Photochemistry of cyclic trisilanes: "Spring-loaded" precursors to methylphenylsilylene

Article abstract of DOI:10.1002/chem.200901249

Two new cyclic trisilanes containing a SiMePh moiety in the central position have been synthesized, and their photochemistries studied by steady-state and laser flash photolysis methods. The yield of the transient silylene SiMePh increases at the expense

Full text of DOI:10.1002/chem.200901249

FULL PAPER
DOI: 10.1002/chem.200901249
Photochemistry of Cyclic Trisilanes:
“Spring-Loaded” Precursors to Methylphenylsilylene
Andrey G. Moiseev, Eugꢀnie Coulais, and William J. Leigh*[a]
Dedicated to Professor Yitzhak Apeloig on the occasion of his 65th birthday
Abstract: Two new cyclic trisilanes con-
silene derivative (formed by photo-
chemical 1,3-silyl migration into the
phenyl ring) as the ring size is de-
creased from 7 to 6, facilitating the
direct detection of the silylene by laser
flash photolysis. The UV/Vis spectrum
of SiMePh in hexanes solution (lmax
270, 505 nm) and absolute rate con-
stants for its reaction with methanol,
taining a SiMePh moiety in the central
position have been synthesized, and
their photochemistries studied by
steady-state and laser flash photolysis
methods. The yield of the transient sily-
lene SiMePh increases at the expense
of that of the corresponding cyclic
=
triethylsilane,
a
terminal alkene,
Keywords: kinetics
·
laser
alkyne, and two aliphatic dienes at
258C are reported.
chemistry · photolysis · silylenes
Introduction
terized by UV/Vis and IR spectroscopy[18–21] and a number
of aspects of their reactivity to be studied under these condi-
tions.[7,19,22–30] The high selectivity toward silylene extrusion
is maintained in solution at room temperature in the case of
dodecamethylcyclohexasilane (1), which has allowed the
simplest dialkylsilylene derivative, SiMe2, to be studied di-
rectly in solution by time-resolved spectro-
Silylenes are important reactive intermediates in the ther-
mal and photochemical reactions of many organosilicon
compounds,[1–3] including those involved in the chemical
vapor deposition of thin silicon films.[4] There has thus been
great theoretical and experimental interest in the study of
their chemical properties and reactivity;[1,2,5–9] great advan-
ces have been made in recent years in the synthesis of deriv-
atives in which steric and/or electronic stabilization of the
divalent silicon center is sufficient to render them isola-
ble,[10–12] and in the development of silylene chemistry for
use in synthetic applications.[13,14]
Many of the early efforts to characterize silylene reactivi-
ty in condensed phases focused on simple (transient) sily-
lenes such as dimethyl-, methylphenyl-, and diphenylsilylene
(SiMe2, SiMePh, and SiPh2, respectively), which are pro-
duced in high yields upon photolysis of oligosilanes such as
1, 2a, and 2b, respectively.[15–17] The corresponding silylenes
are evidently the exclusive transient photoproducts in low
temperature matrixes, which has allowed them to be charac-
scopic methods.[31–34] With the phenylated
trisilanes 2a and 2b, on the other hand, a
second photochemical process—1,3-silyl mi-
gration into the aryl ring, forming a transi-
ent silene derivative (3a,b)—competes
quite significantly with extrusion of the central Si(R)Ph
moiety (see Scheme 1).[15,35,36] This competing reaction,
which proceeds in only modestly lower yields than silylene
extrusion from 2a[35] and 2b,[36] confounded early attempts
to detect and study the chemistry of phenylated silylene de-
rivatives such as SiMePh and SiPh2 in solution by laser flash
photolysis,[37,38] because of strong overlap of the intense
(p,p*) absorption band of the corresponding silene with the
relatively weak, long wavelength (n,p) absorption band that
characterizes the UV/Vis spectra of the silylenes.[39]
As Sakurai and co-workers first showed for alkyltrisilane
derivatives,[40] silylene photoextrusion also proceeds readily
when the trisilane moiety of 2b is incorporated in a cyclic
structure, as in the cyclotrisilane derivatives 4 and 5, respec-
tively.[33,41] These structural modifications result in reduc-
tions of the yields of adducts arising from the corresponding
silene derivatives from its value of about 40% in 2b[36] to
[a] Dr. A. G. Moiseev, E. Coulais, Prof. W. J. Leigh
Department of Chemistry, McMaster University
1280 Main Street West, Hamilton, ON L8S 4M1 (Canada)
Fax : (+1)905-522-2509
E-mail: leigh@mcmaster.ca
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901249.
Chem. Eur. J. 2009, 15, 8485 – 8491
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
8485

Scheme 3.
Scheme 1.
0.2m) and 1,4-dioxane (ca. 0.002m) as internal standard,
with periodic monitoring of the photolysate by 1H NMR
spectroscopy and GC/MS, afforded methoxymethylphenylsi-
lane (8; 82ꢀ3%) and 1,1,2,2-tetramethyldisilacyclohexane
(9; 94ꢀ4%), as shown in Scheme 4. Product yields were de-
about 15% in 5,[41] and to only 3–5% in 4,[33] with corre-
sponding increases in the yields of products derived from si-
lylene extrusion. Laser flash photolysis experiments with 4
afford well-defined UV/Vis spectra of SiPh2 (lmax =290,
515 nm) and its dimer (Si2Ph4; lmax =460 nm) in hexane solu-
tion.[33] Because the interfering absorptions due to the minor
silene co-product are reduced to such minor levels using this
compound as precursor, it is possible to supplement absolute
rate constants for reaction of the silylene with various sub-
strates with detailed spectroscopic and kinetic information
on additional reactive intermediates or transient products
that are involved in some of its reactions.[33,34] Clearly, the
extension of this methodology to allow the photochemical
generation and direct detection
Scheme 4.
