The one- and two-photon photochemistry of benzylsilacyclobutanes, acyclic benzylsilanes, and 1,1,2-triphenylsilacyclobutane

Article abstract of DOI:10.1139/v99-249

The photochemistry of several ∝-silylbenzyl compounds has been investigated in hexane and in methanol solution. Direct photolysis of 1-benzyl-1-methylsilacyclobutane (1) in methanolic hexane solution produces 1-propyl-1methyl-2,3-benzosilacyclobutene (6) in quantitative yield, by a sequential two-photon process involving the photoactive isotoluene derivative 1-methylene-6-(1-methylsilacyclobutyl)-2,4-cyclohexadiene (13a), which has been identified on the basis of its ¹H NMR and UV absorption spectra. In contrast, direct irradiation of 1-benzyl-1-phenylsilacyclobutane (2) under similar conditions results in the formation of a complex mixture of products consistent with the competing formation of 1-benzyl-1-phenylsilene and benzyl- and 1-phenylsilacyclobutyl radicals. The silene is a transient which has been detected directly by laser flash photolysis of 2 (λ_max = 315 nm, τ ～ 4.5 μs). Free radical formation is shown to be due to secondary photolysis of a second primary product, 1-methylene-6-(1-phenylsilacyclobutyl)-2,4-cyclohexadiene (13b), which has also been detected and identified by static UV absorption (λ_max = 335 nm) and ¹H NMR spectroscopy. In a reaction with some analogy to the acid-catalyzed desilylation of allylsilanes, both 13a and 13b can be intercepted in neutral or acidic methanol solution to yield toluene and 1-methyl- or 1-phenyl-1-methoxysilacyclobutane, respectively. Direct photolysis of benzyldimethylphenylsilane (4) also leads to the formation of the corresponding isotoluene derivative, while benzyltrimethylsilane (3) exhibits negligible photoreactivity. The endocyclic benzylsilane 1,1,2-triphenylsilacyclobutane (5) is shown to undergo competing [2 + 2]-cycloreversion and [1,3]-silyl migration to yield a bicyclic isotoluene analogue, which reacts rapidly with methanol to yield the acyclic methoxysilane reported previously to be the main product of photolysis of this silacyclobutane in methanol solution. Relative quantum yields for isotoluene formation from photolysis of 1-4 and absolute rate constants for methanolysis of several of these compounds under neutral and acidic conditions have also been determined.

Full text of DOI:10.1139/v99-249

1459
The one- and two-photon photochemistry of
benzylsilacyclobutanes, acyclic benzylsilanes, and
1,1,2-triphenylsilacyclobutane
William J. Leigh and Thomas R. Owens
Abstract: The photochemistry of several α-silylbenzyl compounds has been investigated in hexane and in methanol so-
lution. Direct photolysis of 1-benzyl-1-methylsilacyclobutane (1) in methanolic hexane solution produces 1-propyl-1-
methyl-2,3-benzosilacyclobutene (6) in quantitative yield, by a sequential two-photon process involving the photoactive
isotoluene derivative 1-methylene-6-(1-methylsilacyclobutyl)-2,4-cyclohexadiene (13a), which has been identified on the
1
basis of its H NMR and UV absorption spectra. In contrast, direct irradiation of 1-benzyl-1-phenylsilacyclobutane (2)
under similar conditions results in the formation of a complex mixture of products consistent with the competing for-
mation of 1-benzyl-1-phenylsilene and benzyl- and 1-phenylsilacyclobutyl radicals. The silene is a transient which has
been detected directly by laser flash photolysis of 2 (λmax = 315 nm, τ ~ 4.5 µs). Free radical formation is shown to be
due to secondary photolysis of a second primary product, 1-methylene-6-(1-phenylsilacyclobutyl)-2,4-cyclohexadiene
1
(13b), which has also been detected and identified by static UV absorption (λmax = 335 nm) and H NMR spectros-
copy. In a reaction with some analogy to the acid-catalyzed desilylation of allylsilanes, both 13a and 13b can be inter-
cepted in neutral or acidic methanol solution to yield toluene and 1-methyl- or 1-phenyl-1-methoxysilacyclobutane,
respectively. Direct photolysis of benzyldimethylphenylsilane (4) also leads to the formation of the corresponding
isotoluene derivative, while benzyltrimethylsilane (3) exhibits negligible photoreactivity. The endocyclic benzylsilane
1,1,2-triphenylsilacyclobutane (5) is shown to undergo competing [2 + 2]-cycloreversion and [1,3]-silyl migration to
yield a bicyclic isotoluene analogue, which reacts rapidly with methanol to yield the acyclic methoxysilane reported
previously to be the main product of photolysis of this silacyclobutane in methanol solution. Relative quantum yields
for isotoluene formation from photolysis of 1–4 and absolute rate constants for methanolysis of several of these com-
pounds under neutral and acidic conditions have also been determined.
Key words: photochemistry, organosilicon, benzylsilane, silacyclobutane, silene, kinetics, isotoluene. 1468
Résumé : On a étudié la photochimie de plusieurs composés α-silylbenzyles en solutions dans l’hexane ou dans le
méthanol. La photolyse directe du 1-benzyl-1-méthylsilacyclobutane (1), en solution méthanolique d’hexane, conduit à
la formation de 1-propyl-1-méthyl-2,3-benzosilacyclobutène (6) avec un rendement quantitatif; la réaction se produit
par un processus séquentiel à deux photons impliquant le dérivé photoréactif isotoluène, 1-méthylène-6-(1-
1
méthylsilacyclobutyl)cyclohexa-2,4-diène (13a) que l’on a identifié sur la base de son spectre RMN du H et de son
spectre d’absorption UV. Par ailleurs, l’irradiation directe du 1-benzyl-1-phénylsilacyclobutane (2) dans des conditions
semblables conduit à un mélange complexe de produits qui pourraient résulter d’une compétition entre la formation des
radicaux 1-benzyl-1-phénylsilène, et benzyl- et 1-phénylsilacyclobutyles. Le silène est une espèce transitoire que l’on
a détectée directement par photolyse éclair à l’aide d’un laser du produit 2 (λmax = 315 nm, τ ~ 4,5 µs). On a
démontré que la formation du radical libre est due à une photolyse secondaire d’un deuxième produit primaire, la
1-méthylène-6-(1-phénylsilacyclobutyl)cyclohexa-2,4-diène (13b) que l’on a aussi détecté et identifié par absorption
1
UV statique (λmax = 335 nm) et spectroscopie RMN du H. Dans une réaction qui présente une certaine analogie
avec la désilylation acidocatalysée des allylsilanes, on peut intercepter les composés 13a ainsi que 13b, en solutions
méthanoliques tant neutre qu’acide, pour obtenir du toluène et respectivement du 1-méthyl- et du 1-phényl-1-
méthoxysilacyclobutane. La photolyse directe du benzyldiméthylphénylsilane (4) conduit aussi à la formation du dérivé
isotoluène correspondant alors que la photoréactivité du benzyltriméthylsilane (3) est négligeable. On a démontré que le
benzylsilane endocyclique, 1,1,2-triphénylsilacyclobutane (5), est soumis à des réactions en compétition de
cycloréversion [2 + 2] et de migration de silyle [1,3] qui conduisent à un analogue isotoluène bicyclique qui réagit
rapidement avec le méthanol pour fournir le méthoxysilane acyclique reconnu antérieurement comme le produit princi-
pal de la photolyse de ce silacyclobutane en solution dans le méthanol. On a aussi déterminé les rendements quantiques
Received July 30, 1999. Published on the NRC Research Press website on November 2, 2000.
