Mechanistic study of the ring-enlargement reaction of (3-oxa-2-silacyclopentyl)methyl radicals into 4-oxa-3-silacyclohexyl radicals. Evidence for a pentavalent silicon-bridging radical transition state in 1,2-rearrangement reactions of β-silyl radicals

Article abstract of DOI:10.1021/ja993239s

A mechanistic study was performed on a novel radical ring-enlargement reaction of (3-oxa-2-silacyclopentyl)methyl radicals into 4-oxa-3-silacyclohexyl radicals. Two pathways, one via a pentavalent silicon-bridging radical transition state (or intermediate

Full text of DOI:10.1021/ja993239s

J. Am. Chem. Soc. 2000, 122, 1343-1351
1343
Mechanistic Study of the Ring-Enlargement Reaction of
(3-Oxa-2-silacyclopentyl)methyl Radicals into 4-Oxa-3-silacyclohexyl
Radicals. Evidence for a Pentavalent Silicon-Bridging Radical
Transition State in 1,2-Rearrangement Reactions of â-Silyl Radicals
Satoshi Shuto,* Isamu Sugimoto, Hiroshi Abe, and Akira Matsuda
Contribution from the Graduate School of Pharmaceutical Sciences, Hokkaido UniVersity, Kita-12, Nishi-6,
Kita-ku, Sapporo 060-0812, Japan
ReceiVed September 7, 1999
Abstract: A mechanistic study was performed on a novel radical ring-enlargement reaction of (3-oxa-2-
silacyclopentyl)methyl radicals into 4-oxa-3-silacyclohexyl radicals. Two pathways, one via a pentavalent silicon-
bridging radical transition state (or intermediate), the other via â-elimination to give a ring-opened silyl radical,
can be postulated. The radical reactions of 1 and 2, which are precursors for a (3-oxa-2-silacyclopentyl)-
methyl radical C′ and a 4-oxa-3-silacyclohexyl radical D′, respectively, showed that the ring-enlargement
rearrangement of C′ into D′ is irreversible. 1H NMR analysis of the radical reactions of 8a and 8b, which have
an asymmetric center at silicon, indicated that the configuration at the silicon atom is retained via a pentavalent
silicon-bridging radical transition state (or intermediate) during the ring-enlargement reaction. Furthermore,
examination of the radical ring-enlargement reaction with a deuterium-labeled substrate 12D showed that the
ring-enlargement reaction did not involve â-elimination to give a ring-opened silyl radical. Based on these
results, we conclude that the ring-enlargement reaction occurs via a pentavalent silicon-bridging radical transition
state (or intermediate). This is the first experimental evidence for such a pentavalent silicon radical, which has
been previously postulated to understand radical reactions of organic silicon compounds.
Introduction
Thus, the selective introduction of both 1-hydroxyethyl and
2-hydroxyethyl groups can be achieved, depending on the
concentration of Bu3SnH in the reaction system, via a 5-exo-
cyclization intermediate E or a 6-endo-cyclization intermediate
F, respectively, after oxidative ring-cleavage by treating the
cyclization products under Tamao oxidation conditions,4 as
shown in Scheme 1.2a,b A vinyl group can also be introduced
by photoirradiating the vinylsilyl ether A in the presence of
(Bu3Sn)2, and then treating the resulting atom-transfer 5-exo-
cyclization product I with fluoride ion.3 We also investigated
the radical cyclization mechanism by deuterium-labeling experi-
ments with Bu3SnD.2a,b The results suggested that the kinetically
favored 5-exo-cyclized radical C, formed from radical B, was
trapped when the concentration of Bu3SnH was high enough to
give E. At lower concentrations of Bu3SnH and higher reaction
temperatures, radical C rearranged into the more stable ring-
enlarged 4-oxa-3-silacyclohexyl radical D, which was then
trapped with Bu3SnH to give F. This radical reaction with a
vinylsilyl tether has been successfully applied to the synthesis
of biologically important branched nucleosides2b-d and C-glyco-
sides.2e
There has been growing interest in radical reactions of silicon-
containing compounds because of their versatility in organic
synthesis.1 Recently, we developed a regio- and stereoselective
method for introducing the 1-hydroxyethyl, 2-hydroxyethyl, and
vinyl groups at the position â to a hydroxyl group in halohydrins
or R-phenylselenoalkanols using an intramolecular radical
cyclization reaction with a dimethyl- or diphenylvinylsilyl group
as a temporary connecting radical-acceptor tether (Scheme 1).2,3
* To whom correspondence and reprint requests should be addressed.
(1) (a) Nishiyama, H.; Kitajima, T.; Matsumoto, M.; Itoh, K. J. Org.
Chem. 1984, 49, 2298-2300. (b) Stork, G.; Kahn, M. J. Am. Chem. Soc.
1985, 107, 500-501. (c) Magnol, E.; Malacria, M. Tetrahedron Lett. 1986,
27, 2255-2256. (d) Koreeda, M.; George, I. A. J. Am. Chem. Soc. 1986,
108, 8098-8100. (e) Tamao, K.; Maeda, K.; Yamaguchi, T.; Ito, Y. J. Am.
Chem. Soc. 1989, 111, 4984-4985. (f) Walkup, R. D.; Kane, R. R.;
Obeyesekere, N. U. Tetrahedron Lett. 1990, 31, 1531-1534. (g) Koreeda,
M.; Hamann, L. G. J. Am. Chem. Soc. 1990, 112, 8175-8177. (h) Myers,
A. G.; Gin, D. Y.; Widdowson, K. L. J. Am. Chem. Soc. 1991, 113, 9661-
9663. (i) Hutchinson, J. H.; Daynard, T. S.; Gillard, J. W. Tetraheron Lett.
1991, 32, 573-576. (j) Koot, W.-J.; van Ginkel, R.; Kranenburg, M.;
Hiemstra, H.; Louwrier, S.; Moolenaar, M. J.; Speckamp, W. N. Tetrahedron
Lett. 1991, 32, 401-404. (k) Stork, G.; Suh, H. S. Kim. G. J. Am. Chem.
Soc. 1991, 113, 7054-7056. (l) Xi, A.; Agback, P.; Plavec, J.; Sandstrom,
A.; Chattopadhyaya, J. B. Tetrahedron, 1992, 58 349-370. (m) Martinez-
Grau, A.; Curran, D. P. J. Org. Chem. 1995, 60, 8332-8333. (n) Bols, M.;
Skrydstrup, T. Chem. ReV. 1995, 95, 1253-1277. (o) Fensterbark, L.;
Malacria, M.; Sieburth, S. M. Synthesis 1997, 813-854.
(2) (a) Shuto, S.; Kanazaki, M.; Ichikawa, S.; Matsuda, A. J. Org. Chem.,
1997, 62, 5676-5677. (b) Shuto, S.; Kanazaki, M.; Ichikawa, S.; Minakawa,
N.; Matsuda, A. J. Org. Chem., 1998, 63, 746-754. (c) Ueno, Y.; Nagasawa,
Y.; Sugimoto, I.; Kojima, N.; Kanazaki, M.; Shuto, S.; Matsuda, A. J. Org.
Chem. 1998, 63, 1660-1667. (d) Sugimoto, I.; Shuto, S.; Mori, S.; Shigeta,
S.; Matsuda, A. Bioorg. Med. Chem. Lett. 1999, 9, 385-388. (e) Yahiro,
Y.; Ichikawa, S.; Shuto, S.; Matsuda, A. Tetrahedron Lett. 1999, 40, 5527-
5531.
As far as we know, this is the first such ring-enlarging
rearrangement reaction of â-silyl radicals reported in the
literature. This rearrangement can also be considered a 1,2-
silicon shift from carbon to carbon in â-silyl carbon-centered
radicals. There is no precedent for this type of silicon shift,5,6
although 1,2-silicon shift reactions of nitrogen-,7 oxygen-,8 and
(3) Sugimoto, I.; Shuto, S.; Matsuda, A. J. Org. Chem. 1999, 64, 7153-
7157.
(4) Tamao, K.; Ishida, N.; Kumada, M. J. Org. Chem. 1983, 48, 2122-
2124.
10.1021/ja993239s CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/04/2000

1344 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Shuto et al.
Scheme 1
Scheme 2
Scheme 3
sulfur-centered9 radicals have been reported. However, the
reaction mechanism is still unclear.
Two possible pathways for the ring-enlargement reaction
(Scheme 2) may be postulated: one via a transition state (or
intermediate) X in which the silicon atom expands its valence
shell to five (path x), the other via â-elimination to give the
ring-opened silyl radical Y (path y), which subsequently
undergoes 6-endo-cyclization to give D. The mechanism of this
radical rearrangement reaction, which may be related mecha-
nistically to the above known radical 1,2-silicon shifts, has been
of great interest to us. In this report, we describe the results of
our mechanistic study demonstrating that the ring-enlargement
reaction occurs via a pentavalent silicon-bridging radical transi-
tion state (or intermediate) X.
