono- and diorganylsi-
lanes [11–14]. The observation of cyclopropane under
the same conditions suggests a direct extrusion of
silylene from SCB.
Ethene, poly(silaethene) being the major and ethyne
and silane being minor products of UV laser photolysis
[9] show that this decomposition is governed by silene
polymerization.
The almost twice higher ethene/propene ratio in the
IR laser induced decomposition (IR multiphoton de-
composition, IRMPD) as compared to the conventional
pyrolyses, as well as the stoichiometry (C/Si ca. 1) of
and H content in the solid deposit, are in line with a
preference of silene formation when surface effects are
excluded, and with silene and methylsilylene dehydro-
genation followed by polymerization of SiCHn (nB4)
species [10].
1. Introduction
¸¹¹¹¹¹¹º
Thermolysis of silacyclobutane (H2SiCH2CH2CH2,
SCB) has attracted much attention [1–6] and the re-
newal of interest in this reaction stems from the suit-
ability of SCB to serve as a precursor for chemical
vapour deposition of silicon carbide [7–10]. The distri-
bution of products and consequently the mechanism of
SCB decomposition is affected by the decomposition
conditions.
The major decomposition pathways in the conven-
tional low pressure (hot wall assisted) pyrolysis (at ca.
0.1 Torr) or pyrolysis in stirred flow (N2 stream) reac-
tor—(i) the cleavage into silene and ethene, and (ii) the
cleavage into silylene and propene [5] via transient
n-propylsilylene and methylsilacyclopropane [6]—were
derived from the observation of volatile silicon non-
containing products (ethene and propene), scavenging
of three silicon-containing, short-lived transients (silene,
methylsilylene and silylene) with butadiene and hexam-
ethylcyclotrisiloxane [1,4], and from kinetic studies [5].
* Corresponding author. Tel.: +42 2 24311498; fax: +42 2
20920661; e-mail: pola@ICPF.CAS.CZ
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(98)00764-5

264
L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
PTFE stopcock and a port for rubber septum for
All these findings support the involvement of the
withdrawal of samples by a syringe. The irradiation
parameters and pressure of the components were se-
lected to match those at which LPD of SCB were
previously examined [10].
initial steps given in Scheme 1 and suggest that the
transient silene, methylsilylene and silylene may relax
by a multitude of consecutive reactions. The pivotal
molecule silene has been unambiguously detected in the
gas phase only very recently by monitoring its milimeter
wave spectrum [15,16]. Time-resolved UV spectra of the
transients in SCB decomposition induced by TEA
(transversely excited atmospheric) CO2, or ArF laser
are in accord with complex kinetics involving at least
two decay channels for silene and the onset of reactions
of other unidentified transients at higher fluences [17].
Although better understanding of the later decompo-
sition stages was attempted through the identification
of minor volatile products and the analysis of the
deposited solid material [7,10], the present state of
knowledge of the SCB decomposition leaves consider-
able uncertainty.
In our preceding papers we reported on the IR and
UV laser induced decomposition of several silacyclobu-
tanes with different substituents at the silicon [18–22]
and on the TEA CO2 or ArF laser induced decomposi-
tion of SCB [9,10,16,17]. We have also examined the
TEA CO2 laser-induced decomposition of SCB under
various decomposition conditions [10]. Extending these
results, we present in this paper GC/MS and GC/FTIR
analyses of the final products¸o¹f¹t¹he¹¹TºEA CO2 laser
induced decomposition of D2SiCH2CH2CH2 (SCB-d2)
and show that these products can be explained by
assuming that the earlier established initial molecular
fissions into silene and silylene are accompanied by
extensive H/D scrambling which is due to 1,2-H(D)
shift occurring in silene and in the course of radical
reactions. These views are supported by additional ex-
periments on the TEA CO2 laser induced decomposi-
tion of SCB and H2S¸iC¹D¹2¹C¹H¹D¹CºD2 (SCB-d5).
Changes in the composition of the irradiated cell
content were monitored by FTIR (Nicolet, model Im-
pact 400) spectroscopy. The withdrawn gaseous sam-
ples were analyzed by gas chromatography (a gas
chromatograph Shimadzu 14 A coupled with Chroma-
topac CR5A data processor), GC-MS (Shimadzu,
model QP 1000 quadrupole spectrometer), and GC-
FTIR (Nicolet, model Impact 400 coupled to a home-
made chromatograph/interface) techniques. For
chromatographic separation of the decomposition
products, a programmed temperature (20–150°C) and a
column packed with Porapak P (1.3 m) were employed.