This paper is dedicated to Professor Adrian Brook in honour of his many contributions to organosilicon chemistry.
W.J. Leigh1 and T.R. Owens. Department of Chemistry, McMaster University, 1280 Main Street West, Hamilton, ON L8S 4M1,
Canada.
1Author to whom correspondence may be addressed. Telephone (905) 525-9140 ext. 23715. e-mail: leigh@mcmaster.ca
Can. J. Chem. 78: 1459–1468 (2000)
© 2000 NRC Canada

1460
Can. J. Chem. Vol. 78, 2000
pour la formation de l’isotoluène par photolyse des composés 1–4 ainsi que les constantes de vitesse absolues de
méthanolyse de plusieurs de ces composés dans des conditions neutres et acides.
Mots clés : photochimie, organosilicium, benzylsilane, silacyclobutane, silène, cinétique, isotoluène.
Introduction
Leigh and
tached to silicon (2). Previous work suggests that benzyl-
Owens
silanes generally exhibit very low photoreactivity upon di-
rect excitation in condensed phases (19). We thus wondered
whether benzylsilacyclobutanes might show similar reactiv-
ity toward silene photoextrusion as is possessed by other
simple silacyclobutane derivatives, with minimal complica-
tions from competing reactions due to the benzylic
substituent. It quickly became clear that this is not the case,
prompting us to expand our studies to include three addi-
tional benzylsilane derivatives, two of which have been studied
previously: benzyltrimethylsilane (3) (19), benzyldimethyl-
phenylsilane (4), and 1,1,2-triphenylsilacyclobutane (5)
(4, 9, 20).
Silacyclobutanes are well-known photochemical and ther-
mal precursors to transient Si=C containing compounds
(“silenes”) (1–8), and are tremendously useful in experi-
ments aimed at detecting these reactive intermediates di-
rectly and studying the kinetics and mechanisms of their
various characteristic reactions (9–17). Photochemical silene
formation proceeds with useful efficiencies from silacyclo-
butanes bearing a wide range of substituents at silicon, rang-
ing from compounds in which the silacyclobutane ring is the
primary chromophore (e.g., 1,1-dialkylsilacyclobutanes) to
those in which the lowest excited state is localized more in
the substituent (e.g., 1-arylsilacyclobutanes). The only known
exception to this generalization is alkoxy-substituted deriva-
tives; substituents of this type, either directly at silicon or on
the aromatic rings of phenylated derivatives, renders the
silacyclobutane moiety inert to photocycloreversion (13, 17).
Recent work in our laboratory has been directed at explor-
ing the versatility of arylsilacyclobutanes as photochemical
silene precursors, through systematic variation of the struc-
ture and photophysical properties of the aryl chromophore
(18). These studies have also afforded new detail on the
mechanism of silene photoextrusion from arylsilacyclo-
butanes. For example, the reaction proceeds from the lowest
excited singlet state, probably via a 1,4-biradical intermedi-
ate that partitions between silene and the ground state of the
silacyclobutane. If this mechanism is correct, then the over-
all quantum yield for reaction will be a function of two fac-
tors. The first is the partitioning of the silacyclobutane
excited singlet state between Si—C bond cleavage and non-
productive decay via internal conversion, intersystem cross-
ing, and fluorescence; this tends to not vary dramatically as
a function of aryl substituent, at least within the series of
compounds that we have studied to date. The second is the
partitioning of the biradical intermediate between silene and
starting material. This does appear to vary as a function of
substituent at silicon, most likely due to substituent-induced
variations in the conformational properties of the biradical.
In this paper, we describe the results of a study of the
photochemistry in solution of two 1-benzylsilacyclobutanes
(1 and 2). These two compounds exemplify two different
photophysical situations – one in which the benzyl group is
the primary chromophore in the molecule (1) and one in
which the role of primary chromophore is shared between it
and the silacyclobutane ring via a second phenyl group at-
Results and discussion
Direct photolysis of 1-benzyl-1-methylsilacyclobutane (1;
0.05 M) in deoxygenated hexane solution containing metha-
nol (0.1 M) with the light from a low pressure mercury lamp
(254 nm) affords the single product 6 in quantitative (>98%)
yield (eq. [1]). The same product was formed, again quanti-
tatively, when the photolysis was conducted in the absence
of methanol.
[1]
In contrast, photolysis of 2 as a deoxygenated 0.05 M so-
lution in hexane containing 0.1 M methanol led to the more
complex mixture of products shown in eq. [2]. Alkoxysilane
7 is the product expected from trapping of 1-benzyl-1-
phenylsilene (12) by methanol, while 8–10 are most likely
the products of hydrogen abstraction or coupling of benzyl
[2]
© 2000 NRC Canada

Leigh and Owens
1461
Fig. 1. UV absorption spectra, recorded by laser flash photolysis of deoxygenated hexane solutions of (a) 1-benzyl-1-
methylsilacyclobutane (1; 0.002 M), 10–15 µs after the pulse and (b) 1-benzyl-1-phenylsilacyclobutane (2; 0.001 M), 3–3.5 µs after the
pulse. The inserts show typical traces, recorded at 335 nm.
and phenylsilacyclobutyl radicals. Several other products
were also formed, in yields which were too low to enable
identification. The mass balance was ~93% at 20% conver-
sion of 2.
Fig. 2. Static UV absorption spectra of a 0.017 M solution of 1
in hexane solution after various time intervals of photolysis with
254-nm light, illustrating the formation of isotoluene 13a (λmax
335 nm).