Scheme 4
Reactions with the Radical Precursors for (3-Oxa-2-
silacyclopentyl)methyl and 4-Oxa-3-silacyclohexyl Radicals.
First, we investigated the reaction of 1 and 2, which are
precursors for a (3-oxa-2-silacyclopentyl)methyl radical C′ and
a 4-oxa-3-silacyclohexyl radical D′, respectively (Scheme 3).10
Such experiments would clearly confirm that 4-oxa-3-sila-
(5) A gas phase 1,2-shift of a trimethylsilyl group from carbon to carbon
in an aromatic â-silyl biradical at >450 °C has been reported: Johnson, G.
C.; Stofko, J. J., Jr.; Lockhart, T. P.; Brown, D. W.; Bergman, R. G. J.
Org. Chem. 1979, 44, 4215-4218.
(6) A gas-phase 1,2-shift of a trimethylsilyl group from silicon to carbon
radical is observed at 600 °C; however, no 1,2-shift occurs when the same
radical is generated in solution: Sakurai, H.; Kishida, T.; Hosomi, A.;
Kumada, M. J. Organomet. Chem. 1967, 8, 65-68.
(7) (a) West, R.; Boudjouk, P. J. Am. Chem. Soc. 1973, 95, 3938-3987.
(b) Harris, J. M.; MacInnes, I.; Walton, J. C.; Maillard, B. J. Organomet.
Chem. 1991, 403, C25-28. (c) Harris, J. M.; Walton, J. C.; Maillard, B.;
Grelier, S.; Picard, J.-P. J. Chem. Soc., Perkin Trans. 2 1993, 2119-2123.
(d) Roberts, B. P.; Vazquez-Persaud, A. R. J. Chem. Soc., Perkin Trans. 2
1995, 1087-1095.
(8) Tsai, Y.-M.; Cherng, C.-D. Tetrahedron Lett. 1991, 32, 3515-3518.
(9) Pitt, C. G.; Fowler, M. S. J. Am. Chem. Soc. 1968, 90, 1928-1930.
(10) Radical reactions of 1 and 2 have been described in a communica-
tion: Sugimoto, I.; Shuto, S.; Matsuda, A. Synlett. 1999, 1766-1768.
cyclohexyl radical D is produced from (3-oxa-2-silacyclopentyl)-
methyl radical C via a novel ring-enlargement reaction (Scheme
1), as previously suggested by deuterium-labeling experiments,2a,b
and also would clarify whether the corresponding reverse
reaction, i.e., ring-contraction of radical D into radical C,
actually occurs.
The precursors 1 and 2 were prepared by the atom-transfer
radical cyclization reaction11 of the vinylsilyl ether 7 as shown
in Scheme 4. (()-1-Phenylseleno-2-indanol (6), prepared from
(()-trans-2-bromo-1-indanol (5) via indanol 1,2-epoxide, was
treated with diphenylvinylchlorosilane, DMAP, and Et3N in
(11) Curran, D. P.; Chang, C. J. Org. Chem. 1986, 54, 3140-3157.

PentaValent Silicon-Bridging Radical
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1345
Scheme 5
toluene to give vinylsilyl ether 7. Photoirradiation of a solution
of 7 and (Bu3Sn)2 (0.01 equiv) in benzene with a high-pressure
mercury lamp at room temperature under an argon atmosphere
gave the expected atom-transfer 5-exo-cyclization product 1 in
59% yield. When the photoreaction was carried out at 80 °C,
the ring-enlarged atom-transfer product 2 was obtained in 21%
yield along with 1 in 36% yield.
Scheme 6
A solution of 1.2 equiv of Bu3SnH and 0.6 equiv of AIBN
in benzene was added slowly over 4 h, using a syringe-pump,
to a solution of 1 in refluxing benzene. The reaction gave the
ring-enlargement product 4 as a major product along with the
directly reduced product 3 (yield 70%, 3:4 ) 9:91, based on its
1H NMR spectrum), after purification by silica gel column
chromatography. On the other hand, when 2 was treated under
conditions identical with those for 1, only the direct reduction
of radical D′ occurred affording 4 in 72% yield as the sole
product; the corresponding ring-contracted product 3 was not
obtained. These results clearly demonstrate that the (3-oxa-2-
silacyclopentyl)methyl radical C′ readily rearranged into the
ring-enlarged 4-oxa-3-silacyclohexyl radical D′ (Scheme 3),
which is consistent with the previous results suggested by
deuterium-labeling experiments.2a,b The results also suggest that
the corresponding reverse reaction, i.e., ring-contraction of D′
into C′, did not occur or was very slow.
The Configuration at the Silicon Atom during the Radical
Ring-Enlargement Reaction. We next investigated the reaction
mechanism of the radical rearrangement by using 8a and 8b as
substrates. Both have a methyl and a phenyl group on the silicon
atom, and are therefore stereoisomers. The configurations at the
silicon atom in the ring-enlargement reaction products derived
from the radical reactions of 8a and 8b should be dominated
by the reaction mechanism shown in Scheme 5. Treatment of
8a (endo-SiMe isomer) or 8b (exo-SiMe isomer) with Bu3SnH/
AIBN would produce radical ia or ib, respectively. Direct
reduction of ia or ib by Bu3SnH gives 9a or 9b, respectively.
If the radical ring-enlargement reaction of ia or ib proceeds via
ring-opened silyl radical iiia or iiib (path y or y′ in Scheme 5),
isomerization at the silicon atom of iiia or iiib should occur, at
least to some extent, before re-cyclization to give a mixture of
6-endo-cyclized radicals iva and ivb. Consequently, a mixture
of 10a (endo-SiMe product) and 10b (exo-SiMe product), which
are stereoisomers at the silicon atom, would be obtained.
Alternatively, the configuration at the silicon atom of 8a or 8b
should be retained during the rearrangement process to give
the ring enlargement product 10a or 10b, respectively, when
the radical rearrangement proceeds via the pentavalent silicon-
bridging radial transition state (or intermediate) iia or iib (path
x or x′ in Scheme 5).
The substrates 8a and 8b were synthesized by a method
similar to the one for the preparation of 1 and 2 (Scheme 6).
(()-1-Phenylseleno-2-indanol (6) was treated with methyl-
phenylvinylchlorosilane, DMAP, and Et3N in toluene to give
vinylsilyl ether 11, which was an inseparable mixture of
stereoisomers at the silicon atom. Photoreaction of 11 in the
presence of (Bu3Sn)2 in benzene gave atom-transfer 5-exo-
cyclization products as a mixture of four stereoisomers, 8a, 8b,
8c, and 8d, in 56% yield (8a:8b:8c:8d ) 48:35:9:8). These
products were successfully separated by HPLC, and their
stereochemistries were confirmed by NOE experiments (Figure
1).12 The major products 8a and 8b were used in the subsequent
experiments.
(12) When the methyl protons and the methyne proton adjacent to the
silicon were irradiated, NOEs as shown in Figure 1 were observed.

1346 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Shuto et al.
Figure 1.
Figure 3. Silyl methyl signals in the 1H NMR spectrum (CDCl3, 500
MHz): a mixture of 9a, 9b, 10a, and 10b (A); products of the radical
reaction of 9a (B); and products of the radical reaction of 9b (C).
9a, respectively, while signals corresponding to the exo-SiMe
products 9b and 10a were not detected at all (Figure 3B).
Similarly, the radical reaction of exo-SiMe substrate 8b gave
the corresponding exo-SiMe products 9b and 10b, as shown in
Figure 3C. These results demonstrate that the configuration at
the silicon atom is retained during the ring-enlargement reaction
and that a ring-opened silicon radical like iiia or iiib is not
produced. The most likely explanation for the configuration-
retaining reaction pathway of the radical ring-enlargement is
that it proceeds via pentavalent silicon-bridging radical transition
states (or intermediates) iia and iib (paths x and x′).
Elucidation of the Mechanism with a Deuterium-Labeled
Substrate. We designed a deuterium-labeled substrate 12D to
further elucidate the reaction pathway of the ring-enlargement
reaction. Our strategy is summarized in Scheme 7. Reaction of
12D with Bu3SnH/AIBN would generate the exo-cyclized
radical vi via the allylic radical v. If the ring-enlargement
reaction of vi proceeds via a pentavalent silicon-bridging radical
transition state (or intermediate) vii, it would produce the ring-
enlargement product 13D, with the deuterium label at the
terminal methylene carbon (path x). On the other hand, if
â-elimination of vi occurs, a mixture of ring-enlargement
products 13D and 14D, the latter of which is deuterium labeled
at the methylene adjacent to the silicon, should be obtained,
since the resulting silyl radical viii can cyclize to both the labeled
and unlabeled terminal methylenes which are regiochemically
equivalent.
