Synthesis and thermal behavior of cis- and trans-1-tert-butyl-4,5-dimethyl- 2-phenyl-2-(trimethylsiloxy)-1-(trimethylsilyl)-1-silacyclohex-4-ene

Article abstract of DOI:10.1016/j.jorganchem.2010.07.031

The cothermolysis of benzoyl(tert-butyl)bis(trimethylsilyl)silane with 2,3-dimethylbutadiene in a sealed tube at 140 °C for 24 h afforded cis- and trans-1-tert-butyl-4,5-dimethyl-2-phenyl-2-(trimethylsiloxy)-1-(trimethylsilyl) -1-silacyclohex-4-ene (2 and 3) in a ratio of approximately 1:1 in 66% combined yield. When cis-silacyclohex-4-ene 2 was heated in a sealed tube at 250 °C for 24 h, dyotropic ring contraction took place to give 1-[(tert-butyl) (trimethylsiloxy)(trimethylsilyl)silyl]-3,4-dimethyl-1-phenylcyclopent-3-ene (4), but not trans-2-tert-butyl-4,5-dimethyl-2-phenyl-1-(trimethylsiloxy)-1- (trimethylsilyl)-1-silacyclohex-4-ene (6). The thermolysis of trans-silacyclohex-4-ene 3 under the same conditions, however, afforded two products, 1-silyl-1-phenylcyclopent-3-ene 4 and trans-1-tert-butyl-4,5-dimethyl- 2-phenyl-1-(trimethylsiloxy)-2-(trimethylsilyl)-1-silacyclohex-4-ene (5). The theoretical calculations were carried out to characterize the transition states and other local minima, and to evaluate the activation energies for the dyotropic rearrangement of 2 to 4 and 6, and 3 to 4 and 5. The energy barriers between 2 and 4, between 3 and 4, and between 3 and 5 were evaluated to be 188, 191, 192 kJ mol^-1, respectively. The energy barrier between 2 and 6, however, was calculated to be 201 kJ mol^-1 or higher. These results are consistent with the experimental finding that the thermal isomerization of 2 affords only 4, but 3 produces both 4 and 5.

Full text of DOI:10.1016/j.jorganchem.2010.07.031

Journal of Organometallic Chemistry 695 (2010) 2499e2505
Contents lists available at ScienceDirect
Journal of Organometallic Chemistry
journal homepage: www.elsevier.com/locate/jorganchem
Synthesis and thermal behavior of cis- and trans-1-tert-butyl-4,5-dimethyl-2-
phenyl-2-(trimethylsiloxy)-1-(trimethylsilyl)-1-silacyclohex-4-ene
,
a
a
b
b,
*
Akinobu Naka a , Shinsuke Ueda , Shu Sakaguchi , Toshiko Miura , Hisayoshi Kobayashi
,
*
Mitsuo Ishikawa a
,
*
a Department of Life Science, Kurashiki University of Science and the Arts, Nishinoura, Tsurajima-cho, Kurashiki, Okayama 712-8505, Japan
b Department of Chemistry and Materials Technology, Kyoto Institute of Technology, Matsugasaki, Kyoto 606-8585, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The cothermolysis of benzoyl(tert-butyl)bis(trimethylsilyl)silane with 2,3-dimethylbutadiene in a sealed
tube at 140 ꢀC for 24 h afforded cis- and trans-1-tert-butyl-4,5-dimethyl-2-phenyl-2-(trimethylsiloxy)-1-
(trimethylsilyl)-1-silacyclohex-4-ene (2 and 3) in a ratio of approximately 1:1 in 66% combined yield.
When cis-silacyclohex-4-ene 2 was heated in a sealed tube at 250 ꢀC for 24 h, dyotropic ring contraction
took place to give 1-[(tert-butyl)(trimethylsiloxy)(trimethylsilyl)silyl]-3,4-dimethyl-1-phenylcyclopent-
3-ene (4), but not trans-2-tert-butyl-4,5-dimethyl-2-phenyl-1-(trimethylsiloxy)-1-(trimethylsilyl)-1-
silacyclohex-4-ene (6). The thermolysis of trans-silacyclohex-4-ene 3 under the same conditions,
however, afforded two products, 1-silyl-1-phenylcyclopent-3-ene 4 and trans-1-tert-butyl-4,5-dimethyl-
2-phenyl-1-(trimethylsiloxy)-2-(trimethylsilyl)-1-silacyclohex-4-ene (5). The theoretical calculations
were carried out to characterize the transition states and other local minima, and to evaluate the acti-
vation energies for the dyotropic rearrangement of 2 to 4 and 6, and 3 to 4 and 5. The energy barriers
Received 18 May 2010
Received in revised form
23 July 2010
Accepted 29 July 2010
Available online 6 August 2010
Keywords:
Silacyclohexene
Dyotropic rearrangement
Thermolysis
between 2 and 4, between 3 and 4, and between 3 and 5 were evaluated to be 188, 191, 192 kJ molꢁ1
,
respectively. The energy barrier between 2 and 6, however, was calculated to be 201 kJ molꢁ1 or higher.
These results are consistent with the experimental finding that the thermal isomerization of 2 affords
only 4, but 3 produces both 4 and 5.
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
During the course of our investigation concerning the thermal
behavior of the silacycloalkenes, we found that the thermolysis of
A wide variety of silenes can be readily produced by the
photolysis, thermolysis, and Peterson type reaction of arylpolysi-
lanes, and many papers concerning the reactions of the silenes thus
formed with various substrates have been published to date [1].