of other reactive silylene deriva-
tives in solution could be quite
termined from the relative slopes of concentration versus
useful.
time plots over the range of 0–25% conversion in 7 (see
Supporting Information). The H NMR spectra also showed
1
In this paper, we report the
synthesis of two new cyclic trisi-
lane derivatives (6 and 7; see
Schemes 2 and 3), and explore
five weak singlets consistent with methoxyl protons, a host
of very weak, complex multiplets scattered throughout the d
3.6–6.6 range, and a series of weak singlets clustered near
d = 0, which we tentatively ascribe to adducts of silene 10
their suitability as photochemical precursors to methylphe-
nylsilylene (SiMePh) in laser flash photolysis studies. The
behavior of the two compounds is compared to those of
1,1,1,2,3,3,3-heptamethyl-2-phenyltrisilane (2a)[39] and to the
cyclic SiPh2 precursors 4[33] and 5.[41] Finally, absolute rate
constants for reaction of SiMePh with a variety of silylene
substrates have been determined for the first time, and are
compared to those reported recently for SiMe2 and SiPh2
under similar conditions.[33,34]
with MeOH; indeed, GC/MS analysis of the reaction mix-
ture verified the presence of at least three minor products
whose retention times and mass spectra are consistent with
structures of this type.[36,42] A second experiment, carried
out with 0.058m 7 and 0.09m MeOH to allow more precise
integrals to be measured for the methoxyl peaks due to the
minor products, indicated them to be formed in a total yield
of about 11ꢀ2%. The yields of 8 and 9 in this experiment
were 62ꢀ4 and 83ꢀ3%, respectively; the experiment also
showed evidence for the formation of a second hydridome-
thoxysilane whose yield relative to the other products in-
creased with photolysis time, characteristic of a secondary
reaction product. The compound was not conclusively iden-
tified.
Results and Discussion
The cyclic trisilanes 6 and 7 were prepared by the routes
shown in Schemes 2 and 3, respectively, in analogous fashion
to our previous syntheses of the corresponding 2,2-diphenyl-
substituted derivatives.[33,34] The two compounds were ob-
tained as colorless oils in yields of 19–31% after purification
by repeated column chromatography.
Steady state photolysis of a deoxygenated solution of 7
(0.051m) in [D12]cyclohexane containing methanol (MeOH;
Similar experiments with trisilacyclohexane 6 afforded 8
(86ꢀ7%) and 1,1,2,2-tetramethyldisilacyclopentane (12;
104ꢀ8%) as the only identifiable products (Scheme 5). As
in the experiments with 7, the spectra also showed several
weak singlets in the methoxyl proton region and a series of
Scheme 2.
8486
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8485 – 8491

Spring-Loaded Precursors to Methylphenylsilylene
FULL PAPER
Scheme 5.
resonances in the d 3.6–6.6 range, which are tentatively as-
cribed to MeOH adducts of silene 11. No attempt was made
to quantify the yields of these products as they were quite
minor and difficult to detect at low conversions; however,
their yields were clearly significantly lower than those of the
corresponding minor products from photolysis of 7.
Laser flash photolysis of a flowed, deoxygenated solution
of 7 (ca. 10ꢁ4 m) in anhydrous hexanes, with the pulses from
a KrF excimer laser (248 nm, ~25 ns, ~100 mJ), afforded
strong transient absorptions throughout the 260–600 nm
monitoring wavelength range. Transient UV/Vis absorption
spectra sampled over various time windows during the first
10 ms after excitation are shown in Figure 1a, along with
three of the transient growth/decay profiles from which the
spectra were extracted; the latter illustrate the transient sig-
nals obtained at monitoring wavelengths of 420, 450, and
520 nm. As expected, the spectra show evidence of at least
three distinct transient species: a short-lived species exhibit-
ing absorption bands in the 260–290 nm and 470–530 nm re-
gions of the spectrum and decaying with a lifetime of t
~1 ms, and two other longer-lived species with lifetimes on
the order of 20–40 ms and absorption maxima centered be-
tween 420 and 450 nm. One of the latter two species is
formed with the laser pulse, while the second grows in
during the first microsecond after excitation, on a similar
timescale as the decay of the short-lived species; this causes
the apparent maximum of the long wavelength absorption
band to shift from 440 to about 430 nm as the spectrum
evolves during the first microsecond after the excitation
pulse. It is thus evident that the absorption maximum of the
promptly-formed species lies somewhat to the red of the
one that grows in subsequent to excitation. We assign these
two long-lived transients to silene 10 and 1,2-dimethyl-1,2-
diphenyldisilene (Si2Me2Ph2, as a mixture of E and Z iso-
mers), respectively, and the short-lived species to SiMePh.
The two disilene isomers have been detected previously by
laser flash photolysis methods, and both exhibit lmax
~415 nm in methylcyclohexane solution at 208C.[43] The
spectrum of 10 is expected to be similar to that of 3a, the
silene derived from acyclic trisilane 2a, which is centered at
lmax =440 nm in hydrocarbon solvents under similar condi-
tions.[39]
*
Figure 1. a) Transient absorption spectra recorded 0.26–0.29 ms ( ), 1.22–
~
1.25 ms (&), and 8.58–8.61 ms ( ) after the laser pulse, by laser flash pho-
tolysis of a deoxygenated hexanes solution of trisilacycloheptane 7 (ca.
