Transformation of cis-epoxy compound to cis-2,3-disubstituted oxane and investigation on propagation step in the ring-expansion reactions of cis,trans-diepoxy systems

Article abstract of DOI:10.1016/S0040-4020(00)00622-0

Conversion of cis-epoxy compounds by successive ring-expansion reaction into trans-fused cyclic ethers was examined from both the initiation step and the propagation step. The ring-expansion reaction of cis-4,5-epoxy compounds containing a leaving group o

Full text of DOI:10.1016/S0040-4020(00)00622-0

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 7123–7137
Transformation of cis-Epoxy Compound to cis-2,3-Disubstituted
Oxane and Investigation on Propagation Step in the
Ring-Expansion Reactions of cis,trans-Diepoxy Systems
*
Nobuyuki Hayashi, Hiroko Noguchi and Sadao Tsuboi
Department of Environmental Chemistry and Materials, Faculty of Environmental Science and Technology, Okayama University, Tsushima,
Okayama 700-8530, Japan
Received 19 June 2000; accepted 10 July 2000
Abstract—Conversion of cis-epoxy compounds by successive ring-expansion reaction into trans-fused cyclic ethers was examined from
both the initiation step and the propagation step. The ring-expansion reaction of cis-4,5-epoxy compounds containing a leaving group on C-1
was attempted as a unit process for the successive reaction. When a chloromesyl group was used as a leaving group, the ring expansion
proceeded to give an oxane derivative (endo-type product) preferentially. On the other hand, investigation of the propagation step was carried
out with respect to epoxy oxane derivatives. It was clarified that the ring-expansion reaction of cis-2,3-disubstituted oxane derivatives
provided spiro acetals as products, not the desired trans-fused cyclic ethers, owing to the reaction pathway triggered by 1,2-hydride
rearrangement. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
compounds. The epoxy groups play opposite roles in these
two methods; in the former case, the epoxy groups work as
an electrophile, in the latter case, as a nucleophile. In the
former method, the cyclization of an ordinary hydroxy
epoxide such as 1 without any directing groups for opening
of the epoxy groups would afford the assembled cyclic
polyethers 3 according to Baldwin’s rules (Scheme 1).9 If
some directing groups are introduced into the substrate, it
might be expected to give a fused cyclic ether by the
successive ring-closure method.10 Actually, Murai’s group
skillfully succeeded in the construction of a trans-fused
tricyclic ether from a triepoxy compound.11 However, as
to the systems with some directing groups, the non-removal
of them from the products would remain as a difficult
problem. Therefore, one of the authors has studied a
novel, successive ring-expansion reaction of epoxy
compounds in order to investigate whether the direction of
the epoxy-opening can be controlled or not with respect to
the substrates without any directing groups.12 The working
hypothesis of the first generation is illustrated in Scheme 2.
trans-Fused polycyclic ethers of marine origins, which are
represented by brevetoxins,1 ciguatoxins,2 maitotoxin3 and
so on, are very interesting compounds owing to the novel
molecular structures and biological activities. Especially,
the unique ladder structures of these compounds have fasci-
nated many organic synthetic chemists. Some groups have
achieved the chemical total syntheses of brevetoxins.4–6 On
the other hand, Nakanishi7 and Shimizu8 have proposed that
this class of compounds might be biosynthesized from
acyclic polyepoxy precursors. If this proposal is syntheti-
cally realized, it has been expected that efficient construc-
tion of the polyethers could be feasible, because, in this
tactic, the stereochemistries of the junctures in the poly-
ethers are supposedly controlled by only those of the
epoxy groups. When this cascade reaction is attempted,
two ways can be chosen; one is the successive ring-closure
reaction of hydroxy polyepoxy compounds, the other is the
successive ring-expansion reaction of polyepoxy
Scheme 1.
Keywords: cyclic ether; ring-expansion reaction; bridged oxonium ion; spiro acetal.
* Corresponding author. Tel.: ϩ81-086-251-8899; fax: ϩ81-086-251-8899; e-mail: stsuboi6@cc.okayama-u.ac.jp
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00622-0

7124
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
Scheme 2.
The first step (initiation step) is the process in which first
bridged oxonium ion 513 would be formed by the intra-
molecular nucleophilic attack of a first trans-epoxy group
to a terminal cationic site. The bridged oxonium ion 5
possesses two electrophilic sites (site a and site b). It was
presumed that site a might be more electrophilic than site b
on account of the strain in the bridged system itself in spite
of no directing groups on the substrate. Accordingly, in the
next propagation step, it was supposed that the second trans-
epoxy group attacks site a on the first oxonium ion 5 to
generate second oxonium ion 6. If the same reaction of
the third trans-epoxy group is followed by trapping of the
last oxonium ion 7 with an external nucleophile (termina-
tion step), it was assumed that the target molecule 8 could be
synthesized. Previous experiments revealed the following.
(1) The ring-expansion reaction of trans-1-bromo-4,5-
epoxide 9 proceeded in endo-mode to give oxane 11
(Scheme 3).14 (2) The intramolecular nucleophilic attack
of the second epoxy group to first oxonium ion 14 (path a
in Scheme 4) was a slower process than that of the counter
anion of oxonium ions 14 (in this case, triflate anion). There-
fore, the one-pot successive ring-expansion reaction of
trans,trans-bromodiepoxide 13 provided cis-fused cyclic
ether 19 via the double inversion of the juncture’s stereo-
chemistry.
It seemed to us that the formation of the juncture by the
double inversion of the stereochemistry might possibly be
utilized for the construction of trans-fused cyclic ethers.
Our new working hypothesis is illustrated in Scheme 5. If
the ring-expansion reaction of cis-polyepoxy groups on 20
is repeated with the double inversion, it is expected that
trans-fused cyclic ether 23 can be formed via the inter-
mediates (21 and 22). In this report, we describe the results
Scheme 3.
Scheme 4.

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7125
Scheme 5.
with regard to this new strategy from the following two
aspects: (1) the single ring-expansion reaction of cis-4,5-
monoepoxy systems containing a leaving group on C-1 as
a unit process for the successive reaction, (2) the
propagation step of a cis,trans-diepoxy system (the simplest
propagation model).
anhydrous conditions in order to exclude the above
influence of water, unfortunately, assembled bicyclic ether
30 was obtained in good yield by the intramolecular nucleo-
philic attack of the terminal silyloxy group. Therefore, so as
to investigate the ring-expansion reaction of cis-epoxy
compounds more exactly, the new substrates (36a, 36b,
and 36c) whose side chains cannot participate to the reactive
sites were prepared as outlined in Scheme 7. The hydroxy
group of alcohol 3110a was protected as the corresponding
TBDPS ether 32. The triple bond in 32 was hydrogenated to
afford cis-olefin 33, whose TBS group was selectively
removed by treatment with pyridinium p-toluenesulfonate
in ethanol. Alcohol 34 was converted to sulfonates (35 and
37), which were oxidized with mCPBA to give cis-epoxides
36a and 36c, respectively. cis-Iodoepoxide 36b was
obtained from 36a by treatment with TBAI.
Ring-Expansion Reaction of a cis-4,5-Epoxy System
Containing a Leaving Group on C-1
In this chapter, the investigation regarding the ring-
expansion reaction of cis-monoepoxy systems is described.
When the ring-expansion reaction of cis-bromo epoxide 24
was previously attempted under the conditions of Scheme 3,
the only product was oxolane derivative 29 (exo-type
product).12d However, the mechanism generating 29 is
unclear (that is to say, it is quite questionable whether 29
is the ring-expanded product or not), because three path-
ways from 24 to 29 as shown in Scheme 6 were assumed.
One is the exo-selective ring-expansion reaction via
oxonium ion 25. Another is the cyclization of diol 26
formed by the hydrolysis of the epoxy group on 24. The
third is 5-exo-cyclization of hydroxy epoxide 27 produced
by the substitution of the bromo group on 24 with water. On
the other hand, when the reaction was carried out under
With respect to these cis-epoxides (36a, 36b, and 36c), the
ring-expansion reactions were examined under the various
conditions shown in Table 1. At first, cis-iodoepoxide 36b
was treated with AgOTf in aqueous THF to afford only exo-
product 40b1 similarly to 24 (entry 1). In this reaction, it was
observed by TLC analysis that hydroxy epoxide 36d
(XˆOH) was temporarily generated and disappeared at
the end point of the reaction. The structure of 36d was
identified as the corresponding acetate 36e (XˆOAc),
Scheme 6.

7126
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
Scheme 7. Reagents and conditions: (a) TBDPSCl, imidazole, CH2Cl2, 25ЊC, 10 min, 92%; (b) H2, Lindlar cat., quinoline, MeOH, 25ЊC, 10 min, 96%; (c)
PPTS, EtOH, 25ЊC, 13 h, 92%; (d) TsCl, Et3N, DMAP, 25ЊC, 2.5 h, 92%; (e) mCPBA, Na2HPO4, 25ЊC, 2.5 h, 74%; (f) TBAI, THF, 65ЊC, 3 h, 81%; (g)
chloromethanesulfonyl chloride, Et3N, DMAP, 0ЊC, 2.5 h, 86%; (h) mCPBA, Na2HPO4, CH2Cl2, 25ЊC, 3.5 h, 76%.
because 36d was readily cyclized to 40b1. Consequently, it
was obvious that the reaction using cis-4,5-epoxy-1-halides
(24 and 36b) in aqueous tetrahydrofuran was not a ring-
expansion reaction, but 5-exo-cyclization reaction of the
hydroxy epoxide intermediate 36d. The above result
suggests that the intramolecular nucleophilic attack of the
cis-epoxy group is a slower process than the intermolecular
attack of an external nucleophile such as water. Accord-
ingly, in order to force the cis-epoxy group to attack the
internal electrophilic site, the external nucleophile for the
bridged oxonium ion 38 must not exist at the starting point
of the reaction. Therefore, the following reactions were
carried out under the conditions without adding an external
nucleophile. In the case of treatment of epoxy iodide 36b
with silver triflate in anhydrous dichloromethane (entry 2),
it was expected that a triflate anion (the counter anion of
oxonium ion 38) was the scavenger of 38. However, the
products were a complex mixture which consisted of the
unknown high-polar products. This result suggested that
the anticipated products (39b2 or 40b2) were quite unstable
because they contain a trifluoromethanesulfonyloxy group
or the intramolecular nucleophilic attack of the cis-epoxy
group could not occur for 36b itself to decompose because
of the low reactivity of the cis-epoxy group. In entry 3, the
use of silver tosylate as a Lewis acid resulted in the substi-
tution of the iodo group into a tosyloxy group. This result in
entry 3 revealed that the nucleophilic ability of the cis-
epoxy group in this system was lower than that of a weak
nucleophile such as a tosylate anion. Therefore, it was
considered that any nucleophiles containing the counter
anions of Lewis acids should not exist in the reaction
system. That is to say, the reaction should depend only on
the nucleophilic ability of the cis-epoxy group of 36 without
adding a Lewis acid in order to form oxonium ion 38, which
should be captured by the X groups eliminated from 36. On
the basis of this concept, 36a (XˆOTs) was heated at reflux
in acetonitrile or nitromethane only to recover the starting
materials quantitatively (entry 4 and 5). It turned out that
the tosyl group did not have enough leaving ability for the
nucleophilic attack of the epoxy group of 36a. Next, the
leaving group was changed to a chloromesyloxy group,
which was reported by Nakata’s group as a better leaving
group for the cyclic ether formation.15 Epoxide 36c was
heated at 83ЊC in various solvents (1,2-dichloroethane,
chlorobenzene, toluene, and benzene) to give two ring-
expanded products (39c and 40c) (entry 6 to 9). The endo-
isomer 39c was major in all cases. Though the ratio of the
endo-isomer increased with decreasing the polarity of
Table 1. Ring-expansion reactions of cis-epoxy systems 36
Entry 36
X
Conditions
Solvent
Nu
Y
39, 40 Yielda (39ϩ40) (%) Ratiob (39:40) Recovered 36 (%)
1
2
3
4
5
6
7
8
9
b
b
b
a
a
c
c
c
c
c
I
I
I
AgOTf, 25ЊC, 3 h
AgOTf, Ϫ30ЊC, 0.5 h CH2Cl2
AgOTs, 83ЊC, 2 h
82ЊC, 2.5 h
101ЊC, 2.5 h
THF/H2O (5:1) YH OH
b1
b2
b3
a
a
c
c
c
c
c
75
0.100
–
0
0
YϪ TfO
Decomposition
–
No reaction
No reaction
76
71
65
37
36
CH3CN
CH3CN
CH3NO2
Cl(CH2)2Cl
PhCl
YϪ TsO
0c
TsO
TsO
YϪ TsO
99
99
17
6.7
43
54
0
YϪ TsO
ClMsO 83ЊC, 48 h
ClMsO 83ЊC, 186 h
ClMsO 83ЊC, 168 h
ClMsO 83ЊC, 192 h
ClMsO 82ЊC, 7.5 h
YϪ ClMsO
YϪ ClMsO
YϪ ClMsO
YϪ ClMsO
YϪ ClMsO
2.0:1
1.9:1
2.7:1
3.2:1
1.1:1
PhCH3
PhH
10
CH3CN
a The yield based on conversion.
b The ratios were determined by 1H NMR analyses.
c The 4,5-epoxy-1-tosylate was obtained in 97% yield by substitution reaction.