Recently, considerable attention has been focused on the formation
and reactions of silacyclobutenes and silacyclohexenes, arising
from cycloaddition of the silenes to alkynes [2e5] and butadienes
[6e8]. On the other hand, it has been reported that these silacy-
cloalkenes readily undergo thermal isomerization to give various
types of the products, whose structures depend highly on the
substituents on the silicon atom and also the adjacent carbon atom
in the ring [2,6b]. Theoretical investigations to elucidate the
mechanism for the formation and reactions of these compounds
have also been carried out [2cee,h,5,6b].
the silacyclohexenes formed by the reactions of the silenes gener-
ated thermally from pivaloyl- and adamantoyltris(trimethylsilyl)
silane with butadienes undergo cleanly dyotropic rearrangement [9]
to give the isomerization products, in which two migrating groups,
trimethylsiloxy and trimethylsilyl groups on the stationary carbon
and silicon atoms in the six-membered ring are mutually exchanged.
In these rearrangements, the trimethylsiloxy group is introduced to
the silicon atom from the opposite side of the leaving trimethylsilyl
group. As the result, two migrating groups are located in a trans
fashion in the resulting products [6b]. Obviously, the driving force of
the rearrangement is thought to be the formation of an energetically
stable SieOeSi bond. For examples, the thermolysis of 2-tert-butyl-
and
2-adamantyl-4,5-dimethyl-2-(trimethylsiloxy)-1,1-bis(trimet
hylsilyl)-1-silacyclohex-4-ene proceeded cleanly to give cis-2-
tert-butyl- and cis-2-adamantyl-4,5-dimethyl-1-(trimethylsiloxy)-
1,2-bis(trimethylsilyl)-1-silacyclo-hex-4-ene, respectively. However,
similar thermolysis of 4,5-dimethyl-2-mesityl-2-(trimethylsiloxy)-
1,1-bis(trimethylsilyl)-1-silacyclohex-4-ene produced a ring-con-
tracted compound, 3,4-dimethyl-1-mesityl-1-[trimethylsiloxybis
* Corresponding authors.
E-mail addresses: anaka@chem.kusa.ac.jp (A. Naka), kobayashi@chem.kit.ac.jp
(H. Kobayashi), ishikawa-m@zeus.eonet.ne.jp (M. Ishikawa).
0022-328X/$ e see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2010.07.031

2500
A. Naka et al. / Journal of Organometallic Chemistry 695 (2010) 2499e2505
Me3Si
OSiMe3
Me3Si
H2C
Si
C
t-Bu
OSiMe3
Me3Si
Me3Si
(Me3Si)3SiCO(t-Bu)
Si
C
CH2
t-Bu
C
C
Me
Me
t-BuC CH
Me3Si
OSiMe3
Me3Si Si
C
C
t-Bu
C
H
t-Bu
(trimethylsilyl)silyl]cyclopent-3-ene.
No
1-silacyclohex-4-ene
HPLC were always shown to be a mixture consisting of 2 and 3.
Although we could not isolate 2 and 3 in a pure form, the mixtures
with different ratio of 2 and 3 could be obtained. Therefore, we
carried out the thermal reactions, using the mixture with
a different ratio of 2 and 3, A (2/3 ¼ 80/20) and B (2/3 ¼ 16/84), as
the starting compound.
derivative was detected in the product. In contrast to the behavior of
this mesityl-substituted derivative, the thermolysis of the
phenyl-substituted derivative, 4,5-dimethyl-2-phenyl-2-(trimethy
lsiloxy)-1,1-bis(trimethylsilyl)-1-silacyclohex-4-ene afforded the 1-
silacyclo-hex-4-ene derivative as the sole product.
In order to learn more about the thermal behavior of the 1-
silacyclohex-4-enes bearing a phenyl group on the carbon atom at
the 2-position in the ring, we have synthesized cis- and trans-1-
tert-butyl-2-phenyl-4,5-dimethyl-2-(trimethylsiloxy)-1-(trimethy-
lsilyl)-1-silacyclohex-4-ene, by the cothermolysis of benzoyl
(tert-butyl)(trimethylsilyl)silane with 2,3-dimethylbutadiene and
investigated their thermal behavior. We have also carried out
theoretical treatment to get more about information on the dyo-
tropic rearrangement.
The structures of 2 and 3 were verified by spectrometric anal-
ysis, as well as by elemental analysis, using the mixtures A and B,
respectively. The locations of the substituents on the silacyclo-
hexenyl ring for 2 and 3 were confirmed by differential NOE
experiments at 300 MHz. Thus, in the differential NOE experiments
for 2, saturation of the trimethylsiloxy protons on the ring carbon
atom at 0.10 ppm led to enhancement of the signal at 0.35 ppm,
attributed to the trimethylsilyl protons, as well as the ring methy-
lene proton at 2.78 ppm and the phenyl ring protons at 7.50 ppm.
Furthermore, irradiation of the trimethylsilyl protons on the silicon
atom in the six-membered ring at 0.35 ppm resulted in a strong
enhancement of the signals at 0.10 and 0.77 ppm, due to the tri-
methylsiloxy protons on the ring carbon and the tert-butyl protons
on the ring silicon, and of the signal at 1.48 ppm, attributable to the
ring methylene protons. These results clearly indicate that the tri-
methylsilyl group on the ring silicon atom and the trimethylsiloxy
group on the adjacent carbon atom for 2 are located in a cis fashion.