10ꢁ4 m); the inset shows transient growth/decay profiles recorded at 420,
450, and 520 nm, while the dashed line shows the difference spectrum ob-
tained from photolysis of the compound in 3-MP at 78 K. b) Spectra re-
corded 64–96 ns ( ), 0.58–0.74 ms (&), and 8.58–8.61 ms ( ) after the
pulse in hexanes containing 3 mm Et3SiH; the inset shows transient
growth/decay profiles recorded at 440 nm and 520 nm, while the dashed
line shows the result of subtracting the 0.58–0.74 ms spectrum from the
64–96 ns spectrum.
~
*
with those wavelength ranges in the room temperature tran-
sient spectrum where the short-lived (t ~1 ms) absorptions
are evident. The silylene absorptions observed in (room
temperature) laser photolysis experiments with 7 are thus
significantly stronger, relative to those of the corresponding
silene, than those obtained in similar experiments with
2a,[39] consistent with the indication from the steady state
photolysis experiments that the silylene is formed in much
higher chemical yield from the cyclic trisilane (see above)
compared to the acyclic derivative.[35,44] As a further result
of the higher yield of SiMePh from 7 than from 2a, clear
evidence for the formation of the corresponding disilene
(Si2Me2Ph2) via silylene dimerization is obtained in laser
photolysis experiments with 7, but could not be using the
acyclic trisilane as the precursor.[39]
Additional support for these assignments is provided by
the transient behavior observed upon addition of 3 mm trie-
thylsilane (Et3SiH) to the solution, which shortened the
transient lifetimes at 270 and 520 nm to about 200 ns and
quenched the growth component of the signal at 420 nm,
but had no effect on the lifetime of the 440 nm absorption.
Figure 1b shows the relevant series of transient absorption
Figure 1a also shows the UV/Vis spectrum of a photo-
lyzed solution of 7 in 3-methylpentane (3-MP) at 78 K, dis-
played as the difference between spectra recorded before
and after brief photolysis of the glassy matrix with 254 nm
light. The spectrum, which exhibits absorption maxima at
270 and 490 nm, is in excellent agreement with those report-
ed previously for SiMePh, generated by photolysis of 2a in
hydrocarbon matrixes at 77–78 K[19,39] and coincides nicely
Chem. Eur. J. 2009, 15, 8485 – 8491
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8487

W. L. Leigh et al.
spectra, focusing on the 320–600 nm region of the spectrum;
the region below 320 nm is strongly distorted by sample
fluorescence in the first 100 ns after excitation and is thus
not shown. Because disilene formation is suppressed in the
presence of the hydridosilane, the position of the prominent
absorption band in the transient spectrum (which is now
comprised solely of that due to silene 10) remains un-
changed throughout its decay. The lifetime of the latter spe-
cies is about 25 ms under these conditions, similar to that ob-
served in the absence of the hydridosilane. The difference
spectrum that results from subtracting the second (0.58–
0.74 ms) spectrum from the first (64–96 ns) exhibits lmax
~500 nm, in good agreement with the spectra obtained from
photolysis of 2a and 7 in 3-MP at 77–78 K.
ference: the contribution to the spectra and growth/decay
profiles from the corresponding silene (11) is reduced by a
factor of roughly 2.5 compared to that observed with 7
under the same conditions, consistent with a correspondingly
lower chemical yield of the silene relative to the silylene
from the six-membered cyclic trisilane, in broad agreement
with the results of the steady state photolysis experiments.
The greater contribution from the silylene to the long wave-
length transient absorption band compared to that observed
with 7 also results in a significantly more prominent contri-
bution to the 1.2 ms spectra (in the absence of Et3SiH; Fig-
ure 2a) from the silylene dimer, Si2Me2Ph2. The difference
spectrum obtained by subtracting the 0.58–0.74 ms spectrum
observed in the presence of 3 mm Et3SiH from that recorded
immediately after the laser pulse is somewhat better defined
than that obtained with 7, and affords a long wavelength ab-
sorption maximum of lmax ~505 nm for SiMePh in hexanes
at 258C (see Figure 2b). The slight red-shift of the absorp-
tion maximum relative to that observed in hydrocarbon ma-
trixes at 77–78 K may be due to differences in the preferred
geometry of the molecule under the two sets of conditions.
There appears to be a slight blue-shift of the absorption
maximum of silene 11 (lmax ~430 nm) compared to that of
the larger-ring analogue 10, the significance of which is not
clear.
kdecay ¼ k0 þ kQ½Qꢂ
ð1Þ
Laser photolysis of a ca. 10ꢁ4 m solution of 6 in anhydrous
hexanes afforded the series of transient spectra shown in
Figure 2a, while those obtained after addition of 3 mm
Et3SiH to the solution are shown in Figure 2b. As can be
seen, the transient spectra and temporal behavior are closely
analogous to those observed with 7 with one important dif-
As might be expected, the increase in the yield of SiMePh
relative to silene throughout the series 2a < 7 < 6 mirrors
that previously observed for the analogous SiPh2 precursors,
2b <5 < 4.[33,41] However, the absorption due to the corre-
sponding silene, which obscures that due to the silylene at
longer wavelengths, is in each case proportionately greater
with the SiMePh precursors. The errors in our chemical
yield measurements are sufficiently large to make it difficult
to tell whether this is due to the yields of silene being sys-
tematically slightly higher relative to those of the silylene in
the SiMePh precursors, or to a difference in the relative ex-
tinction coefficients of the absorptions due to the silene and
the silylene; the possibility that SiMePh is a somewhat
weaker absorber than SiPh2 is certainly reasonable. In any
event, the signals due to SiMePh from both 6 and 7 are suf-
ficiently strong and distinctive compared to the correspond-
ing silenes to allow absolute rate constants for reaction of
the silylene with various substrates to be determined in
straight-forward fashion, by analysis of the pseudo-first
order decay rate constants of the 520 nm transient absorp-
tion as a function of substrate concentration according to
Equation (1). Decays recorded at 520 nm in the absence of
substrate fit reasonably well to second order kinetics, afford-
ing second order decay coefficients of 2k/e520 nm = (6.0ꢀ
0.7)ꢁ107 cmsꢁ1, the same value being obtained from both 6
and 7.
