Tandem Base-Promoted Ring-Opening/Brook Rearrangement/Allylic Alkylation of O-Silyl Cyanohydrins of β-Silyl-α,β-epoxyaldehyde: Scope and Mechanism

Article abstract of DOI:10.1021/jo0352934

Metalated O-silyl cyanohydrins of β-silyl-α,β -epoxyaldehyde have been found to serve as functionalized homoenolate equivalents by a tandem sequence involving base-promoted ring opening of the epoxide, Brook rearrangement, and alkylation of the resulting

Full text of DOI:10.1021/jo0352934

Ta n d em Ba se-P r om oted Rin g-Op en in g/Br ook Rea r r a n gem en t/
Allylic Alk yla tion of O-Silyl Cya n oh yd r in s of
â-Silyl-r,â-ep oxya ld eh yd e: Scop e a n d Mech a n ism
Michiko Sasaki,† Eiji Kawanishi,† Yoshio Nakai,† Tatsuya Matsumoto,†
Kentaro Yamaguchi,‡ and Kei Takeda*,†
Department of Synthetic Organic Chemistry, Graduate School of Medical Sciences, Hiroshima University,
1-2-3 Kasumi, Minami-Ku, Hiroshima 734-8551, J apan, and Chemical Analysis Center, Chiba
University, 1-33 Yayoi-Chou, Inage-Ku, Chiba 263-8522, J apan
takedak@hiroshima-u.ac.jp
Received September 3, 2003
Metalated O-silyl cyanohydrins of â-silyl-R,â-epoxyaldehyde have been found to serve as function-
alized homoenolate equivalents by a tandem sequence involving base-promoted ring opening of
the epoxide, Brook rearrangement, and alkylation of the resulting allylic anion. On the basis of
mechanistic studies involving competitive experiments using the diastereomeric cyanohydrins, we
propose a reaction pathway involving a silicate intermediate 36 formed by a concerted process via
an anti-opening of the epoxide followed by the formation of an O-Si bond.
SCHEME 1
In tr od u ction
About 10 years ago, we reported a new approach to
the synthesis of highly functionalized cyclopentenol 5
using a [3 + 2] annulation that involves a combination
of (â-(phenylthio)acryloyl)silane 1 as the three-carbon
unit and lithium enolate 2 of alkyl methyl ketone as the
two-carbon unit, which relies on the formation of delo-
calized allylic anion 4 via a 1,2-anionic rearrangement
of silicon (Brook rearrangement)1 in the 1,2-adduct 3
followed by internal carbonyl attack by the anion (Scheme
1).2
SCHEME 2
During the course of an extension of this methodology
to asymmetric versions, we had occasion to examine the
reaction of â-silyl-R,â-epoxyacylsilanes 6 with enolate of
alkyl methyl ketone 7. Although we expected the forma-
tion of cyclopentenol derivatives 12 via a 2-fold Brook
rearrangement followed by an intramolecular aldol reac-
tion, the major products were uncyclized enol silyl ether
* Corresponding author.
† Hiroshima University.
‡ Chiba University.
(1) For reviews on the Brook rearrangement, see: (a) Brook, M. A.
Silicon in Organic, Organometallic, and Polymer Chemistry; J ohn
Wiley & Sons: New York, 2000. (b) Brook, A. G.; Bassindale, A. R. In
Rearrangements in Ground and Excited States; de Mayo, P., Ed.;
Academic Press: New York, 1980; pp 149-221. (c) Brook, A. G. Acc.
Chem. Res. 1974, 7, 77-84. For the use of the Brook rearrangement
in tandem bond formation strategies, see: (d) Moser, W. H. Tetrahe-
dron 2001, 57, 2065-2084. Also see: (e) Ricci, A.; Degl’Innocenti, A.
Synthesis 1989, 647-660. (f) Bulman Page, P. C.; Klair, S. S.;
Rosenthal, S. Chem. Soc. Rev. 1990, 19, 147-195. (g) Qi, H.; Curran,
D. P. In Comprehensive Organic Functional Group Transformations;
Katritzky, A. R., Meth-Cohn, O., Rees, C. W., Moody, C. J ., Eds.;
Pergamon: Oxford, 1995; pp 409-431. (h) Cirillo, P. F.; Panek, J . S.
Org. Prep. Proc. Int. 1992, 24, 553-582. (i) Patrocinio, A. F.; Moran,
P. J . S. J . Braz. Chem. Soc. 2001, 12, 7-31.
10 and cyclopropanediol derivative 11, the latter being
formed as a result of an intramolecular allylic attack on
the carbonyl group by a siloxy carbanion generated by
the second Brook rearrangement (8 f 9) (Scheme 2).
On the other hand, we recently found that the reaction
of acryloylsilane 13 with KCN in the presence of MeI and
crown ether afforded a â-methylation product 14 via a
Brook rearrangement (Scheme 3).3
(2) (a) Takeda, K.; Fujisawa, M.; Makino, T.; Yoshii E.; Yamaguchi,
K. J . Am. Chem. Soc. 1993, 115, 9351-9352. (b) Takeda, K.; Ohtani,
Y.; Ando, E.; Fujimoto, K.; Yoshii, E.; Koizumi, T. Chem. Lett. 1998,
1157-1158. (c) Takeda, K.; Yamowaki, K.; Hatakeyama, N. J . Org.
Chem. 2002, 67, 1786-1794.
(3) Takeda, K.; Ohnishi, Y. Tetrahedron Lett. 2000, 41, 4169-4172.
10.1021/jo0352934 CCC: $25.00 © 2003 American Chemical Society
Published on Web 10/29/2003
9330
J . Org. Chem. 2003, 68, 9330-9339

O-Silyl Cyanohydrins of â-Silyl-R,â-epoxyaldehyde
TABLE 1.
SCHEME 3
SCHEME 4
27, yield (%) (E/Z)
RX
from 21a
from 21b
MeI
EtI
i-PrI
PhCH2Br
CH2dCHCH2Br
82 (2.5)
76 (2.9)
58 (2.8)a
86 (2.7)
83 (3.4)
84 (22.0)
74 (28.0)
74 (31.0)
98 (47.0)
87 (40.0)
a
12% yield of 27 (R ) H) was obtained.
CHART 1
via a tandem sequence (23 f 24 f 25 f 26), were
obtained in 82% and 84%, respectively (Table 1). Similar
results were obtained with other alkylating agents.
It is particularly noteworthy that (1) the reactions were
completed within 5 min at -80 °C and alkylation
products of 23, 24, or 25 were not detected and (2) E/ Z
ratios of 27 obtained from the two diastereomers were
markedly different.
These results indicated the possibility of the reaction
of an epoxysilane 15 bearing an anion-stabilizing electron-
withdrawing group at the R-position with an amide base
in the presence of an electrophile. If a tandem process
that involves a base-promoted isomerization of the ep-
oxide (16 f 17), Brook rearrangement (17 f 18), and a
reaction of the resulting allylic anion with an electrophile
(18 f 19 f 20) proceeds well, the epoxysilane 15 would
function as a homoenolate equivalent4 with synthetically
useful functionality (Scheme 4). Although base-promoted
isomerization of epoxides to allylic alcohols has been well
documented,5 the only examples, to the best of our
knowledge, of a tandem sequence involving a ring open-
ing of expoxide followed by Brook rearrangement are two
examples reported by Gonza´lez-Nogai and co-workers,6
who succeeded in the generation of enol silyl ethers via
cleavage of R,â-epoxysilanes with heteroatom nucleo-
philes.
To obtain a better understanding of the reaction
pathway, we first examined the effects of a cation and
base on the E/Z selectivity of the reaction using lithium
hexamethyldisilazide (LHMDS), sodium hexamethyld-
isilazide (NHMDS), and potassium hexamethyldisilazide
(KHMDS) (Table 2). While the use of lithium hexameth-
yldisilazide (LHMDS, 1.0 M in THF) resulted in lower
yields but improvement in E/Z ratios with 21a , compa-
rable yields of 27 were obtained with potassium hexa-
methyldisilazide (KHMDS, 0.5 M in toluene). It is notable
that the increased formation of the Z-derivative with 21a
was observed in the case of the latter base. The best
results, in terms of yield and E/ Z selectivity, were
obtained with sodium hexamethyldisilazide (NHMDS, 1.0
M in THF), allowing the formation of (E)-27 in excellent
yields.
First, we chose O-silyl cyanohydrins of R,â-epoxyalde-
hydes7 21 bearing a nitrile group as the electron-
withdrawing group as a substrate and investigated its
reaction with a base in the presence of alkylating agents
(Chart 1).8
Since increased formation of the (Z)-isomer was ob-
served in the case of LDA and KHMDS, which contain
hexane and toluene, respectively, we decided to investi-
gate the relationship between polarity of the solvent12
and E/Z geometry of 27. The results are summarized in
Table 3. The use of less polar solvents resulted in a
substantial enhancement of Z-selectivity, suggesting that
the nature of the solvent plays an important role in
determination of E/Z selectivity in the reaction.