Figure 2.
The radical reaction of 8a was next carried out. A benzene
solution of Bu3SnH (1.5 equiv) and AIBN (0.6 equiv) was added
slowly over 4 h to a refluxing solution of 8a (0.01 M) in
benzene. A mixture of products 9a, 9b, 10a, and 10b was
obtained in 44% yield (9a:9b:10a:10b ) 22:13:63:2), after
purification by silica gel column chromatography. Each of these
was isolated in pure form by HPLC, and their stereochemistries
were determined by their 1H NMR spectra and NOE experiments
(Figure 2).13 Throughout this experiment, we noticed that
isomerization of the radical reaction products at the silicon atom
occurred during silica gel column chromatography.14,15 There-
fore, we analyzed the reaction products directly by H NMR
without purification, after the reactions of 8a and 8b were carried
out under conditions identical with those described above. The
SiMe signals in the H NMR spectra (-0.1s 0.5 ppm) of the
products derived from the radical reaction of 8a and 8b are
shown in Figures 3B and 3C, and the SiMe signals of a mixture
of 9a, 9b, 10a, and 10b, which are observed at -0.02, 0.47,
0.18, and 0.43 ppm, respectively, are also shown in Figure 3A
as a reference. When 8a, which has an endo-Me group at the
silicon, was treated with BuSn3H/AIBN in benzene, the SiMe
signals of the products were observed at 0.18 and -0.02 ppm,
which were identical with those of the endo-Me groups of the
ring-enlargement product 10a and the directly reduced product
1
1
(13) Although the stereochemistries of 9a and 9b were confirmed by
the NOE data as shown in Figure 2, no NOE was observed when the silyl
methyl protons of 10a or 10b were irradiated. Therefore, the configurations
at the silicon of 10a and 10b were confirmed based on chemical shifts of
the methyl groups attached to the silicon of the compounds having benzo-
[f]-2-oxa-3-silabicyclo[3.3.0]octane or benzo[g]-2-oxa-3-silabicyclo[4.3.0]-
nonane structure synthesized in this study: the signals of the exo-SiMe
protons in 1b, 1d, 9b, and 10b (0.43-0.57 ppm) were observed at lower
fields compared with those of the endo-SiMe protons in 1a, 1c, 9a, and
10a (-0.02-0.18 ppm) presumably due to deshielding by the benzene ring
condensed with the 2-oxa-3-silabicyclo[3.3.0]octane or 2-oxa-3-silabicyclo-
[4.3.0]nonane ring (see Figure 2).
(14) It is known that inversion of configuration at a silicon atom in
organo-silicon compounds readily occurs via pseudorotation: Holmes, R.
R. Chem. ReV. 1990, 90, 17-31.
(15) For example, when pure 9a was passed through a silica gel column
with a solvent system of Et2O-hexane, a mixture of 9a and 9b (ratio, ca.
1:1) was obtained.
The synthesis of the deuterium-labeled substrate 12D as well
as unlabeled 12 is shown in Scheme 8. An R,â-unsaturated ethyl
ester 16, prepared from commercially available ethyl fumarate
15, was reduced with LiAl(OMe)3D16 or DIBAL to give
deuterium-labeled alcohol 17D or unlabeled 17, respectively.
Successive treatment of 17 with PhSeCN/Bu3P17 and TBAF in
THF gave 18, which was further treated with diphenylvinyl-
chlorosilane/DMAP/Et3N in toluene to afford the substrate 12
for the radical reaction. Similarly, the deuterium-labeled sub-
strate 12D was prepared from 17D.
(16) Brown H, C.; Weissman, P. M. J. Am. Chem. Soc. 1965, 87, 5614-
5620.
(17) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976, 41,
1485-1486.

PentaValent Silicon-Bridging Radical
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1347
Scheme 7
Figure 4. 1H NMR spectra of 22D (A) and 20 (B) (CDCl3, 500 MHz).
Scheme 9
Scheme 8
20 as a reference in Figure 4. The spectrum clearly shows that
the protons at the terminal methylene in product 22D were
exclusively replaced by deuteriums and that the regioisomeri-
cally labeled 23D was not detected at all (Scheme 10).
Accordingly, these results suggest that the ring-enlargement
reaction of radical vi is not likely to occur via a ring-opened
silicon radical viii but rather via the pentavalent silicon-bridging
radical transition state (or intermediate) vii in Scheme 7.18
First, the radical reaction of 12 was performed under
thermodynamic conditions; Bu3SnH/AIBN was added slowly
over 4 h to a refluxing solution of 12 in benzene. The reaction
mixture was subsequently treated under Tamao oxidation
conditions, and the resulting 3-hydroxymethyl-4-pentenol was
isolated as the corresponding dibenzoate 20 by treating it with
BzCl/pyridine (Scheme 9). A similar reaction of 12 with Bu3SnD
instead of Bu3SnH gave the corresponding deuterium-labeled
product 21, in which a deuterium was incorporated exclusively
at the 2-position based on the 1H NMR spectrum. These results
confirmed that the radical ring-enlargement reaction occurred
in this system as expected. Therefore, we next investigated the
radical reaction of deuterium-labeled substrate 12D. Thus, 12D
was subjected to the above-mentioned radical reaction with
Bu3SnH/AIBN and the subsequent successive two-step treat-
ments. The product, after purification by HPLC, was analyzed
by 1H NMR, and the spectrum is shown with that of unlabeled
Discussion
We have shown that (3-oxa-2-silacyclopentyl)methyl radicals
rearrange into ring-enlarged 4-oxa-3-silacyclohexyl radicals with
various substrates, i.e., indanols,2a nucleosides,2b,c hexopyranoses,2e
and a 2-butenol derivative used in this study. These results
suggest that this rearrangement reaction is general in (3-oxa-
2-silacyclopentyl)methyl radicals.
(18) The secondary deuterium effect in the cyclization of radical viii
should be considered. It is known that addition of alkyl radicals to an alkene
is slightly facilitated when the terminal CH2d of the alkene is replaced by
CD2d: Feld, A.; Stefani, A. P.; Szwarc, M. J. Am. Chem. Soc. 1962, 84,
4451-4453.

1348 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Shuto et al.
Scheme 10
Scheme 11
Scheme 12
inversion at the silicon center of asymmetric silyl radicals should
be considered, since asymmetric silyl radicals have been reported
to be configurationally stable compared with asymmetric carbon
radicals.20 The rates of inversion at the silicon center of
asymmetric silyl radicals were determined by Ingold and co-
workers to be (3-12) × 109 s-1 at temperatures from 0 to 80
°C.20e On the other hand, 6-endo-cyclizations of pent-4-enylsilyl
radicals have been studied.21 Ingold and co-workers observed
that the rate constant for 6-endo-cyclization of 3-dimethylpent-
4-enylsilyl radical was <109 s-1 at -100 °C and >107 s-1 at
room temperature.21a Considering these results, isomerization
at the silicon atom should occur, at least to some extent, if the
ring-opened silicon radical iiia or iiib is involved in the reaction
process, since 6-endo-cyclization of pent-4-enylsilyl radicals has
been shown to be slower than isomerization at silicon.
The rearrangement reaction appears to be irreversible based
on the study with 2, which is a precursor for a 4-oxa-3-
silacyclohexyl radical. We undertook a theoretical investigation
of the (3-oxa-2-silacyclopentyl)methyl radical C′ and the ring-
enlarged 4-oxa-3-silacyclohexyl radical D′ by computational
methods to compare their stabilities. Geometry optimizations
and single-point energy calculations were performed using PM3
and ab initio calculations at the UHF/STO-3G level, respec-
tively. Two stereoisomers, i.e., endo-Me radical C′1 and exo-
Me radical C′2, should be considered for (3-oxa-2-silacyclopentyl)-
methyl radical C′. The ring-enlarged radical D′ is 11.5 and 6.3
kcal/mol more stable than C′1 and C′2, respectively, based on
the heat of formation (Scheme 11).19 Similar results were
obtained from the calculations on radicals ia, ib, iva, and ivb
with an asymmetric silicon center (Scheme 11). These compu-
tational results clearly support the experimental data indicating
that the ring-enlargement rearrangement is irreversible.