The trans configuration for 3 was also confirmed by differential
NOE experiments at 300 MHz. Thus, saturation of the tert-butyl
protons on the ring silicon atom at 1.08 ppm resulted in enhance-
ment of the signals at 0.06 and 0.13 ppm, attributable to the tri-
methylsiloxy protons and trimethylsilyl protons, as well as the ring
methylene protons.
When the mixture A was heated in a degassed sealed tube at
250 ꢀC for 24 h, two products, 1-[(tert-butyl)(trimethylsiloxy)(tri-
methylsilyl)silyl]-3,4-dimethyl-1-phenylcyclopent-3-ene (4) and
trans-1-tert-butyl-2-phenyl-1-trimethylsiloxy-2-trimethylsilyl-1-
silacyclohex-4-ene (5) were obtained in 43% and 12% yields,
respectively, in addition to the unchanged starting compounds 2
(40%) and 3 (5%) (Scheme 2). No other products were detected in
the reaction mixture by GLC analysis and also by spectrometric
analysis. The true product yield, obtained from each starting
compound 2 or 3, in the reaction mixture was determined on the
basis of the following assumptions: compound 2 undergoes
isomerization to give only 1-silyl-1-phenylcyclopent-3-ene 4, but
compound 3 affords two products, 4 and trans-silacyclohex-4-ene
5, respectively. If this idea is correct, compound 2 produces 4 in
40% yield, and compound 3 affords 4 and 5 in 3% and 12% yields,
respectively.
Me3Si
Me3Si
Me3Si Si
SiMe3
OSiMe3
Me3SiO
H2C
Si
C
R
C
R
CH2
H2C
CH2
C
C
C
C
Me
Me
Me
Me
R = t-Bu, Ad, Ph
R = t-Bu, Ad, Ph
OSiMe3
Me3Si
Me3Si
Me3Si
Me3Si Si
OSiMe3
Si
Mes
C
Mes
C
H2C
CH2
H2C
CH2
C
C
C
C
Me
Me
Me
Me
2. Results and discussion
The cothermolysis of benzoyl(tert-butyl)bis(trimethylsilyl)
silane (1) with 2,3-dimethylbutadiene at 140 ꢀC for 24 h gave
a mixture consisting of cis- and trans-1-tert-butyl-2-phenyl-2-
(trimethylsiloxy)-1-(trimethylsilyl)-1-silacyclohex-4-ene (2 and 3),
as shown in Scheme 1. No other products were detected in the
reaction mixture by spectrometric analysis. The ratio of the two
products was determined to be 1:1 by 1H NMR spectrometric
analysis. Unfortunately, all attempts to separate the product 2 from
3 were unsuccessful. The fractions separated by using the recycling
We also carried out the reaction using the mixture consisting of 2
and 3 in a ratio of 1:1. Thus, the thermolysis of this mixture under the
same conditions afforded the products 4 and 5 in 32% and 30% yields,
respectively, together with the unchanged starting compounds 2

A. Naka et al. / Journal of Organometallic Chemistry 695 (2010) 2499e2505
2501
Me3Si
and trimethylsiloxy protons, a single resonance at 0.79 ppm
attributed to tert-butyl protons, and a resonance at 1.52 ppm due to
the methyl protons on the sp2-ring carbons, as well as methylene
protons and phenyl ring protons. The 13C NMR spectrum of 4
reveals two resonances at 1.5 and 2.4 ppm, attributed to the tri-
methylsilyl carbons, and a resonance at 13.7 ppm, due to the methyl
carbons and a single resonance at 40.3 ppm attributable to the
quarternary sp3-ring carbon, as well as the resonances due to the
tert-butyl carbons, methylene carbons, and phenyl ring carbons.
The 29Si NMR spectrum shows three resonances at ꢁ21.8, 2.7, and
5.6 ppm, as expected. These results are wholly consistent with the
structure proposed for 4.
OSiMe3
OSiMe3
t-Bu
t-Bu
H2C
Ph
CH2
Ph
CH2
Si
C
Me3Si
H2C
Si
C
mixture A
:
=
80 : 20
C
C
C
C
Me
Me
Me
Me
2
3
3
mixture B
:
=
16 : 84
2
The structure of 5 was also confirmed by spectrometric anal-
ysis, as well as by elemental analysis (see Experimental Section),
and its trans configuration was verified by differential NOE
experiments at 300 MHz. Thus, irradiation of the tert-butyl
(Scheme 2) protons on the ring silicon atom at 1.17 ppm led to
enhancement of the signals at 0.09 and 0.14 ppm, attributable to
the trimethylsiloxy protons and the trimethylsilyl protons, as well
as a signal at 0.95 ppm, due to the ring methylene proton. These
results clearly indicate that compound 5 must have the structure
with trans configuration. The thermolysis of B (ratio of 2/3 ¼ 16/
84) under the same conditions proceeded to give the rearranged
product, 4 and 5 in 20% and 59% yields, in addition to the starting
compounds 2 (9%) and 3 (12%). On the basis of the assumption
described above, it may also be considered that compound 2 in the
mixture B produced the product 4 in 7% yield, and compound 3
gave both of the products 4 and 5 in 13% and 59% yields,
respectively.
(25%) and 3 (13%). Again the yields of the products 4 and 5 from each
starting compound are calculated to be 25% for the formation of 4
from 2 in the mixture, and 7% and 30% for 4 and 5 from 3. These
results are consistent with those obtained from the thermolysis of
the mixture A.