In this paper, we describe in full detail the alkylation
reaction communicated previously in a preliminary form.9
Resu lts a n d Discu ssion
Epoxysilanes 21a and 21b were obtained as a diaster-
eomeric mixture by the reaction of TBSCN with epoxy-
aldehyde 22, which was derived from 3-(1-ethoxyethoxy)-
propyne10 via the sequence shown in Scheme 5.11 The
relative stereochemistry in 21a and 21b was determined
on the basis of X-ray analysis of 21b.
When 21a and 21b were treated with LDA (1.1 equiv)
in THF at -80 °C in the presence of MeI (1.2 equiv) for
5 min, R-methylated cyanohydrins 27, products formed
One possible explanation for these results is that the
internal chelated structures 28 and/or 29, which involve
Li-O13 and five-coordinated silicon species,14 respectively,
are immediate precursors to (Z)-27 by alkylation. Thus,
the poorer donor solvents help the intramolecular coor-
dination of the siloxy group, therefore giving rise to the
J . Org. Chem, Vol. 68, No. 24, 2003 9331

Sasaki et al.
TABLE 2.
27, yield (%) (E/Z)
from 21a
from 21b
RX
LHMDSa
KHMDSb
NHMDSa
LHMDSa
KHMDSb
NHMDSa
MeI
EtI
44 (23.0)
24 (16.0)
15 (14.0)
56 (30.0)
45 (31.0)
84 (0.9)
76 (0.7)
42 (2.1)
83 (0.8)
80 (1.1)
96 (40.0)
90 (42.0)
80 (62.0)
98 (65.0)
91 (39.0)
83 (31.0)
64 (28.0)
44 (37.0)
75 (82.0)
80 (89.0)
87 (9.7)
81 (16.0)
73 (83.0)
88 (13.0)
83 (14.0)
98 (E)
89 (42.0)
89 (75.0)
99 (67.0)
92 (41.0)
i-PrI
PhCH2Br
CH2dCHCH2Br
a
b
1.0 M solution in THF was used. 0.5 M solution in toluene was used (THF/toluene ) ca. 2:3).
TABLE 4.
SCHEME 5
LDA
LHMDS
HMPA yield (%) E/Z SM (%) yield (%) E/Z SM (%)
21a
21a
21b
21b
(-)
(+)
(-)
(+)
82
61
84
85
2.5
28.0
22.0
E
44
87
83
81
23.0
19.0
31.0
E
40
26
8
NHMDS
KHMDS
HMPA yield (%) E/Z SM (%) yield (%) E/Z SM (%)
TABLE 3.
21a
21a
21b
21b
(-)
(+)
(-)
(+)
88
84
92
91
55.0
E
47.0
E
84
92
87
84
0.9
15.0
9.7
E
solvent
hexane
diastereomer
yield (%)
E/Z
The effect of the countercation on the E/Z ratio was
then examined using the same solvent system, anticipat-
ing that if chelation such as that of 28 is responsible for
the formation of the (Z)-isomer, the ratio of the (Z)-isomer
would increase as the cation becomes less ionic. The
results obtained using LHMDS and NHMDS are shown
21a
21b
21a
21b
21a
21b
21a
21b
98
78
86
83
84
77
85
84
1.5
6.0
1.0
24.0
1.9
28.0
28.0
52.0
toluene
Et2O
THF
(8) For a base-promoted ring opening of functionalized epoxides, see
(epoxy nitriles): (a) Fleming, F. F.; Wang, Q.; Steward, O. W. J . Org.
Chem. 2001, 66, 2171-2174. (Epoxy amides): (b) Brooks, P. B. Marson,
C. M. Tetrahedron 1998, 54, 9613-9622. (Epoxy esters): (c) Mohr, P.;
Ro¨sslein, L.; Tamm, C. Tetrahedron Lett. 1989, 30, 2513-2516. (d)
Cory, R. M. Ritchie, B. M.; Shrier, A. M. Tetrahedron Lett. 1990, 31,
6789-6792.
formation of (Z)-27. This is supported by the results
showing that in all cases except LHMDS the (Z)-selectiv-
ity in the methylation of 21a ,b with the bases was
significantly lowered upon addition of HMPA, which can
disrupt the chelated structure by solvating the cations
(Table 4).
(9) Takeda, K.; Kawanishi, E.; Sasaki, M.; Takahashi, Y.; Yamagu-
chi, K. Org. Lett. 2002, 4, 1511-1514.
(10) Takeda, K.; Nakajima, A.; Takeda, M.; Yoshii, E.; Zhang, J .;
Boeckman, R. K., J r. Org. Synth. 1999, 76, 199-213.
(11) For preparation of trimethylsilyl derivatives of 7 (R ) Me),
see: Urabe, H.; Matsuka, T.; Sato F. Tetrahedron Lett. 1992, 33, 4179-
4182.
(4) For reviews on homoenolate equivalents, see: (a) Ahlbrecht, H.;
Beyer, U. Synthesis 1999, 365-390. (b) Kuwajima, I.; Nakamura, E.
Top. Curr. Chem. 1990, 155, 1. (c) Hoppe, D. Angew. Chem., Int. Ed.
Engl. 1984, 23, 932-948. (d) Werstiuk, N. H. Tetrahedron 1983, 39,
205-268. (e) Werstiuk, N. H. In Umpoled Synthons; Hase, T. A., Ed.;
J ohn Wiley & Sons: New York, 1987; p 173. Also see: (f) Debal, A.;
Cuvigny, T.; Larcheveˆque, M. Tetrahedron Lett. 1977, 3187-3190.
(5) For reviews on base-promoted isomerization of epoxides, see: (a)
Crandall, J . K.; Apparu, M. Org. React. 1983, 29, 345-443. (b) Satoh,
T. Chem. Rev. 1996, 96, 3303-3325.
(6) (a) Cuadrado, P.; Gonza´lez-Nogai, A. M. Tetrahedron Lett. 2000,
41, 1111-1114. (b) Cuadrado, P.; Gonza´lez-Nogai, A. M. Tetrahedron
Lett. 1997, 38, 8117-8120.
(7) For alkylation of O-silyl cyanohydrins, see: (a) Arseniyadis, S.;
Kyler, K. S.; Watt, D. S. Org. React. 1984, 31, 1-364. (b) Albright, J .
D. Tetrahedron 1983, 39, 3207-3233.
(12) Riddick, J .; Bunger, W. B.; Sakano, T. K. Organic Solvents:
Physical Properties and Methods of Purification, 4th ed.; Wiley-
Interscience: New York, 1986; Technique of Chemistry, Vol. II.
(13) (a) Enda, J .; Kuwajima, I. J . Am. Chem. Soc. 1985, 107, 5495-
5501. (b) Kato, M,; Mori, A.; Oshino, H.; Enda, J .; Kobayashi, K.;
Kuwajima, I. J . Am. Chem. Soc. 1984, 106, 1773-1778.
(14) For reviews of hypervalent silicon compounds, see: (a) ref 1a,
Chapter 4. (b) Kost, D.; Kalikhman, I. Hypervalent Silicon Compounds.
In The Chemistry of Organic Silicon Compounds; Rappoport, Z.,
Apeloig, Y., Eds.; Wiley: Chichester, UK, 1998; Vol. 2, Chapter 23. (c)
Chuit, C.; Corriu, R. J . P.; Reye, C.; Young, J . C. Chem. Rev. 1993, 93,
1371-1448. (d) Deerenberg, S.; Schakel, M.; de Keijzer, A. H. J . F.;
Kranenburg, M.; Lutz, M.; Spek, A. L.; Lammertsma, K. Chem.
Commun. 2002, 348-349.
9332 J . Org. Chem., Vol. 68, No. 24, 2003

O-Silyl Cyanohydrins of â-Silyl-R,â-epoxyaldehyde
TABLE 5.
TABLE 6.
entry
base
diastereomer
yield (%)
E/Z
21
base
LDA
yield (%)
E/Z
sm
3
1
2
3
4
LiN(SiMe3)2
LiN(SiMe3)2
NaN(SiMe3)2
NaN(SiMe3)2
21a
21b
21a
21b
19
18
86
97
0.4
34.0
1.4
67
63
33a
33b
33a
33b
33a
33b
33a
33b
80
88
15
13.0
E
E
LDA
LHMDS
LHMDS
NHMDS
NHMDS
KHMDS
KHMDS
66
16.0
70
93
89
86
79.0
E
2.2
66.0
CHART 2
TABLE 7.
in Table 5. A trend consistent with the expectation of
greater (Z)-selectivity with lithium amide was observed
in the case of 21a (entries 1 and 3) in contrast to the
case of 21b, for which the opposite trend was observed
(entries 2 and 4). Although the reason for the latter case
is not clear, the E/Z selectivity in the reaction does not
seem to strongly depend on the countercation.15
27 (from 21a )
27 (from 21b)
yield (%) E/Z
base
yield (%)
E/Z
Next, we attempted to determine whether a five-
coordinate silicon speices such as 29 is involved in the
reaction pathway by using dimethyl(phenyl)silyl deriva-
tives 30a ,b as a substrate, anticipating increased forma-
tion of the (Z)-isomer owing to stabilization of the silicate
by the phenyl group on the silicon atom (Chart 2).16 When
30a ,b, prepared in the same manner as that for 21, were
treated with NHMDS and MeI, aldehyde derivative 32,
a hydrolyzed product of 31, was obtained after purifica-
tion by column chromatography.