The reactions with substrates 8a and 8b, which have an
asymmetric center at the silicon atom, showed that the config-
uration at the silicon atom is retained during the rearrangement
reaction. This suggests that the radical ring-enlargement pro-
ceeds via pentavalent silicon-bridging radical transition states
(or intermediates) iia and iib (paths x and x′), without producing
a ring-opened silicon radical like iiia or iiib. The rates of
The results with the asymmetric silicon substrates 8a and 8b
were further confirmed by the study with deuterium-labeled
substrate 12D. Therefore, we conclude that the ring-enlargement
reaction is an irreversible process that occurs via a transition
state (or intermediate) X in which the silicon atom expands its
valence shell to five (path x in Scheme 1).
1,2-Silicon shifts of nitrogen-,7 oxygen-,8 and sulfer-centered9
radicals have been reported. It has been postulated that these
rearrangements can occur by a dissociative mechanism or by
an intramolecular process via a pentavalent silicon-bridging
radical transition state (or an intermediate), similar to the one
in this study, as shown in Scheme 12a. Harris et al.7b,c and
Roberts et al.7d investigated the reaction mechanism of 1,2-
(20) (a) Sakurai, H.; Murakami, M.; Kumada, M. J. Am. Chem. Soc.
1969, 91, 519-520. (b) Brook, A. G.; Duff. J. M. J. Am. Chem. Soc. 1969,
91, 2118-2119. (c) Sommer, L. H.; Ulland, L. A. J. Org. Chem. 1972, 37,
3878-3881. (d) Sakurai, H.; Murakami, M.; Kumada, M. Bull. Chem. Soc.
Jpn. 1977, 50, 3384-3387. (e) Chatgilialoglu, C.; Ingold, K. U.; Scainao,
J. C. J. Am. Chem. Soc. 1982, 104, 5123-5127.
(21) (a) Chatgilialoglu, C.; Woynar, H.; Ingold, K. U.; Davies, A. G. J.
Chem. Soc., Perkin Trans. 2 1983, 555-561. (b) Barton, T. J.; Revis, J.
Am. Chem. Soc. 1984, 106, 3802-3805. (c) A. Sarasa, J. P.; Igual, J.; Poblet,
A. M. J. Chem. Soc., Parkin Trans. 2 1986, 861-865. (d) Cai, Y.; Roberts,
B. P. J. Chem. Soc., Perkin Trans. 1 1998, 467-475.
(19) Heat of formation: C′1, -1232.0042536 hartrees; C′2, -1232.0125841
hartrees; D′, -1232.0226381 hartrees; ia, -1043.848567 hartrees; iva,
-1043.857157 hartrees; ib, -1043.842205 hartrees; ivb, -1043.857857
hartrees (1 hartree ) 627.51 kcal/mol).

PentaValent Silicon-Bridging Radical
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1349
(55 mg, 0.45 mmol), and diphenylvinylchlorosilane (1.3 mL, 7.1 mmol)
in toluene (20 mL) was stirred at room temperature for 5 min. Et2O
and water were added, and the resulting mixture was partitioned. The
organic layer was washed with brine, dried (Na2SO4), and evaporated.
The residue was purified by silica gel column chromatography (Et2O/
silicon shifts of silylmethylaminyl radicals by EPR spectroscopy.
Both groups independently suggested that the silicon shifts may
occur via a pentavalent silicon transition state, since free TMS
radicals, which should be produced in the dissociative pathway,
were not detected during the silicon shift reactions by EPR.
On the other hand, novel radical substitution reactions at a
silicon atom have been reported.22 Giese and co-workers
proposed a pentavalent silicon-bridging transition state (or
intermediate) to interpret this kind of radical substitution reaction
(Scheme 12b).22a
While the structure of the pentavalent silicon radicals is yet
unclear, Schiesser and Styles recently reported ab initio mo-
lecular orbital calculations of pentavalent silicon radical transi-
tion states in 1,2-silicon-shift reactions of carbon-, nitrogen-,
and oxygen-centered radicals.23 The calculations suggested that
the transition state adopted tetrahedral geometry at the silicon
and that the structure appeared to resemble coordinated alkenes,
which is consistent with a front-side mechanism for the silicon
shifts. Due to the transition state predicted, configuration at the
silicon should be retained during the shift when the silicon is
an asymmetric center. This may be in agreement with our
present experimental results with 8a and 8b indicating that the
configuration was retained during the rearrangement.
1
hexane, 1:50) to give 7 (2.20 g, 98%) as an oil: H NMR (500 MHz,
CDCl3) δ 7.49-7.14 (m, 19 H), 6.31 (dd, 1 H, J ) 20.1, 14.9 Hz),
6.18 (dd, 1 H, J ) 15.0, 3.8 Hz), 5.76 (dd, 1 H, J ) 20.3, 3.9 Hz),
4.79 (m, 1 H), 4.76 (s, 1 H), 3.13 (dd, 1 H, J ) 16.3, 5.3 Hz), 2.87
(dd, 1 H, J ) 16.3, 1.4 Hz); 13C NMR (125 MHz, CDCl3) δ 141.3,
141.2, 137.3, 135.0, 134.7, 134.3, 134.2, 133.6, 130.0, 129.5, 128.9,
127.8, 127.7, 127.6, 126.8, 125.6, 125.0, 80.3, 54.6, 40.7; EI HRMS
calcd for C29H26OSeSi 498.0916, found 498.0912 (M+). Anal. Calcd
for C29H26OSeSi: C, 70.00; H, 5.27. Found: C, 70.03; H, 5.38.
Synthesis of Radical Precursor 1. A stirred solution of 7 (9.60 g,
19.3 mmol) and (Bu3Sn)2 (0.98 mL, 1.9 mmol) in benzene (100 mL)
was irradiated with a high-pressure mercury lamp (300 W) at room
temperature for 25 h. The solvent was evaporated, and the residue was
purified by silica gel column chromatography (Et2O/hexane, 1:20) to
give 1 (5.66 g, 59%) as an oil: 1H NMR (500 MHz, CDCl3) δ 7.69-
7.00 (m, 19 H), 5.24 (dt, 0.84 H, J ) 5.6, 2.4 Hz), 4.96 (n, 0.16 H),
3.99 (dd, 0.16 H, J ) 7.7, 4.6 Hz), 3.87 (dd, 0.84 H, J ) 4.6, 4.2 Hz),
3.54 (dd, 0.16 H, J ) 11.7, 9.0 Hz), 3.31-2.97 (m, 3.84 H), 2.57 (q,
0.16 H, J ) 8.7 Hz), 2.19 (ddd, 0.84 H, J ) 10.5, 6.6, 2.6 Hz); EI
HRMS calcd for C29H26OSeSi 498.0916, found 498.0910 (M+). Anal.
Calcd for C29H26OSeSi: C, 70.00; H, 5.27. Found: C, 70.18; H, 5.43.
Synthesis of Radical Precursor 2. A stirred solution of 7 (1.00 g,
2.01 mmol) and (Bu3Sn)2 (100 µL, 0.20 mmol) in benzene (20 mL)
was irradiated with a high-pressure mercury lamp (300 W) at 80 °C
for 25 h. The solvent was evaporated, and the residue was purified by
silica gel column chromatography (Et2O/hexane, 1:20) to give 2 (208
mg, 21%; along with 1, 360 mg, 36%) as an oil: 1H NMR (400 MHz,
CDCl3) δ 8.14-6.87 (m, 19 H), 5.11 (q, 0.4 H, J ) 5.0), 4.86 (m, 0.6
H), 4.23 (dt, 0.6 H, J ) 14.4, 2.9 Hz), 4.14 (ddd, 0.4 H, J ) 9.0, 4.9,
4.2 Hz), 3.56 (br s, 0.6 H), 3.44 (dd, 0.4 H, J ) 7.8, 5.4 Hz), 3.11-
2.96 (m, 2 H), 1.79 (m, 1 H, J ) 14.4 Hz), 1.51 (dd, 0.4 H, J ) 15.1,
9.5 Hz), 1.59 (dd, 0.6 H, J ) 14.4, 3.4 Hz); EI HRMS calcd for C29H26-
OSeSi 498.0916, found 498.0898 (M+).
As described above, the reaction mechanisms of these radical
rearrangements of organic silicon compounds can be interpreted
by postulating a pentavalent silicon radical transition state (or
intermediate). This study should be very important, since it
presents the first experimental evidence for the pentavalent
silicon transition state (or intermediate) in radical reactions of
organic silicon compounds.
Experimental Section
Melting points are uncorrected. NMR spectra were recorded at 270,
400, or 500 MHz (1H) and at 100 or 125 MHz (13C), and are reported
in ppm downfield from TMS. Mass spectra were obtained by electron
ionization (EI) or the fast atom bombardment (FAB) method. Thin-
Radical Reaction of 1. To a solution of 1 (100 mg, 0.20 mmol) in
benzene (20 mL) at 80 °C was added a solution of Bu3SnH (65 µL,
0.24 mmol) and AIBN (20 mg, 0.03 mmol) in benzene (5 mL) slowly
over a 4 h period. The solvent was evaporated, and the residue was
purified by silica gel column chromatography (Et2O/hexane, 1:50) to
layer chromatography was performed on Merck coated plate 60F254
.