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7127
thermodynamic stability between 39c and 40c was unclear
from the results shown in Table 2, it was observed that the
recoveries from the mixture of endo/exo (17:1) (entry 1 and
3) were lower than those of the mixture from endo/exo
(1:15) (entry 2 and 4). This result suggests that the endo-
isomer 39c decomposed more rapidly than the exo-isomer
40c. Accordingly, concerning the ring expansion reaction of
36c, the lowering of the endo-preference in a high-polar
solvent is attributed to the endo-selective decomposition.
Furthermore, the net endo-selectivity can be estimated to
be higher than the apparent ratios. In any event, it was
revealed that the formation of the endo-isomer 39c was at
least kinetically favorable, because the skeletal rearrange-
ment of the exo-isomer 40c into the endo-isomer 39c was
not observed under the thermodynamic conditions. The
reason for the lability of endo-isomer 39c will be discussed
in the next chapter.
Figure 1.
solvents, the reaction rates tended to diminish. On the other
hand, although the reaction rate was greatly accelerated in
acetonitrile of a more polar solvent, the marked lowering of
both yield and the ratio of the endo-isomer was observed
(entry 10). The acceleration of reaction rate in the higher
polar solvent is attributed to lowering the activation energy
to oxonium ion 38 by solvation. The structure of 39c was
determined by 1H NMR analyses (NOE experiment and the
H–H coupling constant; J2–3ˆ1.0 Hz) as illustrated in Fig. 1.
1
The structure of 40c was determined by comparing the H
NMR spectrum of 40c with that of the chloromesylate
prepared from the hydroxy oxolane 40b1.
Investigation of Propagation Step in the Ring-Expansion
Reaction
In order to explain the relation between the lowering of the
endo-selectivity and the polarity of the solvent, the follow-
ing three facts can be taken into account as possible reasons:
(1) the conversion of the kinetic endo/exo-ratio; (2) the
skeletal rearrangement of the endo-isomer 39c into
the exo-isomer 40c; (3) the selective decomposition of the
endo-isomer 39c (the endo-selective decomposition).
However, evaluating the net kinetic endo/exo-ratio is
extremely difficult, because these systems possessing a
good leaving group such as a sulfonyloxy group always
involve the possibilities of the above (2) and (3). Therefore,
herein, the relation was discussed except for (1). In order to
investigate thermodynamic stability between the endo-
isomer 39c and the exo-isomer 40c, the experiments
shown in Table 2 were carried out. At first, a 17:1 mixture
of 39c and 40c was heated at 83ЊC in 1,2-dichloroethane
(entry 1). After 48 h, the ratio changed to 13:1 in 67%
recovery. On the other hand, when a 1:15 mixture of 39c
and 40c was exposed under the same conditions, the endo/
exo ratio changed to 1:23 in 85% recovery (entry 2). The
proportion of the endo-isomer 39c was lowered in both
experiments. However, this result can not prove that the
exo-isomer 40c was thermodynamically more stable than
the endo-isomer 39c (in other words, the skeletal rearrange-
ment of 39c into 40c), because the low recoveries cannot
rule out the possibility of the endo-selective decomposition.
Similar experiments were also undertaken in acetonitrile
(entry 3 and 4) to result in decreasing the proportion of
the endo-isomer 39c in a similar tendency as in the above
entry 1 and 2. The recoveries in acetonitrile were lower than
those in 1,2-dichloroethane, and accompanied with
unknown high-polar decomposed products. Although the
In this chapter, the propagation steps in the successive ring-
expansion reactions of the diepoxy systems 42, which is
illustrated as a process from 43 to 44 in Scheme 8, are
discussed. cis,trans-Diepoxides 42 are the simplest model
for obtaining all trans-fused cyclic ethers, because, for the
relative stereochemistry of the two substituent groups (R
and OMsCl) on the final ring in order to be the trans-
form, the stereochemistry of the last epoxy group must be
the trans-form.
At first, the successive reactions with regard to cis,trans-
diepoxides 55 were investigated. A diastereomeric mixture
of cis,trans-epoxides (55a and 55b) was prepared as
outlined in Scheme 9. Allylic alcohol 46, which was
prepared by hydrogenation of propargylic alcohol 45,12d
was converted into allylic bromide 47 with Ph3P and CBr4
in 85% yield. Allylic bromide 47 was coupled with 1-lithio-
3-tetrahydropyranyloxypropyne to afford enyne 48.
Removal of the tetrahydropyranyl group from 48 by
methanolysis gave alcohol 49 in 93% yield. The alcohol
49 was reduced with LiAlH4 to obtain allylic alcohol 50
(88%), which was transformed into epoxide 51 with
TBHP and VO(acac)2 in 85% yield.16 Protection of the
hydroxy group on 51 was followed by removal of the
4-methoxybenzyl group. Alcohol 53 was converted into
sulfonate 54 with chloromethanesulfonylchloride (79%). The
sulfonate 54 was oxidized with m-chloroperbenzoic acid to
give a diastereomeric mixture of diepoxides (55a and 55b).
The mixture of 55a and 55b was heated at reflux in 1,2-
dichloroethane to afford epoxy oxanes (56a and 56b) in 25%
Table 2. The variations of the ratios between the endo-isomer (39c) and the exo-isomer (40c) under thermodynamic conditions
Entry
Initial ratio (39c:40c)a
Conditions
Final ratio (39c:40c)a
Recovery (%)
1
2
17:1
1:15
ClCH2CH2Cl, reflux, 48 h
ClCH2CH2Cl, reflux, 48 h
13:1
1:23
67
85
3
4
17:1
1:15
CH3CN, reflux, 7.5 h
CH3CN, reflux, 7.5 h
5.2:1
1:171
36
42
a The ratios were determined by 1H NMR analyses.

7128
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
Scheme 8.
yield with epoxy oxolanes (17%) and recovered 55a and
55b (25%) (Scheme 10). The other products were unknown
high-polar decomposed products. Epoxy oxanes (56a and
56b) were treated under the same conditions to be gradually
changed into the high-polar decomposed products, which
were not separable. No fused cyclic ethers were detected
at all. It seems that the reason why the ring-expansion reac-
tion stopped before the second ring-expansion occurred is
owing to the distance of both reactive sites (the epoxy group
and the chloromesyloxy group). If the second ring-expan-
sion reaction takes place, the chloromesyloxy group and the
side chain containing the epoxy group must orient axial and
Scheme 9. Reagents and conditions: (a) H2, Lindlar cat., quinoline, pyridine, 25ЊC, 1.5 h, 98%; (b) Ph3P, CBr4, CH2Cl2, 25ЊC, 10 min, 85%; (c) LiCC-
CH2OTHP, THF, Ϫ78 to 25ЊC (28 h), 85%; (d) p-TsOH, MeOH, 25ЊC, 3 h, 93%; (e) LiAlH4, THF, 25ЊC, 14 h, 88%; (f) VO(acac)2, TBHP, benzene, 25ЊC, 1 h,
85%; (g) TBDPSCl, imidazole, CH2Cl2, 25ЊC, 10 min, Ͼ99%; (h) DDQ, CH2Cl2/H2O (10:1), 25ЊC, 1 h, 98%; (i) ClCH2SO2Cl, pyridine, CH2Cl2, 0ЊC, 2 h,
79%; (j) mCPBA, Na2HPO4, CH2Cl2, 25ЊC, 2.5 h, 95%.
Scheme 10.

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7129
Figure 2.
equatorial, respectively, as illustrated in Fig. 2. Because
both reactive sites cannot be in close proximity on account
of the strain arisen in the systems, it is presumed that the
propagation step of the successive ring expansion could not
proceed.
hydroxy group on 60 as a silyl ether, the 4-methoxybenzyl
group was removed with DDQ in aqueous dichloromethane.
The obtained alcohol 62 was converted into chloromesylate
63. The compound 63 was treated with one equivalent of
m-chloroperbenzoic acid to afford monoepoxide 64 in 43%
yield with the monoepoxy isomer (13%) and the diepoxides
(14%). Though epoxide 64 was transformed into cis-oxane
65 (the endo-isomer) in 18% yield by ring-expansion reac-
tion, 65 was unfortunately a minor product. Oxolane 66 (the
exo-isomer) was obtained in 34% yield as a major product
with the unknown high-polar decomposed products. As to
the ring-expansion reaction of epoxide 64, the endo/exo-
ratio was contrary to that of the previous experiments.
Next, in order to resolve this problem, the propagation step
was examined regarding the substrates possessing a longer
side chain (that is, a more flexible side chain). New epoxy
oxanes (67a and 67b) were prepared as outlined in Scheme
11. PCC-oxidation of alcohol 5912d was followed by Wittig
reaction and the reduction with DIBAH to provide allylic
alcohol 60 in 67% yield for 3 steps. After protecting the
Scheme 11. Reagents and conditions: (a) PCC, MS4A, CH2Cl2, 25ЊC, 1.5 h; (b) Ph3PvCHCO2Me, benzene, 24 h; (c) DIBAH, CH2Cl2, Ϫ80ЊC, 1 h, 3 steps
for 67%; (d) TBDPSCl, imidazole, CH2Cl2, 0ЊC, 30 min, 97%; (e) DDQ, CH2Cl2/H2O (10:1), 25ЊC, 2 h, 81%; (f) ClCH2SO2Cl, pyridine, CH2Cl2, 0ЊC, 1.5 h,
92%; (g) mCPBA, Na2HPO4, CH2Cl2, 25ЊC, 30 min, 43%; (h) 83ЊC, 6 h, ClCH2CH2Cl, 65 (18%), 66 (34%); (i) mCPBA, Na2HPO4, CH2Cl2, 25ЊC, 2 h, 72%.
Scheme 12.