=
Flash photolysis of a deoxygenated hexane solution of 1
(0.007 M) with a KrF excimer laser (248 nm) led to the for-
mation of a weakly absorbing species which was stable over
10 ms, the longest timescale able to be monitored with our
system. The UV spectrum recorded 10–15 µs after laser ex-
citation is shown in Fig. 1a, along with a ∆OD vs. time pro-
file. No short-lived transient species could be detected at any
monitoring wavelength between 260–600 nm from this com-
pound. Similar experiments with 2 (0.008 M) led to the re-
sults shown in Fig. 1b: the formation of a long-lived species
which exhibits a similar spectrum to that observed from 1,
on which is superimposed a weak transient absorption which
decayed with pseudo-first order kinetics (τ ~ 4.5 µs). Sub-
traction of the spectrum of the long-lived species from the
spectrum recorded immediately after the laser pulse revealed
an absorption centered at 310–315 nm, which can be associ-
ated with the short-lived transient species. The lifetime of
this transient was shortened upon addition of methanol to
the solution; a plot of kdecay vs. [MeOH] according to eq. [3]
was linear, with slope kMeOH = (9 ± 2) × 107 M–1 s–1. The
spectrum and sensitivity of the lifetime of this transient to
the presence of methanol are consistent with its assignment
to 1-benzyl-1-phenylsilene (12). Neither of the long-lived
species formed from 1 and 2 were affected detectably by the
addition of up to 0.1 M methanol, over the 10 ms timescale.
photolysis for similar periods of time showed the growth of
a broad singlet at δ 3.18 ppm and a number of resonances in
the vinylic region which integrate to a total of six protons
relative to the signal at δ 3.18. These signals grow in during
the early stages of the photolysis and then remain roughly
constant in relative intensity upon continued irradiation, and
can thus be identified as arising from the same initially-
formed product that is responsible for the 335 nm UV
absorption. On the basis of these data, we assign the struc-
ture of this product to 1-methylene-6-(1-methylsilacyclo-
butyl)-2,4-cyclohexadiene (13a; eq. [4]). The spectral data
observed agree very well with those reported for other
[3]
kdecay = ko + kMeOH[MeOH]
The stable species produced in the flash photolysis experi-
ments with 1 and 2 could also be detected by static UV ab-
1
sorption spectroscopy, and by H NMR spectroscopy when
the reactions were carried out in cyclohexane-d12. For exam-
ple, the static UV spectrum of a 0.017 M solution of 1 as a
1
function of photolysis time is shown in Fig. 2. The H NMR
spectra of a cyclohexane-d12 solution of this compound after
© 2000 NRC Canada

1462
Can. J. Chem. Vol. 78, 2000
[4]
isotoluene derivatives (21–24). Compounds of this type are
surprisingly quite stable in non-nucleophilic solvents.
In the case of 1, these experiments indicated 13a to be the
sole primary photolysis product, building up steadily in con-
centration over a conversion range of 0–15% to a level
roughly 6% of that of the starting material, after which it de-
creased in parallel with 1 upon continued photolysis. Com-
pound 6 could only be detected in significant yields after the
formation of 13a had nearly reached its initial plateau, sug-
gesting that it is a secondary product resulting from photo-
lysis of the latter species. This was verified by photolyzing a
solution of 1 with 248 nm light to ca. 10% conversion, and
then continuing the photolysis with 337 nm light; this caused
13a to disappear, with concomitant formation of 6 and no
further consumption of 1, consistent with the fact that the
latter is nonabsorbing above ~280 nm.
tum yield for photolysis of the intermediate is exceedingly
low. This is also indicated by the fact that only slight irre-
versible bleaching of the 335 nm absorption due to 13b
could be detected in these experiments. Flash photolysis ex-
periments carried out with optically matched (at 248 nm) so-
lutions of 1 and 2 indicated that the quantum yield for
formation of 13b from 2 is only slightly higher than that for
formation of 13a from 1 (Φ13b/Φ13a ~ 1.3). This suggests
that the main effect of replacing the methyl substituent in 1
with phenyl (in 2) is to promote silene formation in minor
amounts, via [2 + 2]-cycloreversion of the silacyclobutane
ring. The spectra shown in Fig. 1 were recorded with solu-
tions which were not optically-matched, and with different
excitation laser intensities.
Interestingly, the identity of the second silyl substituent
has a far more substantial effect on the photochemistry of
the corresponding isotoluene derivatives 13a, b. The forma-
tion of 6 from 13a can be formulated in terms of a non-
concerted mechanism involving rearrangement of the 1,4-
biradical intermediate formed by cleavage of the one of the
endocyclic silacyclobutyl Si—C bonds (eq. [5]). If this
mechanism is correct, then the photochemistry of 13a ini-
tially follows a course typical of 1-substituted methylsil-
acyclobutanes (18). Unlike other methylsilacyclobutane-
derived 1,4-biradicals however, that derived from 13a has
available two possible reaction pathways that are both likely
to be less energetically demanding than cleavage to the cor-
responding silene and ethylene: rearrangement to yield the
more stable cyclohexadienyl biradical 14a which can then
aromatize to 6 via intramolecular disproportionation, or
intramolecular abstraction of the cyclohexadienyl proton to
form an ortho-silaxylylene intermediate (14b) followed by
electrocyclic ring closure. We favour the former mechanism,
on the basis of the fact that none of the expected trapping
product of 14b was observed in the photolysis of 1 in 0.3 M
MeOH solution; however, its possible intermediacy in the
formation of 6 from 13a cannot be rigorously ruled out.
The photochemistry of 13b, on the other hand, is evi-
dently initiated by preferential cleavage of the weaker of the
exocyclic silacyclobutyl Si—C bonds, yielding benzyl- and
phenylsilacyclobutyl radicals. To our knowledge, no other
examples of exocyclic Si—C bond cleavage are known in
silacyclobutane photochemistry. One possible explanation for
the difference in the photochemistry of 13a and 13b is that
the latter has available to it a conformer in which a reasonable
degree of π-overlap is possible between the methyl-
enecyclohexadienyl and phenyl rings; this could lead to pref-
erential weakening of the exocyclic Si—C bond due to
charge-transfer interactions between the two moieties in the
lowest excited singlet state. While no evidence for such in-
teractions could be obtained from the UV absorption spectra
Similar experiments with 2 indicated that only alkoxysilane
(7), ethylene, and a product identifiable as isotoluene 13b
are formed upon photolysis in hexane or cyclohexane-d12 so-
lution to low (<ca. 10%) conversions. Again, the isotoluene
1
derivative could be detected by both static UV and H NMR
spectroscopy, and products 8–10 were found to be produced
only after the concentration of 13b rose to ~5% of that of 2.
Alternating 248 and 337 nm irradiation experiments verified
that 8–10 are formed as a result of secondary photolysis of 13b.