Silica gel chromatography was performed with Merck silica gel 5715.
Reactions were carried our under an argon atmosphere. HPLC was
performed with YMC-Pack SIL-06-s-5 (analytical, 4.6 × 250 mm;
preparative, 20 × 250 mm).
1
give a mixture of 3 and 4 (49 mg, 70%, 3:4 ) 9:91 based on its H
NMR) as an oil. From the mixture, 3 (33 mg, 47%) and 4 (3 mg, 4%)
were obtained as oils in a pure form, respectively, by preparative HPLC
separation (AcOEt/hexane, 1:30). 3: 1H NMR (400 MHz, CDCl) δ
7.64-7.11 (m, 14H), 5.26 (dt, 1 H, J ) 6.1, 3.4 Hz), 3.38 (t, 1 H, J )
5.5 Hz), 3.27 (dd, 1 H, J ) 16.6, 6.1 Hz), 3.17 (dd, 1 H, J ) 16.6, 3.4
Hz), 1.96-1.89 (m, 1 H), 1.21 (d, 3 H, J ) 7.8 Hz);13C NMR (100
MHz, CDCl3) δ 144.5, 140.8, 134.7, 134.4, 134.1, 133.1, 129.9, 129.8,
127.8, 127.5, 127.1, 126.5, 125.1, 124.5, 82.6, 57.5, 41.3, 23.7, 15.8;
EI HRMS calcd for C23H22OSi 342.1440, found 342.1428 (M+). Anal.
Calcd for C23H22OSi: C, 80.65; H, 6.47. Found: C, 80.46; H, 6.60. 4:
1H NMR (400 MHz, CDCl3) δ 7.67-7.16 (m, 14 H), 5.00 (ddd, 1 H,
J ) 5.6, 4.2 Hz), 3.25-3.21 (m, 1 H), 3.17 (dd, 1 H, J ) 15.9, 5.6
Hz), 3.08 (dd, 1 H, J ) 16.1, 3.9 Hz), 2.36-2.18 (m, 2 H), 1.25 (ddd,
1 H, J ) 14.9, 9.3, 4.6 Hz), 1.14 (ddd, 1 H, J ) 14.9, 8.6, 4.4 Hz);13C
NMR (100 MHz, CDCl3) δ 143.5, 141.6, 135.8, 135.7, 134.2, 134.0,
129.8, 129.7, 127.9, 127.6, 126.6, 126.4, 125.0, 123.9, 77.6, 47.3, 41.6,
22.0, 6.9; EI HRMS calcd for C23H22OSi 342.1440, found 342.1419
(M+). Anal. Calcd for C23H22OSi: C, 80.65; H, 6.47. Found: C, 80.64;
H, 6.59.
(()-trans-1-Phenylseleno-2-indanol (6). A mixture of (()-trans-
2-bromoindanol (5, 1.07 g, 5.02 mmol) and NaH (0.24 g, 6.0 mmol)
in THF (25 mL) was stirred at room temperature for 2 h. A solution of
NaSePh, prepared from (PhSe)2 (3.12 g, 10.0 mmol), NaBH4 (1.14 g,
30.1 mmol), and EtOH (25 mL),24 was added, and the resulting mixture
was stirred at room temperature for 10 min. Et2O and saturated aqueous
NH4Cl were added, and the resulting mixture was partitioned. The
organic layer was washed with brine, dried (Na2SO4), and evaporated.
The residue was purified by silica gel column chromatography (AcOEt/
hexane, 1:4) to give 6 (1.42 g, 98%) as an oil: 1H NMR (500 MHz,
CDCl3) δ 7.53-7.21 (m, 9 H), 4.66 (d, 1 H, J ) 1.7 Hz), 4.61-4.59
(m, 1 H), 3.30 (dd, 1 H, J ) 16.6, 5.6 Hz), 2.82 (dd, 1 H, J ) 16.6, 2.1
Hz), 1.85 (d, 1 H, J ) 5.9 Hz);13C NMR (125 MHz, CDCl3) δ 140.7,
140.6, 134.2, 129.4, 129.1, 128.0, 127.7, 127.1, 125.7, 125.3, 79.0,
53.9, 40.0; EI HRMS calcd for C15H14OSe 290.0209, found 290.0195
(M+). Anal. Calcd for C15H14OSe: C, 62.29; H, 4.88. Found: C, 62.18;
H, 4.92.
(()-trans-2-O-Diphenylvinylsilyl-1-phenylseleno-2-indanol (7). A
solution of 6 (1.30 g, 4.49 mmol), Et3N (0.95 mL, 6.8 mmol), DMAP
Radical Reaction of 2. Radical precursor 2 (100 mg, 0.20 mmol)
was treated as described above for the reaction of 1 to give 4 (49 mg,
72%), after silica gel column chromatography (Et2O/hexane, 1:50).
(()-trans-2-O-(Methylphenylvinylsily)-1-phenylseleno-2-in-
danol (11). Compound 11 (1.84 g, 89%) was obtained as an oil from
6 (1.37 g, 4.74 mmol) as described above for the synthesis of 7, with
methylphenylvinylchlorosilane (1.3 mL, 7.1 mmol) instead of di-
phenylvinylchlorosilane: 1H NMR (500 MHz, CDCl3) δ 7.48-7.15 (m,
(22) (a) Kulicke, K. J.; Chatgilialoglu, C.; Kopping, B.; Giese, B. HelV.
Chim. Acta 1992, 75, 935-939. (b) Miura, K.; Oshima, K.; Utimoto, K.
Bull. Chem. Soc. Jpn. 1993, 66, 2348-2355. (c) Studer, A. Angew. Chem.,
Int. Ed. 1998, 37, 462-465.
(23) Schiesser, C. H.; Styles M. L. J. Chem. Soc., Perkin Trans. 2 1997,
2335-2340.
(24) Sharpless, K. B.; Lauer, R. F. J. Am. Chem. Soc. 1973, 95, 2697-
2699.

1350 J. Am. Chem. Soc., Vol. 122, No. 7, 2000
Shuto et al.
14 H), 6.15, 6.13 (each dd, each 0.5 H, J ) 19.8, 14.8 Hz), 6.06, 6.05
(each dd, each 0.5 H, J ) 14.8, 4.2 Hz), 5.77, 5.76 (each dd, each 0.5
H, J ) 19.8, 4.2 Hz), 4.69-4.66 (m, 2 H), 3.21-3.16 (m, 1 H), 2.81
(d, 1 H, J ) 16.5 Hz), 0.32, 0.31 (each s, each 1.5H); 13C NMR (100
MHz, CDCl3) δ 141.2, 140.9, 136.0, 135.4, 135.0, 134.61,134.56, 133.9,
129.7, 129.54, 129.52, 128.9, 127.71, 127.65, 127.59, 127.56, 126.7,
125.5, 124.9, 79.8, 54.7, 54.6, 40.69, 40.67, -3.06, -3.09; EI HRMS
calcd for C24H24OSeSi 436.0760, found 436.0756 (M+).
Radical Reaction of 11 for the Preparation of 8a and 8b. A stirred
solution of 11 (2.18 g, 5.01 mmol) and (Bu3Sn)2 (250 µL, 500 µmol)
in benzene (25 mL) was irradiated with a high-pressure mercury lamp
(300 W) at room temperature for 25 h. The solvent was evaporated,
and the residue was purified by silica gel column chromatography (Et2O/
hexane, 1:20) to give a mixture of 8a, 8b, 8c, and 8d (1.23 g, 56%,
8a:8b:8c:8d ) 48:35:9:8, based on its 1H NMR) as an oil. From a part
of the oil (100 mg), 8a (38 mg, 22%), 8b (29 mg, 16%), 8c (5 mg,
3%), and 8d (4 mg, 2%) were obtained in a pure form, respectively,
by preparative HPLC separation (AcOEt/hexane, 1:30).
silica gel column chromatography (Et2O/hexane, 1:50) to give a mixture
of 9a, 9b, 10a, and 10b (63 mg, 44%, 9a:9b:10a:10b ) 22:13:63:2,
1
based on H NMR) as an oil. From the mixture, 9a (10 mg, 7%), 9b
(4 mg, 3%), 10a (28 mg, 20%), and 10b (1 mg, 0.7%) were obtained
in a pure form, respectively, by preparative HPLC separation (AcOEt/
hexane, 1:100).