The decomposition progress was monitored by using
diagnostic IR bands at 2146 (SCB), 1566 (SCB-d2), and
2146 cm−1 (SCB-d5). FTIR spectra of the solid prod-
ucts were recorded for materials deposited on the NaCl
cell windows and on a NaCl plate accommodated in the
cell prior to irradiation.
Samples of SCB-d2 and also of SCB and SCB-d5
were prepared according to previously reported proce-
dures (SCB-d2 and SCB, ref. [23]; SCB-d5, ref. [16]) and
their purity (\99%) was checked by gas chromatogra-
phy. Isotopic purity of SCB-d2 and SCB-d5 (\95%)
was confirmed by GC/MS technique and FTIR spec-
troscopy. The observed mass spectral fragmentations of
SCB and SCB-d2 were identical to those reported ear-
lier [6,18]. Sulfur hexafluoride was purchased from
Fluka. Authentic samples of trimethylsilane, trimethyl-
silane-d1 (CH3)3SiD), ethane-d1 and methane-d1 were
from laboratory stock.
2. Experimental details
Laser-induced homogeneous SF6-photosensitized de-
composition (LPD) of SCB-d2, and also of SCB and
SCD-d5, was carried out using a grating-tuned TEA
CO2 laser (P. Hilendarski Plovdiv University, 1300 M
model) operated on the P(20) line of the 0001“1000
transition (944.2 cm−1). The wavelength was confirmed
by a model 16-A spectrum analyzer (Optical Engineer-
ing). The incident laser beam was rectangular (1.8 and
1.0 cm per side), repetition frequency and incident
fluence in all experiments was 1 Hz and 0.5–0.8 J
cm−2, respectively.
Irradiation of gaseous SCB, SCB-d5 and SCB-d2
(each 3–5 Torr), SF6 (3–6 Torr), N2 (30–40 Torr) was
conducted in a cylindrical (3.6 cm i.d., 10 cm long)
Pyrex cell which was furnished with NaCl windows, a
Scheme 1.

L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
265
Fig. 3. Product distribution (in mol mol−1 of SCB-d2 decomposed) in
LPD of SCB-d2. Product designation: " ethene, ꢀ propene, ꢁ
methylsilane (mostly D2HC–SiD2H), X dimethylsilane (SiC2D5H3).
19% decomposition—SCB-d2 (4.6 Torr), SF6 (3 Torr), N2 (55 Torr),
2000 pulses; 27% decomposition—SCB-d2 (3.3 Torr), SF6 (3 Torr),
N2 (55 Torr), 5000 pulses; 43% decomposition—SCB-d2 (4.1 Torr),
SF6 (6.0 Torr), N2 (38 Torr), 60 pulses; 57% decomposition—SCB-d2
(4.0 Torr), SF6 (4.0 Torr), N2 (38 Torr), 1500 pulses; 94% decomposi-
tion—SCB-d2 (3.0 Torr), SF6 (3.0 Torr), N2 (38 Torr), 2000 pulses;
Fig. 1. Typical GC-MS trace of the mixture of products obtained by
LPD of SCB-d2. Column Porapak P. Peak identification: 1, methane,
SF6, N2 and ethene (C2H4); 2, methylsilane (mostly D2HC–SiD2H);
3, propene (C3H5D); 4, dimethylsilane (SiC2D5H3); 5, butane
(C4H9D); 6, propylsilane (SiC3D3H7 and SiC3D4H6); 7, SCB-d2; 8,
trisilane (Si3H5D3); 9, disilapentane (mostly Si2C3H6D6); 10, methyld-
isilacyclopentene (mostly Si2C4H6D4); 11, 1,3,5-trisilacyclohexane
(Si3C3D2H10) and 1,2,4-trisilacyclopentane (Si3C2H7D3). The formu-
lae in the brackets designates only the isotopomer with highest
number of D, though others with less D are also present.
all runs with fluence of 0.5 J cm−2
.
decomposition progress. This is documented in Fig. 3.