7130
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
to the non-bridged oxonium ion might promote the decom-
position of the endo-isomer 65.
Conclusion
Figure 3.
In this report, in order to construct the trans-fused cyclic
ethers, the successive ring-expansion reaction of the
diepoxy systems containing the cis-epoxy group was
investigated from both the unit reaction using the cis-
monoepoxy system 36 without a directing group and the
propagation step in the successive reaction. With respect
to the ring-expansion reaction of 36, when the chloro-
mesyloxy group was used as a leaving group, the ring-
expansion reaction proceeded preferentially in endo-
fashion. It was speculated that the endo-product was the
kinetically favorable compound as well as in the case of
the trans-epoxy isomer. The ratio of the endo-isomer
increased with decreasing the polarity of solvents. As to
the propagation step, it was clear that the cis-2,3-disubsti-
tuted oxane system, which was the ring-expanded product of
cis-epoxy compound, underwent readily 1,2-hydride
rearrangement. Consequently, epoxy oxanes (67a and
67b) were not converted into the desired trans-fused cyclic
ethers, but into spiro acetal 70 by the ring expansion
reaction of the epoxy group. It was assumed that this ring-
expansion reaction proceeded via the non-bridged oxonium
ion and the bridged oxonium ion. It was suggested that this
novel property of the cis-2,3-disubstituted oxane system
affected greatly also the endo/exo-ratio of the ring-
expansion reaction of 36c.
This phenomenon will be discussed below. The compound
65 was oxidized with m-chloroperbenzoic acid to obtain a
diastereomeric mixture of the epoxy oxanes (67a:67bˆ1:1)
in 76% yield. The mixture of 67a and 67b was heated at
reflux in 1,2-dichloroethane to carry out the ring-expansion
reaction. However, the products were not the desired fused
cyclic ethers, but spiro acetals 70 (ca. 31%) and 71 (ca. 9%),
unexpectedly.17 Although the carbon skeleton and the C2–
C3’s relative stereochemistry of 70 were determined by 1H
NMR (H–H coupling constant; J2–3ˆ10.0 Hz), 13C NMR
(C-6ˆ95.22 ppm) and H–H COSY, the stereochemistry of
the acetal moiety could not be clarified. The structure of 71
could not be established except for only the planar structure
1
determined by H NMR and H–H COSY, because of the
small amount of 71. It is assumed that 70 and 71 were
generated via non-bridged oxonim ion 68 and bridged
oxonium ion 69 as illustrated in Scheme 12. Production of
a spiro acetal from a cis-disubstituted hydroxy oxepane via a
non-bridged oxonium ion is reported by Martin et al.,18
although the proposed reaction mechanism is different
from ours. Whether 70 and 71 were generated from 67a or
67b is not important, because at the point of formation of
non-bridged oxonium ion 68, all asymmetric centers on 67’s
ring disappear. It is presumed that this 1,2-hydride
rearrangement of the cis-2,3-disubstituted oxane system
was promoted by the following two stereoelectronic effects
shown in Fig. 3: (1) the overlap between the lone pair orbital
of an axial position and the antibonding orbital of an
adjacent C–H bond, and (2) the overlap between the bond-
ing orbital of a C–H bond and the antibonding orbital of an
adjacent C–OMsCl bond. Although the desired fused cyclic
ether could not be obtained, this formation of spiro acetal 70
via the 1,2-hydride shift is interesting, because this property
of the cis-2,3-disubstituted oxane system seems to explain
the lability of the oxanes (39c and 56) and the exo-prefer-
ential ring expansion of the epoxides 64. Oxolane 40c (the
isomer of 39c) is much less subject to the above stereo-
electronic effects than oxane 39c, because the 40c’s flexible
conformation prevents the corresponding orbitals from
sufficient overlapping. Therefore, 39c would be converted
into the corresponding non-bridged oxonium ion more
rapidly than 40c to decompose faster than 40c on account
of the instability of the non-bridged oxonium ion. It is
considered that the formation of the non-bridged oxonium
ion is greatly accelerated in a high-polar solvent such as
acetonitrile by solvation. This assumption seems quite
reasonable, because it can explain why the ratio of the
endo-isomer decreased with increasing the polarity of the
solvents (in other words, why the endo-isomer decomposed
more rapidly than the exo-isomer). It is assumed that the
decomposition of 56a and 56b in dichloroethane by a
prolonged reaction time took place by a similar mechanism.
Although there is no obvious experimental evidence at all
regarding the exo-preferential ring expansion of the
epoxides 64, the participation of the intramolecular olefin
Experimental
Solvents and reagents were dried and distilled before use.
Dichloromethane, 1,2-dichloroethane, benzene, acetonitrile,
nitromethane, and triethylamine were distilled from calcium
hydride. Tetrahydrofuran was distilled from sodium benzo-
phenone ketyl. Normal reagent-grade solvents were used for
flash chromatography, preparative thin-layer chromato-
graphy (PTLC), and extraction.
All reactions were monitored by thin-layer chromatography
(TLC) with precoated silica gel (SiO2) plates (MERCK,
Silica gel 60 F254 Art. 1.05554). Visualization was
achieved via ultraviolet light and a 5.6% ethanolic p-anisal-
dehyde solution containing 5.6% of concentrated sulfuric
acid-heat. For flash chromatography was utilized SiO2
(MERCK, Silica gel 60 1.09385. 0925). For PTLC were
utilized precoated SiO2 plates (MERCK, Silica gel 60
F254 1.05744 or 1.05715). HPLC were run with a HITACHI
HPLC Pump L-7100 equipped with a HITACHI UV
detector L-7400. For HPLC was utilized SiO2 column
(Inertsil ODS-3, GL Sciences Inc.).
The NMR spectra were recorded on Varian model
VXR200S or VXR500 spectrometers in chloroform-d1
(CDCl3) or benzene-d6 (C6D6). Infrared (IR) spectra were
obtained on a JASCO model FT/IR-5000 infrared spectro-
photometer in neat state. Chemical shifts (d) are reported
with tetramethylsilane (TMS) (dˆ0.00 ppm) or benzene
(dˆ7.20 ppm) as internal standards. Splitting patterns are

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7131
designated as ‘s, d, t, q, qui, and m’; these symbols indicate
‘singlet, doublet, triplet, quartet, quintet, and multiplet’,
respectively. High-resolution mass spectra (HR-MS) were
obtained on a JEOL model JMS600 mass spectrometer
under electron ionization (EI) and chemical ionization
(CI), a JEOL model JMS-SX102A mass spectrometer
under field desorption (FD) condition.
824 cmϪ1; HR-FD-MS calcd for C22H31O2Si (MϩϩH)
355.2168, found 355.2113.
(2Z)-1-(tert-Butyldiphenylsilyloxy)-6-(p-toluenesulfonyl-
oxy)-2-hexene (35). To a solution of alcohol 34 (0.851 g,
2.40 mmol) in CH2Cl2 (24 mL) were added Et3N (1.00 mL,
7.20 mmol), tosyl chloride (0.549 g, 2.88 mmol), and
4-dimethylaminopyridine (29.3 mg, 0.240 mmol) at 0ЊC.
The mixture was stirred at 25ЊC for 2.5 h, diluted with
diethyl ether (140 mL), and washed with satd aq. NH4Cl
(100 mL). The organic layer was separated, washed with
satd aq. NaHCO3 (100 mL) and brine (100 mL), dried
over MgSO4, and concentrated in vacuo. The residue was
purified by flash chromatography (SiO2, hexane–EtOAc,
20:1) to give sulfonate 35 (1.13 g, 92%) as a colorless oil:
Rfˆ0.49 (hexane–EtOAc, 4:1); 1H NMR (500 MHz,
CDCl3) d 7.73 (2H, d, Jˆ8.0 Hz), 7.68–7.66 (4H, m),
7.44–7.35 (6H, m), 7.29 (2H, d, Jˆ8.0 Hz), 5.63–5.58
(1H, m), 5.30–5.25 (1H, m), 4.18 (2H, dd, Jˆ0.5, 6.5 Hz),
3.93 (2H, t, Jˆ6.0 Hz), 2.42 (3H, s), 1.87 (2H, q, Jˆ7.5 Hz),
1.62–1.57 (2H, m), and 1.04 (9H, s); IR (neat) 3074, 3020,
All reactions were carried out under anhydrous conditions
and nitrogen atmosphere, unless otherwise noted.
6-(tert-Butyldimethylsilyloxy)-1-(tert-butyldiphenylsilyl-
oxy)-2-hexyne (32). To a solution of propargylic alcohol 31
(1.54 g, 6.74 mmol) in CH2Cl2 (34 mL) were added
imidazole (1.14 g, 16.8 mmol) and tert-butyldiphenylsilyl
chloride (1.93 mL, 7.41 mmol) at 0ЊC. The mixture was
stirred at 25ЊC for 10 min, diluted with Et2O (150 mL),
and washed with H2O (50 mL). The organic layer was
separated, washed with satd aq. NaHCO3 (50 mL), and
brine (50 mL), dried over MgSO4, and concentrated in
vacuo. The residue was purified by flash chromatography
(SiO2, hexane–EtOAc, 100:1) to give silyl ether 32 (2.98 g,
2934, 2862, 1475, 1429, 1243, 1112, 1075, and 824 cmϪ1
;
1
92%) as a colorless oil: Rfˆ0.71 (hexane–EtOAc, 4:1); H
HR-EI-MS calcd for C25H27O4SSi (MϩϪtBu) 451.1399,
NMR (200 MHz, CDCl3) d 7.75–7.69 (4H, m), 7.48–7.33
(6H, m), 4.30 (2H, t, Jˆ2.2 Hz), 3.66 (2H, t, Jˆ6.0 Hz),
2.25 (2H, tt, Jˆ2.2, 7.2 Hz), 1.72–1.59 (2H, m), 1.07 (9H,
s), 0.89 (9H, s), and 0.05 (6H, s); IR (neat) 3076, 3054,
2958, 2934, 1473, 1431, 1390, 1377, 1363, 1257, 1145,
found 451.1395.
(2Rء,3Rء)-1-(tert-Butyldiphenylsilyloxy)-6-(p-toluenesul-
fonyloxy)-2,3-epoxyhexane (36a). To a solution of olefin
35 (627 mg, 1.23 mmol) in CH2Cl2 (25 mL) cooled to 0ЊC
were added Na2HPO4 (1.88 g, 6.17 mmol) and m-chloro-
perbenzoic acid (608 mg, 2.47 mmol) and the solution was
stirred at 25ЊC for 3 h, and satd aq. NaHCO3 (5 mL) and satd
aq. Na2S2O3 (5 mL) was added at 0ЊC. The mixture was
extracted with Et2O (140 mL) and the Et2O layer was
washed with satd aq. Na2S2O3 (50 mL), satd aq. NaHCO3
(3×50 mL) and brine (50 mL), dried over MgSO4, and
concentrated in vacuo. The residue was purified by flash
chromatography (SiO2, hexane–EtOAc, 20:1) to give
epoxide 36a (481 mg, 74%) as a colorless oil: Rfˆ0.40
1110, 1073, 1000, 973, 940, 837, 777, 739, and 702 cmϪ1
;
HR-CI-MS calcd for C28H43O2Si2 (MϩϩH) 467.2802, found
467.2834.
(2Z)-6-(tert-Butyldimethylsilyloxy)-1-(tert-butyldiphenyl-
silyloxy)-2-hexene (33). To a solution of alkyne 32 (2.88 g,
5.99 mmol) in MeOH (35 mL) were added quinoline
(0.22 mL) and Pd/CaCO2 (0.288 g) at 25ЊC and the mixture
was stirred at 25ЊC for 10 min under hydrogen atmosphere
and filtered through Celite. The filtrate was concentrated in
vacuo. The residue was purified by flash chromatography
(SiO2, hexane–EtOAc, 100:1) to give alkene 33 (2.76 g,
1
(hexane–EtOAc, 4:1); H NMR (500 MHz, CDCl3) d 7.75
(2H, d, Jˆ8.0 Hz), 7.68–7.66 (4H, m), 7.45–7.38 (6H, m),
7.30 (2H, d, Jˆ8.0 Hz), 4.04 (1H, dt, Jˆ10.0, 6.0 Hz), 3.98
(1H, dt, Jˆ10.0, 6.0 Hz), 3.76 (1H, ddd, Jˆ1.5, 5.0,
12.0 Hz), 3.68 (1H, ddd, Jˆ3.0, 5.0, 12.0 Hz), 3.14–3.11
(1H, m) 2.88 (1H, dt, Jˆ8.5, 5.0 Hz), 2.42 (3H, s), 1.83–
1.73 (2H, m), 1.53–1.46 (1H, m), 1.33–1.27 (1H, m), and
1.05 (9H, s); IR (neat) 3074, 2962, 2934, 1601, 1473, 1431,
1363, 1178, 1112, 971, 932, 824, 743, 706, 663, and
613 cmϪ1; HR-EI-MS calcd for C25H27O5SSi (MϩϪtBu)
467.1348, found 451.1351.
1
96%) as a colorless oil: Rfˆ0.67 (hexane–EtOAc, 8:1); H
NMR (500 MHz, CDCl3) d 7.70–7.65 (4H, m), 7.43–7.35
(6H, m), 5.64–5.57 (1H, m), 5.44–5.39 (1H, m), 4.26 (2H,
dd, Jˆ0.5, 6.0 Hz), 3.52 (2H, t, Jˆ6.5 Hz), 1.92 (2H, q,
Jˆ7.5 Hz), 1.48 (2H, quin, Jˆ7.5 Hz), 1.04 (9H, s), 0.85
(9H, s), and 0.00 (6H, s); IR (neat) 3074, 3020, 2958, 2934,
2862, 1473, 1429, 1390, 1363, 1257, 1110, 835, 777, 741,
and 702 cmϪ1; HR-CI-MS calcd for C28H45O2Si2 (MϩϩH)
469.2958, found 469.2913.
(4Z)-6-(tert-Butyldiphenylsilyloxy)-4-hexen-1-ol (34). To
a solution of silyl ether 33 (2.72 g, 5.63 mmol) in ethanol
(38 mL) was added pyridinium p-toluenesulfonate (0.141 g,
0.563 mmol) and the mixture was stirred at 25 ЊC for 12 h.
The solvent was removed in vacuo. The residue was purified
by flash chromatography (SiO2, hexane–EtOAc, 10:1) to
give alcohol 34 (1.84 g, 92%) as a colorless oil: Rfˆ0.11
(hexane–EtOAc, 8:1); 1H NMR (500 MHz, CDCl3) d 7.70–
7.66 (4H, m), 7.44–7.36 (6H, m), 5.66–5.61 (1H, m), 5.47–
5.42 (1H, m), 4.25 (2H, dd, Jˆ0.5, 6.0 Hz), 3.66 (1H, t,
Jˆ6.5 Hz), 3.59–3.54 (2H, m), 2.03–1.98 (2H, m), 1.55
(2H, quin, Jˆ6.5 Hz), and 1.04 (9H, s); IR (neat) 3074,
3020, 2934, 2862, 1475, 1429, 1243, 1112, 1075, and
(2Rء,3Rء)-1-(tert-Butyldiphenylsilyloxy)-6-iodo-2,3-epoxy-
hexane (36b). To a solution of tosylate 36a (54.8 mg,
0.104 mmol) in THF (3 mL) was added tetrabutylammo-
nium iodide (46.1 mg, 0.120 mmol) and the solution was
stirred at 65ЊC for 3 h. The solvent was removed in vacuo.
The residue was purified by flash chromatography (SiO2,
hexane–EtOAc, 100:1) to give iodide 36b (40.3 mg, 81%)
as a colorless oil; Rfˆ0.60 (hexane–EtOAc, 4:1); 1H NMR
(200 MHz, CDCl3) d 7.72–7.67 (4H, m), 7.49–7.35 (6H,
m), 3.83 (1H, dd, Jˆ5.6, 11.4 Hz), 3.72 (1H, dd, Jˆ5.4,
11.4 Hz), 3.23–3.11 (3H, m), 3.18–3.08 (1H, m), 2.95
(1H, ddd, Jˆ5.4, 5.6, 7.6 Hz), 2.05–1.87 (2H, m), 1.62–
1.39 (2H, m), and 1.07 (9H, s); IR (neat) 3074, 2962,