The formation of 13a, b as the major primary products of
photolysis of 1 and 2 indicates that the photochemistry of
these compounds is in both cases dominated by reactivity
specific to the benzyl chromophore rather than the silacy-
clobutane moiety, unlike the case with other silacyclobutanes
that have been studied. Silene formation competes signifi-
cantly only in 2, where the silacyclobutane ring bears a sec-
ond phenyl chromophore which competes effectively with
the benzyl group for direct light absorption. The reactive ex-
cited states in both 1 and 2 are most likely the lowest singlet
states, since our flash photolysis experiments indicate that
13a, b are both formed within the duration of the laser pulse
(~25 ns), consistent with either a concerted sigmatropic rear-
rangement or a singlet radical-pair recombination mecha-
nism for their formation. No evidence for the competing
formation of benzyl radicals in the primary photochemical
event could be obtained by flash photolysis in either case,
which provides strong evidence against a mechanism for the
formation of 13 which involves the triplet excited states of 1
and 2. For that matter, no evidence for benzyl radical forma-
tion could be obtained either in two-laser flash photolysis
experiments with 2, where the second (337 nm) laser pulse
was delivered 5–10 µs after 13b was formed by the primary
(248 nm) pulse. Since product studies clearly indicate that
free benzyl radicals must be formed upon secondary
photolysis of 13b, this result merely suggests that the quan-
© 2000 NRC Canada

Leigh and Owens
1463
[5]
[6]
[7]
of the two compounds, the vinylic resonances in the 1H
NMR spectrum of 13b appear at consistently lower fields
than those in the spectrum of 13a, indicative of significant
deshielding of the vinyl protons by the phenyl ring. Neither
compound appears to undergo the 4π electrocyclic ring clo-
sure that is typical of other isotoluene derivatives that have
been studied (24), to any detectable extent.
mixture therefore varied depending on the time taken be-
tween photolysis and GC analysis. The yields of 8 and 11a
increased to a combined yield of ~73% upon photolysis of a
methanol solution of 1 containing 0.1 M HCl, and as ex-
pected (27), the relative yields of 15 and 16 changed to
1.4:1. Evaporation of the solvent after photolysis of a hex-
ane solution of 1 to ca. 10% conversion and redissolution of
the mixture in methanol confirmed that the formation of
toluene (8) and 1-methoxy-1-methylsilacyclobutane (11a)
are due to reaction of 13a with the alcohol. Monitoring of
this reaction by UV spectroscopy revealed that the
isotoluene disappears with pseudo-first order kinetics and a
rate constant of k–13a = (1.10 ± 0.05) × 10–3 s–1 in neutral
methanol solution at 25°C. In methanol containing 0.1 M
HCl, disappearance of 13a was too rapid for a rate constant
to be measured by conventional UV absorption spectros-
copy.
Steady state photolysis of 2 in deoxygenated methanol so-
lution led to the formation of two main products, toluene (8)
and disiloxane 17, and only traces of bibenzyl (9) and the
silene-derived methoxysilane 7 (eq. [8]). The mass balance
in this experiment was ~92% after ca. 40% conversion of 2.
Disiloxane 17 most likely arises from condensation of 11b
due to the presence of small amounts of water in the solvent,
a process which can be expected to proceed more readily in
11b than in 11a under neutral (as well as acidic) conditions
(28). As before, evaporation of the solvent after photolysis
of a hexane solution of 2 to ca. 10% conversion and
redissolution of the mixture in methanol resulted in the
disappearance of the 335 nm UV absorption due to 13b and
the formation of 8 and 17. The reaction is complete within
~5 s under these conditions, but was found to be too slow
for a rate constant to be obtained by flash photolysis. Thus,
while an absolute rate constant for the reaction of 13b with
methanol could not be easily determined, it is clearly at least
To gain additional evidence for the primary involvement
of 13a, b in the photochemistry of 1 and 2, we examined the
photolyses of the latter two compounds in neat methanol
solution. One of the common properties of isotoluene deriva-
tives is their sensitivity towards acid-catalyzed aroma-
tization, the exact course of which (isomerization vs. solvolysis)
depends on the nature of the substituents at the bis-allylic
position of the methylenecyclohexadienyl ring (21, 24).
Judging from the well-known reactivity of allylsilanes to-
ward acid-catalyzed desilylation (25, 26), the present com-
pounds should be expected to react readily with aliphatic
alcohols to yield toluene and the corresponding methoxy-
silacyclobutane. Thus, it seemed very likely that the forma-
tion of the secondary products of photolysis of 1 and 2
might be suppressed in methanol solution due to interception
of the primary products (13) by reaction with the solvent.
This might also explain the formation of methoxyphenyl-
silacyclobutane (11b) in trace amounts in the photolysis of 2
in hexane containing 0.1 M methanol (eq. [2]).
As expected, photolysis of 1 in neat methanol produced 6
(69%), toluene (8; 7%), and 1-methoxy-1-methylsilacyclo-
butane (11a; 11%), as shown in eq. [6]. The reaction mixture
was actually somewhat more complicated than this, because
like other benzosilacyclobutenes (27), 6 reacts rapidly in
methanol solution, in this case to yield two methoxysilanes
in a 8:1 ratio. These were tentatively identified as 15 and 16,
1
respectively, on the basis of H NMR and GC/MS evidence
(eq. [7]). The actual amount of 6 detected in the photolysis
© 2000 NRC Canada

1464
Can. J. Chem. Vol. 78, 2000
[8]
[9]
[10]
two orders of magnitude greater than that for reaction of 13a
under similar conditions. The much higher reactivity of 13b
toward dark reaction with methanol more effectively lowers
its steady-state concentration during the photolysis of 2 in
this solvent, resulting in much lower yields of secondary
photolysis products relative to what is observed in hydrocar-
bon solvents, compared to 13a.
While the photoisomerization of 1 and 2 to the corre-
sponding isotoluene derivatives has precedent in the photo-
chemistry of other benzyl compounds (24), it has not been
reported previously for benzylsilanes. Benzyltrimethylsilane
(3) was first reported to be inert to photolysis in solution
(20), but Kira et al. (19) later found that extended photolysis
of the compound in hydrocarbon solution leads to the ineffi-
cient formation of ortho-tolyltrimethylsilane (18a), which
they suggested to arise from a minor reaction of benzyl and
trimethylsilyl radicals produced by homolysis of the benzyl
carbon-silicon bond in the benzylsilane. More recently, spec-
troscopic evidence for the formation of free benzyl radicals
upon photolysis of 3 in an ethanol matrix at 77 K has been
reported by Hiratsuka et al. (29).
fied as the isotoluene derivative 13d on the basis of its UV
and H NMR spectra. This compound, the only detectable
1
product at low (<5%) conversion of 4, is formed with a sim-
ilar quantum yield to that for formation of 13b from 2. It
also behaves similarly to 13b with respect to secondary
photolysis and reaction with methanol. For example, eqs. [9]
and [10] show the product mixtures obtained from photo-
lysis of 4 in deoxygenated hexane and methanol solutions,
respectively, to 40–50% conversion of the starting material.