*
Figure 2. a) Transient absorption spectra recorded 0.26–0.29 ms ( ), 1.22–
~
1.25 ms (&), and 8.58–8.61 ms ( ) after the laser pulse, by laser flash pho-
Indeed, addition of sub-millimolar concentrations of
MeOH, 4,4-dimethyl-1-pentene (DMP), isoprene, 2,3-di-
methyl-1,3-butadiene (DMB), or 3,3-dimethyl-1-butyne
(TBE) to hexane solutions of 6 or 7 caused the lifetime of
the short-lived component of the transient signals at 520 nm
to shorten and quenched the formation of the disilene ab-
tolysis of a deoxygenated hexanes solution of trisilacyclohexane 6 (ca.
10ꢁ4 m); the inset shows transient growth/decay profiles recorded at 420,
*
450, and 520 nm. b) Spectra recorded 64–96 ns ( ), 0.58–0.74 ms (&), and
~
8.58–8.61 ms ( ) after the pulse in hexanes containing 3 mm Et3SiH; the
inset shows transient growth/decay profiles recorded at 440 and 520 nm,
while the dashed line shows the result of subtracting the 0.58–0.74 ms
spectrum from the 64–96 ns spectrum.
8488
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8485 – 8491

Spring-Loaded Precursors to Methylphenylsilylene
FULL PAPER
sorption, in a similar fashion as that observed upon addition
of Et3SiH. Plots of the pseudo-first order rate constants for
decay of the silylene versus substrate concentration ([Q])
were linear in each case (see Figure 3), and analysis of the
only very minor effects on the absolute reactivities of simple
silylenes in solution.[34]
Conclusions
Photolysis of the well-known silylene precursors
1,1,1,2,3,3,3-heptamethyl-2-phenyl- and 1,1,1,3,3,3-hexam-
ethyl-2,2-diphenyltrisilane affords the corresponding sily-
lenes from extrusion of the central Si(R)Ph unit, along with
slightly lower yields of the silene resulting from [1,3]-trime-
thylsilyl migration into an ortho-position of the phenyl
ring(s).[15,35,36] While the competition from the latter reaction
did little to hamper any of the early, product-oriented stud-
ies of various aspects of silylene reactivity using these pre-
cursors, it thwarted early attempts to detect SiMePh and
SiPh2 directly in solution by time-resolved UV/Vis spectros-
copy and study their reactivity in a quantitative way. The
problem can be solved by tying the ends of the three-silicon
atom chains together to incorporate the trisilane moiety in a
six- or seven-membered ring; this structural constraint has
the effect of increasing the yield of photoextrusion of the
central silylene unit at the expense of photoisomerization,
allowing the silylenes to be detected and studied with much-
reduced interference from the nearby, instrinsically more in-
tense absorptions due to the silene co-product. The effect
may be due to a combination of strain-related factors, such
as the compression of the Si-Si-Si bond angle that presuma-
bly results from tying the ends together with a 3- or 4-
carbon chain (“spring-loading” the system toward silylene
extrusion), and the fact that the silenes produced by the
photoisomerization process are necessarily contained in
(presumably somewhat strained) medium-size rings. The se-
lectivity toward silylene extrusion over photoisomerization
is markedly higher for the trisilacyclohexanes than for the
corresponding seven-membered ring compounds. For equal
ring sizes, it also appears to be somewhat higher for the 2,2-
diaryl systems than for the 2-methyl-2-aryl systems, based
on the relative complexities of the transient UV/Vis spectra
obtained in laser photolysis experiments with the series of
compounds.
Figure 3. Plots of the pseudo-first order rate constants for decay of
SiMePh (kdecay) versus [Q], for quenching by isoprene (ꢃ), 2,3-dimethyl-
!
1,3-butadiene (DMB, &), 3,3-dimethyl-1-butyne (TBE, ), 4,4-dimethyl-
~
^
*
1-pentene (DMP, ), MeOH ( ), and Et3SiH ( ) in deoxygenated hex-
anes at 258C.
data according to Equation (1) afforded the second-order
reaction rate constants (kQ) listed in Table 1. The values ob-
tained for quenching by Et3SiH and MeOH are in excellent
agreement with those reported by us earlier using 2a as pre-
cursor.[39] Furthermore, the ratios of the absolute rate con-
stants for reaction of SiMePh with MeOH and DMB rela-
tive to that with Et3SiH are in good agreement with the
values for EtOH and DMB that were reported a number of
years ago by Hawari et al., based on the results of competi-
tive trapping experiments.[45]
Table 1 also lists the corresponding rate constants for re-
action of SiMe2 and SiPh2 with the same substrates under
similar conditions, from our previous time-resolved studies
of these silylene derivatives.[33,34] The values of the rate con-
stants determined for reaction of SiMePh with the six sub-
strates are in all cases quite similar to those reported previ-
ously for SiMe2 and SiPh2, as should be expected consider-
ing the relatively small differences in the reactivities of the
latter two species toward the specific substrates that have
been studied so far. This both verifies and extends our earli-
er conclusion that phenyl-for-methyl substitution results in
The most important advantage to be gained from reduc-
ing the contribution from the undesired silene co-product to
the transient spectra is that it facilitates the detection and
characterization of reactive intermediates or products that
may be involved in various silylene reactions, as we have
shown previously for the reactions of SiPh2 with alcohols, al-
koxysilanes, CCl4, and oxygen.[34] Efforts to extend this ap-
proach to the study of other silylene reactions and to other
reactive silylenes of fundamental and potentially practical
importance are in progress.