In fact, the idea is based on the results obtained from the
dyotropic rearrangement of the 2-tert-butyl- and 2-adamantyl-2-
(trimethylsiloxy)-1,1-bis(trimethylsilyl)-1-silacyclohex-4-ene deri-
vatives. We have found that the dyotropic rearrangement of these
compounds proceeds to give cis-1-(trimethylsiloxy)-1,2-bis(tri-
methylsilyl)-1-silacyclohex-4-ene derivatives, via the transition
state involving a pentacoordinate silicon atom [6b]. In this rear-
rangement, a trimethylsiloxy shift from the carbon atom at the 2-
position in the silacyclohexenyl ring to the ring silicon atom takes
place from the opposite side of the leaving trimethylsilyl group. As
a result, both migrating trimethylsiloxy and trimethylsilyl groups
from the stationary carbon and silicon atoms in the silacyclohex-
enyl ring are located at the trans position. In compound 2, the
trimethylsilyl group on the ring silicon atom and the trimethylsi-
loxy group on the adjacent phenyl-substituted ring carbon atom
are located at the cis position. Consequently, the formation of
trans-silacyclohex-4-ene 5 in the thermal isomerization of 2 must
involve inversion of the substituents on the ring silicon or adjacent
carbon atom with scission of the siliconecarbon bond in the
silacyclohexenyl ring. This process seems to be energetically
unfavorable. In fact, theoretical treatment indicates that the
energy of the TS for the reaction of 2 giving compound 5 is higher
than that of the reaction leading to the product 4.
One might consider the formation of a product, trans-2-tert-butyl-
4,5-dimethyl-2-phenyl-1-(trimethylsiloxy)-1-(trimethylsilyl)-1-sila-
cyclohex-4-ene (6), inwhich a tert-butyl group on the ring silicon and
a trimethylsiloxygroup on the carbon atom at the 2-position in the 1-
silacyclohexenyl ring in 2 mutually interchange their positions.
However, no product 6 was detected in the resulting mixture by
spectrometric analysis. The theoretical study shown below also
indicates that the formation of 6 is energetically unfavorable.
Products 4 and 5 were isolated by column chromatography. The
structure of 4 was confirmed by spectrometric analysis, as well as
by elemental analysis. The 1H NMR spectrum of 4 shows two
resonances at 0.23 and 0.26 ppm due to the trimethylsilyl protons
The present dyotropic rearrangement indicates that cis isomer
2
undergoes isomerization to give the ring-contracted
compound, silyl-substituted cyclopent-3-ene 4, and trans isomer
3 affords trans-silacyclohex-4-ene 5 with the exchange of the
Me3SiO and Me3Si group as a major product, along with 4 as
a minor product.
Theoretical calculations. To obtain further information about the
energy and structure changes in the dyotropic rearrangement of 2
and 3 to 4, of 3 to 5, and of 2 to 6, respectively, shown in Scheme 2,
theoretical calculations based on the density functional theory
(DFT) method were carried out. First, we adopted “simplified
models”, where all the methyl groups in the starting compounds
and also in the products are replaced by the hydrogen atoms and
the phenyl group by a methyl group, respectively. To distinguish
real models and the corresponding simplified models, a prime is
added to each number of the latter compounds. So the simplified
models analogues to 2, 3, 4, 5, and 6, are labeled as 2’, 3’, 4’, 5’, and
6’, respectively. For each reaction, the transition state (TS) was
characterized, and then the intrinsic reaction coordinate (IRC)
analysis was carried out at the TS for both directions. The IRC
calculation was restricted in the neighborhood of the TS. At the
endpoint, the IRC was followed by normal optimization runs to
confirm the continuity of the reaction coordinate to the reactant
and product. Second, LMs and TSs were reexamined with the real
models where the molecular structures are identical with those
Me3Si
OSiMe3
OSiMe3
t-Bu
Me3Si
Me3Si Si
t-Bu
H2C
Ph
Ph
CH2
O
C
Si
C
Me3Si
H2C
Si
C
H2C
CH2
Me
140°C, 24 h
+
Ph
CH2
+
C
C
Me
C
C
C
C
t-Bu
Me
Me
Me
Me
1
2
3
Scheme 1. Synthesis of 2 and 3.

2502
A. Naka et al. / Journal of Organometallic Chemistry 695 (2010) 2499e2505
Me3Si
t-Bu Si
above. Fig. 2(d) shows the TS which is formed in homolytic scission
of the silicon-carbon bond. The electronic structure for this TS was
obtained by the spin polarized solution.