Interestingly, a product tentatively identified as ortho-
tolyldimethylphenylsilane (18b) could be detected in signifi-
cant yield by GC/MS analysis of the hexane photolyzate. It
is barely resolved from 4 under our GC/MS conditions, and
coelutes with 4 under those employed for our GC/FID analy-
ses. A closer inspection of the product mixture from
photolysis of 2 in hexane revealed that an analogous iso-
meric product is formed in this case as well, but in yields of
less than 10%. These products are most likely formed by a
free radical pathway initiated by hydrogen abstraction from
13b/d by benzyl radicals, since they are not formed in de-
tectable yields in methanol solution, where the steady state
concentration of 13 is substantially lower due to reaction
with the solvent. As was found with 13b, dissolution of 13d
in neutral methanol results in its immediate disappearance
and the concomitant formation of 8 and 20. It is possible
that the higher photoreactivity of 4 compared to 3 with re-
spect to isotoluene formation is due to similar factors as
those suggested above to explain the different photo-
reactivities of 13a, b. However, we are at a loss to explain
why the methylsilacyclobutyl derivative 1 is also more reac-
tive than 3 toward this photorearrangement.
We also find that the parent benzylsilane exhibits very low
photoreactivity in solution at room temperature. Extended
photolysis of 3 in hexane with 254 nm light produces very
weak UV absorptions on the long-wavelength edge of the
The results reported above provide a reasonable explana-
tion for one or two unusual aspects of the photochemistry of
2-phenylsilacyclobutanes (5) which have been reported pre-
viously (4, 9, 20). Both 1,1-dimethyl-2-phenyl- and 1,1,2-
triphenylsilacyclobutane (5a and 5b, respectively), yield sty-
rene and the corresponding silene 21a, b upon photolysis in
methanol solution, as evidenced by the formation of the
methoxysilane (22) expected from trapping of the silene by
the solvent (eq. [11]) (4, 20). Also formed are the
methoxysilanes 23a, b, which were originally ascribed to
nucleophilic trapping of the 1,4-biradical/zwitterion formed
1
absorption band due to 3, but the H NMR spectrum of the
mixture shows no evidence of vinylic resonances consistent
with the formation of isotoluene 13c in detectable amounts.
The quantum yield for formation of this product from 3 is
thus no greater than ~5% of that for the formation of 13a
from 1 under similar conditions.
On the other hand, photolysis of benzyldimethylphenyl-
silane (4) in hexane results in the fairly efficient formation
of a long-wavelength absorbing species which can be identi-
© 2000 NRC Canada

Leigh and Owens
1465
[11]
[12]
[13]
by cleavage of a single carbon—silicon bond in the excited
state of the silacyclobutane (eq. [11]. The formation of these
products was offered as the first evidence for a stepwise re-
action mechanism for the formal [2 + 2]-photocycloreversion
of silacyclobutanes (4, 20).
Our results for 1–4 suggest that a more likely explanation
for the formation of 23 is that it arises from reaction of
methanol with the isotoluene derivative (24) formed by
[1,3]-silyl migration into the ortho-position of the 2-phenyl
ring in 5 (eq. [12]). The isotoluenes 24a, b would be ex-
pected to be stable enough that they should build up to sig-
nificant concentrations, at least in non-nucleophilic solvents.
However, their secondary photolysis should result mainly in
reversion to the biradical precursor, so no additional prod-
ucts that might have signalled their presence would be ex-
pected.
In fact, we have already published evidence consistent
with this interpretation, although we failed to recognize it
at the time (9). Laser flash photolysis of 5b in isooctane
leads to the formation of a transient species identified as
1,1-diphenylsilene (21b), whose absorptions (λmax = 325 nm)
are superimposed on those due to a species which is stable
over the timescale of our experiment and exhibits
λmax ~ 335 nm. We originally reported that this species did
not persist over a timescale of several minutes, but this con-
clusion was based on examination of photolyzed solutions
containing small amounts of methanol. In fact, inspection of
photolyzed hexane or cyclohexane-d12 solutions of 5b in the
absence of added methanol reveals that the species is quite
stable, and exhibits a 1H NMR spectrum consistent with
isotoluene 24b. The compound reacts within ~5 s upon dis-
solution in neutral methanol to yield the previously reported
methoxysilane 23.
The three phenylsilyl-substituted isotoluenes (13b, 13d,
and 24b) all react with methanol under neutral conditions at
least two orders of magnitude faster than the trialkylsilyl de-
rivative 13a. This difference in reactivity, which has been
documented previously in other systems (25, 30), suggests
that desilylation proceeds by an SN2 mechanism under these
conditions, probably with simultaneous protonation at the
exocyclic double bond of the leaving group (eq. [13]).
On the other hand, little difference in reactivity as a func-
tion of substitution at silicon is observed in acidic methanol
solution. For example, 13a, 13b, 13d, and 24b were all
found (by laser flash photolysis) to have lifetimes in the
12–17 ms range in methanol containing 0.3 M HCl. This
suggests that under acidic conditions, the rate-determining
step for reaction is protonation of the exocyclic double bond,
© 2000 NRC Canada

1466
Can. J. Chem. Vol. 78, 2000
Table 1. Pseudo-first order rate constants for decay of
isotoluenes in methanol and methanol containing 0.3 M HCl at
23°C.
ization detector, a conventional splitless injector (heated),
and a DB-1 fused silica capillary column (15 m × 0.2-mm;
Chromatographic Specialties, Inc.). Low resolution mass
spectra were recorded using a Hewlett–Packard 5890 gas
chromatograph equipped with a HP-5971A mass selective
detector and a DB-5 fused silica capillary column (30 m ×
0.25 mm; Chromatographic Specialties, Inc.). High-
resolution electron impact mass spectra and exact masses
were determined using a VGH ZABE mass spectrometer.
Radial chromatographic separations were carried out us-
ing a Chromatotron® (Harrison Research, Inc.), with 2 or
4 mm silica gel 60 thick layer plates using hexane as the
eluant.
Hexanes and methanol (BDH Omnisolv), cyclohexane-d12
(Cambridge Isotope Laboratories), benzyl bromide
(Aldrich), benzyltrimethylsilane (3, Aldrich), dimethyl-
phenylsilane (19, Aldrich), and chlorodimethylphenylsilane
(Aldrich) were used as received from the suppliers.