When the same reaction was carried out using tert-
butyldiphenylsilyl derivatives 33a ,b, more stable toward
hydrolysis, the formation of (E)-34 was, unexpectedly,
increased with all bases resulted, presumably because
of increasing steric repulsion between the tert-butyl-
diphenylsilyl group and the cynohydrin moiety (Table 6).
Consequently, no evidence of the participation of silicate
species 29 could be obtained in these experiments.
Although the intermediacy of chelated species such as
28 is consistent with the solvent effect on the E/ Z ratio
of the product, it cannot explain the fact that 27 was
obtained from the two diastereomers 21a ,b in different
E/ Z ratios in the cases of LDA and KHMDS, unless it is
assumed that (E)- and (Z)-26 are formed from the
starting cyanohydrins at a different extent depending on
the stereochemistry of the starting cyanohydrins and on
the solvent polarity and that they collapse to products
LDA
76
36a
86
78
2.9
39.0
38.0
0.3
69
68
85
66
38.0
54.0
124.0
12.0
LHMDS
NHMDS
KHMDS
a
56% of 21a was recovered.
before an establishment of the equilibrium. To determine
whether alkylation occurs before equilibration, methyl
iodide was added after addition of the base. The E/ Z
ratios changed only slightly (Table 7), suggesting two
possibilities.
One possibility is that 26 is an immediate precursor
for the alkylation, isomerization between (E)- and (Z)-
26 does not occur under the conditions, and the solvent
effect for E/ Z selectivity operates in earlier steps.
Consequently, the intermediacy of the chelated species
28 can be ruled out. The other possibility is that 26 is
not an immediate precursor and the alkylation can occur
in a concerted manner from intermediates that preserve
stereochemical information originating from the starting
materials and form in a ratio depending on the solvent
polarity. The former possibility can be ruled out if the
E/ Z isomerization between (E)- and (Z)-26 is proved to
occur under the conditions.
To verify whether isomerization between (E)- and (Z)-
26 occurs, carbanions were generated by deprotonation
of (E)- and (Z)-35, which were prepared by quenching the
reaction of 21 with NHMDS in THF and with KHMDS
in Et2O/toluene, respectively, by acetic acid (Table 8). No
or only a slight E/Z-isomerization was observed regard-
less of the order of addition of the reagents when (E)-35
and (Z)-35 were treated with a base and methyl iodide
under conditions similar to those for 21.
(15) (a) Stork, G.; Boeckman, R. K., J r. J . Am. Chem. Soc. 1973, 95,
2016-2017. (b) Fleming, F. F.; Shook, B. C. Tetrahedron 2002, 58,
1-23.
(16) Brook and co-workers reported the rate of C-to-O migration of
a silyl group in the reaction of diphenyl(trimethylsilyl)carbinol with
diethylamine. Replacement of methyl on the trimethylsilyl group by
phenyl accelerates the 1,2-migration (the Brook rearrangement), which
is believed to proceed via a pentavalent silicon intermediate, by a factor
of about 6. Brook A. G.; LeGrow G. E.; MacRae D. M. Can. J . Chem.
1967, 45, 239-253.
Although the above results cannot rule out the pos-
sibility of intermediacy of 26, we next considered the
J . Org. Chem, Vol. 68, No. 24, 2003 9333

Sasaki et al.
SCHEME 6
TABLE 8.
Fleming and co-workers reported that the lithium
amide base-promoted ring opening of â,γ-epoxynitrile
proceeds by syn-elimination via a six-membered transi-
tion state in which the lithium ion coordinates the oxygen
atom of the epoxide, on the basis of the slow ring opening
of a substrate in which intramolecular chelation is
geometrically precluded.8a To determine the mode of ring
opening of the epoxide, we designed the following com-
petition experiment that relies on that the conformational
preference in syn- and anti-elimination acts in an op-
posite sense in the reaction of 21a and 21b. Thus, while
the transition state from 21a is more favorable than that
from 21b in the syn-elimination in terms of less repulsive
interactions between H-4 and the O-silyl cyanohydrin
moiety (A value17 for OSiMe3, 0.74; for CN, 0.2), in the
case of anti-elimination, the transition state from 21b is
more favorable. When a mixture of 21a (1 equiv) and 21b
(1equiv) in THF was treated with LDA (1 equiv) in the
presence of MeI (1 equiv) at -80 °C for 5 min, a 1.0:0.7
mixture of 21a and 21b was obtained in 40% yield
together with 35% of 27 (E/Z ) 6.6), indicating that 21b
is more reactive than is 21a (Table 9). These results are
consistent with anti-elimination, thus supporting the
pathway described above. Furthermore, the concerted
anti-deprotonation and ring opening process was sup-
ported by the fact that the reactivities of 21a and 21b
were not affected by the addition of HMPA, which can
disrupt the chelated structure by solvating the lithium
cation.
27
35
yield
(%)
yield
(%)
35
base
LDA
solvent
order of addition
E/Z
E
E
E
E
E
E
E
E
Z
Z
Z
Z
Z
Z
Z
Z
THF/hexane (1) CH3I, (2) base
THF/hexane (1) base, (2) CH3I
90
76
41
46
93
81
31
75
31
41
E
LDA
58.4
E
LHMDS THF
LHMDS THF
NHMDS THF
NHMDS THF
KHMDS THF/toluene (1) CH3I, (2) base
KHMDS THF/toluene (1) base, (2) CH3I
LDA
LDA
LHMDS THF
LHMDS THF
NHMDS THF
NHMDS THF
(1) CH3I, (2) base
(1) base, (2) CH3I
(1) CH3I, (2) base
(1) base, (2) CH3I
47
47
E
E
E
0.04
E
0.04
0.01
6
39
8
39
18
87
THF/hexane (1) CH3I, (2) base
THF/hexane (1) base, (2) CH3I
(1) CH3I, (2) base
(1) base, (2) CH3I
(1) CH3I, (2) base
(1) base, (2) CH3I
26
30
87
76
0.05
0.02
0.02
0.01
57
57
4
KHMDS THF/toluene (1) CH3I, (2) base
KHMDS THF/toluene (1) base, (2) CH3I
8
latter possibility and propose a pathway in which a
pentacoordinate silicate species (E)- and (Z)-36 are
formed first via addition of oxyanion, generated by ring
opening of the epoxide, to the silicon center and then
undergo alkylation in a concerted manner to provide (E)-
and (Z)-27 (Scheme 6). Silicate (E)-36 can be formed via
syn-elimination of 21a or anti-elimination of 21b, and
(Z)-36 can be formed via anti-elimination of 21a or syn-
elimination of 21b. After rotation around the C3-C4
bond in such a way that the C4-Si bond can adopt a
coplanar arrangement with the π orbitals of the double
bond, (E)- and (Z)-36 are converted into silicates (E)-
36a ,b and (Z)-36a ,b, respectively, among which only (Z)-
36a can form a internally chelated structure with a metal
cation. Consequently, if the chelated intermediate is
responsible for the formation of (Z)-27, the mode of ring-
opening should be anti because (Z)-27 is formed in a
higher ratio from 21a than from 21b.
Further support for the proposed anti-elimination was
also provided by reactions using cis-epoxysilanes, in
which difference between in reactivities of the two
diastereomers was expected to increase because of more
severe steric repulsion between the siloxy and silyl groups
in a transition state for anti-elimination from 37a (Table
10). Thus, in all cases, reaction with 37a was dramati-
cally slowed compared to that with 37b.
(17) Eliel, E.; Wilen, S. H. In Stereochemistry of Organic Compounds;
J ohn Wiley & Sons: New York, 1994; p 696.
9334 J . Org. Chem., Vol. 68, No. 24, 2003

O-Silyl Cyanohydrins of â-Silyl-R,â-epoxyaldehyde
TABLE 9.