[2R*,3R*,4R*,5S*]-Indano[1,2-d]-2,3-dimethyl-2-phenyl-1,2-
oxasilolane (9a): 1H NMR (500 MHz, CDCl3) δ 7.56-7.19 (m, 9
H), 5.13 (dt, 1 H, J ) 5.3, 2.2 Hz), 3.34 (t, 1 H, J ) 4.4 Hz), 3.21 (dd,
1 H, J ) 16.6, 2.2 Hz), 3.14 (dd, 1 H, J ) 16.6, 2.0 Hz), 1.65 (dq, 1
H, J ) 7.8, 3.5 Hz), 1.04 (d, 3 H, J ) 7.8 Hz), -0.02 (s, 3 H); NOE
(CDCl3, 400 MHz) irradiated Si-Me, observed H-3 (0.8%); irradiated
H-3, observed H-4 (2.9%); irradiated 3-Me, observed H-4 (2.4%); the
assignments were in agreement with the H-H COSY spectrum;13C
NMR (100 MHz, CDCl3) δ 144.6, 141.3, 135.4, 133.8, 129.7, 127.7,
127.0, 126.5, 125.2, 124.4, 82.2, 57.6, 41.5, 23.8, 15.9, -1.4; EI HRMS
calcd for C18H20OSi 280.1283, found 280.1293 (M+). Anal. Calcd for
C18H20OSi: C, 77.09; H, 7.19. Found: C, 77.00; H, 7.26.
[2S*,3R*,4R*,5S*]-Indano[1,2-d]-2,3-dimethyl-2-phenyl-1,2-
oxasilolane (9b): 1H NMR (500 MHz, CDCl3) δ 7.25-6.96 (m, 9
H), 5.07 (dt, 1 H, J ) 5.4, 2.4 Hz), 3.34 (t, 1 H, J ) 4.5), 3.21 (dd, 1
H, J ) 16.6, 5.4 Hz), 3.15 (dd, 1 H, J ) 16.6, 2.1 Hz), 1.69 (dq, 1 H,
J ) 7.8, 3.5 Hz), 1.33 (d, 3 H, J ) 7.9 Hz), 0.46 (s, 3 H); NOE (CDCl3,
400 MHz) irradiated Si-Me, observed 3-Me (1.0%), H-4 (0.2%), H-5
(0.3%); irradiated H-3, observed H-4 (3.4%); irradiated 3-Me, observed
H-4 (2.9%); the assignments were in agreement with the H-H COSY
spectrum;13C NMR (100 MHz, CDCl3) δ 144.4, 141.2, 136.2, 133.3,
129.3, 127.3, 127.0, 126.5, 125.0, 124.7, 81.9, 57.8, 41.2, 23.6, 15.7,
-4.4; EI HRMS calcd for C18H20OSi 280.1283, found 280.1303 (M+).
[2R*,5R*,6S*]-Indano[1,2-e]-2-methyl-2-phenyl-1,2-oxasilinane
(10a): 1H NMR (500 MHz, CDCl3) δ 7.63-7.17 (m, 9 H), 4.88 (ddd,
1 H, J ) 5.8, 3.7 Hz), 3.19-3.14 (m, 2 H), 3.03 (dd, 1 H, J ) 16.1,
3.6 Hz), 2.26-2.15 (m, 2 H), 1.00-0.89 (m, 2 H), 0.18 (s, 3 H);13C
NMR (125 MHz, CDCl3) δ 143.9, 141.6, 138.1, 133.3, 129.6, 127.9,
126.7, 126.4, 125.0, 123.9, 77.1, 47.3, 41.8, 21.9, 7.7, -1.0; EI HRMS
calcd for C18H20OSi 280.1283, found 280.1296 (M+). Anal. Calcd for
C18H20OSi: C, 77.09; H, 7.19. Found: C, 77.28; H, 7.28.
[2R*,3R*,4R*,5S*]-Indano[1,2-d]-2-methyl-2-phenyl-3-(phenyl-
selenomethyl)-1,2-oxasilolane (8a): 1H NMR (500 MHz, CDCl3) δ
7.58-7.00 (m, 14 H), 5.13 (ddd, 1 H, J ) 5.3, 1.8 Hz), 3.85 (dd, 1 H,
J ) 4.4, 3.4 Hz), 3.22 (dd, 1 H, J ) 16.6, 5.3 Hz), 3.14 (dd, 1 H, J )
16.6 Hz), 2.92 (dd, 1 H, J ) 12.2, 6.8 Hz), 2.89 (dd, 1 H, 12.2, 10.7
Hz), 1.82 (ddd, 1 H, J ) 10.3, 6.8, 3.0 Hz), 0.00 (s, 3 H); NOE (CDCl3,
400 MHz) irradiated Si-Me, observed H-3 (0.8%); irradiated H-3,
observed H-4 (3.3%); the assignments were in agreement with the H-H
COSY spectrum; 13C NMR (100 MHz, CDCl3) δ 144.0, 141.2, 134.4,
133.9, 133.1, 130.1, 130.0, 129.0, 128.0, 127.1, 126.9, 126.6, 125.1,
124.3, 82.1, 55.0, 41.5, 31.1, 30.4, -1.1; EI HRMS calcd for C24H24-
OSeSi 436.0760, found 436.0766 (M+). Anal. Calcd for C24H24OSeSi:
C, 66.19; H, 5.55. Ffound: C, 66.02; H, 5.64.
[2S*,3R*,4R*,5S*]-Indano[1,2-d]-2-methyl-2-phenyl-3-(phenyl-
selenomethyl)-1,2-oxasilolane (8b): 1H NMR (500 MHz, CDCl3) δ
7.52 (m, 14 H), 5.05 (ddd, 1 H, J ) 5.0, 1.7 Hz), 3.83 (m 1 H), 3.38
(dd, 1 H, J ) 11.8, 6.2 Hz), 3.22 (dd, 1 H, J ) 11.7, 4.9 Hz), 3.20-
3.15 (m, 2 H), 1.92 (ddd, 1 H, J ) 11.4, 6.2, 2.4 Hz), 0.57 (s, 3 H);
NOE (CDCl3, 400 MHz) irradiated SiMe, observed 3-CH2Se (0.9%,
0.5%), H-5 (0.3%); irradiated H-3, observed H-4 (3.8%); the assign-
ments were in agreement with the H-H COSY spectrum; 13C NMR
(100 MHz, CDCl3) δ 143.8, 141.2, 135.3, 133.4, 133.2, 129.8, 129.5,
129.1, 127.4, 127.1, 127.0, 126.6, 125.0, 124.7, 81.7, 55.5, 41.3, 31.0,
30.5, -4.1; EI HRMS calcd for C24H24OSeSi 436.0760, found 436.0753
(M+). Anal. Calcd for C24H24OSeSi: C, 66.19; H, 5.55. Found: C,
66.18; H, 5.70.
[2R*,3S*,4R*,5S*]-Indano[1,2-d]-2-methyl-2-phenyl-3-(phenyl-
selenomethyl)-1,2-oxasilolane (8c): 1H NMR (500 MHz, CDCl3) δ
7.58-7.20 (m, 14 H), 4.95 (ddd, 1 H, J ) 4.9, 2.0 Hz), 3.91 (dd, 1 H,
J ) 7.8, 5.0 Hz), 3.62 (dd, 1 H, J ) 11.4, 9.0 Hz), 3.33 (dd, 1 H, J )
11.4, 9.0 Hz), 3.15 (dd, 1 H, J ) 16.7, 4.9 Hz), 3.13 (dd, 1 H, 16.6,
1.1 Hz), 2.19 (ddd, 1 H, J ) 9.1 Hz), 0.04 (s, 3 H); NOE (CDCl3, 400
MHz) irradiated SiMe, observed 3-CH2Se (0.4%, 0.5%); irradiated H-3,
observed H-4 (10.4%), H-6 (2.1%); the assignments were in agreement
with the H-H COSY spectrum; 13C NMR (100 MHz, CDCl3) δ 142.8,
141.3, 136.0, 133.7, 132.7, 130.1, 130.0, 129.0, 127.8, 127.2, 126.9,
126.4, 126.2, 125.5, 82.6, 52.2, 41.7, 31.2, 26.4, -3.9.