7132
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
2934, 1473, 1429, 1392, 1363, 1261, 1228, 1112, 824, 741,
and 704 cmϪ1; HR-EI-MS calcd for C18H20IO2Si (MϩϪtBu)
423.0278, found 423.0276.
m), 3.94–3.90 (1H, m), 3.87–3.82 (1H, m), 3.79–3.74 (1H,
m), 3.70 (2H, d, Jˆ5.5 Hz), 3.61–3.58 (1H, m), 1.93–1.84
(3H, m), 1.71–1.66 (1H, m), and 1.06 (9H, s); IR (neat)
3447, 3070, 2930, 2857, 1589, 1427, 1390, 1362, 1113,
938, 823, 741, and 703 cmϪ1; HR-FD-MS calcd for
C22H31O3SSi (MϩϩH) 371.2067, found 371.2028.
(2Z)-1-(tert-Butyldiphenylsilyloxy)-6-chloromethanesul-
fonyloxy-2-hexene (37). To a solution of alcohol 34
(50.5 mg, 0.140 mmol) in CH2Cl2 (1.4 mL) were added
Et3N (60.0 mL, 0.430 mmol), chloromethanesulfonyl
chloride (20.0 mL, 0.210 mmol), and 4-(dimethyl)amino-
pyridine (1.7 mg, 0.014 mmol) at 0ЊC. The mixture was
stirred at 0ЊC for 2.5 h, diluted with Et2O (40 mL), and
washed with satd aq. NH4Cl (20 mL). The organic layer
was separated, washed with satd aq. NaHCO3 (20 mL) and
brine (20 mL), dried over MgSO4, and concentrated in
vacuo. The residue was purified by flash chromatography
(SiO2, hexane–EtOAc, 20:1) to give sulfonate 37 (57.6 mg,
(2Sء,3Sء)-2-[(tert-Butyldiphenylsilyloxy)methyl]-3-chlor-
omethanesulfonyloxyoxane (39c) and (2Rء,10Rء)-2-[20-
(tert-butyldiphenylsilyloxy)-10-chloromethanesulfonyl-
oxy]ethyloxolane (40c). General Procedure. A solution of
epoxide 36c (19.4 mg, 0.0402 mmol) in 1,2-dichloroethane
(1.3 mL) was heated at reflux for 48 h. The solvent was
removed in vacuo. The residue was purified by PTLC
(SiO2, hexane–acetone, 4:1) to give oxane 39c (8.1 mg,
42%), oxolane 40c (4.1 mg, 21%), and recovered 36c
(3.3 mg, 17%). 39c: a colorless oil; Rfˆ0.50 (developed
twice with hexane–acetone, 4:1); 1H NMR (500 MHz,
C6D6) d 7.84–7.79 (4H, m, Ph), 7.30–7.20 (6H, m, Ph),
5.01 (1H, broad s, C3H), 4.04 (1H, dd, Jˆ7.5, 10.5 Hz,
1
86%) as a colorless oil: Rfˆ0.44 (hexane–EtOAc, 4:1); H
NMR (500 MHz, CDCl3) d 7.69–7.66 (4H, m), 7.45–7.37
(6H, m), 5.71–5.66 (1H, m), 5.40–5.35 (1H, m), 4.52 (1H,
d, Jˆ20.0 Hz), 4.48 (1H, d, Jˆ20.0 Hz), 4.30 (2H, t,
Jˆ7.0 Hz), 4.25 (2H, d, Jˆ6.0 Hz), 2.03–1.98 (2H, m),
1.75 (2H, quin, Jˆ7.0 Hz), and 1.05 (9H, s); IR (neat)
3071, 3017, 2931, 2857, 1472, 1428, 1371, 1252, 1176,
1112, 938, 880, 824, 742, and 704 cmϪ1; HR-CI-MS calcd
for C23H32ClO4SSi (MϩϩH) 467.1479, found 467.1474.
0
0
C1 H), 4.00 (1H, dd, Jˆ6.0, 10.5 Hz, C1 H), 3.83 (1H, d,
Jˆ12.5 Hz, SO2CH2Cl), 3.70 (1H, d, Jˆ12.5 Hz,
SO2CH2Cl), 3.67–3.63 (1H, m, C6Ha), 3.33 (1H, ddd,
Jˆ1.0, 6.0, 7.5 Hz, C2H), 2.93–2.88 (1H, m, C6Hb), 2.17–
2.12 (1H, m, C4H), 1.87 (1H, tq, Jˆ4.5, 13.5 Hz, C5H), 1.23
(9H, s, tBu), 1.16–1.09 (1H, m, C4H), and 0.83–0.77 (1H,
m, C5H); 13C NMR (125 MHz, C6D6) d 136.02, 135.96,
133.69, 133.45, 130.16, 130.14, 128.29, 78.39, 77.13,
67.59, 63.23, 53.77, 28.92, 27.08, 20.23, and 19.39;19 IR
(neat) 3076, 3020, 2960, 2862, 1377, 1180, 1094, 905,
824, 795, 741, and 704 cmϪ1; HR-FD-MS calcd for
C19H22ClO5SSi (MϩϪtBuϩH) 425.0642, found 425.0648.
40c: a colorless oil; Rfˆ0.54 (developed twice with
hexane–acetone, 4:1); 1H NMR (500 MHz, CDCl3) d
7.69–7.65 (4H, m), 7.47–7.39 (6H, m), 4.77 (1H, d,
Jˆ12.5 Hz), 4.74 (1H, d, Jˆ12.5 Hz), 4.68–4.65 (1H, m),
4.14–4.10 (1H, m), 3.91 (1H, dd, Jˆ5.5, 12.0 Hz), 3.88–
3.76 (3H, m), 2.01–1.84 (2H, m), 1.65–1.57 (2H, m), and
1.07 (9H, s); IR (neat) 2930, 2857, 1472, 1427, 1390, 1362,
1113, 823, 740, and 702 cmϪ1; HR-CI-MS calcd for
C23H32ClO5SSi (MϩϩH) 483.1428, found 483.1421.
(2Rء,3Rء)-1-(tert-Butyldiphenylsilyloxy)-6-chloromethane-
sulfonyloxy-2,3-epoxyhexane (36c). To a solution of olefin
37 (243 mg, 0.520 mmol) in CH2Cl2 (12 mL) cooled to 0ЊC
were added Na2HPO4 (437 mg, 3.07 mmol) and m-chloro-
perbezoic acid (379 mg, 1.54 mmol) and the solution was
stirred at 25ЊC for 3.5 h, and satd aq. NaHCO3 (4 mL) and
satd aq. Na2S2O3 (4 mL) was added at 0ЊC. The mixture was
extracted with Et2O (80 mL) and the Et2O layer was washed
with satd aq. Na2S2O3 (60 mL), satd aq. NaHCO3
(3×60 mL) and brine (60 mL), dried over MgSO4, and
concentrated in vacuo. The residue was purified by flash
chromatography (SiO2, hexane–EtOAc, 20:1) to give
epoxide 36c (481 mg, 76%) as a colorless oil; Rfˆ0.40
(hexane–EtOAc, 4:1); 1H NMR (500 MHz, CDCl3) d
7.69–7.67 (4H, m), 7.46–7.39 (6H, m), 4.57 (1H, d,
Jˆ12.5 Hz), 4.54 (1H, d, Jˆ12.5 Hz), 4.43 (1H, dt,
Jˆ10.0, 6.5 Hz), 4.38 (1H, dt, Jˆ10.0, 6.5 Hz), 3.82 (1H,
dd, Jˆ6.0, 11.5 Hz), 3.73 (dd, Jˆ5.5, 11.5 Hz), 3.19–3.16
(1H, m), 2.97 (1H, dt, Jˆ8.5, 4.5 Hz), 1.96–1.90 (2H, m),
1.66–1.59 (1H, m), 1.44–1.37 (1H, m), and 1.06 (9H, s); IR
(neat) 3071, 2956, 2858, 1655, 1589, 1560, 1542, 1508,
1472, 1428, 1372, 1253, 1177, 1112, 936, 881, 824, 742,
704, and 612 cmϪ1; HR-FD-MS calcd for C19H22ClO5SSi
(MϩϪtBu) 425.0676, found 425.0624.
(2Rء,3Rء)-6-Acetoxy-1-(tert-butyldiphenylsilyloxy)-2,3-
epoxyhexane (36e). To a solution of iodide 36b (78.2 mg,
0.163 mmol) in THF/H2O (5.4 mL, 5:1) was added AgOTf
(50.2 mg, 0.195 mmol) at 25ЊC. After stirring at 25ЊC for
1 h, AgOTf (50.2 mg, 0.195 mmol) was added again. The
solution was stirred at 25ЊC for 1 h, and poured into Et2O
(30 mL). The mixture was washed with satd aq. NaHCO3
(15 mL) and brine (15 mL). The organic layer was dried
over MgSO4, and concentrated in vacuo. To a solution of
the residue (79.5 mg) in CH2Cl2 (1.6 mL) were added
pyridine (0.0527 mL, 0.651 mmol), acetic anhydride
(0.0307 mL, 0.326 mmol), and 4-(dimethylamino)pyridine
(2.0 mg, 0.016 mmol) at 0ЊC. The solution was stirred at
25ЊC for 30 min, and was poured into Et2O (35 mL). The
mixture was washed with satd aq. NaHCO3 (15 mL) and
brine (15 mL). The organic layer was dried over MgSO4,
and concentrated in vacuo. PTLC (SiO2, hexane–EtOAc,
3:1) gave acetate 36e (13.7 mg, 20% for 2 steps) as a color-
less oil: Rfˆ0.29 (hexane–EtOAcˆ4:1); 1H NMR
(500 MHz, CDCl3) d 7.69–7.67 (4H, 5m), 7.45–7.37 (6H,
m), 4.09–4.01 (2H, m), 3.80 (1H, dd, Jˆ6.0, 11.5 Hz), 3.73
(2Rء,10Rء)-2-[20-(tert-Butyldiphenylsilyloxy)-10-hydroxy]-
ethyloxolane (40b). To a solution of 36b (27.8 mg,
0.0579 mmol) in THF/H2O (1.8 mL, 5:1) was added silver
triflate (AgOTf) (22.3 mg, 0.0869 mmol) at three times
every 1 h at 25ЊC. After stirred for 3 h from the first addition
of AgOTf, the reaction mixture was poured into satd aq.
NaHCO3 (10 mL) at 25ЊC and extracted with Et2O
(40 mL). The Et2O layer was washed with brine (10 mL),
dried over MgSO4, and concentrated in vacuo. PTLC (SiO2,
hexane–EtOAc, 4:1) gave oxolane 40b (16.2 mg, 75%) as a
colorless oil: Rfˆ0.29 (hexane–EtOAc, 4:1); 1H NMR
(500 MHz, CDCl3) d 7.68–7.66 (4H, m), 7.45–7.37 (6H,