Residual pressure in the irradiated cell after freezing the
volatile compounds in a trap cooled with liquid nitro-
gen indicated the presence of some H2, D2 or HD. The
GC/MS trace reveals ethene as a major product, which
indicates that mostly silene gives rise to formation of
observed organosilicon products. The volatile
organosilicon products were identified by combining
the potential of GC/MS and GC/FTIR techniques as
singly, and multi (di-, tri- and tetra) deuterium-substi-
tuted methylsilane, dimethylsilane, propylsilane, disi-
lapentane, methyldisilacyclopentene, 1,3,5-trisilacyclo-
hexane, 1,2,4-trisilacyclopentane and trisilane (Table 1,
Fig. 2). Although we are unable to give quantities of
the particular isotopomers, the GC/MS and GC/FTIR
analyses (more specifically mass fragmentation and IR
absorptivity at wC–H, wSi–H and wSi–D) allow us to deter-
mine these species qualitatively. Absorption bands due
to wC–D (their partial overlap with those of wSi–H) do
not hamper our analysis, since (i) the wC–D band of RD
(R–CH3, C2H5) has its Q branch or maximum at 2200
cm−1 which is well separated from the wSi–H absorption
bands of all the volatile products (Fig. 2), and (ii)
absorptivity of the wC–D band of R–D (R–CH3, C2H5)
is by ca. 2 orders of magnitude lower than that of wSi–H
as checked with the RD and (CH3)3Si–H as model
compounds.
3. Results and discussion
The LPD of SCB-d2 induced with SCB-d2 (3–5
Torr), SF6 (3–6 Torr), N2 (38–55 Torr) mixtures at
incident laser fluence 0.5–0.8 J cm−2 and 50–5000
pulses affords a mixture of products of which some are
blends of several isotopomers. A typical GC/MS trace
and FTIR spectra for each peak of the trace are given
in Figs. 1 and 2, respectively. As reported previously by
us with the CO2 laser induced multiphoton decomposi-
tion of SCB [10], also the relative amounts of products
formed by LPD of SCB-d2 are, within the range of the
irradiating conditions, practically independent of the
The relative amounts (in mol mol−1 of SCB decom-
posed) of these products formed from SCB-d2 within
the decomposition progress 10–50%, methane (traces),
ethene (only C2H4, 0.45–0.55); methylsilanes (mostly
D2HC–SiD2H, 0.08–0.1), propene (only C3H5D, ca.
Fig. 2. FTIR spectra of the products of LPD of SCB-d2 separated by
gas chromatography (column Porapak P). The numbering of the
products is the same as in Fig. 1.

266
L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
Table 1
Mass and IR Spectra of GC-separated Fractions from SCB-d2 Decompositiona
Product
MS spectra m/z (relative intensity, %)
IR spectrab
A(wC–H):A(wSi–H
)
A(wSi–D):A(wSi–H)
Ethene
Methylsilane
14(3), 25(16), 26(93), 27(100), 28(100), 29(7)
28(40), 32(40), 44(35), 45(36), 46(100), 47(42), 48(62), 49(14),
50(7)
—
0.23
—
2.28
Propene
28(28), 39(51), 40(58), 41(54), 42(100), 43(76), 44(16), 45(4),
46(4)
—
—
Dimethylsilane
28(52), 44(85), 45(94), 60(82), 61(90), 62(100), 63(51), 64(22),
65(4)
0.31
1.44
Butane
Propylsilane
28(63), 29(44), 42(53), 43(100), 44(74), 58(11), 59(21)
28(30), 43(73), 44(92), 45(64), 61(28), 62(100), 74(27), 75(19),
76(5), 77(7)
28(20), 42(5), 43(10), 44(15), 45(8, 46(2), 86(14), 87(40),
88(72), 89(100), 90(57), 91(34), 92(40), 93(13), 94(3), 95(2)
28(26), 43(31), 44(22), 76(33), 77(55), 78(38), 88(52), 100(9),
101(12), 102(28), 103(51), 104(53), 105(60), 106(100), 107(71),
108(25), 109(15), 110(12)
—
5.9
—
5.9
Trisilanec
—
1.95
2.87
Disilapentane
2.06
Methyldisilacyclo-pentene
28(31), 44(38), 70(46), 74(39), 86(36), 90(28), 100(59), 101(99), 0.21
102(71), 103(20), 114(28), 115(78), 116(98), 117(100), 118(49)
28(25), 43(57), 44(100), 71(27), 78(22), 89(97), 90(68), 104(18), 2.22
117(12), 119(19), 120(30), 121(11), 132(8), 134(8)
0.35
2.47
1,3,5-Trisilacyclo-hexane
+1,2,4-Trisilacyclo-pentene
a
Obtained on the basis of GC-MS and GC-FTIR spectroscopy.