Tetrahydrofuran (BDH) was distilled from sodium. 1-
Chloro-1-methylsilacyclobutane (31), 1-chloro-1-phenylsil-
acyclobutane (32), and benzyldimethylphenylsilane (4) (33)
were prepared as previously described. 1,1,2-Triphenylsil-
acyclobutane (5) (9), 1-methoxy-1-methylsilacyclobutane
(11a) (17), 1-phenylsilacyclobutane (10) (18), and dimethyl-
methoxyphenylsilane (20) (18) were available from previous
studies. Benzylmethoxymethylphenylsilane (7) was prepared
by reaction of benzylmethylphenylsilane with methanol in
the presence of PdCl2 (17).
kdecay (s–1)
13a
13b
13d
24b
MeOH
0.0011±0.0001 >0.2
>0.2
>0.2
0.3 M HCl–MeOH 85 ± 1
61 ± 2 77 ± 1 58 ± 4
the rate of which is not expected to depend on substitution at
silicon. The resulting β-silyl-substituted cyclohexadienyl
carbenium ion then undergoes rapid methanol-assisted
desilylation (30) to yield toluene and the corresponding
methoxysilane. The pseudo-first-order rate constants for de-
cay of the four isotoluenes in methanol and methanol con-
taining 0.3 M HCl are summarized in Table 1.
Summary and conclusions
The direct photolysis of benzylsilanes in solution results
mainly in the formation of the corresponding 6-
silylisotoluene derivatives, which are sufficiently stable in
hydrocarbon solvents that they build up in concentration and
start to compete with the precursor for light absorption after
conversions of only 5–6%. Secondary photolysis of the
isotoluene derivatives is responsible for the formation of
most of the products of benzylsilane photolysis that are de-
tected chromatographically. Isotoluene formation results for-
mally from a photochemically allowed [1,3]-sigmatropic
rearrangement, although a singlet radical pair recombination
mechanism cannot be ruled out, and is generally more effi-
cient in benzyldialkylphenylsilanes than in benzyltrialkyl-
silanes. Phenyl substitution at silicon also affects the
photochemistry of the corresponding 6-silylisotoluene deriv-
ative, promoting Si—C bond homolysis to yield benzyl and
silyl free radicals. The isotoluene derivatives undergo rapid
desilylation in neutral or acidic methanol solution, which al-
lows the formation of secondary photolysis products to be
suppressed. The rate of solvolysis under neutral conditions is
at least two orders of magnitude greater for the phenylsilyl
derivatives, suggesting that it proceeds by an SN2 mecha-
nism in neutral methanol. The rates are accelerated by up to
five orders of magnitude in methanol containing 0.3 M HCl
and show no dependence on silyl substituent, consistent with
protonation being the rate determining step for isotoluene
desilylation under acidic conditions. The photochemical
[1,3]-rearrangement also occurs in “masked” benzylsilanes
such as 1,1,2-triphenylsilacyclobutane and accounts for the
formation of silacyclobutane-ring-opening products in the
photolyses of 2-phenylsilacyclobutanes in methanol solution.
1-Benzyl-1-methylsilacyclobutane (1) (34) was synthe-
sized as follows. Magnesium filings (2.55 g, 0.105 g-atom)
were placed in a 500 mL 2-neck round bottom flask fitted
with an addition funnel, reflux condenser with nitrogen inlet,
and magnetic stirrer. The entire apparatus was flame-dried
under nitrogen and after cooling, anhydrous diethyl ether
(100 mL) and 1-chloro-1-methylsilacyclobutane (4.0 mL,
2.75 g, 0.0299 mol) were added. Benzyl bromide (3.3 mL,
2.29 g, 0.0136 mol) was dissolved in anhydrous ether
(50 mL) and added dropwise over a 30 min period with stir-
ring. The mixture was stirred for 48 h at room temperature,
during which time it turned a light gray color. Ice water
(30 mL) was added slowly, and the resulting mixture was
then filtered. The filtrate was placed in a 500 mL separatory
funnel, the aqueous layer was removed, and the ether solu-
tion was washed with cold water (3 × 50 mL). After drying
with anhydrous magnesium sulfate and filtering, the solvent
was removed on a rotary evaporator to afford a slightly yel-
low oil (2.01 g). Compound 1 (1.04 g, 0.006 mol, 29%) was
isolated as a colourless oil in >99% purity by radial chroma-
tography on silica gel using hexane as eluant, and identified
1
on the basis of the following spectroscopic data: H NMR,
δ = 0.289 (s, 3H), 1.03 (m, 2H), 1.09 (m, 2H), 2.08 (m, 2H),
2.36 (s, 2H), 7.13 (m, 3H), 7.26 (d, 1H), 7.28 (d, 1H); 13C
NMR, δ = –1.79, 13.8, 17.85, 22.68, 26.56, 31.61, 124.16,
128.02, 128.29, 128.60, 139.40; 29Si NMR, δ = 18.85; 3082(s),
3062(s), 3027(s), 2964(m), 2926(w), 2798(m), 1934(m),
1868(m), 1801(m), 1600(s), 1494(s), 1452(s), 1410(m),
1250 (s), 1206(s), 1152(s), 1121(m), 1055(s), 1031(s),
903(s), 873(s), 755(s), 698(m), 653(s); UV (λmax) = 220,
260 nm; MS (EI), m/z (I) = 176 (27), 148 (100), 133 (68),
119 (18), 107 (17), 105 (28), 91 (40), 85 (95), 65 (38), 43
(47); HRMS calcd. for C11H16Si: 176.1022; found: 176.1025.
Experimental
1H, 13C, and 29Si NMR spectra were recorded using
Bruker AC200, Bruker AC300, or DRX500 NMR spectrom-
eters. Infrared spectra were recorded on a BioRad FTS-40
FT/IR spectrometer and are reported in wavenumbers (cm–1).
Ultraviolet absorption spectra were recorded on Cary–Varian
Model 50 or Model 400 spectrometers. Gas chromatographic
separations employed
chromatograph equipped with a 3396A integrator, flame ion-
a
Hewlett–Packard 5890 gas
© 2000 NRC Canada

Leigh and Owens
1467
1-Benzyl-1-phenylsilacyclobutane (2) was synthesized
and purified by a similar procedure to the one given above
for 1. It was isolated as a colourless oil and identified on the
(o-tolyl CH3), and 3.41 (OCH3) and resonances assignable to
the n-propyl group (δ 0.90 (t, 2H), 0.95 (t, 3H), 1.38 (m,
2H)), and the following mass spectrum: m/z (I) = 208 (4),
193 (2), 165 (100), 151 (12), 135 (33), 119 (12), 105 (18),
91 (11), 59 (20). Both the photolysis and dark reaction mix-
tures also contained a second minor isomer in the same yield
relative to 15 (1:8) which was tentatively identified as
benzylmethylmethoxy(1-propyl)silane (16, 21%) on the ba-
sis of its mass spectrum: m/z (I) = 208 (17), 165 (15), 151
(10), 133 (12), 117 (90), 91 (30), 75 (100), 59 (30)). The lat-
ter clearly shows the fragmentation pattern expected for a
benzylsilane.