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7133
(1H, dd, Jˆ5.0, 11.5 Hz), 3.17 (1H, dt, Jˆ4.5, 5.5 Hz), 3.70
(2H, d, Jˆ5.5 Hz), 2.97 (1H, dt, Jˆ7.0, 4.5 Hz), 2.00 (3H,
s), 1.82–1.69 (2H, m), 1.54–1.38 (2H, m), and 1.06 (9H, s);
IR (neat) 3074, 3054, 2962, 2934, 2896, 2862, 1742, 1475,
1431, 1392, 1367, 1243, 1112, 824, 743, and 613 cmϪ1; HR-
EI-MS calcd for C20H23O4Si (MϩϪtBu) 355.1365, found
355.1365.
4.29 (1H, dt, Jˆ15.0, 2.0 Hz), 4.19 (1H, dt), Jˆ15.0,
2.0 Hz), 3.86–3.80 (1H, m), 3.81 (3H, s), 3.54–3.50 (1H,
m), 3.44 (2H, t, Jˆ6.5 Hz), 2.99–2.98 (2H, m), 2.13 (2H, q,
Jˆ7.0 Hz), 1.87–1.78 (1H, m), and 1.75–1.51 (8H, m); IR
(neat) 2944, 2862, 1613, 1586, 1516, 1456, 1363, 1303,
1249, 1203, 1176, 1100, 1025, 973, 946, 903, 872, 818,
and 708 cmϪ1; HR-EI-MS calcd for C22H30O4 (Mϩ)
358.2144, found 358.2163.
(2Z)-6-(40-Methoxybenzyloxy)-2-hexen-1-ol (46). To a
solution of propargylic alcohol 45 (1.45 g, 6.19 mmol) in
pyridine (47 mL) were added quinoline (0.13 mL) and Pd/
CaCO2 (0.145 g) at 25ЊC and the mixture was stirred at 25ЊC
for 1.5 h under hydrogen atmosphere and filtered through
Celite. The filtrate was concentrated in vacuo. The residue
was purified by flash chromatography (SiO2, hexane–
EtOAc, 3:1) to give allylic alcohol 46 (1.44 g, 98%) as a
(5Z)-9-(40-Methoxybenzyloxy)-5-nonen-2-yn-1-ol (49). To
a solution of tetrahydropyranylether 48 (1.67 g, 4.65 mmol)
in MeOH (25 mL) was added p-toluenesulfonic acid mono-
hydrate (4.0 mg, 0.023 mmol) at 25ЊC. The solution was
stirred at 25ЊC for 3 h, and Et3N (0.65 mL, 4.65 mL) was
added. The mixture was concentrated in vacuo. The residue
was purified by flash chromatography (SiO2, hexane–
EtOAc, 5:1) to give alcohol 49 (1.19 g, 93%) as a colorless
1
colorless oil: H NMR (500 MHz, CDCl3) d 7.26 (2H, d,
1
Jˆ8.5 Hz), 6.88 (2H, d, Jˆ8.5 Hz), 5.63–5.69 (1H, m), 5.53
(1H, dt, Jˆ10.5, 7.5 Hz), 4.42 (2H, t, Jˆ6.0 Hz), 3.81 (3H,
s), 3.46 (2H, t, Jˆ7.0 Hz), 2.20 (2H, q, Jˆ7.0 Hz), 1.68 (2H,
qui, Jˆ7.0 Hz), and 1.55 (1H, t, Jˆ6.0 Hz); IR (neat) 3356,
3014, 2938, 2866, 1613, 1586, 1516, 1464, 1356, 1303,
1249, 1176, 1098, 1036, and 820 cmϪ1; HR-EI-MS calcd
for C14H20O3 (Mϩ) 236.1412, found 236.1404.
oil: Rfˆ0.58 (hexane–EtOAcˆ1:1); H NMR (500 MHz,
CDCl3) d 7.27 (2H, d, Jˆ8.5 Hz), 6.88 (2H, d, Jˆ8.5 Hz),
5.49–5.41 (2H, m), 4.43 (2H, s), 4.23 (2H, dt, Jˆ6.0,
2.5 Hz), 3.81 (3H, s), 3.45 (2H, t, Jˆ6.0 Hz), 2.97–2.96
(2H, m), 2.15 (2H, q, Jˆ6.5 Hz), 1.70–164 (2H, m), and
1.64 (1H, t, Jˆ6.0 Hz); IR (neat) 3414, 2936, 2866, 1613,
1586, 1516, 1458, 1365, 1303, 1249, 1176, 1098, 1033, 820,
and 708 cmϪ1; HR-EI-MS calcd for C17H22O3 (Mϩ)
274.1569, found 274.1541.
(2Z)-1-Bromo-6-(40-methoxybenzyloxy)-2-hexene (47).
To a solution of allylic alcohol 46 (2.81 g, 11.9 mmol) in
CH2Cl2 (119 mL) were added triphenylphosphine (3.75 g,
14.3 mmol) and carbon tetrabromide (5.94 g, 17.9 mmol) at
0ЊC and the mixture was stirred at 25ЊC for 10 min. Satd aq.
NaHCO3 (70 mL) was added and the aqueous layer was
extracted with Et2O (3×90 mL). The combined organic
layer was washed by brine (100 mL), dried over MgSO4,
and concentrated in vacuo. The residue was purified by flash
chromatography (SiO2, hexane–EtOAc, 100:1) to give
bromide 47 (3.01 g, 85%) as a colorless oil: Rfˆ0.53
(hexane–EtOAcˆ4:1); 1H NMR (500 MHz, CDCl3) d
7.27 (2H, d, Jˆ8.5 Hz), 6.88 (2H, d, Jˆ8.5 Hz), 5.75 (1H,
dtt, Jˆ10.5, 1.5, 8.5 Hz), 5.58 (1H, dt, Jˆ10.5, 7.5 Hz), 4.44
(2H, s), 4.00 (2H, d, Jˆ8.5 Hz), 3.81 (3H, s), 3.46 (2H, t,
Jˆ6.0 Hz), 2.24 (2H, dq, Jˆ1.5, 7.5 Hz), and 1.68 (2H, qui,
Jˆ6.0 Hz); IR (neat) 2938, 2860, 1613, 1586, 1516, 1464,
1365, 1303, 1249, 1207, 1174, 1100, 1036, 820, and
(2E,5Z)-9-(40-Methoxybenzyloxy)-2,5-nonadien-1-ol (50).
A solution of propargylic alcohol 49 (0.431 g, 1.57 mmol)
in THF (5 mL) was added at 0ЊC dropwise via cannular to a
suspension of LiAlH4 (0.119 mg, 3.14 mmol) in THF
(10 mL), and the mixture was stirred at 25ЊC for 14 h. The
reaction was quenched at 0ЊC with aq. 1 M HCl. The
mixture was extracted with EtOAc (150 mL). The organic
layer was washed with aq. 1 M HCl (50 mL), satd aq.
NaHCO3 (50 mL) and brine (50 mL). The solution was
dried over MgSO4, and concentrated in vacuo. The residue
was purified by flash chromatography (SiO2, hexane–
EtOAc, 4:1) to give allylic alcohol 50 (0.382 g, 88%) as a
colorless oil: Rfˆ0.47 (hexane–EtOAcˆ1:1); 1H NMR
(500 MHz, CDCl3) d 7.26 (2H, d, Jˆ8.5 Hz), 6.88 (2H, d,
Jˆ8.5 Hz), 5.62–5.49 (2H, m), 5.47–5.38 (2H, m), 4.43
(2H, s), 4.09 (2H, t, Jˆ5.0 Hz), 3.81 (3H, s), 3.44 (2H, t,
Jˆ6.5 Hz), 2.79 (2H, t, Jˆ6.0 Hz), 2.13 (2H, q, Jˆ7.0 Hz),
1.69–163 (2H, m), and 1.30 (1H, t, Jˆ5.0 Hz); IR (neat)
3368, 2938, 2864, 1613, 1586, 1516, 1464, 1365, 1303,
1249, 1176, 1098, 1036, 973, and 820 cmϪ1; HR-EI-MS
calcd for C17H24O3 (Mϩ) 276.1725, found 276.1723.
754 cmϪ1
;
HR-EI-MS calcd for C14H19O2Br (Mϩ)
298.0568, found 298.0543.
(5Z)-9-(40-Methoxybenzyloxy)-1-(oxan-20-yloxy)-5-nonen-
2-yne (48). To a solution of 3-tetrahydropyranyloxypropyne
(0.655 g, 4.67 mmol) in THF (14 mL) cooled to Ϫ78ЊC was
added dropwise n-BuLi (3.05 mL of a 1.53 M solution in
hexane, 4.67 mmol), and the mixture was stirred at Ϫ78ЊC
for 30 min. A solution of bromide 47 (1.16 g, 3.89 mmol) in
THF (6 mL) was then added dropwise at Ϫ78ЊC. The solu-
tion was stirred at 25ЊC for 29 h, and satd aq. NH4Cl was
added at 0ЊC. The mixture was extracted with EtOAc
(250 mL). The organic layer was washed with satd aq.
NaHCO3 (50 mL) and brine (50 mL), dried over MgSO4,
and concentrated in vacuo. The residue was purified by
flash chromatography (SiO2, hexane–EtOAc, 20:1) to give
acetylenic compound 48 (3.01 g, 85%) as a colorless oil:
Rfˆ0.44 (hexane–EtOAcˆ4:1); 1H NMR (500 MHz,
CDCl3) d 7.26 (2H, d, Jˆ8.5 Hz), 6.88 (2H, d, Jˆ8.5 Hz),
5.48–5.41 (2H, m), 4.80 (1H, t, Jˆ3.5 Hz), 4.43 (2H, s),
(2Sء,3Sء,5Z)-9-(40-Methoxybenzyloxy)-2,3-epoxy-5-nonen-
1-ol (51). To a solution of allylic alcohol 50 (0.298 g,
1.08 mmol) in benzene (11 mL) were added VO(acac)2
(28.6 mg, 0.108 mmol) and t-butylhydroperoxide (0.95 mL
of a 1.37 M solution in toluene, 1.30 mmol) at 25ЊC. The
mixture was stirred at 25ЊC for 1 h, and diluted with EtOAc
(150 mL). The solution was washed with aq. 1 M HCl
(50 mL), satd aq. NaHCO3 (50 mL), satd aq. Na2S2O3
(50 mL), and brine (50 mL), dried over MgSO4, and concen-
trated in vacuo. The residue was purified by flash chromato-
graphy (SiO2, hexane–EtOAc, 2:1) to give epoxide 51
(0.270 g, 85%) as a colorless oil: Rfˆ0.40 (hexane–
1
EtOAcˆ1:1); H NMR (500 MHz, CDCl3) d 7.26 (2H, d,
Jˆ8.5 Hz), 6.88 (2H, d, Jˆ8.5 Hz), 5.56–5.51 (1H, m),