b A (absorptivity) in the region of 2765–3065 cm−1 (wC–H), 1976–2275 cm−1 (wSi–H) and 1492–1656 cm−1 (wSi–D).
c Together with traces of a compound tentatively assigned to 1,3-disilacyclobutane-dn.
0.1); dimethylsilanes (0.03); butanes (C4H10, C4H9D,
0.006), propylsilanes-d2–4 (0.05), trisilanes (0.06–0.08);
a disilapentane (0.05–0.07); a methyldisilacyclopentene
(ca. 0.04); 1,3,5-trisilacyclohexane and 1,2,4-trisilacy-
clopentane (0.06–0.07) can be reconciled with reactions
in Scheme 2. They indicate that the dominating initial
routes (1%) and (2%) are accompanied by a number of
other steps and the importance of these reactions is
independent of the amount of SCB-d2 decomposed.
Methylsilane-d4 and propylsilane-d4 are likely pro-
duced via silylene chain processes discovered [11,24] by
O’Neal and Ring (Reaction 7).
The observed methylsilane and propylsilane isoto-
pomers with Si–H bonds (Fig. 2) seem to support this
type of reaction with HD or H2. HD can be provided
via short chain radical processes which occur to a small
extent as with e.g. the thermolysis of methylsilane [26].
The fact that both HD and H2 (together with D2) can
be inferred [10] to be produced by dehydrogenation of
SiCHnDm (n+m=4), is supported by the FTIR spec-
tra of the solid deposits (see later).
The other, higher-molecular weight organosilicon
products might have been explained as being formed
from recombination of silene (D2SiꢀCH2), methylsi-
lylene (DH2C(D)Si:) and silylene (D2Si:), or as formed
by a sequence of the insertion of silylenes into the Si–D
bonds and dehydrogenation steps. However, the D-en-
richment (Table 1) and contributions of not only Si–D
but also of Si–H vibrations (Fig. 2) observed for the
isotopomeric mixtures of these products are not com-
patible with exclusively these reactions, but they reveal
that an important channel in the SCB-d2 decomposition
is H/D scrambling. These H/D exchange reactions are
also clearly manifested by the FTIR spectra of the
gaseous (Fig. 4) and of solid (Fig. 5) organosilicon
products deposited on the inside of the cell.
¸¹¹¹¹¹º
RDSi:+SCB-d2“D2RSi–(D)SiCH2CH2H2
“RSiD3+:Si(CH2)3
(7)
(R=DH2C, DH2CCH2CH2)
Considering, however, that H2 is produced upon
conventional [6,7] thermal decomposition of SCB and
that its important source in IRMPD of SCB is [10]
dehydrogenation of SiCH4 (H2SiꢀCH2 and (CH3)HSi:)
species, we may admit that an alternative (although
slower [25]) route for formation of methylsilane-d4 and
propylsilane-d4 can be the reaction of methylsilylene-d2
and propylsilylene-d2 with D2 (Reactions 8a,b).
The FTIR spectrum of all the volatile organosilicon
products derived by subtracting the contributions of
ethene, propene and SCB-d2 from the spectrum of the
irradiated mixture of SCB-d2 SF6-N2 (Fig. 4a) reveals
the wC–H absorption band at 2948–2985 cm−1 and the
wSi–H band at 2145 cm−1, the FTIR spectrum obtained
[D2SiꢀCH2“D(DH2C)Si:]+D2“(DCH2)SiD3
DH2CCH2CH2(D)Si:+D2“DH2CCH2CH2SiD3
(8a)
(8b)

L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
267
Scheme 2.
by adding the contributions (Fig. 2) of all the volatile
organosilicon products separated by GC/FTIR (Fig.