OSiMe3
Ph
C
Me3Si
OSiMe3
t-Bu
H2C
Si
C
Ph
CH2
All the TS energies are compared for both simplified and real
models as shown in Fig. 3. With the real models, the energies for
TS(2 to 4), TS(3 to 4), TS(3 to 5), TS(2 to 6), and TS(2 to biradical
species (BS)) arising from homolytic scission of the ring sili-
conecarbon bond, are calculated to be 188, 191, 192, 201 and
204 kJ molꢁ1, respectively. Comparing the five TS energies, the
latter two are larger than the former three. For more quantitative
discussion, we used “the threshold energy” which is defined as
the difference between the highest energy among TS(2 to 4), TS(3
to 4), and TS(3 to 5) and the lower energy between TS(2 to 6) and
TS(2 to BS). The threshold energy was calculated to be 6 and
9 kJ molꢁ1 for the simplified and real models, and this difference
is too small to explain the result observed experimentally. We
carried out the frequency analysis to obtain the Gibbs free
energy, and the MP2 calculation. In the free energy calculation,
TS(2 to BS) was more stable than the TSs for the allowed reac-
tions, and the threshold energy could not be improved to explain
the experimental results at all. TS(2 to BS) has an open structure
with the SieC bond length of 3.8 or 4.2 Å for the simplified and
real model, and the major spin density localizes on the Si and C
atoms. The relative stability for TS(2 to BS) with the real model
increased remarkably compared to the simplified model. We feel
that the stability of radical species may be overestimated with
the B3LYP functional, and the entropy correction becomes more
significant for the open structures like TS(2 to BS). We estimated
the MP2 energy for the TSs optimized by the B3LYP functional.
The energy for TS(2 to BS) was calculated to be much higher than
those for the other TSs. The threshold energy also increased to
19 kJ molꢁ1. These results agree well with the experimental
finding that 3 produces both 4 and 5, but 2 affords 4, as the sole
product.
250°C, 24 h
H2C
CH2
Me
C
C
C
C
Me
Me
Me
4
2
Ph
C
Me3SiO
250°C, 24 h
Me3Si
H2C
Si
t-Bu
CH2
C
C
Me
Me
6
OSiMe3
t-Bu
Ph
Me3SiO
Me3Si
H2C
Si
C
Ph
CH2
t-Bu
H2C
Si
C
SiMe3
250°C, 24 h
4
+
CH2
C
C
C
C
Me
Me
Me
Me
3
5
Scheme 2. Thermolysis of 2 and 3.
used in the experiments. The activation energy is defined as energy
of the TS measured from the corresponding reactant. The calcula-
tions were carried out with the hybrid B3LYP method [10,11] and
the 6-31G(d) basis sets, which are implemented in the Gaussian 03
software [12].
Fig. 1 shows the structures and relative energies for four LMs,
and 2’, 3’, 4’, and 5’, and for two TSs between 2’ and 4’, and between
3’ and 5’. In the skeletal ring contraction of 2’e4’, the siloxy group
on the carbon atom in the six-membered ring migrates to the
adjacent silicon atom, and simultaneously, the methyleneesilicon
bond in the six-membered ring is broken, and then the methylene
group attacks the carbon atom adjacent to the silicon to form the
five-membered ring. On the other hand, in the isomerization
reaction of 3’ to 5’, the siloxy group on the carbon atom and the silyl
group on the silicon atom migrate mutually to exchange their
bonding atoms.
Fig. 2 shows the other TSs and LMs relating to the isomeri-
zation reactions illustrated in Scheme 2. The TS in the reaction of
3’ leading to 4’ is shown in Fig. 2(a). The structure of the product
is the same as 4’ described above, but it is the enantiomer. The
isomerization reaction of 2’ leading to the silacyclohex-4-ene,
similar to that of 3’ to 5’, however, does not afford 5’, but leads to
a new product 6’. The TS in the conversion of 2’ to 6’, denoted as
TS(2’ to 6’) and the virtual product 6’ are shown in Fig. 2(b) and
(c), respectively. The energy for the TS(2’ to 6’) is apparently
higher than those for the three reactions mentioned above. This
result is consistent with the fact that the product 6 was not
detected experimentally (The energetics is discussed later,
again.).
In conclusion, the thermolysis of cis-2-trimethylsiloxy-1-
trimethylsilyl-1-silacyclohex-4-ene
2
at 250 ꢀC underwent
dyotropic rearrangement to give the ring-contracted product,
silyl-substituted cyclopent-3-ene 4 as the sole product. No
product, trans-2-tert-butyl-1-trimethylsiloxy-1-silacyclohex-4-
ene 6, arising from mutual exchange of the trimethylsilyl group
and tert-butyl group on the silacyclohexenyl ring, was detected
in the reaction mixture. Similar thermolysis of trans-2-trime-
thylsiloxy-1-trimethylsilyl-1-silacyclohex-4-ene 3 afforded two
products, 4 and trans-1-trimethylsiloxy-2-trimethylsilyl-1-sila-
cyclohex-4-ene 5.
The theoretical calculation showed that 3 produced both 4
and 5, but 2 afforded only 4 through the comparison of the TS
energies, which was consistent with the experimental results.
The border line whether the reaction is allowed or forbidden
is rather subtle. However, our most reliable estimate, which
was obtained by using the B3LYP optimized structure and the
MP2 single point energy, indicated the threshold energy of
19 kJ molꢁ1
.
3. Experimental
A possible route for the isomerization of 2’ leading to the
formation of 5’ involves homolytic scission of the siliconecarbon
bond in the six-membered ring. Once the siliconecarbon bond in
the ring was cleaved, the methyl and silyl groups on the silicon
atom, and the methyl and siloxy groups on the carbon atom can
rotate freely. Recombination of the siliconecarbon bond after
rotation of the methyl and silyl groups on the silicon obviously
produces 3’. Thus, the resulting 3’ isomerizes to give 5’ as described
3.1. General procedure
All reactions of the 1-silacyclohex-3-ene derivatives were
carried out in a degassed sealed glass tube (1.0 ꢂ 15 cm). NMR
spectra were recorded on JNM-LA300 spectrometer and JNM-
LA500 spectrometer. Low-resolution mass spectra were measured
on a JEOL Model JMS-700 instrument. Column chromatography
was performed by using Wakogel C-300 (WAKO). Gel permeation

A. Naka et al. / Journal of Organometallic Chemistry 695 (2010) 2499e2505
2503
Fig. 1. Optimized structures for 2’ (a), TS(2’ to 4’) (b), 4’ (c), 3’ (d), TS(3’ to 5’) (e), and 5’ (f) with simplified models. 2’, 3’, 4’, and 5’ are simplified models corresponding to 2, 3, 4, and
5, respectively. E means the relative energy of TS and product measured from the corresponding reactant.