7134
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
5.44–5.39 (1H, m), 4.43 (2H, s), 3.89 (1H, ddd, Jˆ2.5, 6.5,
12.5 Hz), 3.81 (3H, s), 3.62 (1H, ddd, Jˆ4.5, 7.0, 12.5 Hz),
3.44 (2H, t, Jˆ6.5 Hz), 2.99 (1H, dt, Jˆ2.5, 5.5 Hz), 2.95
(1H, dt, Jˆ4.5, 2.5 Hz), 2.40 (1H, dt, Jˆ15.0, 5.5 Hz), 2.32
(1H, dt, Jˆ15.0, 5.5 Hz), 2.13 (2H, q, Jˆ7.0 Hz), 1.69–164
(2H, m), and 1.30 (1H, m); IR (neat) 3408, 2938, 2864,
1613, 1516, 1464, 1365, 1303, 1249, 1176, 1098, 1035,
and 820 cmϪ1; HR-EI-MS calcd for C17H24O4 (Mϩ)
292.1675, found 292.1642.
NaHCO3 (20 mL) and brine (20 mL), dried over MgSO4,
and concentrated in vacuo. The residue was purified by
flash chromatography (SiO2, hexane–EtOAc, 10:1 to 5:1)
to give sulfonate 54 (0.495 g, 79%) as a colorless oil:
Rfˆ0.44 (hexane–EtOAc, 4:1); 1H NMR (500 MHz,
CDCl3) d 7.69–7.66 (4H, m), 7.45–7.37 (6H, m), 5.52–
5.45 (2H, m), 4.59 (2H, s), 4.41 (2H, t, Jˆ6.5 Hz), 3.80
(1H, dd, Jˆ3.5, 12.0 Hz), 3.73 (1H, dd, Jˆ4.5, 12.0 Hz),
2.95–2.93 (1H, m), 2.85 (1H, dt, Jˆ2.0, 5.0 Hz), 2.33 (2H, t,
Jˆ5.0 Hz), 2.21–2.17 (2H, m), 1.71 (1H, t, Jˆ5.5 Hz),
1.88–1.83 (2H, m), and 1.05 (9H, s); IR (neat) 2960,
2934, 2862, 1473, 1429, 1373, 1255, 1178, 1114, 938,
882, 824, 795, 743, and 704 cmϪ1; HR-EI-MS calcd for
C22H26O5SiSCl (MϩϪtBu) 465.0959, found 465.0954.
(7Sء,8Sء,4Z)-9-(tert-Butyldiphenylsilyloxy)-1-(40-methoxy-
benzyloxy)-7,8-epoxy-4-nonene (52). To a solution of
alcohol 51 (0.359 g, 1.23 mmol) in CH2Cl2 (7.5 mL) cooled
to 0ЊC were added imidazole (0.201 g, 2.95 mmol) and tert-
butyldiphenylsilyl chloride (0.385 mL, 1.48 mmol). The
solution was stirred at 0ЊC for 10 min, diluted with Et2O
(70 mL). The mixture was washed with satd aq. NH4Cl
(20 mL), satd aq. NaHCO3 (20 mL), and brine (20 mL),
dried over MgSO4, and concentrated in vacuo. The residue
was purified by flash chromatography (SiO2, hexane–
EtOAc, 20:1) to give silyl ether 52 (0.652 g, Ͼ99%) as a
colorless oil: Rfˆ0.78 (hexane–EtOAcˆ1:1); 1H NMR
(500 MHz, CDCl3) d 7.73–7.66 (4H, m), 7.44–7.36 (6H,
m), 7.25 (2H, d, Jˆ8.5 Hz), 6.87 (2H, d, Jˆ8.5 Hz), 5.54–
5.49 (1H, m), 5.42–5.37 (1H, m), 4.42 (2H, s), 3.80 (3H, s),
3.78 (1H, dd, Jˆ3.5, 12.0 Hz), 3.72 (1H, dd, Jˆ4.5,
12.0 Hz), 3.44 (2H, t, Jˆ6.5 Hz), 2.93–2.91 (1H, m), 2.83
(1H, dt, Jˆ2.0, 5.5 Hz), 2.37 (1H, dt, Jˆ13.5, 5.5 Hz), 2.26
(1H, dt, Jˆ13.5, 5.5 Hz), 2.12 (2H, q, Jˆ7.0 Hz), 1.69–1.63
(2H, m), and 1.05 (9H, s); IR (neat) 2934, 2860, 1613, 1516,
1464, 1429, 1114, 824, 741, and 702 cmϪ1; HR-FD-MS
calcd for C33H42O4Si (Mϩ) 530.2852, found 530.2827.
(2Sء,8Sء,5Rء,6Sء)-1-(tert-Butyldiphenylsilyloxy)-9-chloro-
methanesulfonyloxy-2,3,5,6-diepoxy-nonane (55a) and
(2Rء,8Rء,5Rء,6Sء)-1-(tert-butyldiphenylsilyloxy)-9-chloro-
methanesulfonyloxy-2,3,5,6-diepoxynonane (55b). To a
solution of olefin 54 (76.6 mg, 0.147 mmol) in CH2Cl2
(3 mL) cooled to 0ЊC were added Na2HPO4 (0.104 g,
0.735 mmol) and m-chloroperbenzoic acid (90.7 mg,
0.368 mmol, purity 70%). The solution was stirred at 25ЊC
for 2.5 h. Satd aq. NaHCO3 (2 mL) and satd aq. Na2S2O3
(2 mL) were then added at 0ЊC. The mixture was extracted
with EtOAc (35 mL). The organic layer was washed with
satd aq. Na2S2O3 (10 mL), satd aq. NaHCO3 (10 mL) and
brine (10 mL), dried over MgSO4, and concentrated in
vacuo. The residue was purified by flash chromatography
(SiO2, hexane–EtOAc, 5:1 to 2:1) to give a mixture of
diepoxides 55a and 55b (75.7 mg, 95%) as a colorless oil:
Rfˆ0.31 (hexane–EtOAc, 2:1); 1H NMR (500 MHz,
CDCl3) d 7.69–7.67 (4H×2, m), 7.45–7.38 (6H×2, m),
4.64 (1H×2, Jˆ13.0 Hz), 4.59 (1H×2, Jˆ13.0 Hz), 4.53–
4.44 (2H×2, m), 3.84 (1H, dd, Jˆ4.0, 12.0 Hz), 3.83 (1H,
dd, Jˆ4.0, 12.0 Hz), 3.77 (1H, dd, Jˆ4.0, 12.0 Hz), 3.76
(1H, dd, Jˆ5.0, 12.0 Hz), 3.13–3.10 (1H, m), 3.06–3.02
(1H×2, m), 3.01–2.95 (5H, m), 2.05 (2H×2, m), 1.92–
1.87 (2H, m), 1.84–1.73 (2H, m), 1.64–1.51 (2H, m), and
1.06 (9H×2, s); IR (neat) 2960, 2862, 1736, 1591, 1473,
1431, 1375, 1255, 1180, 917, 824, 795, 741, 704, 613, and
547 cmϪ1; HR-FD-MS calcd for C26H36O6SiSCl (MϩϩH)
538.1612, found 539.1682.
(7Sء,8Sء,4Z)-9-(tert-Butyldiphenylsilyloxy)-7,8-epoxy-4-
nonen-1-ol (53). To a solution of 4-methoxybenzyl ether 52
(0.652 g, 1.23 mmol) in CH2Cl2/H2O (39.6 mL, 10:1)
cooled to 0ЊC was added dichlorodicyanobenzoquinone
(0.420 g, 1.85 mmol) and the solution was stirred at 25ЊC
for 1 h. Satd NaHCO3 was added at 0ЊC and the mixture was
extracted with Et2O (120 mL). The organic layer was
washed with water (3×40 mL), dried over MgSO4, concen-
trated in vacuo. The residue was purified by flash chromato-
graphy (SiO2, hexane–EtOAc, 20:1 to 2:1) to give alcohol
53 (0.491 g, 98%) as a colorless oil: Rfˆ0.58 (hexane–
1
EtOAcˆ1:1); H NMR (500 MHz, CDCl3) d 7.68–7.66
Ring expansion reaction of 55a and 55b. A solution of the
mixture of diepoxide 55a and 55b (30.7 mg, 0.0569 mmol)
in 1,2-dichloroethane (1.9 mL) was heated at reflux for 22 h.
The solvent was removed in vacuo. The residue was purified
by PTLC (SiO2, hexane–acetone, 3:1) to give a mixture of
oxanes 56a and 56b (7.7 mg, 25%) as a colorless oil:
(4H, m), 7.45–7.37 (6H, m), 5.55–5.49 (1H, m), 5.45–
5.40 (1H, m), 3.78 (1H, dd, Jˆ4.0, 12.0 Hz), 3.74 (1H,
dd, Jˆ4.5, 12.0 Hz), 3.63 (2H, q, Jˆ5.5 Hz), 2.97–2.95
(1H, m), 2.86 (1H, dt, Jˆ2.0, 5.0 Hz), 2.35 (2H, broad t,
Jˆ6.0 Hz), 2.21–2.08 (2H, m), 1.71 (1H, t, Jˆ5.5 Hz),
1.66–1.58 (2H, m), and 1.05 (9H, s); IR (neat) 3352,
2934, 2862, 1473, 1429, 1392, 1363, 1114, 1060, 866,
824, 797, 741, and 704 cmϪ1; HR-FD-MS calcd for
C25H34O3Si (Mϩ) 410.2277, found 410.2256.
1
Rfˆ(hexane–EtOAc, 2:1); H NMR (500 MHz, CDCl3) d
7.69–7.67 (4H×2, m), 7.45–7.38 (6H×2, m), 4.84 (1H,
broad s), 4.78 (1H, broad s), 4.72 (1H, d, Jˆ12.5 Hz),
4.64 (2H, s), 4.61 (1H, d, Jˆ12.5 Hz), 3.81 (1H, dd,
Jˆ3.5, 12.0 Hz), 3.79 (1H, dd, Jˆ3.5, 12.0 Hz), 3.73 (1H,
dd, Jˆ4.0, 12.0 Hz), 3.71 (1H, dd, Jˆ5.0, 12.0 Hz), 3.67
(1H, dd, Jˆ4.5, 9.0 Hz), 3.57–3.54 (1H, m), 3.53 (1H, dt,
Jˆ2.5, 11.0 Hz), 3.48 (1H, dt, Jˆ2.5, 12.5 Hz), 3.03–2.97
(2H×2, m), 2.41–2.34 (1H×2, broad d, Jˆ14.5 Hz), 2.13
(1H, ddd, Jˆ3.0, 9.0, 14.5 Hz), 2.09–1.79 (5H, m), 1.53–
1.45 (4H, m), 1.051 (9H, s), and 1.048 (9H, s); IR (neat)
3076, 3018, 2960, 2934, 2862, 1473, 1431, 1375, 1255,
1212, 1180, 1112, 1089, 824, 795, 741, 704, 613, and
(7Sء,8Sء,4Z)-9-(tert-Butyldiphenylsilyloxy)-1-chloro-
methanesulfonyloxy-7,8-epoxy-4-nonene (54). To a solu-
tion of alcohol 53 (0.491 g, 1.20 mmol) in CH2Cl2 (12 mL)
were added pyridine (0.388 mL, 4.80 mmol), chloro-
methanesulfonyl chloride (0.214 mL, 2.40 mmol) at 0ЊC.
After stirring at 0ЊC for 2 h, satd aq. NaHCO3 was added.
The mixture was extracted with Et2O (70 mL). The organic
layer was washed with satd aq. NH4Cl (20 mL), satd aq.