4b) reveals absorption bands of wC–H at 2962, wSi–H at
2159 and wSi–D at 1567 cm−1, but no wC–D absorption
band at 2200 cm−1. The failure to observe a wC–D
absorption band in the spectrum of the volatile
organosilicon products reflects lower oscillator strength
of C–D compared to the C–H bond and a higher
content of D(Si) than D(C) atoms.
The FTIR spectrum of the solid deposits from SCB-
d2 (Fig. 5b) reveals wC–H (2861–2919), wSi–H (2120),
wSi–D (1543), wSi–C–Si (1013) and wSi–C (826 cm−1) ab-
sorption bands where AwSi–H\AwSi–D. The spectrum of
the solid deposit from SCB-d5 (Fig. 5), given for the
sake of comparison, reveals a similar pattern of absorp-
tion bands with an additional broad shoulder at 2238
cm−1 which may indicate the occurrence of wC–D ab-
sorption band. It appears that the relative absorptivities
of wC–H, wSi–H, wSi–D and wSi–C–Si as well as wSi–C bands
in the deposits from SCB-d2 and SCB-d5 are very alike,
Fig. 4. FTIR spectrum of gaseous silicon-containing products result-
ing from LPD of SCB-d2 obtained by (a) subtracting contributions of
ethene, propene and SCB-d2 (peaks designated ꢀ relate to a fraction
of SF6), and by (b) adding contributions of all the silicon-containing
products as separated (Table 1) by the GC-FTIR technique. Irradia-
tion conditions: SCB-d2 (4 Torr), SF6 (4 Torr), N2 (38 Torr),
fluence=0.5 J cm−2, 30% decomposition.
Fig. 5. FTIR spectra of the deposits from SCB (a), SCB-d2 (b) and
SCB-d5 (c). Irradiation conditions with SCB, SCB-d2 and SCB-d5 as
in Fig. 4.

268
L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
Table 2
D2SiꢀCH2lDHSiꢀCHDlH2SiꢀCD2
the latter possibly taking place [4,34] via intermediary
methylsilylenes DH2C(D)Si: and D2HC(H)Si:.
(10)
FTIR spectra of the solid depositsa
Solid from
wSi–H
,
wSi–D
,
N
Si–D/NSi–H
NC–H/NSi–H
cm−1 cm−1
We have observed that (i) SCB-d2 becomes H(Si)
enriched in LPD occurring in the absence of hydrogen,
and (ii) it does not become H(Si) enriched when LPD
takes place in excess of hydrogen. Thus the irradiation
of SCB-d2 SF6 (each 4 Torr) in H2 (63 Torr) driven to
12% conversion leads to ca. 0.5% incorporation of H
into SCB-d2, while that of SCB-d2 SF6 (each 4 Torr) in
N2 (30 Torr) driven to 25–35% conversion results in ca.
6% incorporation of hydrogen. We judge that the
greater H(Si)-enrichment of SCB-d2 in the absence of
H2 rather than in its presence is due to the fact that the
addition of DHSiꢀCHD or H2SiꢀCD2 (rearranged
D2SiꢀCH2) (10, -1%) to ethene is more facile than the
reaction of silacyclobutylidene with H2 (-4%). The excess
of the high thermal conductivity hydrogen decreases the
temperature in the hot zone of the irradiated system
[35,36] and will also decrease the role of (i) elimination
[10] of H2, HD or D2 from D2SiꢀCH2, HDSiꢀCHD and
H2SiꢀCH2 and (ii) polymerization of D2SiꢀCH2,
HDSiꢀCHD and H2SiꢀCH2 (which are generated in
excess of H2 in lower concentrations) making these
silenes more available for the addition to ethene. As-
suming further that (i) both 1,1-D2 elimination (4%) and
2+2 cycloreversion (1%) are reversible (and well docu-
mented [37]) processes, and (ii) only silenes (D2SiꢀCH2,
HDSiꢀCHD and H2SiꢀCD2) but neither silylenes-d2 nor
1- or 2-silavinylidenes can add to ethene to yield silacy-
clobutane, the findings on the H(Si) SCB-d2 enrichment
suggest that the H/D exchange at the silicon occurs via
1,2-H(D) shifts in silene (Reaction 10) and not via
1,2-H shifts in 1-silavinylidene and silyne (Reaction (4%)
in Scheme 2). We therefore conclude that of both
molecular paths for H/D exchange at the silicon, route
(10) is more plausible.