TABLE 11.
yield (%)
21a:21b
yield (%)
E/Z
HMPA
21
27
(-)
(+)
40
67
1.00:0.70
1.00:0.76
35
26
6.6
25.0
yield (%)
yield (%)
base
(E)-27
(E)-34
(E)-27/(E)-34
21b
33b
LDA
LiNEt2
LTMP
22
19
22
7
6
15
3.1
3.2
1.5
18
24
19
40
39
30
understood by assuming that O-Si bond formation is
involved in the rate-determining transition state and is
more favorable in the tert-butyldimethylsilyl goup than
in the tert-butyldiphenylsilyl counterpart, due to a steric
factor rather than an electronical one, which would reflect
the stabilization of the silicate ion by the phenyl group
on the silyl group. Increased steric repulsion between a
base and the silyl group in the proton abstraction, which
is another possible factor explaining the result, proved
to be less important on the basis of competitive experi-
ments using less and more hindered bases, LiNEt2 and
lithium 2,2,6,6-tetramethylpiperidide (LTMP). Thus, re-
action with LiNEt2 resulted in almost the same ratio of
(E)-27/(E)-34, while the ratio was decreased in the
reaction with LTMP. Therefore, the partial formation of
an O-Si bond in the transition state can make the syn-
elimination less favorable in the case of R-silyl epoxide
relative to substrates lacking an R silyl group.
TABLE 10.
Finally, we carried out similar reactions using 38 and
39 to evaluate the role of the siloxy group in 21 and the
participation of chelation involving a siloxy group such
as (Z)-36a . Preparation of 38 lacking a siloxy group and
39, a substrate in which a double bond between epoxysi-
lane and cyanohydrin moiety was introduced, is shown
in Scheme 7.
Reaction of 38 with NHMDS in the presence of MeI
afforded dimethyl derivative 41 in addition to mono-
methyl derivative 40 (Table 12). The formation of the
dimethyl derivative was suppressed by addition of MeI
after treatment with NHMDS, suggesting that a second
deprotonation and methylation are very fast processes.
When 39 was subjected to the methylation reaction
under the same conditions to those used for 21, the
reaction proceeded in a manner similar to that of 21 to
give 43 in an excellent yield (Table 13). Solvent-
dependence of the E/Z ratio, which is lower than that for
21, is partly attributed to the chelation structure 44.
The results described above for 38 and 39 suggests that
the siloxy group of the cyanohydrin moiety does not have
a great effect on the reactivity or reaction course of 21.
The key question that needed to be addressed is why
the mode of elimination in the case of 21 is anti, which
is in sharp contrast to the widely accepted chelation-
assisted syn-elimination mechanism for a base-promoted
ring opening. We speculated that the intramolecular
chelated structure for the syn-elimination is less favor-
able because partial formation of an oxygen-silicon bond
occurs in the transition state leading to 36 from 21. To
test this, we carried out a competitive experiment using
21b and 33b, which bear electronically and sterically
different silyl groups, anticipating that if the formation
of the oxygen-silicon bond occurs after complete forma-
tion of the double bond, a difference between reactivities
of 21b and 33b would not be observed (Table 11). When
21b (1 equiv) and 33b (1 equiv) were treated with LDA
(1.4 equiv) in the presence of CH3I (1.4 equiv) in THF
for 5 min, methylated derivatives (E)-27 and (E)-34 were
obtained in a ratio of 3.1:1 together with recovery of the
starting materials 21b (18%) and 33b (40%), indicating
that 21b is more reactive than 33b. This result can be
J . Org. Chem, Vol. 68, No. 24, 2003 9335

Sasaki et al.
SCHEME 7
Regarding the mechanism of the reaction, we propose
a reaction pathway that involves a silicate intermediate
36 obtained by a concerted process via an anti-opening
of the epoxide followed by the formation of an O-Si bond.
Silicates (E,Z)-36a ,b are transformed into alkylation
products (E,Z)-27 via concerted alkylation of rotamers
36a ,b or via allyl anion intermediates 26 (Scheme 8).
Although further study using optically active substrates
is needed to elucidate the detailed mechanism involving
stereochemistry of the Brook rearrangement and the SE′
reaction from (E,Z)-36a ,b to 27, the results of this study
provide a consistent picture of the reaction pathways for
the tandem process.
Exp er im en ta l Section
3-(ter t-Bu tyld im eth ylsilyl)-2,3-p r op en -1-ol. To a cooled
(-80 °C) solution of 3-(1-ethoxyethoxy)-1-propyne (25.0 g, 195
mmol) in THF (200 mL) was added dropwise a solution of LDA
prepared from diisopropylamine (31.5 mL, 224 mmol) and
n-BuLi (2.66 M in hexane, 80.7 mL, 215 mmol) in THF (150
mL) over 50 min. The solution was stirred at the same
temperature for 30 min before addition of tert-butyldimethyl-
silyl chloride (32.3 g, 215 mmol) in THF (80 mL). After being
stirred at the same temperature for 10 min, the reaction
mixture was allowed to warm to 20 °C. The mixture was
diluted with saturated aqueous NaHCO3 solution (200 mL) and
then extracted with Et2O (150 mL × 3). The combined organic
phases were washed with saturated brine (200 mL), dried, and
concentrated to give crude silylated compound (49.6 g). The
product was used in the following step without further
purification.
TABLE 12.
yield (%) (E/Z)
conditions
40
41
42
To a solution of the above compound in acetone-H2O (70:
30, 250 mL) was added p-toluenesulfonic acid monohydrate
(5.6 g, 29.3 mmol). After being refluxed for 70 min, the reaction
mixture was diluted with saturated aqueous NaHCO3 solution
(200 mL) and extracted with Et2O (100 mL × 3). The combined
organic phases were washed with saturated brine, dried, and
concentrated to give crude 3-(tert-butyldimethylsilyl)-2-propyn-
1-ol (32.8 g). The product was used in the following step
without further purification.
(1) CH3I, (2) base, 5 min
(1) base, 5 min, (2) CH3I
26 (0.7)
51 (0.6)
32 (6.0)
0 (-)
9 (0.2)
5 (24.0)
TABLE 13.
To a cooled (ice-water) solution of Red-Al (65% in toluene,
82.3 g, 268 mmol) in Et2O (115 mL) was added dropwise a
solution of the above compound (30.0 g) in Et2O (115 mL) over
70 min. After being stirred at the same temperature for 15
min, the cooling bath was removed, and stirring was continued
for 90 min. After addition of 3% aqueous H2SO4 solution (200
mL), the mixture was filtered through a pad of Celite. The
filtrate was separated, and the aqueous phase was extracted
with Et2O (150 mL × 3). Combined organic phases were
successively washed with water (100 mL) and saturated brine
(100 mL), dried, and concentrated. The residual oil was
distilled under reduced pressure to give the title compound
(24.0 g, 78%): bp 62 °C/0.15 mmHg, a colorless clear oil; Rf )
base
LDA
HMPA
yield (%)
E/Z
-
+
-
+
-
+
-
+
87
82
91
84
97
83
92
81
9.8
14.0
16.5
6.7
16.5
21.0
7.2
LDA
LHMDS
LHMDS
NHMDS
NHMDS
KHMDS
KHMDS
0.28 (hexane/AcOEt ) 5:1); IR (film) 3318 cm-1 1H NMR δ
.
0.03 (6H, s, SiMe2), 0.87 (9H, s, t-Bu), 1.71 (1H, br s, OH),
4.18 (1H, dd, J ) 4.4, 1.7 Hz, H-1), 5.90 (1H, dt, J ) 18.8, 1.7
Hz, H-2), 6.19 (1H, dt, J ) 18.8, 4.4 Hz, H-2); 13C NMR δ -6.00
(SiMe2), 16.6 (CMe3), 26.6 (CMe3), 65.8 (C-1), 126.8 (C-3), 146.4
(C-2); HRMS calcd for C9H20OSi 172.1283, found 172.1322.
(2R *,3R *)-3-(t er t -Bu t yld im e t h ylsilyl)-2,3-e p oxyp r o-
p a n ol. To a cooled (ice-water) solution of 3-(tert-butyldi-
methylsilyl)-2-propen-1-ol (24.0 g, 139 mmol) and Na2HPO4‚
H2O (59.8 g, 167 mmol) in CH2Cl2 (278 mL) was added
m-CPBA (77% purity, 38.0 g, 167 mmol). After the cooling bath
was removed, the reaction mixture was stirred at room
temperature for 12 h. The mixture was diluted with saturated
aqueous NaHCO3 solution (250 mL) and separated, and the
aqueous phase was extracted with CH2Cl2 (150 mL × 3).
Combined organic phases were washed with saturated brine
(150 mL), dried, and concentrated. The residual oil was
7.9
Con clu sion s
We have found that O-silyl cyanohydrins of â-silyl-R,â-
epoxyaldehyde can function as a highly functionalized
homoenolate equivalent via a tandem sequence involving
base-promoted ring opening, Brook rearrangement, al-
lylic rearrangement, and alkylation.