[2S*,3S*,4R*,5S*]-Indano[1,2-d]-2-methyl-2-phenyl-3-(phenyl-
selenomethyl)-1,2-oxasilolane (8d): 1H NMR (500 MHz, CDCl3) δ
7.48-6.87 (m, 14 H), 4.89 (ddd, 1 H, J ) 4.7, 4.3 Hz), 3.86 (dd, 1 H,
J ) 7.9, 4.5 Hz), 3.53 (dd, 1 H, J ) 11.2, 6.5 Hz), 3.22 (d, 1 H, J )
16.6 Hz), 3.17 (dd, 1 H, J ) 16.6, 4.7 Hz), 3.10 (dd, 1 H, J ) 10.8
Hz), 2.18 (ddd, 1 H, J ) 9.6, 7.8 Hz), 0.57 (s, 3 H); NOE (CDCl3, 400
MHz) irradiated Si-Me, observed H-3 (1.3%); irradiated H-3, observed
H-4 (10.0%), H-5 (2.8%); the assignments were in agreement with the
H-H COSY spectrum; 13C NMR (100 MHz, CDCl3) δ 143.0, 141.5,
134.1, 133.8, 132.7, 130.0, 129.5, 129.1, 127.3, 127.2, 126.9, 126.5,
126.3, 125.4, 82.5, 52.9, 41.4, 32.8, 26.7, -1.0.
[2S*,5R*,6S*]-Indano[1,2-e]-2-methyl-2-phenyl-1,2-oxasilinane
(10b): 1H NMR (500 MHz, CDCl3) δ 7.35-7.16 (m, 9 H), 4.93 (ddd,
1 H, J ) 6.0, 4.5 Hz), 3.21 (dd, 1 H, J ) 12.3, 6.0 Hz), 3.14 (dd, 1 H,
J ) 16.0, 6.0 Hz), 2.98 (dd, 1 H, J ) 16.0, 4.3 Hz), 2.22-2.18 (m, 2
H), 1.04 (ddd, 1 H, J ) 14.3, 8.3, 6.0 Hz), 0.77 (ddd, 1 H, J ) 14.6,
8.1, 5.6 Hz), 0.43 (s, 3 H);13C NMR (125 MHz, CDCl3) δ 143.8, 141.5,
133.3, 129.5, 127.7, 126.7, 126.5, 125.0, 124.0, 77.2, 47.2, 41.4, 22.2,
8.3, -1.7; EI HRMS calcd for C18H20OSi 280.1283, found 280.1270
(M+). Anal. Calcd for C18H20OSi: C, 77.09; H, 7.19. Found: C, 77.08;
H, 7.34
1
Radical Reactions of 8a and 8b and H NMR Analysis of the
Products. To a solution of 8a or 8b (22 mg, 0.051 mmol) in benzene
(5 mL) at 80 °C was added a solution of Bu3SnH (60 µL, 0.22 mmol)
and AIBN (5 mg, 0.03 mmol) in benzene (5 mL) slowly over a 4 h
period. The solvent was evaporated, and the residue was dissolved in
1
CDCl3 and its H NMR spectrum (500 MHz) was measured.
(()-Ethyl trans-4-(tert-Butyldiphenylsilyloxy)-2-butenoate (16).
A mixture of 15 (10.0 g, 69.4 mmol) and BH3 (1 M in THF, 70 mL,
70 mmol) in THF (30 mL) was stirred at room temperature for 12 h.
After aqueous AcOH (50%, 2 mL) was added, the solvent was
evaporated, and the residue was partitioned between AcOEt and aqueous
NaHCO3 (saturated). The organic layer was washed with brine, dried
(Na2SO4), and evaporated to give an oil. A mixture of the resulting
oil, imidazole (6.81 g, 100 mmol), and TBDPSCl (13.0 mL, 50.0 mmol)
in DMF (100 mL) was stirred at room temperature for 3 h. Et2O and
water were added, and the resulting mixture was partitioned. The
organic layer was washed with brine, dried (Na2SO4), and evaporated.
The residue was purified by silica gel column chromatography (Et2O/
hexane, 1:50) to give 16 (7.73 g, 30%) as an oil: 1H NMR (270 MHz,
CDCl3) δ 7.73-7.35 (m, 10 H), 6.98 (ddd, 1 H, J ) 15.8, 3.3, 3.3
Hz), 6.26 (ddd, 1 H, J ) 15.8, 2.6, 2.0 Hz), 4.34 (dd, 2 H, J ) 3.3, 2.0
Hz), 4.22 (q, 2 H, J ) 7.3), 1.31 (t, 3 H, J ) 7.3 Hz), 1.08 (s, 9 H).
(()-trans-4-(tert-Butyldiphenylsilyloxy)-2-butenol (17). To a solu-
tion of 16 (2.06 g, 5.59 mmol) in THF (20 mL) was added slowly
Radical Reactions of 8a and Purification of the Products 9a, 9b,
10a, and 10b. To a solution of 8a (220 mg, 0.51 mmol) in benzene
(50 mL) at 80 °C was added a solution of Bu3SnH (200 µL, 0.74 mmol)
and AIBN (50 mg, 0.30 mmol) in benzene (10 mL) slowly over a 4 h
period. The solvent was evaporated, and the residue was purified by

PentaValent Silicon-Bridging Radical
J. Am. Chem. Soc., Vol. 122, No. 7, 2000 1351
DIBAH (0.95 M in hexane, 14.1 mL, 13.4 mmol) at -78 °C, and the
resulting mixture was stirred at the same temperature for 10 min. Et2O
and water were added, and the resulting mixture was partitioned. The
organic layer was washed with 1 N HCl and brine, dried (Na2SO4),
and evaporated. The residue was purified by silica gel column
chromatography (AcOEt/ hexane ) 1:5) to give 17 (1.80 g, 99%) as
an oil: 1H NMR (270 MHz, CDCl3) δ 7.69-7.34 (m, 10 H), 5.91 (ddd,
1 H, J ) 15.2, 5.3, 1.3 Hz), 5.79 (dt, 1 H, J ) 15.2, 4.6 Hz), 4.23-
4.21 (m 2 H), 4.14 (m, 2 H), 1.26 (t, 1 H, J ) 6.0 Hz), 1.06 (s, 9 H);
FAB HRMS (NaI) calcd for C20H26NaO2Si 349.1600, found 349.1603
(MNa+).
from 18D (75 mg, 0.33 mmol) as described above for the synthesis of
7: 1H NMR (400 MHz, CDCl3) δ 7.59-7.21 (m, 15 H), 6.44 (dd, 1
H, J ) 20.3, 14.9 Hz), 6.26 (dd, 1 H, J ) 14.9, 3.9 Hz), 5.87 (dd, 1 H,
J ) 20.3, 3.7 Hz), 5.81 (d, 1 H), 5.52 (dt, 1 H, J ) 15.1, 5.4 Hz), 4.21
(dd, 2 H, J ) 5.4, 1.5 Hz); 13C NMR (100 MHz, CDCl3) δ 137.0,
134.9, 133.9, 133.4, 133.4, 133.2, 131.2, 129.9, 128.8, 127.8, 127.0,
126.8, 63.7; EI HRMS calcd for C24H22D2OSeSi 438.0885, found
438.0897 (M+).
(()-5-Benzoyloxy-3-(benzoyloxymethyl)-1-pentene (20). To a re-
fluxing solution of 12 (118 mg, 0.27 mmol) in benzene (30 mL) was
added a solution of Bu3SnH (110 µL, 0.41 mmol) and AIBN (23 mg,
0.14 mmol) in benzene (5 mL) slowly over 4 h. The solvent was
evaporated to give an oil. A mixture of the obtained oil (430 mg, 0.50
mmol), aqueous H2O2 (30%, 570 µL, 5.0 mmol), KF (291 mg, 5.0
mmol), and KHCO3 (125 mg, 1.3 mmol) in MeOH/THF (1:1, 2 mL)
was stirred at room temperature for 26 h. After aqueous Na2S2O3 (1
M, 5 mL) was added, the resulting insoluble materials were filtered
off, and the filtrate was evaporated. A mixture of the residue and BzCl
(600 µL, 5.2 mmol) in pyridine (5 mL) was stirred at room temperature
for 1 h. Et2O and water were added, and the resulting mixture was
partitioned. The organic layer was washed with brine, dried (Na2SO4),
and evaporated. The residue was purified by silica gel column
chromatography (AcOEt/hexane, 1:19) to give crude 20 (68 mg), which
was further purified by HPLC (AcOEt/hexane, 1:19) to give pure 20
(27 mg, 17%) as an oil: 1H NMR (400 MHz, CDCl3) δ 8.05-7.42
(m, 10 H), 5.77 (ddd, 1 H, J ) 17.6, 10.7, 8.8 Hz), 5.22 (d, 1 H, J )
16.6 Hz), 5.19 (dd, 1 H, J ) 10.7, 2.0 Hz), 4.45 (ddd, 1 H, J ) 11.6,
5.9 Hz), 4.39-4.29 (m, 3 H), 2.82-2.74 (m, 1 H), 2.14-2.06 (m, 1
H), 1.91-1.83 (m, 1 H); 13C NMR (100 MHz, CDCl3) δ 166.4, 166.3,
137.4, 132.9, 132.8, 130.14, 130.07, 129.48, 129.47, 128.30, 128.29,
117.7, 67.4, 62.7, 40.5, 30.2; EI HRMS calcd for C20H20O4 324.1361,
found 324.1346 (M+).