LPD of SCB
LPD of SCB-d2 2117
LPD of SCB-d5 2122
2115
—
1537
1537
—
0.60
0.56
1.1
2.3
2.2
a Obtained upon irradiation of silacycle (3–5 Torr), SF6 (3–6 torr),
N2 (30–40 Torr) mixtures.
which is consistent with similar distribution of H and D
atoms between Si centers in both deposits.
Both spectra, and also that of the deposit from SCB
(Fig. 5a) show that all deposits contain less hydrogen
than poly(silene), (H2Si–CH2)n (for IR spectrum see
ref. [27]) and they are compatible [28–33] with
SiC:H(D) and SiC:H films. Provided that (i) the Si–H
and Si–D bonds are in the same environment of other
(C, Si) atoms and that the integrated absorptivity of
wSi–D and wSi–H is inversely proportional to the reduced
mass [33], and that (ii) the number (N) of the C–H and
Si–H bonds can be assessed from the integrated ab-
sorptivities of these bonds using an absorption cross
section of 0.74×10−21 cm2 for the wC–H band and
0.71×10−20 cm2 for the wSi–H band [29–31], we esti-
mate the relative content of the Si–H and Si–D bonds,
and that of the C–H and Si–H bonds as shown in
Table 2. The NSi–D/NSi–H ratios for LPD of SCB-d2
(0.6) and SCB-d5 (0.6) are equal; the same applies to
the NC–H/NSi–H ratio which is 2.3 for the LPD of
SCB-d2 and 2.2 for the LPD of SCB-d5.
We can thus infer that all these films are formed by
the sequence of reactions assumed by us in our previous
work [10] among those dehydrogenation of SiCHnDm
(n+m=4) species (Reaction 9), i.e. silene, and
methylsilylene (and also silylene) are important steps.
HDSi:, D2SiꢀCH2, (DCH2)DSi:—[–H2(–D2)]
An alternative explanation of the D/H exchange at
the silicon via reaction sequence (11) starting from step
(-2%) involving silylene k-insertion into the C–H(D)
bond can be ruled out on the following grounds.
“H(D)/Si/C(solid)
(9)
Assuming (i) a major role of Reaction (1%) and a
minor role of Reaction (2%) (ethene:propene ratio being
ca. 5), and (ii) the polymerization of D2SiꢀCH2,
D(DH2C)Si: and HDSi: as dominant paths forming the
solid deposit, the FTIR spectra of the deposits from
SCB-d2 and SCB-d5 should possess prevailing absorp-
tion bands either due to wSi–D or wSi–H vibrations,
respectively. However, they are (almost equally) consti-
tuted by both of Si–H and Si–D bands. The high
content of Si–H bonds in the products of LPD of
SCB-d2 (and of Si–D bonds in the products of LPD of
SCB-d5) cannot be explained in terms of Reactions (1%)
and (3%), but can be only reconciled in terms of (i)
Reaction (6%) producing the SiCH2 species which un-
dergo H migration(s) from carbon to silicon, and/or (ii)
1,2-H(D) shift in the intermediate silene (10),
It is known that intermolecular insertions of silylenes
into C–H bonds is by at least two and four orders of
magnitude slower than reaction of silylene into H2 or
Si–H bonds, respectively [25], and that the former have
[38–40], contrarily to the latter [25], sizeable activation
energy. Although silacyclic intermediates produced via
intramolecular insertions into C–H bonds are accepted
intermediates for alkylsilylene decomposition [40–42],
the intramolecular insertions into C–H bonds leading
to three membered ring formation are favoured over
larger ring production [41]. Indeed, only formation of
silacyclopropane through i-insertion, but not forma-

L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
269
tion of silacyclobutane via k-insertion (-2%) has been
References
proved in the examination of the flash vacuum pyroly-
sis of SCB-d2 [6]. Apart from that, the reaction se-
quence (11) should yield D(C) enriched silacyclobutane
which should decompose into D-enriched ethene. Our
observation of the only ethene isotopomer, C2H4 (Table
1) serves as unequivocal proof that path (11) does not
take place.
[1] C.M. Golino, R.D. Bush, L.H. Sommer, J. Am. Chem. Soc. 97
(1975) 7371.