Table 1. UV/Vis absorption maxima and absolute rate constants (in units
of 109 mꢁ1 sꢁ1) for quenching of transient silylenes by various substrates in
hexanes solution at 258C.
[b,c]
[b,c]
Substrate
SiMePh[a]
SiMe2
SiPh2
lmax [nm]
MeOH
Et3SiH
DMP[d]
isoprene
DMB[e]
TBE[f]
270, 505
16.3ꢀ0.8
2.96ꢀ0.05
7.0ꢀ0.3
465
290, 515
13ꢀ2
18ꢀ2
3.6ꢀ0.3
3.3ꢀ0.2
7.9ꢀ1.6
13.6ꢀ1.4
14.5ꢀ1.3
8.3ꢀ0.7
11.7ꢀ0.5
19.8ꢀ0.7
15.9ꢀ0.7
17.8ꢀ0.6
19.5ꢀ0.6
12.9ꢀ0.8
9.5ꢀ0.2
Experimental Section
[a] This work. [b] Ref. [33]. [c] Ref. [34]. [d] 4,4-Dimethyl-1-pentene.
[e] 2,3-Dimethyl-1,3-butadiene. [f] 3,3-Dimethyl-1-butyne.
1H and 13C NMR spectra were recorded on Bruker AV200 or AV600
spectrometers in deuterated chloroform or [D12]cyclohexane and were
Chem. Eur. J. 2009, 15, 8485 – 8491
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8489

W. L. Leigh et al.
referenced to the residual solvent proton and 13C signals, respectively,
while 29Si spectra were recorded on the AV600 using the HMBC pulse se-
quence and were referenced to an external solution of tetramethylsilane.
GC/MS analyses were carried out on a Varian Saturn 2200 GC/MS/MS
128.4, ꢁ3.5, ꢁ9.1 ppm. The compound was dissolved in anhydrous THF
(12 mL) and used in entirety in the next step of the sequence.
To an oven-dried 100 mL round-bottom flask fitted with a reflux con-
denser, magnetic stirrer, and nitrogen inlet were placed magnesium turn-
ings (0.54 g, 0.02 mol) and anhydrous THF (12 mL). A solution of 1,4-di-
chlorobutane (1.2 mL, 0.01 mol) in anhydrous THF (15 mL) was added
dropwise via syringe over 2 h, the resulting mixture was stirred under
reflux for 3 h, and then cooled to room temperature. The THF solution
of 1,3-dichloro-1,1,2,3,3-pentamethyl-2-phenyltrisilane was then added
dropwise via syringe to the stirred solution over 1 h, and the resulting
mixture was stirred at room temperature for an additional 20 h. The reac-
tion mixture was quenched with 5% aqueous ammonium chloride
(100 mL), and then extracted with diethyl ether (3ꢁ100 mL). The com-
bined organic phases were washed with water (3ꢁ100 mL), dried with
magnesium sulphate, filtered, and the solvent was removed under re-
duced pressure to afford a colorless oil. The material was purified twice
by column chromatography on silica gel using hexanes as eluant, to yield
7 as a colorless oil (0.61 g, 19%). 1H NMR (600 MHz, CDCl3): d=7.42
(dd, J=1.8, 7.5 Hz, 2H), 7.30 (m, 3H), 1.70 (m, 4H), 0.81 (m, 4H), 0.43
(s, 3H), 0.12 (s, 6H), 0.045 (s, 6H); 13C NMR (150 MHz, CDCl3): d=
137.9, 134.9, 127.8, 25.3, 14.3, ꢁ2.1, ꢁ2.5, ꢁ7.9 ppm; 29Si NMR (119 MHz,
CDCl3): d=ꢁ12.45 (SiMe2), ꢁ47.1 ppm (SiMePh); IR (film): n˜ = 3067
(m), 2950 (s), 2904 (s), 2850 (m), 1485 (m), 1427 (m), 1245 (s), 1097 (m),
871 (m), 833 (s), 778 (s), 732 (s), 699 cmꢁ1 (s); MS (70 eV). m/z (%): 292
[M]+ (42), 277 (12), 233 (11), 215 (41), 178 (100), 163 (75), 135 (47), 121
(8), 105 (18), 97 (10), 73 (16), 59 (17), 43 (30); HRMS (70 eV): m/z:
calcd for C15H28Si3: 292.1499; found: 292.1491.
system equipped with
a VF-5 ms capillary column (30 mꢁ0.25 mm;
0.25 mm; Varian, Inc.). High-resolution electron impact mass spectra and
exact masses were determined on a Micromass TofSpec 2E mass spec-
trometer using electron impact ionization (70 eV). Infrared spectra were
recorded as thin-films on potassium bromide plates using a Bio-Rad
FTS-40 FTIR spectrometer. Low temperature UV/Vis spectrophotometry
employed 2ꢁ1ꢁ1 cm cuvettes constructed from Suprasil quartz tubing
(Vitro Dynamics, Inc.) and an Oxford Optistat liquid nitrogen cryostat
equipped with an Oxford ITC601 temperature controller.