D
chromatography separation was performed with a Model LC-908
Recycling Preparative HPLC (Japan Analytical Industry Co. Ltd.).
benzoylchloride in 50 mL of THF at ꢁ78 ꢀC by the use of dropping
funnel. The mixture was stirred for 30 min at ꢁ78 ꢀC and allowed
to stand at room temperature for 12 h. After the mixture was
hydrolyzed with water, the organic layer was separated and the
aqueous layer was extracted with ether. The solvent ether was
evaporated and the residue was chromatographed on silica gel
eluting with hexane:ethyl acetate/1:1, to give 3.095 g (56% yield)
of 1: Anal. Calcd for C17H32OSi3: C, 60.64; H, 9.58. Found: C, 60.33;
3.2. Synthesis of benzoyl(tert-butyl)bis(trimethylsilyl)silane (1)
In a 200 mL two-necked flask was placed 5.035 g (18.7 mmol)
of tert-butyltris(trimethylsilyl)silane and 2.265 g (20.2 mmol) of
potassium tert-butoxide. To this was added 100 mL of THF at room
temperature. The solution turned yellow immediately. The
mixture was stirred for 43 h at room temperature. The resulting
solution was added slowly to a solution of 2.624 g (18.7 mmol) of
H, 9.44. MS m/z 336 (Mþ); 1H NMR
d (CDCl3) 0.26 (s, 18H, Me3Si),
1.15 (s, 9H, t-Bu), 7.41e7.53 (m, 3H, phenyl ring protons), 7.69 (d,
2H, phenyl ring protons, J ¼ 7.1 Hz); 13C NMR
d (CDCl3) 1.5 (Me3Si),

2504
A. Naka et al. / Journal of Organometallic Chemistry 695 (2010) 2499e2505
Fig. 2. Optimized structures for TS(3’ to 4’) (a), TS(2’ to 6’) (b), virtual product 6’ (c), and TS(2’ to BS), i.e., TS for the SieC bond breaking (d). See the caption of Fig. 1, and 6’ is
a simplified model corresponding to 6.
21.8 (CMe3), 30.9 (Me3C), 127.1, 128.2, 132.1, 144.8 (phenyl ring
140 ꢀC for 24 h. The reaction mixture was chromatographed on
silica gel using hexane as the eluent to give a mixture of products
2 and 3 (0.352 g, 66% combined yield). The ratio of the isomers
was calculated to be approximately 1:1 by the 1H NMR spectro-
metric analysis. All attempts to separate one isomer from the
other were unsuccessful. But the isolated mixture showed
different ratios of the isomers by the use of the recycling HPLC.
For 2: Anal. Calcd for C23H42OSi3: C, 65.96; H, 10.11. Found: C,
carbons), 237.7 (CO); 29Si NMR
(Me3SiSi).
d
(CDCl3) ꢁ25.9 (SiSiMe3), ꢁ15.1
3.3. Synthesis of 2 and 3
A mixture of 0.461 g (1.37 mmol) of 1 and 0.392 g (4.78 mmol)
of 2,3-dimethyl-1,3-butadiene was heated in a sealed tube at
65.99; H, 10.00. MS m/z 418 (Mþ); 1H NMR
d (CDCl3) 0.10 (s, 9H,
450
Me3Si), 0.35 (s, 9H, Me3Si), 0.77 (s, 9H, t-Bu), 1.48 (br s, 2H, ring
protons), 1.71 (s, 3H, Me), 1.81 (s, 3H, Me), 2.78 (d, 1H, ring proton,
J ¼ 16.8 Hz), 2.96 (d, 1H, ring proton, J ¼ 16.8 Hz), 7.00e7.52 (m,
simplified
real(B3LYP/SCF)
400
350
300
250
200
150
100
50
real(B3LYP/Free E)
real(MP2)
5H, phenyl ring protons); 13C NMR
d (CDCl3) 1.3, 3.1 (Me3Si), 16.3
(CH2Si), 19.8 (CMe3), 21.5, 23.7 (Me), 28.5 (Me3C), 48.9 (CH2), 76.9
(CO), 125.4, 125.6, 125.9, 126.0, 127.6, 149.2 (phenyl ring and
olefinic carbons); 29Si NMR
d
(CDCl3) ꢁ18.4 (SiSiMe3), ꢁ9.2
(Me3SiSi), 11.9 (OSiMe3). For 3: MS m/z 418 (Mþ); 1H NMR
(CDCl3) 0.06 (s, 9H, Me3Si), 0.13 (s, 9H, Me3Si), 1.08 (s, 9H, t-Bu),
d
1.29 (d, 1H, ring proton, J ¼ 16.5 Hz), 1.65 (s, 3H, Me), 1.78 (s, 3H,
Me), 1.80 (d, 1H, ring proton, J ¼ 16.5 Hz), 2.68 (d, 1H, ring proton,
J ¼ 17.4 Hz), 2.88 (d, 1H, ring proton, J ¼ 17.4 Hz), 6.99e7.51 (m,
0
TS(2 to 4)
TS(3 to 4)
TS(3 to 5)
TS(2 to 6)
TS(2 to BS)
TS structure
5H, phenyl ring protons); 13C NMR
d (CDCl3) 0.8, 2.7 (Me3Si), 16.5
Fig. 3. Comparison of the five TS energies evaluated for the simplified and real models.