[2] A.K. Maltsev, V.N. Khabasheku, O.M. Nefedov, Dokl. Akad.
Nauk. SSSR 247 (1979) 383.
[3] N. Auner, J. Grobe, Z. Anorg. Allg. Chem. 459 (1979) 15.
[4] R.T. Conlin, R.S. Gill, J. Am. Chem. Soc. 105 (1983) 618.
[5] I.M.T. Davidson, A. Fenton, S. Ijadi-Maghsoodi, et al., J.
Organomet. 3 (1984) 1593.
We admit, however, that the H/D scrambling is
partly due to radical reactions, since these reactions are
proved by the identification of methane, butane,
dimethylsilane and perhaps disilapentane and methyld-
isilacyclopentene among the products. The rate con-
stants at room temperature of the silylene insertion or
addition and that of the H-abstraction by silylene from
hydrocarbons (methane, ethane) were thought to be ca.
10−10–10−11 cm3 molecule−1 s−1 [25,43] and ca. (1+
0.5) 10−14 cm−3 molecule−1 s−1 [44], respectively.
However, it was admitted that the latter values might
correspond to reactions of silylene with impurities, and
that the real values of silylene H-abstraction at room
temperature must be significantly lower than 10−14
[6] T.J. Barton, N. Tillman, J. Am. Chem. Soc. 109 (1987) 6711.
[7] A.D Johnson, J. Perrin, J.A. Mucha, D.E. Ibbotson, J. Phys.
Chem. 97 (1993) 12937.
[8] A.D Johnson, J.A. Mucha, D.E. Ibbotson, Proc. Electrochem.
Soc. 92–18 (1992) 507.
&
[9] J. Pola, Z. Bastl, J. Subrt, R. Taylor, J. Mater. Chem. 5 (1995)
1345.
[10] Z. Bastl, H. Bu¨rger, R. Fajgar, et al., Appl. Organomet. Chem.
10 (1996) 83.
[11] M.A. Ring, H.E. O’Neal, S.F. Rickborn, B.A. Sawrey,
Organometallics 2 (1983) 1891.
[12] B.A. Sawrey, H.E. O’Neal, M.A. Ring, D. Coffey, Int. J. Chem.
Kinet. 16 (1984) 801.
[13] S.F. Rickborn, M.A. Ring, H.E. O’Neal, D. Coffey, Int. J.
Chem. Kinet. 16 (1984) 289.
[14] D.M. Rayner, R.P. Steer, P.A. Hackett, C.L. Wilson, P. John,
Chem. Phys. Lett. 123 (1986) 449.
[15] S. Baileux, M. Bogey, J. Breidung, et al., Angew. Chem. Int. Ed.
Engl. 35 (1996) 2513.
[16] S. Bailleux, M. Bogey, J. Demaison, et al., J. Chem. Phys. 106
(1997) 10016.
[17] S. Dhanya, A. Kumar, R.K. Vatsa, R.D. Saini, J.P. Mittal, J.
Pola, J. Chem. Soc Faraday Trans. 92 (1996) 179.
[18] J. Pola, V. Chvalovsky´, E.A. Volnina, L.E. Gusel’nikov, J.
Organomet. Chem. 341 (1988) 13.
cm−3 molecule−1 −1. Assuming the typical preexpo-
s
nential factor ca. 10−10 cm3 molecule−1 s−1 and the
activation energy of ca. 8–10 kcal mol−1 for bimolecu-
lar reactions implies that the reaction constant of the
addition of silylene at ca. 1000 K (effective temperature
in the laser hot zone) is only several times higher than
the reaction constant for the H abstraction by silylene.
This estimation is in favour of the occurrence of H-ab-
straction reactions from the main H(C) containing
products (the parent compound or mostly ethene,
propene-d1) which can further initiate short chain radi-
cal reactions resulting in the H/D scrambling (Reac-
tions 12 and 13).
[19] M. Sedla´cˆkova´, J. Pola, E.A. Volnina, L.E Gusel’nikov, J. Anal.
Appl. Pyrolysis 14 (1988) 345.
[20] J. Pola, E.A. Volnina, L.E. Gusel’nikov, J. Organomet. Chem.
391 (1990) 275.