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7135
547 cmϪ1; HR-FD-MS calcd for C26H35O6ClSiS (MϩϩH)
539.1736, found 539.1713.
dichlorodicyanobenzoquinone (0.334 g, 1.47 mmol) and
the mixture was stirred at 25ЊC for 2 h. Satd aq. NaHCO3
(30 mL) was added at 0ЊC and the mixture was extracted
with Et2O (150 mL). The organic layer was washed with
water (3×50 mL) and brine (50 mL), dried over MgSO4,
and concentrated in vacuo. The residue was purified by
flash chromatography (SiO2, hexane–EtOAc, 20:1 to 5:1)
to give alcohol 62 (0.327 g, 81%) as a colorless oil: Rfˆ0.33
(hexane–EtOAcˆ3:1); 1H NMR (500 MHz, CDCl3) d
7.70–7.67 (4H, m), 7.43–7.36 (6H, m), 5.66 (1H, dt,
Jˆ15.0, 5.0 Hz), 5.56 (1H, dt, Jˆ15.0, 5.0 Hz), 5.42–5.32
(2H, m), 4.15 (2H, d, Jˆ5.0 Hz), 3.65 (2H, t, Jˆ6.0 Hz),
2.15–2.06 (6H, m), 1.66–160 (2H, m), and 1.05 (9H, s); IR
(neat) 3322, 3074, 3052, 2934, 1591, 1473, 1381, 1189,
1112, 1054, 971, 824, 739, 702, and 613 cmϪ1; HR-FD-
MS calcd for C26H37O2Si (MϩϩH) 409.2563, found
409.2584.
(2E,6Z)10-(40-Methoxybenzyloxy)-2,6-decadien-1-ol (60).
To a solution of pyridinium chlorochromate (2.59 g,
12.0 mmol) in CH2Cl2 (52 mL) were added molecular
sieves 4A powder (2.7 g) and a solution of alcohol 59
(1.58 g, 5.98 mmol) in CH2Cl2 (6 mL) at 25ЊC. After stirring
at 25ЊC for 1.5 h, Et2O (100 mL) and MgSO4 (24 g) was
added. The mixture was stirred at 25ЊC for 10 min, filtered
through Florisil, and concentrated in vacuo. To a solution of
the residue in benzene (50 mL) was added methyl (tri-
phenylphosphoranyliden)acetate (2.4 g, 7.18 mmol) at
25ЊC. After stirring at 25ЊC for 24 h, the solvent was
removed in vacuo. The residue was purified by flash
chromatography (SiO2, hexane–EtOAc, 10:1) to give a
mixture of the desired ester and an unknown compound.
To a solution of this mixture in CH2Cl2 (50 mL) cooled at
Ϫ78ЊC was added dropwise diisobutylaluminum hydride
(15.8 mL of a 0.95 M in hexane, 15.0 mmol). After stirring
at Ϫ78ЊC for 1 h, Et2O (100 mL), water (1.2 mL), and aq.
4 M NaOH (0.6 mL) were added at 0ЊC. The mixture was
stirred at 25ЊC until white precipitates were sedimented.
After the addition of MgSO4, the mixture was filtered
through Celite. The filtrate was concentrated in vacuo.
The residue was purified by flash chromatography (SiO2,
hexane–EtOAc, 3:1) to give allylic alcohol 60 (1.17 g,
67% for 3 steps) as a colorless oil: Rfˆ0.24 (hexane–
(2E,6Z)-1-(tert-Butyldiphenylsilyloxy)-10-chloromethane-
sulfonyloxy-2,6-decadiene (63). To a solution of alcohol 62
(0.106 mg, 0.260 mmol) in CH2Cl2 (2.6 mL) were added
pyridine (0.109 mL, 1.35 mmol), chloromethanesulfonyl
chloride (0.0603 mL, 0.676 mmol) at 0ЊC. After stirring at
0ЊC for 1.5 h, satd aq. NaHCO3 was added and the mixture
was extracted with Et2O (35 mL). The organic layer was
washed with satd aq. NH4Cl (10 mL), satd aq. NaHCO3
(10 mL) and brine (10 mL), dried over MgSO4, and concen-
trated in vacuo. The residue was purified by flash chromato-
graphy (SiO2, hexane–EtOAc, 25:1) to give sulfonate 63
(0.124 g, 92%) as a colorless oil: Rfˆ0.51 (hexane–
1
EtOAc, 3:1); H NMR (500 MHz, CDCl3) d 7.26 (2H, d,
Jˆ8.5 Hz), 6.88 (2H, d, Jˆ8.5 Hz), 5.72–5.61 (2H, m),
5.42–5.33 (2H, m), 4.43 (2H, s), 4.09 (2H, d, Jˆ5.5 Hz),
3.81 (3H, s), 3.44 (2H, t, Jˆ6.5 Hz), 2.18–2.06 (6H, m), and
1.68–162 (2H, m); IR (neat) 3374, 3008, 2938, 2860, 2364,
1613, 1516, 1458, 1365, 1303, 1249, 1176, 1098, 1036,
1006, 971, and 820 cmϪ1; HR-EI-MS calcd for C18H26O3
(Mϩ) 290.1882, found 290.1902.
1
EtOAc, 3:1); H NMR (500 MHz, CDCl3) d 7.70–7.67
(4H, m), 7.44–7.36 (6H, m), 5.69–5.64 (1H, m), 5.57
(1H, dt, Jˆ15.0, 4.0 Hz), 5.48–5.43 (1H, m), 5.37–5.32
(1H, m), 4.57 (2H,s), 4.40 (2H, t, Jˆ6.5 Hz), 4.16 (2H, d,
Jˆ5.0 Hz), 2.18 (2H, q, Jˆ8.0 Hz), 2.12–2.08 (4H, m),
1.84 (2H, qui, Jˆ6.5 Hz), and 1.05 (9H, s); IR (neat)
3018, 2934, 2860, 1473, 1431, 1375, 1178, 1112, 1052,
(2E,6Z)-1-(tert-Butyldiphenylsilyloxy)-10-(40-methoxy-
benzyloxy)-2,6-decadiene (61). To a solution of alcohol 60
(0.596 g, 2.05 mmol) in CH2Cl2 (20 mL) cooled to 0ЊC were
added imidazole (0.355 g, 4.92 mmol) and tert-butyl-
diphenylsilyl chloride (0.640 mL, 2.46 mmol). The mixture
was stirred at 0ЊC for 30 min, and poured into Et2O
(150 mL). The solution was washed with satd aq.
NaHCO3 (50 mL) and brine (50 mL), dried over MgSO4,
and concentrated in vacuo. The residue was purified by
flash chromatography (SiO2, hexane–EtOAc, 100:1 to
50:1) to give silyl ether 61 (1.05 g, 97%) as a colorless
1000, 967, 938, 880, 824, 741, 704, and 611 cmϪ1
;
HR-FD-MS calcd for C23H28O4ClSiS (MϩϪtBu)
463.1166, found 463.1172.
(2E,6Rء,7Sء)-1-(tert-Butyldiphenylsilyloxy)-10-chloro-
methanesulfonyloxy-6,7-epoxy-2-decene (64). To a solu-
tion of olefin 63 (0.124 g, 0.238 mmol) in CH2Cl2 (2 mL)
cooled to 0ЊC were added Na2HPO4 (67.6 mg, 0.476 mmol)
and m-chloroperbenzoic acid (28.8 mg, 0.238 mmol, purity
70%). The solution was stirred at 25ЊC for 30 min. Satd aq.
NaHCO3 (2 mL) and satd aq. Na2S2O3 (2 mL) were then
added at 0ЊC. The mixture was extracted with EtOAc
(38 mL). The organic layer was washed with satd aq.
Na2S2O3 (10 mL), satd aq. NaHCO3 (10 mL) and brine
(10 mL), dried over MgSO4, and concentrated in vacuo.
The residue was purified by flash chromatography (SiO2,
hexane–EtOAc, 25:1 to 6:1) to give epoxide 64 (54.9 mg,
1
oil: Rfˆ0.62 (hexane–EtOAcˆ3:1); H NMR (500 MHz,
CDCl3) d 7.69–7.67 (4H, m), 7.43–7.35 (6H, m), 7.25
(2H, d, Jˆ8.5 Hz), 6.87 (2H, d, Jˆ8.5 Hz), 5.66 (1H, dt,
Jˆ15.0, 5.0 Hz), 5.56 (1H, dt, Jˆ15.0, 5.0 Hz), 5.42–5.33
(2H, m), 4.42 (2H, s), 4.15 (2H, d, Jˆ5.0 Hz), 3.79 (3H, s),
3.44 (2H, t, Jˆ6.5 Hz), 2.13–2.05 (6H, m), 1.65 (2H, qui,
Jˆ6.5 Hz), and 1.05 (9H, s); IR (neat) 3004, 2934, 2860,
1613, 1589, 1516, 1473, 1464, 1429, 1363, 1303, 1249,
1174, 1112, 1040, 971, 824, 741, 704, and 611 cmϪ1; HR-
FD-MS calcd for C34H44O3Si (Mϩ) 528.3060, found
528.3079.
1
43%) as a colorless oil: Rfˆ0.37 (hexane–EtOAc, 3:1); H
NMR (500 MHz, CDCl3) d 7.69–7.67 (4H, m), 7.44–7.36
(6H, m), 5.69 (1H, ddt, Jˆ1.5, 16.0, 6.0 Hz), 5.60 (1H, dt,
Jˆ16.0, 5.0 Hz), 4.63 (1H, d, Jˆ13.0 Hz), 4.53 (1H, d,
Jˆ13.0 Hz), 4.51 (1H, dt, Jˆ10.0, 6.0 Hz), 4.46 (1H, dt,
Jˆ10.0, 6.5 Hz), 4.17 (2H, dd, Jˆ1.5, 5.0 Hz), 2.97–2.93
(2H, m), 2.28–2.14 (2H, m), 2.03–1.97 (2H, m), 1.81–1.75
(2H, m), 1.62–1.52 (2H, m), and 1.06 (9H, s); IR
(neat) 2960, 2934, 1473, 1429, 1375, 1255, 1180, 1114,
(4Z,8E)-10-(tert-Butyldiphenylsilyloxy)-4,8-decadien-1-ol
(62). To a solution of ether 61 (0.520 g, 0.983 mmol) in
CH2Cl2/H2O (33 mL, 10:1) cooled to 0ЊC was added

7136
N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
1052, 936, 882, 824, 743, 704, and 611 cmϪ1; HR-FD-MS
calcd for C27H38O5ClSiS (MϩϩH) 537.1898, found
537.1905.
and 704 cmϪ1; HR-FD-MS calcd for C27H38O6ClSiS
(MϩϩH) 553.1846, found 553.1833.
Spiro acetal (70). A solution of 67 (8.4 mg, 0.0152 mmol)
in 1,2-dichloroethane (3 mL) was heated at reflux for 21 h.
The solvent was removed in vacuo. The residue was purified
by PTLC (SiO2, hexane–EtOAc, 3:1) to give the mixture of
70 and 71 (3.5 mg, 70:71ˆ3.4:1, 42%). The pure 70
(2.4 mg) was obtained by preparative HPLC (hexane, flow
rate 0.5 mL/min, retention time 10.5 min) as a colorless oil;
1H NMR (500 MHz, CDCl3) d 7.76–7.12 (4H, m), 7.45–
7.36 (6H, m), 4.94 (1H, dt, Jˆ5.5, 10.0 Hz), 4.61 (1H, d,
Jˆ12.5 Hz), 4.48 (1H, d, Jˆ12.5 Hz), 3.97 (1H, dd, Jˆ3.5,
11.5 Hz), 3.90 (1H, dd, Jˆ1.5, 11.5 Hz), 3.72 (1H, ddd,
Jˆ1.5, 3.5, 10.0 Hz), 3.62–3.58 (1H, m), 3.56 (1H, dt,
Jˆ2.5, 11.0 Hz), 2.28–2.18 (2H, m), 1.86 (1H, dt,
Jˆ13.0, 4.0 Hz), 1.82 (1H, dt, Jˆ13.5, 3.0 Hz), 1.67–1.48
(6H, m), and 1.09 (9H, s); 13C NMR (125 MHz, CDCl3) d
135.94, 135.55, 133.55, 133.01, 129.72, 129.67, 127.69,
127.55, 95.22, 77.85, 70.69, 62.73, 60.74, 53.72, 34.91,
34.25, 26.01, 24.93, 19.35, and 18.60; IR (neat) 3076,
3052, 3018, 2936, 2860, 1464, 1431, 1379, 1274, 1230,
1181, 1114, 1046, 1021, 982, 965, 849, 822, 797, 737,
704, 634, and 605 cmϪ1; HR-FD-MS calcd for C27H38O6Cl-
SiS (MϩϩH) 553.1846, found 553.1845.
(2Sء,3Sء,30E)-3-Chloromethanesulfonyloxy-2-[50-(tert-
butyldiphenylsilyloxy)-30-pentenyl]-oxane (65) and
(2Rء,10Rء,4E)-2-[10-chloromethanesulfonyloxy-60-(tert-
butyldiphenylsilyloxy)-40-hexenyl]oxolane (66). A solu-
tion of epoxide 64 (73.7 mg, 0.137 mmol) in 1,2-dichloro-
ethane (5 mL) was heated at reflux for 6 h. The solvent was
removed in vacuo. The residue was purified by PTLC (SiO2,
hexane–EtOAc, 3:1) to give epoxide 65 (12.9 mg, 18%) and
66 (25.0 mg, 34%). 65: a colorless oil; Rfˆ0.47 (hexane–
EtOAc, 3:1); 1H NMR (500 MHz, CDCl3) d 7.68–7.66 (4H,
m), 7.44–7.36 (6H, m), 5.63 (1H, dt, Jˆ15.0, 6.0 Hz), 5.58
(1H, dt, Jˆ15.0, 4.5 Hz), 4.79 (1H, broad s), 4.64 (1H, d,
Jˆ13.0 Hz), 4.61 (1H, d, Jˆ13.0 Hz), 4.05–4.01 (1H, m),
3.55 (1H, dt, Jˆ2.0, 12.0 Hz), 3.39 (1H, dd, Jˆ5.0, 9.0 Hz),
2.38–2.32 (1H, m), 2.22–2.08 (2H, m), 2.02 (1H, tq, Jˆ4.5,
13.5 Hz), 1.83–1.75 (2H, m), 1.60–1.52 (1H, m), 1.51–1.46
(1H, m), and 1.05 (9H, s); IR (neat) 3074, 3050, 2958, 2860,
1715, 1591, 1473, 1464, 1431, 1361, 1255, 1178, 1112,
1050, 919, 824, 741, and 704 cmϪ1; HR-FD-MS calcd for
C27H38O5ClSiS (MϩϩH) 537.1898, found 537.1879. 66: a
colorless oil; Rfˆ0.53 (hexane–EtOAc, 3:1); 1H NMR
(500 MHz, CDCl3) d 7.69–7.66 (4H, m), 7.44–7.36 (6H,
m), 5.66 (1H, dt, Jˆ15.5, 5.5 Hz), 5.62 (1H, dt, Jˆ15.5,
4.0 Hz), 4.90 (1H, d, Jˆ12.0 Hz), 4.62 (1H, d, Jˆ
12.0 Hz), 4.62 (1H, dt, Jˆ4.0, 9.0 Hz), 4.17–4.16 (2H,
m), 3.99 (1H, q, Jˆ8.0 Hz), 3.88–3.81 (2H, m), 2.32–
2.24 (1H, m), 2.22–2.16 (1H, m), 2.04–1.89 (3H, m),
1.78–1.72 (1H, m), 1.71–1.64 (1H, m), 1.59–1.51 (1H,
m), and 1.05 (9H, s); IR (neat) 3074, 3030, 2960, 2934,
2862, 1473, 1462, 1429, 1365, 1180, 1112, 1069, 919,
824, 741, and 704 cmϪ1; HR-FD-MS calcd for C23H28O5Cl-
SiS (MϩϪtBu) 479.1116, found 479.1127.
Acknowledgements
We thank Prof. Dr Akio Murai, Dr Kenshu Fujiwara, and
Dr Tetsuo Tokiwano (Hokkaido University) for the mass
spectroscopic analyses of our synthetic compounds. We
are grateful to SC-NMR Laboratory of Okayama University
for NMR experiment.
(2Sء,3Sء,30Rء,40Rء)-3-Chloromethanesulfonyloxy-2-[50-
(tert-butyldiphenylsilyloxy)-30,40-epoxy-pentenyl]oxane
(67a) and (2Sء,3Sء,30Sء,40Sء)-3-chloromethanesulfonyl-
oxy-2-[50-(tert-butyl-diphenylsilyloxy)-30,40-epoxypente-
nyl]oxane (67b). To a solution of olefin 65 (11.5 mg,
0.0221 mmol) in CH2Cl2 (1.5 mL) cooled to 0ЊC were
added Na2HPO4 (15.6 mg, 0.0111 mmol) and m-chloro-
perbenzoic acid (13.6 mg, 0.0553 mmol, purity 70%). The
solution was stirred at 25ЊC for 2 h. Satd aq. NaHCO3
(2 mL) and satd aq. Na2S2O3 (2 mL) were then added at
0ЊC. The mixture was extracted with EtOAc (38 mL). The
organic layer was washed with satd aq. Na2S2O3 (10 mL),
satd aq. NaHCO3 (10 mL) and brine (10 mL), dried over
MgSO4, and concentrated in vacuo. The residue was
purified by PTLC (SiO2, hexane–EtOAc, 2:1) to give a dia-
stereomeric mixture of epoxides 67a and 67b (8.8 mg, 72%)
as a colorless oil; Rfˆ0.23 (hexane–EtOAc, 3:1); 1H NMR
(500 MHz, CDCl3) d 7.68–7.66 (4H×2, m), 7.45–7.37
(6H×2, m), 4.87–4.74 (1H×2, m), 4.63 (2H, s), 4.62 (2H,
s), 4.03–3.99 (2H, m), 3.77 (1H, dd, Jˆ3.5, 12.0 Hz), 3.75
(2H, d, Jˆ4.0 Hz), 3.71 (1H, dd, Jˆ5.0, 12.0 Hz), 3.50–
3.44 (2H×2, m), 2.93–2.89 (1H×2, m), 2.83–2.79 (1H×2,
m), 2.38–2.31 (1H×2, m), 2.06–1.96 (1H×2, m), 1.86–1.74
(6H, m), 1.72–1.61 (4H, m), 1.48 (2H×2, broad d,
Jˆ9.5 Hz), and 1.05 (9H×2, s); IR (neat) 2958, 2862,
1473, 1431, 1375, 1178, 1114, 1091, 919, 824, 797, 741,
References
1. (a) Lin, Y.-Y.; Risk, M.; Ray, S. M.; Van Engen, D.; Clardy, J.;
Golik, J.; James, J. C.; Nakanishi, K. J. Am. Chem. Soc. 1981, 103,
6773. (b) Shimizu, Y.; Chou, H.-N.; Bando, H.; Duyne, G. V.;
Clardy, J. J. Am. Chem. Soc. 1986, 108, 514. (c) Shimizu, Y.;
Bando, H.; Chou, H.-N.; Duyne, G. V.; Clardy, J. C. J. Chem.
Soc. Chem. Commun. 1986, 1656. (d) Pawlak, J.; Tempesta,
M. S.; Golik, J.; Zagorski, M. G.; Lee, M. S.; Nakanishi, K.;
Iwashita, T.; Gross, M. L.; Tomer, K. B. J. Am. Chem. Soc.
1987, 109, 1144. (e) Prasad, A. V. K.; Shimizu, Y. J. Am. Chem.
Soc. 1989, 111, 6476.
2. (a) Murata, M.; Legrand, A. M.; Ishibashi, Y.; Yasumoto, T.
J. Am. Chem. Soc. 1989, 111, 8929. (b) Murata, M.; Legrand,
A. M.; Ishibashi, Y.; Fukui, M.; Yasumoto, T. J. Am. Chem. Soc.
1990, 112, 4380. (c) Suzuki, T.; Sato, O.; Hirama, M.; Yamamoto,
Y.; Murata, M.; Yasumoto, T.; Harada, N. Tetrahedron Lett. 1991,
32, 4505. (d) Satake, M.; Morohashi, A.; Oguri, H.; Oishi, T.;
Hirama, M.; Harada, N.; Yasumoto, T. J. Am. Chem. Soc. 1997,
119, 11325.
3. (a) Yokoyama, A.; Murata, M.; Oshima, Y.; Iwashita, T.;
Yasumoto, T. J. Biochem. 1988, 104, 184. (b) Murata, M.;
Iwashita, T.; Yokoyama, A.; Sasaki, M.; Yasumoto, T. J. Am.
Chem. Soc. 1992, 114, 6594. (c) Murata, M.; Naoki, H.; Iwashita,
T.; Matsunaga, S.; Sasaki, M.; Yokoyama, A.; Yasumoto, T. J. Am.
Chem. Soc. 1993, 115, 2060. (d) Murata, M.; Naoki, H.;