1
basis of the following spectroscopic data: H NMR, δ = 0.90
(s), 0.98 (t), 1.33 (m, 2H), 1.96 (m, 2H), 2.64 (t, 2H), 7.17
(m, 3H), 7.29 (m, 3H), 7.46 (m, 2H), 7.63 (m, 2H); 13C
NMR, δ = 12.88, 18.05, 22.68, 25.32, 31.59, 124.40, 127.88,
128.30, 128.36, 129.51, 133.67, 137.20,138.74; 29Si
NMR, δ = 12.00; 3067(w), 3025(m), 2969(m), 2924(w),
1954(m), 1881(m), 1816(m), 1600(s), 1493(s), 1428(s),
1395(m), 1207(s), 1184(s), 1152(m), 1117(m), 1056(s),
1030(s), 999(s), 924(s), 902(s), 858(m), 820(s), 801(s),
766(m), 735(s), 697(w); UV (λmax) = 262, 222; MS (EI), m/z
(I) = 238 (33), 210 (25), 161 (25), 147 (48), 132 (60), 105
(100), 93 (60), 91 (67), 65 (25), 53 (55); HRMS calcd. for
C16H18Si: 238.1179; found: 238.1177.
The products of steady state photolysis of 2–4 in hexane,
hexane containing 0.1 M methanol, or neat methanol were
identified by GC coinjection with authentic samples. In the
few cases where these were not available, tentative identifi-
cation was made on the basis of GC/MS. For example
photolysis of 2 in hexane yielded an isomer in ~10% yield,
which was barely resolved from 2 under our GC/MS condi-
tions. It exhibited the following mass spectral data: MS, m/z
(I) = 226 (7), 211 (100), 195 (7), 181 (10), 165 (15), 148
(67), 133 (15), 105 (22), 91 (10). An analogous product was
formed in ~35% yield from photolysis of 4 in hexane solu-
tion, and tentatively identified as dimethylphenyl(o-tolyl)sil-
ane (18b) m/z: (I) = 226 (7), 211 (100), 195 (7), 181 (10),
165 (15), 148 (67), 133 (15), 105 (22), 91 (10). Disiloxane
17 from photolysis of 2 in methanol solution was identified
Compound 6 was isolated from a semi-preparative scale
photolysis of a hexane solution of 1-benzyl-1-methyl-
silacyclobutane (1, 0.22 g, 1.25 mmol) in hexane (20 mL).
The solution was placed in a quartz photolysis tube, sealed
with a rubber septum, deoxygenated with dry argon, and
then irradiated in a Rayonet reactor fitted with seven RPR-
2537 lamps The photolysis was monitored periodically by
GC, and terminated after ca. 95% conversion of 1 (59 hs).
After evaporation of the solvent, the single product was iso-
lated by radial chromatography, further purified by silica gel
column chromatography (hexane as eluent), and identified as
1-methyl-1-propyl-3,4-benzosilacyclobutene (6; 0.023 g,
0.125 mmol, 10%) on the basis of the following spectro-
1
on the basis of its H NMR (δ 1.45 (m, 8H), 1.67 (m, 2H),
2.06 (m, 2H), 7.41 (m, 6H), 7.61 (m, 4H)) and mass spec-
trum: m/z (I) = 310 (3), 282 (25), 269 (100), 254 (50), 241
(78), 227 (27), 197 (36), 195 (28), 181 (39), 119 (27), 105
(59), 93 (16), 78 (19) (35).
1
scopic data: H NMR, δ = 0.43 (s, 3H), 0.93 (t, 2H), 0.97 (t,
3H), 1.48 (m, 2H), 2.11 (d, 1H), 2.14 (d, 1H), 7.13 (d, 1H),
7.16 (d, 1H), 7.24 (d, 1H), 7.30 (d, 1H); 13C NMR, δ =
–2.173, 17.39, 17.66, 18.35, 19.08, 126.12, 126.82, 130.30,
130.57, 145.71, 150.95; 29Si NMR, δ = 10.82; IR (neat),
3078(m), 2969(m), 2938(w), 2875(m), 2360(m), 2313(m),
1718(s), 1703(s), 1671(s), 1625(s), 1546(s), 1531(s),
1500(s), 1422(s), 1407(m), 1375(s), 1281(s), 1250(m),
1125(m), 1063(m), 1047(m), 812(s), 750(w); MS (EI), m/z
(I) = 176(8), 161(3), 134(14), 119(15), 105(4), 85(43),
56(76), 57(74), 43(91), 41(100); HRMS calcd. for C11H16Si:
176.1022; found: 176.0995.
Steady state photolysis experiments were carried out in a
Rayonet reactor fitted with 6–10 RPR-2537 lamps, on argon-
saturated solutions contained in quartz tubes. In some cases,
KrF excimer (248 nm) or nitrogen (337 nm) lasers, operated
at 0.5 Hz with the intensity reduced using neutral density fil-
ters, were employed as excitation sources. Product yields
were estimated in small scale runs, from the GC peak areas
relative to that of an internal standard (usually decane).
These were not corrected for differences in FID response
factors.
The isotoluene derivatives 13a, b, d, and 24b were de-
1
tected by UV absorption and H NMR spectroscopy on the
crude mixtures from 254-nm photolysis of 1–5 (0.01–0.02
M) to 5–10% conversion in hexane or cyclohexane-d12. The
four compounds exhibited identical UV spectra (λmax
=
1
335 nm; see Fig. 1 for 13a) and the following H NMR
spectra: 13a (from 1; silacyclobutyl Hs excluded), δ = 0.49
(s, 3H), 3.18 (m, 1H), 4.12 (s, 1H), 4.46 (s, 1H), 5.74 (m,
3H), 5.97 (d, 1H, J = 8.7 Hz); 13b (from 2; phenyl and
silacyclobutyl Hs excluded), δ = 3.34 (m, 1H), 4.42 (s, 1H),
4.58 (s, 1H), 5.60 (m, 1H), 5.72 (m, 2H), 5.94 (d, 1H, J =
6.3 Hz); 13d (from 4; phenyl Hs excluded), δ = 3.33 (m,
1H), 4.41 (s, 1H), 4.76 (s, 1H), 5.82 (m, 1H), 5.86 (m, 2H),
6.14 (d, 1H, J = 7.4 Hz); 24b (from 5; phenyl Hs excluded),
δ = 3.20 (m, 1H), 4.55 (m, 1H), 5.75 (m, 1H), 5.94 (m, 2H),
6.08 (d, 1H, J = 7.5 Hz). Photolysis of a 0.01 M solution of
3 under the same conditions resulted in the appearance of
weak edge absorptions in the 300–340 nm range, even after
photolysis for roughly 5 times longer than the other com-
pounds. The 1H NMR spectrum of the photolysate (after
evaporation of the solvent and redissolution in cyclohexane-
d12) showed no evidence of absorptions in the 3–6 ppm
range.