9336 J . Org. Chem., Vol. 68, No. 24, 2003

O-Silyl Cyanohydrins of â-Silyl-R,â-epoxyaldehyde
SCHEME 8
subjected to column chromatography (silica gel 350 g, elution
with hexane/AcOEt ) 3:1) to give the title compound (23.3 g,
89%) as a colorless clear oil: Rf ) 0.35 (hexane/AcOEt ) 2:1);
IR (neat) 3425, 1283 cm-1; 1H NMR δ-0.04 and 0.02 (each 3H,
s, SiMe2), 0.96 (9H, s, t-Bu), 1.73 (1H, dd, 7.1, 5.9 Hz, OH),
2.34 (1H, d, J ) 3.7 Hz, H-3), 3.02 (1H, ddd, J ) 4.6, 3.7, 2.4
Hz, H-2), 3.60 (1H, ddd, J ) 12.5, 7.1, 4.6 Hz, H-1), 3.99 (1H,
ddd, J ) 12.5, 5.9, 2.4 Hz, H-1); 13C NMR δ -8.34 (SiMe2),
16.6 (CMe3), 26.5 (CMe3), 46.3 (C-3), 55.5 (C-2), 63.3 (C-1).
Anal. Calcd for C9H20O2Si: C, 57.40; H, 10.70. Found: C, 57.19;
H, 10.74.
3H, s, SiMe2Bu), 0.93(9H, s, t-Bu), 0.97(9H, s, t-Bu), 2.31 (1H,
d, J ) 3.4 Hz, H-4), 3.11 (1H, dd, J ) 5.9, 3.4 Hz, H-3), 4.23
(1H, d, J ) 5.9 Hz, H-2); 13C NMR δ-8.6 and -7.9 (SiMe2),
-5.1 and -5.0 (OSiMe2), 16.7 (CMe3), 18.3 (OSiCMe3), 25.7
(CMe3), 26.6(OSiCMe3), 47.1 (C-4), 56.7 (C-3), 66.2 (C-2), 117.2
(CN); MS (APCI-LC/MS) 345 (M + NH4). Anal. Calcd for
C
16H33NO2Si2: C, 58.66; H, 10.15; N, 4.28. Found: C, 58.37;
H, 10.13; N, 4.37. 21b: plates (hexane); mp 38.3-40.5 °C; Rf
) 0.28 (hexane/Et2O ) 20:1); IR (KBr) 1257 cm-1 1H NMR
;
δ-0.03 and 0.03 (each 3H, s, SiMe2Bu), 0.14 and 0.19 (each
3H, s, SiMe2Bu), 0.91 (9H, s, t-Bu), 0.97 (9H, s, t-Bu), 2.37 (1H,
d, J ) 3.2 Hz, H-4), 3.09 (1H, dd, J ) 4.4, 3.2 Hz, H-3), 4.39
(1H, d, J ) 4.4 Hz, H-2); 13C NMR δ -8.4 and -8.1 (SiMe2),
-5.1 and -5.1 (OSiMe2), 16.8 (CMe3), 18.3 (OSiCMe3), 25.6
(CMe3), 26.6 (OSiCMe3), 48.0 (C-4), 55.9 (C-3), 64.1 (C-2), 117.7
(CN); MS (APCI-LC/MS) 345 (M + NH4). Anal. Calcd for
(2R*,3R*)-3-(ter t-Bu t yld im et h ylsilyl-2,3-ep oxyp r op a -
n a l (22). To a cooled (ice-water) solution of 3-(tert-butyldi-
methylsilyl)-2,3-epoxypropanol (7.38 g, 39.18 mmol), DMSO
(55.9 mL, 0.79 mol), and NEt3 (44.2 mL, 10.32 mol) in CH2Cl2
(96 mL) was added SO3‚pyridine (98%, 14.6 g, 90.1 mmol).
After being stirred at the same temperature for 1 h, the
mixture was diluted with hexanes-Et2O (1:1, 100 mL). Phases
were separated, and the aqueous phase was extracted with
Et2O-hexane (1:1, 100 mL × 3). Combined organic phases
were successively washed with water (100 mL) and 1 M
hydrochloric acid (100 mL), dried, and concentrated. The
residual oil was subjected to column chromatography (silica
gel 200 g, elution with hexane/AcOEt ) 3:1) to give 22 (6.0 g,
82%) as a pale yellow oil: Rf ) 0.35 (hexane/Et2O ) 5:1); IR
C
16H33NO2Si2: C, 58.66; H, 10.15; N, 4.28. Found: C, 58.35;
H, 10.29; N, 4.28.
Gen er a l P r oced u r e for Alk yla tion of 21: Rea ction of
21b w ith MeI a n d NHMDS. This procedure is representative
for the alkylation. To a cooled (-80 °C) solution of 21b (100
mg, 0.305 mmol) and MeI (23 µL, 0.369 mmol) in THF (0.7
mL) was added a solution of NHMDS (1.02 M in THF, 0.330
mL, 0.337 mmol). After being stirred at the same temperature
for 5 min, the mixture was diluted with saturated aqueous
NH4Cl solution (10 mL) and extracted with Et2O (10 mL × 3).
Combined organic phases were washed with saturated brine
(10 mL), dried, and concentrated. The residual oil was sub-
jected to column chromatography (silica gel 10 g, elution with
hexane/Et2O ) 20:1) to give 27 (R ) Me) (102 mg, 98%).
1
(film) 1730, 1252 cm-1; H NMR δ0.00 and 0.04 (each 3H, s,
SiMe2), 0.97 (9H, s, t-Bu), 2.56 (1H, d, J ) 3.4 Hz, H-3), 3.15
(1H, dd, J ) 6.6, 3.4 Hz, H-2), 8.82 (1H, d, J ) 6.6 Hz, CHO);
13C NMR δ -8.2 and -8.1 (SiMe2), 17.0 (CMe3), 26.6 (CMe3),
46.6 (C-3), 56.2 (C-2), 199.2 (CHO); HRMS calcd for C5H9O2Si
(M+ - C4H9) 129.0372, found 129.0343.
(E)-27 (R ) Me). For separation of E/Z isomers, MPLC
(elution with hexane/Et2O ) 60:1) was used to provide a
colorless clear oil: Rf ) 0.36 (hexane/Et2O ) 30:1); IR (film)
(1R*,2S*,3S*)- a n d (1R*,2R*,3R*)-2-(ter t-Bu tyld im eth -
ylsiloxy)-4-(ter t-bu tyld im eth ylsilyl)-3,4-ep oxybu ta n en i-
tr ile (21a ,b). To a solution of 22 (214 mg, 1.15 mmol) in
CH2Cl2 (2.3 mL) were added KCN (15 mg, 0.23 mmol), n-Bu4-
PBr (78 mg, 0.23 mmol), and TBSCN (97%, 201 mg, 1.38
mmol). After being stirred at room temperature for 40 min,
the mixture was diluted with saturated aqueous NaHCO3
solution (10 mL) and separated, and the aqueous phase was
extracted with CH2Cl2 (10 mL × 3). Combined organic phases
were washed with saturated brine (10 mL), dried, and con-
centrated. The residual oil was subjected to column chroma-
tography (silica gel 45 g, elution with hexane/Et2O ) 20:1) to
give 21a (162 mg, 43%) and 21b (166 mg, 44%). The relative
stereochemistry in 21a and 21b was determined on the basis
of X-ray analysis of 21b. 21a : plates (hexane); mp 32.4-35.8
1
1662, 1472, 1258, 1195, 1114 cm-1; H NMR δ 0.16 and 0.21
(each 3H, s, SiMe2), 0.17 (6H, s, SiMe2), 0.88 (9H, s, t-Bu), 0.92
(9H, s, t-Bu), 1.64 (3H, s, CH3), 5.07 (1H, d, J ) 12.0 Hz, H-3),
6.79 (1H, d, J ) 12.0 Hz, H-4); 13C NMR δ -5.1, -5.0, -3.4,
and -2.9 (SiMe2), 18.1 and 18.4 (CMe3), 25.7 and 25.7 (CMe3),
31.9 (CH3), 67.7 (C-2), 113.4 (C-3), 121.3 (CN), 144.0 (C-4);
HRMS calcd for C16H32NO2Si2,(M+ - CH3), 326.1972, found
326.1929. (Z)-27 (R ) Me). For separation of E/Z isomers,
MPLC (elution with hexane/CH2Cl2 ) 6:1) was used yo provide
a colorless clear oil: Rf ) 0.36 (hexane/Et2O ) 30:1); IR (film)
1
658, 1472, 1258, 1146, 1124, 1092 cm-1; H NMR δ 0.18 and
0.22 (each 3H, s, SiMe2), 0.19 (6H, s, SiMe2), 0.88 (9H, s, t-Bu),
0.96 (9H, s, t-Bu), 1.72 (3H, s, CH3), 4.51 (1H, d, J ) 6.4 Hz,
H-3), 6.21 (1H, d, J ) 6.1 Hz, H-4); 13C NMR δ -5.3, -5.2,
-3.4, and -3.2 (SiMe2), 18.1 and 18.3 (CMe3), 25.7 (CMe3), 30.6
1
°C; Rf ) 0.33 (hexane/Et2O ) 20:1); IR (KBr) 1253 cm-1; H
NMR δ-0.04 and 0.05 (each 3H, s, SiMe2Bu), 0.17and 0.19 (each
J . Org. Chem, Vol. 68, No. 24, 2003 9337

Sasaki et al.