&
[21] D. Cukanova´, J. Pola, J. Organomet. Chem. 453 (1993) 17.
[22] M. Urbanova´, E.A. Volnina, L.E. Gusel’nikov, J. Pola, J.
Organomet. Chem. 509 (1996) 73.
[23] J. Laane, J. Am. Chem. Soc. 89 (1967) 1144.
[24] S.F. Rickborn, M.A. Ring, H.E. O’Neal, Int. J. Chem. Kinet. 16
(1984) 1371.
[25] J.M. Jasinski, R. Becerra, R. Walsh, Chem. Rev. 95 (1995) 1203.
[26] P.S. Neudorfl, E.M. Lown, I. Safarik, A. Jodhan, O.P. Strausz,
J. Am. Chem. Soc. 109 (1987) 5780.
’
’
D(H2DC)Si:+RH“D(H2DC)HSi +R
(12a)
(12b)
(13a)
(13b)
’
’
D(H2DC)HSi “D(H)Si:+ CH2D
’
’
DHSi:+RH“DH2Si +R
[27] H.-J. Wu, L.V. Interrante, Polym. Prep. (Am. Chem. Soc. Div.
Polym. Chem.) (1991) 32.
’
DH2Si “H2Si:+D
[28] H. Ru¨bel, B. Schroder, W. Fuhs, J. Krauskopf, T. Rupp, K.
Bethge, Phys. Stat. Sol. (B) 139 (1987) 131.
[29] E. Gat, M.A. El Khakani, M. Chaker, et al., J. Mater. Res. 7
(1992) 2478.
[30] K. Mui, D.K. Basa, F.W. Smith, Phys. Rev. (B) 35 (1987) 8089.
[31] D.K. Basa, F.W. Smith, Thin Solid Films 192 (1990) 121.
[32] A. Jean, M. Chaker, Y. Diawara, et al., J. Appl. Phys. 72 (1992)
3110.
[33] M.H. Brodsky, M. Cardona, J. Cuomo, J. Phys. Rev. (B) 16
(1977) 3556.
In conclusion, we show that SCB-d2 decomposition is
not only controlled by initial silylene and silene extru-
sions but also by extensive H/D scrambling reactions.
Apart from the earlier demonstrated 1,2-hydrogen shift
from Si to C producing n-propylsilylene [6], we reveal
that other modes of H/D scrambling occur via the
1,2-H(D) shift in silene and also by radical chain reac-
tions initiated by H-abstraction by silylenes.
[34] G. Maier, G. Mihm, H.P. Reisenauer, D. Littman, Chem. Ber.
117 (1984) 2369.
[35] D.K. Russell, Chem. Soc. Rev. 19 (1990) 407.
Acknowledgements
[36] J. Pola, M. Farkac' ova´, P. Kuba´t, A. Trka, J. Chem. Soc.
Faraday Trans. I 80 (1984) 1499.
The work was supported by the grant of the
Academy of Sciences of the Czech Republic (Grant No.
A4072509).
[37] G. Raabe, J. Michl, Chem. Rev. 85 (1985) 419.
[38] B.A. Sawrey, H.E. O’Neal, M.A. Ring, D. Coffey, Int. J. Chem.
Kinet. 16 (1984) 31.

270
L. Khachatryan et al. / Journal of Organometallic Chemistry 566 (1998) 263–270
[39] I.M.T. Davidson, R. Scampton, J. Organomet. Chem. 271 (1984)
249.
[40] I.M.T. Davidson, J. Organomet. Chem. 437 (1992) 1.
[41] M.A. Ring, H.E. O’Neal, in: E.R. Corey, J.Y. Corey, P.P.
Gaspar (Eds.), Silicon Chemistry, Horwood, Chichester, 1988, p.
427.
[42] I.M.T. Davidson, in: E.R. Corey, J.Y. Corey, P.P. Gaspar
(Eds.), Silicon Chemistry, Horwood, Chichester, 1988, p. 173.
[43] J.M. Jasinski, in: A.R. Bassindale, P.P. Gaspar (Eds.), Frontiers
of Organosilicon Chemistry, Royal Society Chemistry, Cam-
bridge, UK, 1991, p. 112.
[44] J.M. Jasinski, J.O. Chu, J. Chem. Phys. 88 (1988) 1678.
.