(()-5-Benzoyloxy-3-(benzoyloxymethyl)-[4-2H]pentene (21). Com-
pound 21 (11 mg, 17%) was obtained as an oil from 12 (97 mg, 0.20
mmol) as described above for the synthesis of 20, with Bu3SnD instead
of Bu3SnH:1H NMR (400 MHz, CDCl3) δ 8.05-7.42 (m, 10 H), 5.77
(ddd, 1 H, J ) 17.1, 10.3, 8.5 Hz), 5.22 (ddd, 1 H, J ) 17.1, 1.5, 1.0
Hz), 5.19 (dd, 1 H, J ) 10.3, 1.5 Hz), 4.44 (dd, 1 H, J ) 11.2, 6.8
Hz), 4.38-4.29 (m, 3 H), 2.81-2.74 (m, 1 H), 2.27-2.05 (m, 1 H),
1.91-1.82 (m, 0.28H); EI HRMS calcd for C20H19DO4 325.1423, found
325.1450 (M+).
(()-5-Benzoyl-3-(benzoyloxymethyl)-[1,1-2H2]-pentene (22D). Com-
pound 22D (13 mg, 19%) was obtained as an oil from 12D (93 mg,
0.21 mmol) as described above for the synthesis of 20: 1H NMR (400
MHz, CDCl3) δ 8.80 (m, 10 H), 5.76 (d, 1 H, J ) 8.5 Hz), 4.44 (ddd,
1 H, J ) 11.0, 6.6, 5.6 Hz), 4.39-4.29 (m, 3 H), 2.82-2.74 (m, 1 H),
2.14-2.06 (m 1H), 1.90-1-81 (m, 1 H);13C NMR (100 MHz, CDCl3)
δ 166.3, 166.2, 137.0, 132.83, 132.77, 130.0, 129.9, 129.39, 129.37,
128.21, 128.20, 117.1 (m), 67.3, 62.6, 40.3, 30; EI HRMS calcd for
C20H18D2O4 326.1485, found 326.1502 (M+).
(()-trans-4-tert-Butyldiphenylsilyloxy[1,1-2H2]-2-butenol (17D).
A solution of LiAl(OMe)3D was prepared by adding MeOH (1.34 mL,
33.1 mmol) to a solution of LiAlD4 (1.0 M in THF, 11.0 mL, 11.0
mmol) at 0 °C. To a solution of 17 (1.69 g, 4.59 mmol) in THF (20
mL) was added slowly the prepared LiAl(OMe)3D solution at -78 °C,
and the resulting mixture was stirred at 0 °C for 10 min. Et2O and
water were added, and the resulting mixture was partitioned. The
organic layer was washed with 1 N HCl and brine, dried (Na2SO4),
and evaporated. The residue was purified by silica gel column
chromatography (AcOEt/hexane ) 1:5) to give 17D (1.10 g, 3.35 mmol,
73%) as an oil: 1H NMR (400 MHz, CDCl3) δ 7.69-7.35 (m, 10H),
5.90 (d, 1 H, J ) 15.6 Hz), 5.78 (dt, 1 H, J ) 15.4, 4.4 Hz), 4.22 (dd,
2 H, J ) 4.4, 1.7 Hz), 1.73 (br s, 1 H), 1.06 (s, 9 H);13C NMR (100
MHz, CDCl3) δ 135.4, 133.5, 130.7, 129.6, 128.6, 127.6, 63.7, 62.5
(m), 26.9, 19.3; FAB HRMS (NaI) calcd for C20H24D2NaO2Si 351.1723,
found 351.1727 (MNa+).
(()-trans-4-Phenylseleno-2-butenol (18). To a solution of PhSeCl
(0.58 g, 3.0 mmol) in THF (5 mL) was added slowly a solution of
TMSCN (0.48 mL, 3.6 mmol) in THF (5 mL) at room temperature,
and the resulting mixture was evaporated to give PhSeCN as a residue.25
To a solution of 17 (356 mg, 1.09 mmol) and the prepared PhSeCN in
THF (6 mL) was added Bu3P (0.75 mL, 3.0 mmol) slowly. The resulting
mixture was evaporated, and the residue was purified by silica gel
column chromatography (Et2O/hexane ) 1:50) to give TBDPS ether
of trans-4-phenylseleno-2-butenol as an oil. A mixture of the obtained
oil and TBAF (1 M in THF, 1.0 mL, 1.0 mmol) in THF (3 mL) was
stirred at room temperature for 2 h. The solvent was evaporated, and
the residue was purified by silica gel column chromatography (AcOEt/
hexane, 1:4) to give 18 (129 mg, 52%) as an oil:.1H NMR (270 MHz,
CDCl3) δ 7.56-7.20 (m, 5 H), 5.88-5.77 (m, 1 H, J ) 15.2, 4.9, 1.3
Hz), 5.53 (dt, 1 H, J ) 15.2, 5.9 Hz), 4.03 (t, 2 H, J ) 5.9 Hz), 3.52
(d, 2 H, J ) 7.9), 1.12 (t, 1 H, J ) 5.9 Hz);13C NMR (67.8 MHz,
CDCl3) δ 133.7, 131.6, 128.9, 128.2, 127.3, 62.9, 29.3; EI HRMS calcd
for C10H12OSe 228.0063, found 228.0031 (M+). Anal. Calcd for C10H12-
OSe: C, 52.87; H, 5.32. Found: C, 52.81; H, 5.37.
(()-trans-4-Phenylseleno[4,4-2H2]-2-butenol (18D). Compound
18D (131 mg, 26%) was obtained as an oil from 17D (712 mg, 2.17
mmol) as described above for the synthesis of 18: 1H NMR (270 MHz,
CDCl3) δ 7.51-7.25 (m, 5 H), 5.82 (d, 1 H, J ) 15.2 Hz), 5.53 (dt, 1
H, J ) 15.2, 5.9 Hz), 4.05-4.01 (m, 2 H), 1.15 (t, 1 H, J ) 5.9 Hz);13C
NMR (100 MHz, CDCl3) δ 133.6, 131.5, 128.8, 127.9, 127.2, 62.9; EI
HRMS calcd for C10H10D2OSe 230.0177, found 230.0185 (M+)
(()-trans-1-Diphenylvinylsilyloxy-4-phenylseleno-2-butene (12).
Compound 12 (287 mg, 94%) was obtained as an oil from 18 (159
mg, 0.700 mmol) as described above for the synthesis of 7: 1H NMR
(400 MHz, CDCl3) δ 7.59-7.22 (m, 15 H), 6.44 (dd, 1 H, J ) 20.3,
14.9 Hz), 6.26 (dd, 1 H, J ) 14.9, 3.9 Hz), 5.87 (dd, 1 H, J ) 20.3,
3.9 Hz), 5.87-5.80 (m 1 H), 5.52 (dt, 1 H, J ) 15.1, 5.2 Hz), 4.20
(dd, 2 H, J ) 5.1, 0.7 Hz), 3.51 (d, 2 H, J ) 7.8 Hz);13C NMR (100
MHz, CDCl3) δ 137.1, 134.9, 133.9, 133.4, 133.2, 131.2, 130.0, 128.8,
127.8, 127.0, 126.9, 63.7, 29.5; EI HRMS calcd for C24H24OSeSi
436.0761, found 436.0743 (M+). Anal. Calcd for C24H24OSeSi: C,
66.19; H, 5.55. Found: C, 66.26; H, 5.73.
Computations. Initial starting guess geometries were generated using
MM2* in MacroModel version 5.0 software. Geometry optimizations
and single-point energy calculations were performed using PM3 and
UHF/STO-3G in Spartan version 4.0 software.
Acknowledgment. This investigation was supported in part
by a Grant from the Ministry of Education, Science, Sports,
and Culture of Japan. We are grateful to Miss M. Kanazaki
and Drs. N. Nishizono and K. Takenuki for their helpful
discussions.
Supporting Information Available: 1H NMR spectra of
2, 11, 8c, 8d, 9b, 12D, 16, 17, 17D, 18D, 21, and 22D (PDF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
(()-trans-1-Diphenylvinylsilyloxy-4-phenylseleno[4,4-2H2]-2-
butene (12D). Compound 12D (95 mg, 67%) was obtained as an oil
(25) Tomoda, S.; Takeuchi, Y.; Nomura, Y. Chem. Lett. 1981, 1069-
1070.
JA993239S