For real models, the SCF energy, Gibbs free energy through the frequency analysis, and
MP2 energy are shown. The MP2 energy was estimated by a single point calculation
with the B3LYP optimized structure. Primes for simplified models are omitted for
clarity.
(CH2Si), 20.5 (CMe3), 21.1, 23.7 (Me), 29.1 (Me3C), 49.4 (CH2), 77.5
(CO), 125.2, 125.3, 125.5, 126.6, 127.8, 149.4 (phenyl ring and
olefinic carbons); 29Si NMR
(Me3SiSi), 10.8 (OSiMe3).
d
(CDCl3) ꢁ19.5 (SiSiMe3), ꢁ10.1

A. Naka et al. / Journal of Organometallic Chemistry 695 (2010) 2499e2505
2505
3.4. Thermolysis of 2 and 3 in a ratio of 80:20
Acknowledgment
Compounds 2 and 3 (0.0138 g, 0.0329 mmol) were heated in
a sealed tube at 250 ꢀC for 24 h. The reaction mixture was analyzed
by 1H NMR spectrum as being the products 4 (43%) and 5 (12%), in
addition to the starting compounds 2 (40%) and 3 (5%). Products 4
and 5 were isolated by column chromatography. For 4: Anal. Calcd
for C23H42OSi3: C, 65.96; H, 10.11. Found: C, 66.01; H, 9.98. MS m/z
We thank Shin-Etsu Co., Ltd. for the gift of chlorosilanes. This
work was supported by the Asahi Glass Foundation.
References
[1] (a) A.G. Brook, J. Organomet. Chem. 300 (1986) 21;
(b) A.G. Brook, in: S. Patai, Z. Rappoport (Eds.), The Chemistry of Organic
Silicon Compounds, Wiley, New York, 1989 (Chapter 15);
(c) A.G. Brook, M.A. Brook, Adv. Organomet. Chem. 39 (1996) 71;
(d) A.G. Brook, in: Z. Rappoport, Y. Apeloig (Eds.), The Chemistry of Organic
Silicon Compounds, vol. 2, Wiley, Chichester, 1998 (Chapter 21).
[2] (a) A. Naka, M. Ishikawa, S. Matsui, J. Ohshita, A. Kunai, Organometallics 15
(1996) 5759;
418 (Mþ); 1H NMR
d (CDCl3) 0.23 (s, 9H, Me3Si), 0.26 (s, 9H, Me3Si),
0.79 (s, 9H, t-Bu), 1.52 (s, 6H, Me), 2.87 (d, 1H, ring proton,
J ¼ 16.5 Hz), 2.93 (br s, 2H, ring protons), 3.17 (d, 1H, ring proton,
J ¼ 16.5 Hz), 7.02 (tt, 1H, phenyl ring proton, J ¼ 7.3 Hz, 1.2 Hz), 7.20
(t, 2H, phenyl ring protons, J ¼ 7.3 Hz), 7.39 (dd, 2H, phenyl ring
protons, J ¼ 7.3 Hz, 1.2 Hz); 13C NMR
d (CDCl3) 1.5, 2.4 (Me3Si), 13.7
(Me), 23.1 (CMe3), 27.7 (Me3C), 40.3 (quarternary ring carbon), 47.6,
49.8 (CH2), 124.4, 127.4, 127.7, 129.2, 129.7, 149.2 (phenyl ring and
(b) A. Naka, M. Ishikawa, J. Organomet. Chem. 611 (2000) 248;
(c) A. Naka, J. Ikadai, S. Motoike, K. Yoshizawa, Y. Kondo, S.-Y. Kang, et al.,
Organometallics 21 (2002) 2033;
(d) K. Yoshizawa, Y. Kondo, S.-Y. Kang, A. Naka, M. Ishikawa, Organometallics
21 (2002) 3271;
(e) Y. Shiota, M. Yasunaga, A. Naka, M. Ishikawa, K. Yoshizawa, Organome-
tallics 23 (2004) 4744;
(f) A. Naka, H. Ohnishi, J. Ohshita, J. Ikadai, A. Kunai, M. Ishikawa, Organo-
metallics 24 (2005) 5356;
(g) A. Naka, S. Motoike, N. Senba, J. Ohshita, A. Kunai, K. Yoshizawa, et al.,
Organometallics 27 (2008) 2750;
(h) A. Naka, N. Senba, S. Motoike, H. Fujimoto, T. Miura, H. Kobayashi, et al.,
Organometallics 28 (2009) 5641.
olefinic carbons); 29Si NMR
d
(CDCl3) ꢁ21.8 (Me3SiSi), 2.7 (SiSiMe3),
5.6 (OSiMe3). For 5: Anal. Calcd for C23H42OSi3: C, 65.96; H, 10.11.