Dichloromethylphenylsilane,
dimethylchlorosilane,
allylACTHNGUTERNN(UG chloro)-
dimethylsilane, acetone, and deuterated solvents were used as received
from the suppliers. Methanol was distilled. Hexanes, diethyl ether, and
tetrahydrofuran were dried by passage through activated alumina under
nitrogen using a Solv-Tek solvent purification system (Solv-Tek, Inc.).
1,3-Bis(chlorodimethylsilyl)propane was synthesized by the method of
Kumada and co-workers and exhibited 1H and 13C NMR spectra similar
to those reported.[46] 1,1,2,3,3-Pentamethyl-2-phenyltrisilane was prepared
by stirring a xylenes solution of dichloromethylphenylsilane and two
equivalents each of chlorodimethylsilane and sodium-potassium alloy
under nitrogen for 20 h at room temperature. The reaction mixture was
quenched with 1:1 EtOH/acetic acid, followed by an aqueous workup
and column chromatography to yield the material as a colorless oil in
21% yield; it was identified by comparison of its 1H and 13C NMR and
mass spectra to the published data.[47]
Steady state photolysis: The experiments were carried out in quartz
NMR tubes using a Rayonet photochemical reactor (Southern New Eng-
land Ultraviolet Co.) equipped with a merry-go-round apparatus and two
1,1,2,3,3-Pentamethyl-2-phenyl-1,2,3-trisilacyclohexane (6): An oven-
dried 15 mL two-necked round-bottom flask was fitted with a condenser,
argon inlet, and magnetic stir bar, and then purged with argon. THF
(4.0 mL) and lithium powder (30% dispersion in mineral oil; 0.20 g,
8.8 mmol Li) were added, and the flask was cooled in an ice bath. A solu-
tion of 1,3-bis-(chlorodimethylsilyl)propane (0.51 g, 2.2 mmol) and di-
chloromethylphenylsilane (0.42 g, 2.2 mmol) in THF (2 mL) was then
added dropwise via syringe to the stirred mixture over 30 min, after
which the ice bath was removed and the mixture was stirred at room
temperature for 3 h. The course of the reaction was monitored frequently
by TLC, to avoid dephenylation of the desired product.[33] The reaction
mixture was quenched with water (10 mL) and extracted with diethyl
ether (3ꢁ30 mL). The combined organic fractions were dried over anhy-
drous sodium sulfate, filtered, and the solvent was removed on the rotary
evaporator to yield a colorless oil (0.9 g). Column chromatography on
silica gel with hexanes as eluant afforded a colorless oil (0.2 g, 33%),
which was identified as compound 6 on the basis of the following data:
1H NMR (600 MHz, CDCl3): d=7.45 (dd, J=1.8, 7.2 Hz, 2H), 7.31 (m,
3H), 1.81 (m, 2H), 0.84 (m, 2H), 0.75 (m, 2H), 0.43 (s, 3H), 0.13 (s, 6H),
0.09 ppm (s, 6H); 13C NMR (150 MHz, CDCl3): d=137.4, 135.0, 128.0,
127.9, 19.1, ꢁ2.7, ꢁ2.9, ꢁ9.0 ppm; 29Si NMR (119 MHz, CDCl3): d=
ꢁ16.4 (SiMe2), ꢁ50.3 ppm (SiMePh); IR (film): n˜ = 3067 (m), 2950 (s),
2893 (s), 1485 (m), 1462 (m), 1428 (m), 1411 (m), 1245 (s), 1112 (m),
1099 (m), 1028 (m), 938 (m), 902 (m), 831 (s), 780 (s), 759 (s), 732 (m),
698 (s), 591 cmꢁ1 (m); MS (70 eV): m/z (%): 278 (100) [M]+, 263 (5), 233
(5), 191 (10), 178 (31), 163 (40), 135 (51), 128 (17), 105 (15), 73 (17), 43
(22); HRMS (70 eV): m/z: calcd for C14H26Si3: 278.1342; found: 278.1348.
RPR-2537 lamps (254 nm). Solutions of
6 or 7 (0.051–0.061m) in
[D12]cyclohexane containing MeOH (0.09–0.2m) and 1,4-dioxane (3–
5 mm) were sealed with rubber septa, deoxygenated for 20 min with a
slow stream of dry argon, and then irradiated for 1–2 min intervals, fol-
lowing the course of the photolyses by 600 MHz 1H NMR spectroscopy.
The mixtures were also analyzed before and after ca. 30% conversion by
GC/MS. Methoxymethylphenylsilane (8),[48] 1,1,2,2-tetramethyldisilacy-
clohexane (9),[41,49] and 1,1,2,2-tetramethyldisilacyclopentane (12)[33,49]
1
were identified by comparison of the H NMR and mass spectra to those
of authentic samples and/or literature data. Product yields were deter-
mined from the relative slopes of concentration vs time plots constructed
for the starting trisilanes and the products based on integration of the
NMR spectra. Concentration vs time plots and the 1H NMR spectrum of
the mixture from photolysis of 7 (0.051m) in [D12]cyclohexane containing
0.2m MeOH to ca. 20% conversion are shown in Figures S1 and S3, re-
spectively. The corresponding plots and spectrum for photolysis of 6
(0.061m) under the same conditions are shown in Figures S2 and S4.