N. Hayashi et al. / Tetrahedron 56 (2000) 7123–7137
7137
Matsunaga, S.; Satake, M.; Yasumoto, T. J. Am. Chem. Soc. 1994,
116, 7098.
4. Total synthesis of brevetoxin A, Nicolaou, K. C.; Yang, Z.; Shi,
G.-q., Gunzer, J. L.; Agrios, K. A.; Gu¨arter, P. Nature 1998, 392,
264.
Lett. 1996, 487. (g) Fujiwara, K.; Tokiwano, T.; Murai, A.,
Tetrahedron Lett. 1995, 36, 8063. (h) Fujiwara, K.; Mishima, H.;
Amano, A.; Tokiwano, T.; Murai, A. Tetrahedron Lett. 1998, 39,
393. See also: (i) Janda, K. D.; Shevlin, C. G.; Lerner, R. A.
Science 1993, 259, 490. (j) Janda, K. D.; Shevlin, C. G.; Lerner,
R. A. J. Am. Chem. Soc. 1995, 117, 2659. (k) Mori, Y.; Yaegashi,
K.; Furukawa, H. J. Am. Chem. Soc. 1996, 118, 8158.
11. Tokiwano, T.; Fujiwara, K.; Murai, A. Synlett 2000, 335.
12. (a) Hayashi, N.; Fujiwara, K.; Murai, A. Chem. Lett. 1996,
341. (b) Hayashi, N.; Fujiwara, K.; Murai, A. Synlett 1997, 7,
793. (c) Hayashi, N.; Fujiwara, K.; Murai, A. Tetrahedron Lett.
1996, 37, 6173. (d) Hayashi, N.; Fujiwara, K.; Murai, A.
Tetrahedron 1997, 37, 12425.
13. The pioneering studies on the formation of cyclic ethers via the
bridged oxonium ion has been reported. (a) Nakata, T.; Schmid, G.;
Vranesic, B.; Okigawa, M.; Smith-Palmer, T.; Kishi, Y. J. Am.
Chem. Soc. 1978, 100, 2933. (b) Ting, P. C.; Bartlett, P. A.
J. Am. Chem. Soc. 1984, 106, 2668. (c) Michael, J. P.; Ting,
P. C.; Bartlett, P. A. J. Org. Chem. 1985, 50, 2416. (d) Alvarez,
E.; Diaz, M. T.; Perez, R.; Ravelo, J. L.; Regueiro, A.; Vera, J. A.;
Zurita, D.; Martin, J. D. J. Org. Chem. 1994, 59, 2848 and
references cited therein. See also: (e) Nakata, T.; Nomura, S.;
Matsukura, H. Tetrahedron Lett. 1996, 37, 213. (f) Nakata, T.;
Nomura, S.; Matsukura, H.; Morimoto, M. Tetrahedron Lett.
1996, 37, 217.
5. Total synthesis of brevetoxin B, (a) Nicolaou, K. C.;
Theodorakis, E. A.; Rutjes, F. P. J. T.; Tiebes, J.; Sato, M.;
Untersteller, E.; Xiao, X.-Y. J. Am. Chem. Soc. 1995, 117, 1171.
(b) Nicolaou, K. C.; Rutjes, F. P. J. T.; Theodorakis, E. A.; Tiebes,
J.; Sato, M.; Untersteller, E. J. Am. Chem. Soc. 1995, 117, 1173.
(c) Nicolaou, K. C.; Hwang, C.-K.; Duggan, M. E.; Nugiel, D. A.;
Abe, Y.; Reddy, K. B.; DeFrees, S. A.; Reddy, D. R.; Awartani,
R. A.; Conley, S. R.; Rutjes, F. P. J. T.; Theodorakis, E. A. J. Am.
Chem. Soc. 1995, 117, 10227. (d) Nicolaou, K. C.; Theodorakis,
E. A.; Rutjes, F. P. J. T.; Tiebes, J.; Sato, M.; Tiebes, J.; Xiao,
X.-Y.; Hwang, C.-K.; Duggan, M. E.; Yang, Z.; Couladouros,
E. A.; Sato, F.; Shin, J.; He, H.-M.; Bleckman, T. J. Am. Chem.
Soc. 1995, 117, 10239. (e) Nicolaou, K. C.; Rutjes, F. P. J. T.;
Theodorakis, E. A.; Tiebes, J.; Sato, M.; Untersteller, E. J. Am.
Chem. Soc. 1995, 117, 10252.
6. Total synthesis of hemibrevetoxin B, (a) Nicolaou, K. C.;
Reddy, K. R.; Skokotas, G.; Sato, F.; Xiao, X.-Y.; Hwang, C.-K.
J. Am. Chem. Soc. 1992, 114, 7935. (b) Nicolaou, K. C.; Reddy,
K. R.; Skokotas, G.; Sato, F.; Xiao, X.-Y.; Hwang, C.-K. J. Am.
Chem. Soc. 1993, 115, 3558. (c) Kadota, I.; Jung-Youl, P.;
Koumura, N.; Polaud, G.; Matsukawa, Y.; Yamamoto, Y.
Tetrahedron Lett. 1995, 36, 5777. (d) Morimoto, M.; Matsukura,
H.; Nakata, T. Tetrahedron Lett. 1996, 37, 6365. (e) Mori, Y.;
Yaegashi, K.; Furukawa, H. J. Am. Chem. Soc. 1997, 119, 4557.
7. Lee, M. S.; Qin, G.-w.; Nakanishi, K.; Zagorski, M. J. Am.
Chem. Soc. 1989, 111, 6234.
14. In this paper, the terms of endo and exo are defined by the
attaching position of the nucleophile on the products: the product
containing the nucleophile on the cyclic framework is the endo-
product; on the other hand, the product possessing it on the side
chain is the exo-product.
15. Shimizu, T.; Hiranuma, S.; Nakata, T. Tetrahedron Lett. 1996,
37, 6145.
16. Sharpless, K. B.; Michaelson, R. C. J. Am. Chem. Soc. 1973,
95, 6136.
8. Prasad, A. V. K.; Shimizu, Y. J. Am. Chem. Soc. 1989, 111,
6476.
9. Baldwin, J. E. J. Chem. Soc., Chem. Commun. 1976, 734.
10. (a) Nicolaou, K. C.; Prasad, C. V. C.; Somers, P. K.; Hwang,
C.-K. J. Am. Chem. Soc. 1989, 111, 5330. (b) Nicolaou, K. C.;
Prasad, C. V. C.; Somers, P. K.; Hwang, C.-K. J. Am. Chem. Soc.
1989, 111, 5335. (c) Suzuki, T.; Sato, O.; Hirama, M. Tetrahedron
Lett. 1990, 31, 4747. (d) Mukai, C.; Ikeda, Y.; Sugimoto, Y.;
Hanaoka, M. Tetrahedron Lett. 1994, 35, 2179. (e) Mukai, C.;
Sugimoto, Y.; Ikeda, Y.; Hanaoka, M. Tetrahedron Lett. 1994,
35, 2183. (f) Matsukura, H.; Morimoto, M.; Nakata, T. Chem.
17. The yields of 70 and 71 were calculated from the ratio of 70/71
1
in the H NMR spectrum. A part of the 70 could be obtained as a
pure sample by HPLC. On the other hand, the 71 could not be
separated perfectly.
18. Alvarez, E.; Rodriguez, M. L.; Zurita, D.; Martin, J. D.
Tetrahedron Lett. 1991, 32, 2253.
19. One of the signals in phenyl group is not recorded, because the
signal overlaps with the signal of benzene-d6.