Found: C, 65.86; H, 9.79. MS m/z 418 (Mþ); 1H NMR
(CDCl3) 0.09 (s,
d
9H, Me3Si), 0.14 (s, 9H, Me3Si), 0.95 (d, 1H, ring proton, J ¼ 17.1 Hz),
1.05 (d, 1H, ring proton, J ¼ 17.1 Hz), 1.17 (s, 9H, t-Bu), 1.60 (s, 3H,
Me), 1.67 (s, 3H, Me), 2.61 (d, 1H, ring proton, J ¼ 18.3 Hz), 2.83 (d,
1H, ring proton, J ¼ 18.3 Hz), 7.03 (tt, 1H, phenyl ring proton,
J ¼ 7.3 Hz, 1.2 Hz), 7.22 (t, 2H, phenyl ring protons, J ¼ 7.3 Hz), 7.36
[3] (a) K.K. Milnes, M.C. Jennings, K.M. Baines, J. Am. Chem. Soc. 128 (2006) 2491;
(b) K.K. Milnes, K.M. Baines, Can. J. Chem. 87 (2009) 307.
[4] (a) N. Auner, J. Mohseni-Ala, Inorg. Chim. Acta 359 (2006) 4677;
(b) D.C. Yan, M.D. Thomson, M. Backer, M. Bolte, R. Hahn, R. Berger, et al.,
Chem. Eur. J. 15 (2009) 8625.
[5] H. Tong, A.M. Eklof, P.G. Steel, H. Ottosson, J. Mol. Struct. THEOCHEM 811
(2007) 153.
[6] (a) M. Ishikawa, S. Matsui, A. Naka, J. Ohshita, Main Group Chem. 1 (1996)
219;
(dd, 2H, phenyl ring protons, J ¼ 7.3 Hz, 1.2 Hz); 13C NMR
d (CDCl3)
0.1, 2.3 (Me3Si), 17.9 (CH2Si), 20.17 (CMe3), 20.22, 24.2 (Me), 28.3
(Me3C), 29.6 (quarternary ring carbon), 39.1 (CH2), 123.4, 124.9,
125.8, 127.4, 127.8, 145.9 (phenyl ring and olefinic carbons); 29Si
NMR
d
(CDCl3) ꢁ1.8 (ring Si), 3.5 (Me3SiC), 7.8 (OSiMe3).
(b) A. Naka, S. Ueda, J. Ohshita, A. Kunai, T. Miura, H. Kobayashi, et al.,
Organometallics 27 (2008) 2922.
[7] (a) M.B. Berry, R.J. Griffiths, M.J. Sanganee, P.G. Steel, D.K. Whelligan,
Org. Biomol. Chem. 2 (2004) 2381;
3.5. Thermolysis of 2 and 3 in a ratio of 50:50
Compounds 2 and 3 (0.2587 g, 0.618 mmol) were heated in
a sealed tube at 250 ꢀC for 24 h. The reaction mixture was analyzed
by 1H NMR spectrum as being the products 4 (32%) and 5 (30%), in
addition to the starting compounds 2 (25%) and 3 (13%). Products 4
and 5 were isolated by column chromatography. All spectral data
for 4 and 5 were identical with those of the authentic sample
described above.
(b) M.J. Sanganee, P.G. Steel, D.K. Whelligan, Org. Biomol. Chem.
(2004) 2393;
(c) J.D. Sellars, P.G. Steel, Org. Biomol. Chem. 4 (2006) 3223;
(d) N.I. Hughes, R.D.C. Pullin, M.J. Sanganee, J.D. Sellars, P.G. Steel,
M.J. Turner, Org. Biomol. Chem. 5 (2007) 2841;
2
(e) R.D.C. Pullin, J.D. Sellars, P.G. Steel, Org. Biomol. Chem. 5 (2007) 3201.
[8] H. Ottosson, P.G. Steel, Chem. Eur. J. 12 (2006) 1576.
[9] (a) M.T. Reetz, Angew. Chem. Int. Ed. Engl. 11 (1972) 129;
(b) M.T. Reetz, Tetrahedron 29 (1973) 2189;
Adv. Organomet. Chem. 16 (1977) 33;
(c) J. Fassler, S. Bienz, Organometallics 13 (1994) 4704.(d) M. Kira, T. Iwamoto,
in: Z. Rappoport, Y. Apeloig (Eds.), In the Chemistry of Organic Silicon
Compounds, vol. 3, Whiley, Chichester, 2001 (Chapter 16);
(e) I. El-Sayed, T. Guliashvili, R. Hazell, A. Gogoll, H. Ottosson, Org. Lett.
3.6. Thermolysis of 2 and 3 in a ratio of 16:84
4
(2002) 1915;
Compounds 2 and 3 (0.0110 g, 0.0263 mmol) were heated in
(f) S.G. Cho, J. Organomet. Chem. 510 (1996) 25;
(g) Y.M. Yu, S.Y. Feng, J. Phys. Chem. 110 (2006) 12463.
[10] A.D. Becke, J. Chem. Phys. 98 (1993) 1372.
a sealed tube at 250 ꢀC for 24 h. The reaction mixture was analyzed
1
by H NMR spectrum as being the products 4 (20%) and 5 (59%), in
[11] P.M.W. Gill, B.G. Johnson, J.A. Pople, Int. J. Quantum Chem., Symp. 26 (1992)
319.
[12] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, et al., Gaussian 03, Revision C.02. Gaussian, Inc., Wallingford,
CT, 2004.
addition to the starting compounds 2 (9%) and 3 (12%). Products 4
and 5 were isolated by column chromatography. All spectral data for
4 and 5 were identical with those of the authentic samples described
above.