Laser flash photolysis: The experiments employed the pulses from a
Lambda Physik Compex 120 excimer laser, filled with F2/Kr/Ne (248 nm;
~25 ns; 100ꢀ5 mJ) mixtures, and a Luzchem Research mLFP-111 laser
flash photolysis system, modified as described previously.[50] The system
employs signal averaging (of 5–30 individual experiments, carried out a
repetition rate of ca. 0.3 Hz) for the acquisition of transient absorbance-
time profiles at selected monitoring wavelengths. Sample solutions were
prepared at concentrations such that the absorbance at the excitation
wavelength (248 nm) was between ca. 0.7 and 0.9. The solutions were
pumped continuously through a thermostatted 7ꢁ7 mm Suprasil flow cell
connected to a calibrated 250 mL reservoir using a Masterflex 77390 peri-
staltic pump fitted with Teflon tubing (Cole-Parmer Instrument Co.), at
an appropriate flow rate to ensure reproducible signal strengths and
decay kinetics from one laser shot to the next (typically 3–5 mLminꢁ1).
The calibrated reservoir was fitted with a glass frit to allow bubbling of
argon gas through the solution for at least 30 min prior to and then
throughout the duration of each experiment. The glassware, sample cells,
and transfer lines used for these experiments were stored in a 858C
vacuum oven when not in use, and the oven was vented with dry nitrogen
just prior to assembling the sample-handling system at the beginning of
an experiment. Reagents were added directly to the reservoir by microli-
1,1,2,3,3-Pentamethyl-2-phenyl-1,2,3-trisilacycloheptane (7): 1,1,2,3,3-Pen-
tamethyl-2-phenyltrisilane (2.58 g, 0.01 mol) was placed in a nitrogen-
flushed 50 mL round-bottom flask, and carbon tetrachloride (30 mL,
47.5 g, 0.31 mol) and 2,2’-azo-bis(2-methylpropionitrile) (0.02 g,
0.122 mmol) were added. A reflux condenser, magnetic stirrer, and nitro-
gen inlet were attached and the mixture was refluxed for 14 h until TLC
analysis (silica gel, hexanes) indicated reaction to be complete. The
excess carbon tetrachloride was removed on the rotary evaporator to
yield a colorless oil, which was further pumped on under high vacuum
for 1 h. The 1H and 13C NMR spectra of the material were consistent
with
1,3-dichloro-1,1,2,3,3-pentamethyl-2-phenyltrisilane.
1H NMR
(200 MHz, CDCl3): d
=
7.49 (m, 2H), 7.36 (m, 3H), 0.6 (s, 12H),
0.57 ppm (s, 3H); 13C NMR (50 MHz, CDCl3): d = 134.8, 132.7, 129.3,
8490
www.chemeurj.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 8485 – 8491

Spring-Loaded Precursors to Methylphenylsilylene
FULL PAPER
ter syringe as aliquots of standard solutions. Solution temperatures were
measured with a Teflon-coated copper/constantan thermocouple inserted
in the sample compartment in close proximity to the sample flow cell,
and are considered accurate to within 28C.
[20] V. A. Korolev, O. M. Nefedov, Adv. Phys. Org. Chem. 1995, 30, 1–
61.
[21] G. Maier, A. Meudt, J. Jung, H. Pacl in The Chemistry of Organic
Silicon Compounds (Eds.: Z. Rappoport, Y. Apeloig), Wiley, New
York, 1998, pp. 1143–1185.
[22] W. Ando, K. Hagiwara, A. Sekiguchi, Organometallics 1987, 6,
2270–2271.
[23] G. R. Gillette, G. H. Noren, R. West, Organometallics 1987, 6, 2617–
2618.
[24] A. Sekiguchi, K. Hagiwara, W. Ando, Chem. Lett. 1987, 209–210.
[25] T. Akasaka, S. Nagase, A. Yabe, W. Ando, J. Am. Chem. Soc. 1988,
110, 6270–6272.
[26] W. Ando, A. Sekiguchi, K. Hagiwara, A. Sakakibara, H. Yoshida,
Organometallics 1988, 7, 558–559.
Transient decay rate constants were calculated by linear least squares
analysis of log(DA) vs time data or by non-linear least squares analysis of
the absorbance-time profiles using the Prism 5.0 software package
(GraphPad Software, Inc.) and the appropriate user-defined fitting equa-
tions. Rate constants were calculated by linear least squares analysis of
decay rate-concentration data that incorporated at least 7 points and
spanned as large a range in transient decay rate as possible. Errors are
quoted as twice the standard error from the least squares analysis.
[27] G. R. Gillette, G. H. Noren, R. West, Organometallics 1989, 8, 487–
491.
Acknowledgements
[28] A. Patyk, W. Sander, J. Gauss, D. Cremer, Angew. Chem. 1989, 101,
920–922; Angew. Chem. Int. Ed. Engl. 1989, 28, 898–900.
[29] J. Belzner, H. Ihmels, Adv. Organomet. Chem. 1998, 43, 1–42.
[30] H. Bornemann, W. Sander, J. Am. Chem. Soc. 2000, 122, 6727–6734.
[31] G. Levin, P. K. Das, C. Bilgrien, C. L. Lee, Organometallics 1989, 8,
1206–1211.
We thank the Natural Sciences and Engineering Research Council of
Canada for financial support of this work, and Ms. Stephanie Ladowski
for technical assistance.
[32] M. Yamaji, K. Hamanishi, T. Takahashi, H. Shizuka, J. Photochem.
Photobiol. A 1994, 81, 1–6.
[33] A. G. Moiseev, W. J. Leigh, Organometallics 2007, 26, 6268–6276.
[34] A. G. Moiseev, W. J. Leigh, Organometallics 2007, 26, 6277–6289.