(CH3), 66.0 (C-2), 111.8 (C-3), 121.9 (CN), 140.8 (C-4); HRMS
rated, and the aqueous phase was extracted with CH2Cl2 (10
mL × 3). Combined organic phases were washed with satu-
rated brine (20 mL), dried, and concentrated. The residual oil
was subjected to column chromatography (silica gel 200 g,
elution with hexane/Et2O ) 20:1) to give 37a (945 mg, 61%)
and 37b (344 mg, 22%). The relative stereochemistry in 37a
and 37b was determined on the basis of X-ray analysis of 37a .
37a : plates (hexane); mp 34.0 °C; Rf ) 0.50 (hexane/Et2O )
10:1); IR (KBr) 1257 cm-1; 1H NMR δ 0.09 and 0.14 (each 3H,
s, SiMe2Bu), 0.17 (6H, s, SiMe2Bu), 0.94 (9H, s, t-Bu), 1.00 (9H,
s, t-Bu), 2.48 (1H, d, J ) 5.3 Hz, H-4), 3.48 (1H, dd, J ) 7.8,
5.3 Hz, H-3), 4.02 (1H, d, J ) 7.8 Hz, H-2); 13C NMR δ -6.2
and -6.0 (SiMe2), -5.0 and -4.8 (OSiMe2), 17.0 (CMe3), 18.3
(OSiCMe3), 25.7 (CMe3), 26.6(OSiCMe3), 47.8 (C-4), 59.4 (C-
3), 64.8 (C-2), 117.6 (CN); HRMS calcd for C12H24NO2Si2, (M+
- C4H9), 270.1346, found 270.1361. 37b: colorless clear oil;
Rf ) 0.42 (hexane/Et2O ) 10:1); IR (firm) 1256 cm-1; 1H NMR
δ 0.08 and 0.09 (each 3H, s, SiMe2Bu), 0.20and 0.28 (each 3H,
s, SiMe2Bu), 0.92 (9H, s, t-Bu), 0.98 (9H, s, t-Bu), 2.48 (1H, d,
J ) 4.6 Hz, H-4), 3.40 (1H, dd, J ) 7.8, 4.6 Hz, H-3), 4.11 (1H,
d, J ) 7.8 Hz, H-2); 13C NMR δ -6.4 and -6.3 (SiMe2), -4.5
and -4.2 (OSiMe2), 17.0 (CMe3), 18.2 (OSiCMe3), 25.7 (CMe3),
26.5(OSiCMe3), 50.6 (C-4), 57.8 (C-3), 63.1 (C-2), 118.4 (CN).
Anal. Calcd for C16H33NO2Si2: C, 58.66; H, 10.15; N, 4.28.
Found: C, 58.30; H,10.40; N, 4.17.
Com p etitive Meth yla tion Rea ction of 21b a n d 33b
w ith Ba se/MeI. To a cooled (-80 °C) solution of 21b and 33b
(21b/33b ) 1.00:1.00, 79.8 mg, 0.205 mmol) and MeI (9.0 µL,
0.144 mmol) in THF (1.75 mL) was added a solution of LDA
(0.5 M, 288 µL, prepared from diisopropylamine (147 µL, 1.05
mmol) and n-BuLi (2.30 M in hexane, 435 µL, 1.00 mmol) in
THF (1.42 mL)). After being stirred at the same temperature
for 5 min, the mixture was diluted with saturated aqueous
NH4Cl solution (10 mL) and extracted with Et2O (10 mL × 3).
Combined organic phases were washed with saturated brine
(10 mL), dried, and concentrated. The residual oil was sub-
jected to column chromatography (silica gel 12 g, elution with
hexane: Et2O ) 25:1) to give a mixture (70.3 mg) of (E)-27
(22%), (E)-34 (7%), 21b (18%), and 33b (40%).
calcd for C16H32NO2Si2 (M+ - CH3) 326.1972, found 326.1929.
(2R*,3S*,4S*)- a n d (2R*,3R*,4R*)-2-(ter t-Bu tyld im eth -
ylsiloxy)-4-(ter t-bu tyld ip h en ylsilyl)-3,4-ep oxybu ta n en i-
tr ile (33a ,b). To a cooled (ice-water) solution of 3-(tert-
butyldiphenylsilyl-2,3-epoxypropanal (2.00 g, 6.44 mmol) in
CH2Cl2 (12.8 mL) were added KCN (20 mg, 0.31 mmol), n-Bu4-
PBr (55 mg, 0.16 mmol), and TBSCN (97%, 954 mg, 6.56
mmol). After stirring at the same temperature for 10 min, the
cooling bath was removed and stirring was continued for 2 h.
The mixture was diluted with hexane, filtrated through a plug
of Al2O3, and concentrated. The residue was purified by column
chromatography (silica gel 70 g, elution with hexane/Et2O )
15:1) to give 33 (2.50 g, 86%, 33a /33b ) 1.00:1.24). For
separation of diastereomers, MPLC (elution with hexane/ether
) 25:1) was used. The relative stereochemistry in 33a and 33b
was determined on the basis of X-ray analysis of 33b. 33a :
plates (hexane); mp 70.0-71.0 °C; Rf ) 0.38 (hexane/Et2O )
10:1); IR (KBr) 1250, 1117, 852, 841 cm-1; 1H NMR δ 0.13 and
0.18 (each 3H, s, SiMe2Bu), 0.91(9H, s, t-Bu), 1.20 (9H, s, t-Bu),
2.90 (1H, d, J ) 3.2 Hz, H-4), 2.96 (1H, dd, J ) 5.3, 3.2 Hz,
H-3), 4.44 (1H, d, J ) 5.3 Hz, H-2), 7.33-7.46 (6H, m, Ph),
7.57-7.59 (2H, m, Ph), 7.63-7.65 (2H, m, Ph); 13C NMR δ-5.1
and -5.0 (SiMe2), 18.3 (CMe3), 18.7 (OSiCMe3), 25.7 (CMe3),
28.0 (OSiCMe3), 46.1 (C-4), 56.3 (C-3), 66.7 (C-2), 117.2 (CN),
128.0, 128.2, 130.0, 130.2, 131.6, 131.7, 136.2, 136.2 (Ph); MS
(APCI-LC/MS) 451 (M+). Anal. Calcd for C26H37NO2Si2 C,
69.13; H, 8.26; N, 3.10. Found: C, 68.78; H, 8.45; N, 3.29.
33b: plates (hexane); mp 76.5-77.0 °C; Rf ) 0.36 (hexane/
Et2O ) 10:1); IR (KBr) 1268, 1259, 1114, 1096, 851, 841 cm-1
;
1H NMR δ 0.15 and 0.18 (each 3H, s, SiMe2Bu), 0.92(9H, s,
t-Bu), 1.20 (9H, s, t-Bu), 2.92 (1H, d, J ) 3.2 Hz, H-4), 2.99
(1H, dd, J ) 4.1, 3.2 Hz, H-3), 4.47 (1H, d, J ) 4.1 Hz, H-2),
7.33-7.47 (6H, m, Ph), 7.56-7.58 (2H, m, Ph), 7.62-7.64 (2H,
m, Ph); 13C NMR δ -5.1 (SiMe2), 18.3 (CMe3), 18.8 (OSiCMe3),
25.7 (CMe3), 28.0 (OSiCMe3), 46.6 (C-4), 56.0 (C-3), 64.6 (C-2),
117.7 (CN), 128.0, 128.2, 130.1, 130.2, 131.6, 131.7, 136.1,
136.2(Ph); HRMS calcd for C26H37NO2Si2 451.2363, found
451.2371.
Isom er iza tion of (E,Z)-35. To a cooled (-80 °C) solution
of (E)-35 (50 mg, 0.153 mmol) and MeI (6.0M in THF, 31 µL,
0.184 mmol) in THF (0.404 mL) was added a solution of
NHMDS (0.95 M in THF, 0.117 mL, 0.168 mmol). After being
stirred at the same temperature for 5 min, the mixture was
diluted with saturated aqueous NH4Cl solution (10 mL) and
extracted with Et2O (5 mL × 3). Combined organic phases were
washed with saturated brine (10 mL), dried, and concentrated.
The residual oil was subjected to column chromatography
(silica gel 5 g, elution with hexane/Et2O ) 40:1) to give (E)-27
(R ) Me) (48.4 mg, 93%).