The products of photolysis of 1 in deoxygenated methanol
solution containing 0.1 M HCl were identified by GC/MS
and GC coinjection with authentic samples. The major prod-
uct of dark reaction of 6 with methanol was tentatively iden-
tified as methylmethoxy(o-tolyl)(1-propyl)silane (15; 76%)
Nanosecond laser flash photolysis experiments employed
the pulses (248 nm; ca. 25 ns; 70–140 mJ) from a Lambda
Physik Compex 120 excimer laser filled with F2–Kr
mixtures in neon, and a microcomputer-controlled detection
system (36). The laser beam was unfocussed, and its
intensity was reduced to 10–15 mJ at the cell using a series
1
by H NMR and GC/MS, on the crude mixture obtained
from stirring a sample of 6 (0.1 g) in methanol (5 mL) for
fourteen hours at 25°C, followed by evaporation of the sol-
1
vent and redissolution in CDCl3. The H NMR spectrum of
the compound exhibited singlets at δ 0.41 (SiCH3), 2.46
© 2000 NRC Canada

1468
Can. J. Chem. Vol. 78, 2000
14. C. Kerst, R. Boukherroub, and W.J. Leigh. J. Photochem.
Photobiol. A, 110, 243 (1997).
15. C.J. Bradaric and W.J. Leigh. Organometallics, 17, 645 (1998).
16. W.J. Leigh, C. Kerst, R. Boukherroub, T.L. Morkin, S.
Jenkins, K. Sung, and T.T. Tidwell. J. Am. Chem. Soc. 121,
4744 (1999).
of stainless steel wire meshes as neutral density filters. Solu-
tions were prepared at concentrations such that the
absorbance at the excitation wavelength was ca. 0.7 (~8 ×
10–4 M), and were flowed continuously through a 3 × 7 mm
Suprasil flow cell connected to a calibrated 100 mL reser-
voir. Nitrogen was bubbled continuously through the reser-
voir throughout the experiments.
17. W.J. Leigh, R. Boukherroub, and C. Kerst. J. Am. Chem. Soc.
120, 9504 (1998).
18. W.J. Leigh, R. Boukherroub, C.J. Bradaric, C. Cserti, and J.M.
Schmeisser. Can. J. Chem. 77, 1136 (1999).
19. M. Kira, H. Yoshida, and H. Sakurai. J. Am. Chem. Soc. 107,
7767 (1985).
20. P.B. Valkovich, T.I. Ito, and W.P. Weber. J. Org. Chem. 39,
3543 (1974).
21. W.J. Bailey and R.A. Baylouny. J. Org. Chem. 27, 3476 (1962).
22. D. Hasselmann and K. Loosen. Angew. Chem. Int. Ed. Engl.
17, 606 (1978).
23. K.R. Kopecky and M.-P. Lau. J. Org. Chem. 43, 525 (1978).
24. F.L. Cozens, A.L. Pincock, J.A. Pincock, and R. Smith. J. Org.
Chem. 63, 434 (1998).
Acknowledgements
We wish to thank D. Hughes for his assistance with high-
resolution NMR spectra, the McMaster Regional Centre for
Mass Spectrometry for high-resolution mass spectra and
exact mass determinations, and Dr. R. Boukherroub and
E.C. Lathioor for technical assistance. The financial support
of the Natural Sciences and Engineering Research Council
(NSERC) of Canada is also gratefully acknowledged.
References
25. E.W. Colvin. Silicon in organic synthesis. London,
Butterworths. 1981. pp. 97–124.
26. T. Morita, Y. Okamoto, and H. Sakurai. Tetrehedron Lett. 835
(1980).
27. C. Eaborn, D.R.M. Walton, and M. Chan. J. Organomet.
Chem. 9, 251 (1967).
28. L.H. Sommer. Stereochemistry, mechanism, and silicon.
McGraw–Hill, New York. 1965. p. 132.
29. H. Hiratsuka, Y. Kadokura, H. Chida, M. Tanaka, S.
Kobayashi, T. Okutsu, M. Oba, and K. Nishiyama. J. Chem.
Soc. Faraday Trans. 92, 3035 (1996).
30. C.S.Q. Lew and R.A. McClelland. J. Am. Chem. Soc. 115,
11516 (1993).
31. R. Damrauer, R.A. Davis, D.R. Burley, M. Karni, and J.L.
Goodman. J. Organomet. Chem. 43, 121 (1972).
32. G. Bertrand, J. Dubac, P. Mazerolles, and J. Ancelle. Nouv. J.
Chim. 6, 381 (1982).
33. A. Hosomi and H. Sakurai. Bull. Chem. Soc. Jpn. 45, 248
(1972).
1. L.E. Gusel’nikov and M.C. Flowers. J. Chem. Soc. Chem.
Commun. 864 (1967).
2. P. Boudjouk and L.H. Sommer. J. Chem. Soc. Chem.
Commun. 54 (1973).
3. N. Auner and J. Grobe. J. Organomet. Chem. 197, 147 (1980).
4. P. Jutzi and P. Langer. J. Organomet. Chem. 202, 401 (1980).
5. L.E. Gusel’nikov, N.S. Nametkin, and V.M. Vdovin. Acc.
Chem. Res. 8, 18 (1975).
6. G. Raabe and J. Michl. Chem. Rev. 85, 419 (1985).
7. A.G. Brook. In The chemistry of organic silicon compounds.
Edited by S. Patai and Z. Rappoport. John Wiley and Sons,
New York. 1989. pp. 965–1005.
8. A.G. Brook and M.A. Brook. Adv. Organomet. Chem. 39, 71
(1996).
9. W.J. Leigh, C.J. Bradaric, and G.W. Sluggett. J. Am. Chem.
Soc. 115, 5332 (1993).
10. C.J. Bradaric and W.J. Leigh. J. Am. Chem. Soc. 118, 8971
(1996).
11. W.J. Leigh, C.J. Bradaric, C. Kerst, and J.H. Banisch.
Organometallics, 15, 2246 (1996).
34. V.M. Vdovin, N.S. Nametkin, and P.L. Grinberg. Dokl. Akad.
Nauk. SSSR, 150, 799 (1963).
35. A.I. Mikaya, N.V. Ushakov, V.G. Zaikin, and V.M. Vdovin.
Zh. Obshch. Khim. 54, 1008 (1984).
36. W.J. Leigh, M.S. Workentin, and D. Andrew. J. Photochem.
Photobiol. A, 57, 97 (1991).
12. R.K. Vatsa, A. Kumar, P.D. Naik, H.P. Upadhyaya, U.B.
Pavanaja, R.D. Saini, J.P. Mittal, and J. Pola. Chem. Phys.
Lett. 255, 129 (1996).
13. C.J. Bradaric and W.J. Leigh. Can. J. Chem. 75, 1393 (1997).
© 2000 NRC Canada