[35] M. Ishikawa, K. I. Nakagawa, R. Enokida, M. Kumada, J. Organo-
met. Chem. 1980, 201, 151–163.
[36] T. Miyazawa, S. Y. Koshihara, C. Liu, H. Sakurai, M. Kira, J. Am.
Chem. Soc. 1999, 121, 3651–3656.
[37] P. P. Gaspar, B. H. Boo, S. Chari, A. K. Ghosh, D. Holten, C. Kirma-
ier, S. Konieczny, Chem. Phys. Lett. 1984, 105, 153–157.
[38] P. P. Gaspar, D. Holten, S. Konieczny, J. Y. Corey, Acc. Chem. Res.
1987, 20, 329–336.
[39] W. J. Leigh, A. G. Moiseev, E. Coulais, F. Lollmahomed, M. S.
Askari, Can. J. Chem. 2008, 86, 1105–1117.
[40] H. Sakurai, Y. Kobayashi, Y. Nakadaira, J. Am. Chem. Soc. 1971, 93,
5272.
[41] A. G. Moiseev, W. J. Leigh, J. Am. Chem. Soc. 2006, 128, 14442–
14443.
[42] W. J. Leigh, G. W. Sluggett, J. Am. Chem. Soc. 1994, 116, 10468–
10476.
[43] A. Sekiguchi, I. Maruki, H. Sakurai, J. Am. Chem. Soc. 1993, 115,
11460–11466.
[44] M. Ishikawa, K. I. Nakagawa, M. Ishiguro, F. Ohi, M. Kumada, J.
Organomet. Chem. 1978, 152, 155–174.
[45] J. A. Hawari, M. Lesage, D. Griller, W. P. Weber, Organometallics
1987, 6, 880–882.
[46] M. Kumada, K. Tamao, T. Takubo, M. Ishikawa, J. Organomet.
Chem. 1967, 9, 43–45.
[47] J. Y. Corey, D. M. Kraichely, J. L. Huhmann, J. Braddock-Wilking,
A. Lindeberg, Organometallics 1995, 14, 2704–2717.
[48] W. J. Leigh, R. Boukherroub, C. J. Bradaric, C. C. Cserti, J. M.
Schmeisser, Can. J. Chem. 1999, 77, 1136–1147.
[49] T. J. Barton, A. Revis, J. Am. Chem. Soc. 1984, 106, 3802–3805.
[50] W. J. Leigh, C. R. Harrington, I. Vargas-Baca, J. Am. Chem. Soc.
2004, 126, 16105–16116.
[1] P. P. Gaspar in Reactive Intermediates, Vol. 3 (Eds.: M. Jones, Jr.,
R. A. Moss), Wiley, New York, 1985, pp. 333–427.
[2] P. P. Gaspar, R. West in The Chemistry of Organic Silicon Com-
pounds, Vol. 2 (Eds.: Z. Rappoport, Y. Apeloig), Wiley, New York,
1998, pp. 2463–2568.
[3] M. Kira, T. Miyazawa in The Chemistry of Organic Silicon Com-
pounds (Eds.: Z. Rappoport, Y. Apeloig), Wiley, New York, 1998,
pp. 1311–1337.
[4] J. M. Jasinski, B. S. Meyerson, B. A. Scott, Annu. Rev. Phys. Chem.
1987, 38, 109–140.
[5] I. Safarik, V. Sandhu, E. M. Lown, O. P. Strausz, T. N. Bell, Res.
Chem. Intermed. 1990, 14, 105–131.
[6] Y. Apeloig, R. Pauncz, M. Karni, R. West, W. Steiner, D. Chapman,
Organometallics 2003, 22, 3250–3256.
[7] S. E. Boganov, V. I. Faustov, M. P. Egorov, O. M. Nefedov, Russ.
Chem. Bull. 2004, 53, 960–979.
[8] R. Becerra, R. Walsh, Phys. Chem. Chem. Phys. 2007, 9, 2817–2835.
[9] J. Olꢂh, T. Veszprꢃmi, F. DeProft, P. Geerlings, J. Phys. Chem. A
2007, 111, 10815–10823.
[10] N. Tokitoh, R. Okazaki, Coord. Chem. Rev. 2000, 210, 251–277.
[11] N. J. Hill, R. West, J. Organomet. Chem. 2004, 689, 4165–4183.
[12] N. Tokitoh, W. Ando in Reactive Intermediate Chemistry (Eds.:
R. A. Moss, M. S. Platz, M. Jones, Jr.), Wiley, New York, 2004,
pp. 651–715.
[13] A. K. Franz, K. A. Woerpel, Acc. Chem. Res. 2000, 33, 813–820.
[14] H. Ottosson, P. G. Steel, Chem. Eur. J. 2006, 12, 1576–1585.
[15] M. Ishikawa, M. Kumada, Adv. Organomet. Chem. 1981, 19, 51–95.
[16] V. J. Tortorelli, M. Jones, Jr., S. Wu, Z. Li, Organometallics 1983, 2,
759–764.
[17] A. G. Brook in The Chemistry of Organic Silicon Compounds (Eds.:
S. Patai, Z. Rappoport), Wiley, New York, 1989, pp. 965–1005.
[18] T. J. Drahnak, J. Michl, R. West, J. Am. Chem. Soc. 1979, 101, 5427–
5428.
[19] M. J. Michalczyk, M. J. Fink, D. J. De Young, C. W. Carlson, K. M.
Welsh, R. West, J. Michl, Silicon Germanium Tin Lead Compd.
1986, 9, 75–80.
Received: May 11, 2009
Published online: July 16, 2009
Chem. Eur. J. 2009, 15, 8485 – 8491
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
8491