Com p etitive Meth yla tion Rea ction of 21a a n d 21b
w ith LDA/MeI. To a cooled (-80 °C) solution of 21a and 21b
(21a /21b ) 1.00:1.04, 121 mg, 0.369 mmol) and MeI (11.5 µL,
0.184 mmol) in THF (1.17 mL) was added a solution of LDA
(0.30 mL, 0.184 mmol) prepared from diisopropylamine (269
µL, 1.92 mmol) and n-BuLi (2.05 M in hexane, 898 µL,1.84
mmol) in THF (1.84 µL). After being stirred at the same
temperature for 5 min, the mixture was diluted with saturated
aqueous NH4Cl solution (10 mL) and extracted with Et2O (10
mL × 3). Combined organic phases were washed with satu-
rated brine (10 mL), dried, and concentrated. The residual oil
was subjected to column chromatography (silica gel 12 g,
elution with hexane/Et2O ) 30:1) to give 27 (R ) Me) (44.6
mg, 35%) and 21 (48.4 mg, 40%, 21a /21b ) 1.00:0.70).
(3R*,4R*)-4-(ter t-Bu tyld im eth ylsilyl)-2,3-ep oxybu ta n e-
n itr ile (38). To a cooled (ice-water) solution of pyridine (1,-
12 mL, 13.9 mmol) in CH2Cl2 (29 mL) was added dropwise
trifluoromethansulfonic anhydride (2.34 mL, 13.9 mmol) over
10 min. The reaction mixture was stirred at the same tem-
perature for 10 min before dropwise addition of (2R*,3R*)-3-
(tert-butyldimethylsilyl)-2,3-epoxypropanol (2.50 g, 13.3 mmol)
in CH2Cl2 (59 mL). The mixture was allowed to warm to room
temperature over 30 min, diluted with water (100 mL), and
extracted with CH2Cl2 (30 mL × 3). Combined organic phases
were washed with saturated brine (50 mL), dried, and con-
centrated. To a cooled (ice-water) solution of the residual oil
(4.00 g) in CH2Cl2 (31 mL) was added tetrabutylphosphonium
bromide (1.05 g, 3.10 mmol) and KCN (1.22 g, 18.7 mmol). The
mixture was stirred at the same temperature for 10 min and
at room temperature for 3 h before addition of saturated
aqueous sodium bicarbonate solution (30 mL) and extracted
with Et2O (30 mL × 3). Combined organic phases were washed
with saturated brine (30 mL), dried, and concentrated. The
residual oil was subjected to column chromatography (silica
gel 50 g, elution with hexane/CH2Cl2 ) 5:6) to give the title
compound (1.22 g, 48%) as a pale yellow oil: Rf ) 0.13 (hexane/
AcOEt ) 10:1); IR (film) 2255, 1468, 1254 cm-1 1H NMR δ
;
-0.08 and -0.02 (each 3H, s, SiMe2), 0.93 (9H, s, t-Bu), 2.23
(1H, d, J ) 3.4 Hz, H-4), 2.66 (1H, dd, J ) 17.1, 4.9 Hz, H-2),
2.75 (1H, dd, J ) 17.1, 4.5 Hz, H-2), 3.00 (1H, m, H-3); 13C
NMR δ -8.3, -8.16 (SiMe2), 16.7 (CMe3), 22.8 (C-2), 26.6
(CMe3), 49.9 and 50.3 (C-3, C-4), 116.0 (C-1); HRMS calcd for
C6H10ONSi (M+ - C4H9) 140.0531, found 140.0503.
Meth yla tion of 38. To a cooled solution of 38 (100 mg,
0.507 mmol) and MeI (0.103 mL, 0.608 mmol) in THF (1.37
mL) was added dropwise NHMDS (0.91 M in THF, 0.557 mL,
0.507 mmol). After being stirred at the same temperature for
(2R*,3S*,4R*)- a n d (2R*,3R*,4S*)-2-(ter t-Bu tyld im eth -
ylsiloxy)-4-(ter t-bu tyld im eth ylsilyl)-3,4-ep oxybu ta n en i-
t r ile (37a ,b). To a solution of 3-tert-butyldimethylsilyl-2,3-
epoxypropanal (885 mg, 4.75 mmol) in CH2Cl2 (11.4 mL) were
added KCN (62 mg, 0.95 mmol), n-Bu4PBr (322 mg, 0.95
mmol), and TBSCN (97%, 830 mg, 5.70 mmol). After being
stirred at room temperature for 2 h, the mixture was diluted
with saturated aqueous NaHCO3 solution (20 mL) and sepa-
9338 J . Org. Chem., Vol. 68, No. 24, 2003

O-Silyl Cyanohydrins of â-Silyl-R,â-epoxyaldehyde
5 min, the mixture was diluted with saturated aqueous NH4-
Cl solution (10 mL) and extracted with Et2O (10 mL × 3).
Combined organic phases were washed with saturated brine
(10 mL), dried, and concentrated. The residual oil was sub-
jected to column chromatography (silica gel 10 g, elution with
hexane/AcOEt ) 10:1) to give a mixture (74 mg) of 40, 41, and
42, which were separated by MPLC (elution with hexane/
AcOEt ) 20:1).
NaHCO3 solution (20 mL). The phases were separated, and
the aqueous phase was extracted with Et2O (5 mL × 2). The
combined organic phases were washed with saturated aqueous
NaHCO3 solution (20 mL), dried, and concentrated. The
residual oil was subjected to column chromatography (silica
gel, 40 g; elution with hexane/AcOEt ) 10:1) to give 39 (541
mg, 94%) as a 1:1 mixture of diastereomers as colorless plates
(hexane) as a mixture of diastereomers: mp 41-43 °C; Rf )
0.33 (hexane/AcOEt ) 15:1); IR (film) 1469, 1255, 1105 cm-1
;
(E)-40: a colorless oil; Rf ) 0.28 (hexane/AcOEt ) 10:1); IR
1H NMR δ -0.04, -0.03, 0.01 and 0.018 (each 3H, s, SiMe2),
0.15, 0.15, 0.18 and 0.18 (each 3H, s, SiMe2), 0.91 and 0.91
(each 9H, s, t-Bu), 0.96 and 0.96 (each 9H, s, t-Bu), 2.22 and
2.23 (each 1H, d, J ) 3.4 Hz, H-6), 3.20 and 3.20 (each 1H,
dd, J ) 3.4, 7.3 Hz, H-5), 4.97 and 4.99 (each 1H, d, J ) 5.4
Hz, H-2), 5.70 and 5.72 (each 1H, dd, J ) 7.3, 15.4 Hz, H-4),
5.92 and 5.92 (each 1H, dd, J ) 5.4, 15.4 Hz, H-3); 13C NMR
δ -8.3, -8.2, -8.2, -8.1, -5.0, -4.9, -4.9 and -4.9 (SiMe2),
16.8 and16.8 (CMe3), 18.3 and 18.3 (OSiCMe3), 25.7, 25.7, 26.7
and 26.7 (CMe3), 52.0 and 52.1 (C-6), 54.1 and 54.1 (C-5), 61.8
and 62.1 (C-2), 118.2 and 118.3 (CN), 128.2 and 128.4 (C-3),
134.4 and 134.6 (C-4). Anal. Calcd for C18H35NO2Si2: C, 61.13;
H, 9.98; N, 3.96. Found: C, 61.25; H, 9.89; N, 3.83.
(film) 2240, 1664 cm-1 1H NMR δ 0.15 (6H, s, SiMe2), 0.91
;
(9H, s, t-Bu), 1.38 (3H, d, J ) 7.1 Hz, 2-CH3), 3.21 (1H, dq, J
) 7.1, 7.2 Hz, H-2), 4.94 (1H, dd, J ) 12.0, 7.2 Hz, H-3), 6.52
(1H, dd, J ) 12.0,1.0 Hz, H-4); 13C NMR δ -5.15 (SiMe2), 18.4
(CMe3), 20.3 (C-2), 24.2 (CH3), 25.6 (CMe3), 107.4 (C-3), 121.8
(C-1), 143.6 (C-4); HRMS calcd for C7H12ONSi (M+ - C4H9)
154.0688, found 154.0662. (Z)-40: a colorless oil; Rf ) 0.32
(hexane/AcOEt ) 10:1); IR (film) 2242, 1656 cm-1; 1H NMR δ
0.15 and 0.16 (each 3H, s, SiMe2), 0.92 (9H, s, t-Bu), 1.35 (3H,
d, J ) 7.1 Hz, 2-CH3), 3.73 (1H, m, H-2), 4.48 (1H, dd, J )
8.5, 5.6 Hz, H-3), 6.29 (1H, dd, J ) 5.6, 1.0 Hz, H-4); 13C NMR
δ -5.3 and -5.2 (SiMe2), 18.3 (CMe3), 19.1 (CH3), 20.3 (C-2),
25.6 (CMe3), 105.7 (C-3), 122.5 (C-1), 141.7 (C-4); HRMS calcd
for C7H12ONSi (M+ - C4H9) 154.0688, found 154.0656. (E)-
41, (Z)-41: a colorless oil; Rf ) 0.28 (hexane/AcOEt ) 10:1);
IR (film) 2236, 1659, 1468 cm-1; 1H NMR ((E)-41) δ 0.16 (6H,
s, SiMe2), 0.92 (9H, s, t-Bu), 1.43 (6H, s, CH3), 4.96 (1H, d, J
) 12.0 Hz, H-3), 6.61 (1H, d, J ) 12.0 Hz, H-4); ((Z)-41) δ
0.18 (6H, s, SiMe2), 0.97 (9H, s, t-Bu), 1.49 (6H, s, CH3), 4.32
(1H, d, J ) 5.9 Hz, H-3), 6.24 (1H, J ) 5.9 Hz, H-4). 13C NMR
δ -5.3, -5.1 (SiMe2), 18.3 and 18.5 (CMe3), 25.7 and 25.8 (CH3),
28.0 and 28.8(CMe3), 29.6 and 31.3 (C-2), 110.1 and 114.0 (C-
Gen er a l P r oced u r e for Meth yla tion of 39. To a cooled
(-80 °C) solution of 39 (112 mg, 0.317 mmol) and MeI (24 µL,
0.380 mmol) in THF (1.58 mL) was added a solution of
NHMDS (0.88 M in THF, 396 µL, 0.348 mmol). After being
stirred at the same temperature for 5 min, the mixture was
diluted with saturated aqueous NH4Cl solution (10 mL) and
extracted with Et2O (5 mL × 2). Combined organic phases were
washed with saturated brine (10 mL), dried, and concentrated.
The residual oil was subjected to column chromatography
(silica gel, 10 g; elution with hexane/AcOEt ) 20:1) to give 43
(113 mg, 97%). For separation of E/Z isomers, MPLC (elution
with hexane/CH2Cl2 ) 6:1) was used. (E)-43: a colorless oil;
Rf ) 0.37 (hexane/AcOEt ) 20:1); IR (film) 1655, 1620, 1469,
3), 124.0 (CN), 141.7 and 142.0 (C-4); HRMS calcd for C8H14
-
ONSi (M+ - C4H9) 168.0844, found 168.0835. (E)-42:
a
colorless oil; Rf ) 0.26 (hexane/AcOEt ) 10:1); IR (film) 2251,
1
1668 cm-1; H NMR δ 0.15 (6H, s, SiMe2), 0.91 (9H, s, t-Bu),
1
1301, 1257, 1204, 1109 cm-1; H NMR δ 0.13 and 0.20 (each
2.97 (2H, dd, J ) 6.6, 1.5 Hz, H-2), 4.92 (1H, dt, J ) 12.0, 6.6
Hz, H-3), 6.49 (1H, dt, J ) 11.7, 1.5 Hz, H-4); 13C NMR δ -5.1
(SiMe2), 16.1 (C-2), 18.4 (CMe3), 25.7 (CMe3), 99.2 (C-3), 118.4
(C-1), 145.0 (C-4); HRMS calcd for C6H10ONSi (M+ - C4H9)
140.0531, found 140.0503. (Z)-42: a colorless oil; Rf ) 0.33
3H, s, SiMe2), 0.18 (6H, s, SiMe2), 0.88 (9H, s, t-Bu), 0.92 (9H,
s, t-Bu), 1.63 (3H, s, CH3), 5.38 (1H, d, J ) 15.1 Hz, H-3), 5.67
(1H, dd, J ) 11.7, 11.2 Hz, H-5), 6.39 (1H, dd, J ) 15.1, 11.2
Hz, H-4), 6.66 (1H, d, J ) 11.7 Hz, H-6); 13C NMR δ -5.1,
-5.0, -3.4 and -2.9 (SiMe2), 18.1 and 18.4 (CMe3), 25.7
(CMe3), 31.1 (CH3), 70.3 (C-2), 111.3 (C-5), 121.3 (CN), 127.5
(C-3), 128.3 (C-4), 147.8 (C-6); HRMS calcd for C19H37NO2Si2
367.2363, found 367.2383. (Z)-43: a colorless oil; Rf ) 0.37
(hexane/AcOEt ) 20:1); IR (film) 1652, 1611, 1469, 1411, 1259,
1103, 1061 cm-1; 1H NMR δ 0.17 and 0.22 (each 3H, s, SiMe2),
0.17 (6H, s, SiMe2), 0.90 (9H, s, t-Bu), 0.94 (9H, s, t-Bu), 1.65
(3H, s, CH3), 5.16 (1H, dd, J ) 10.7, 5.6 Hz, H-5), 5.50 (1H, d,
J ) 15.6 Hz, H-3), 6.31 (1H, d, J ) 5.6 Hz, H-6), 6.93 (1H, dd,
J ) 10.7, 15.6 Hz, H-4); 13C NMR δ -5.3, -5.2, -3.4 and -3.1
(SiMe2), 18.2 and 18.5 (CMe3), 25.7 and 25.8 (CMe3), 30.9 (CH3),
69.9 (C-2), 108.4 (C-5), 121.3 (CN), 124.3 (C-4), 128.8 (C-3),
143.0 (C-6); HRMS calcd for C18H34NO2Si2 (M+ - CH3)
352.2128, found 352.2147.
(hexane/AcOEt ) 10:1); IR (film) 2251, 1660, 1259, 1121 cm-1
;
1H NMR δ 0.16 (6H, s, SiMe2), 0.93 (9H, s, t-Bu), 3.12 (2H, dd,
J ) 7.1, 1.5 Hz, H-2), 4.51 (1H, dt, J ) 7.1, 5.6 Hz, H-3), 6.37
(1H, dt, J ) 5.6, 1.5 Hz, H-4); 13CNMR δ-5.2 (SiMe2), 12.4 (C-
2), 18.3 (CMe3), 25.7 (CMe3), 97.5 (C-3), 118.9 (C-1), 143.2 (C-
4); HRMS calcd for C6H10ONSi (M+ - C4H9) 140.0531, found
140.0503.
5-(ter t-Bu tyld im eth ylsilyl)-4,5-ep oxy-(E)-2-p en ten a l. A
solution of epoxy aldehyde 22 (317 mg, 1.70 mmol) and
(triphenylphosphoranylidene)acetaldehyde (97%, 801 mg, 2.55
mmol) in CH3CN (3.2 mL) was refluxed for 25 min before
concentration. The residue was filtered using Et2O, and the
filtrate was concentrated. The residual oil was subjected to
column chromatography (silica gel, 12 g; elution with hexane/
CH2Cl2 ) 1:1) to give the title compound (345 mg, 95%) as a
colorless oil: Rf ) 0.35 (hexane/AcOEt ) 6:1); IR (film) 1693,
1469, 1254, 1095 cm-1; 1H NMR δ -0.02 and 0.04 (each 3H, s,
SiMe2), 0.96 (9H, s, t-Bu), 2.34 (1H, d, J ) 3.4 Hz, H-5), 3.37
(1H, dd, J ) 6.6, 3.4 Hz, H-4), 6.40 (1H, dd, J ) 15.6, 7.1 Hz,
H-2), 6.47 (1H, dd, J ) 15.6, 6.6 Hz, H-3), 9.55 (1H, d, J ) 7.1
Hz, H-1); 13C NMR δ -8.3 and -8.1 (SiMe2), 16.9 (CMe3), 26.6
(CMe3), 53.0 (C-5), 53.6 (C-4), 133.6 (C-3), 155.6 (C-2), 192.7
(CHO); HRMS calcd for C11H19O2Si (M+ - H) 211.1154, found
211.1135.
2-(ter t-Bu tyldim eth ylsiloxy)-6-(ter t-bu tyldim eth ylsilyl)-
5,6-ep oxy-(E)-3-h exen en itr ile (39). To a cooled (ice-water)
solution of 5-(tert-butyldimethylsilyl)-4,5-epoxy-(E)-2-pentenal
(345 mg, 1.62 mmol), KCN (21 mg, 0.325 mmol), and tetrabu-
tylphosphonium bromide (110 mg, 0.325 mmol) in CH2Cl2 (3.5
mL) was added TBSCN (97%, 284 mg, 1.95 mmol). The
reaction mixture was allowed to warm to room temperature,
stirred for 20 min, and then quenched by saturated aqueous
Ack n ow led gm en t. This research was supported, in
part, by a Grant-in-Aid for Scientific Research from the
J apanese Ministry of Education, Sciences, Sports and
Culture, the Uehara Memorial Foundation, and the
Naito Foundation. We also thank the Research Center
for Molecular Medicine, Faculty of Medicine, Hiroshima
University, and the Natural Science Center for Basic
Research and Development (N-BARD), Hiroshima Uni-
versity, for the use of their facilities.
Su p p or tin g In for m a tion Ava ila ble: Full experimental
details and characterization data for all new compounds
described in the text. This material is available free of charge
via the Internet at http://pubs.acs.org.
J O0352934
J . Org. Chem, Vol. 68, No. 24, 2003 9339