Diastereoselective rhodium(II)-catalyzed sulfonium ylide formation-[2,3]-sigmatropic rearrangement reaction of chiral non-racemic allylic sulfides

Article abstract of DOI:10.1016/S0957-4166(03)00079-X

The tandem sulfonium ylide formation-[2,3]-sigmatropic rearrangement reaction of chiral non-racemic secondary allylic sulfides, (E)-9 and (Z)-10, is found to proceed with high diastereocontrol. The C-5 stereocenter bearing the sulfide group is essential for high diastereoselectivity in the reaction. Transition state conformers are proposed to explain the high diastereoselectivity in the formation of the diastereomeric products, 18a and 18b. The method is applied to the synthesis of (R)-4-(4-chlorophenyl)-2-butyrolactone. Modest enantioselectivity (63% ee) was achieved and this is attributed to partial racemization during the formation of the secondary allylic sulfide 22.

Full text of DOI:10.1016/S0957-4166(03)00079-X

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 897–909
Diastereoselective rhodium(II)-catalyzed sulfonium ylide
formation-[2,3]-sigmatropic rearrangement reaction of chiral
non-racemic allylic sulfides
Andrew G. H. Wee,* Qing Shi, Zhongyi Wang and Kimberly Hatton†
Department of Chemistry and Biochemistry, University of Regina, Regina, Saskatchewan, S4S 0A2, Canada
Received 4 December 2002; accepted 2 January 2003
Abstract—The tandem sulfonium ylide formation-[2,3]-sigmatropic rearrangement reaction of chiral non-racemic secondary allylic
sulfides, (E)-9 and (Z)-10, is found to proceed with high diastereocontrol. The C-5 stereocenter bearing the sulfide group is
essential for high diastereoselectivity in the reaction. Transition state conformers are proposed to explain the high diastereoselec-
tivity in the formation of the diastereomeric products, 18a and 18b. The method is applied to the synthesis of (R)-4-(4-
chlorophenyl)-2-butyrolactone. Modest enantioselectivity (63% ee) was achieved and this is attributed to partial racemization
during the formation of the secondary allylic sulfide 22. © 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
asymmetric induction strategies involving a chiral cata-
lyst and either 1 (R1=chiral auxiliary) or 2 (R2=chiral
auxiliary).5b,d More importantly, the results from the
studies of Katsuki, Hashimoto and Wang5b,e,f showed
that the diastereoselectivity of the reactions of primary
allylic sulfides is not catalyst dependent, which strongly
suggested that the [2,3]-sigmatropic rearrangement
involved a free sulfonium ylide intermediate as opposed
to metal-stabilized one.5c Enantioselectivity, on the
other hand, was determined by the degree of enantio-
differentiation of the two lone pairs of electrons on the
sulfur atom in the sulfide moiety during the transfer of
the carbene ligand from the chiral metallocarbenoid
intermediate to the sulfide unit.
The [2,3]-sigmatropic rearrangement1 of allylic sulfo-
nium ylide intermediates,2 is an important and widely
used carbonꢀcarbon bond forming reaction in organic
synthesis. The rearrangement reaction involves a con-
certed, orbital symmetry-allowed, suprafacial process
that generally proceeds with high stereoselectivity.
The use of asymmetric induction3 to effect enantioselec-
tivity and diastereoselectivity during bond formation in
the allylic sulfonium ylide-[2,3]-sigmatropic rearrange-
ment reaction is a topic of ongoing interest4 and to this
end, a number of approaches, including the use of
allylic sulfides possessing a chiral auxiliary group on the
sulfur atom,4a–c the use of sulfonium ylides with a
stereochemically defined center either at the sulfonium
atom4d or at the carbon bearing the sulfide moiety4e,f
have been devised. Recent investigations5 have been
directed at the chiral metallocarbenoid-mediated
tandem sulfonium ylide-[2,3]-sigmatropic rearrange-
ment of allylic sulfides (Eq. (1)). The rearrangement
reactions were generally found to proceed with modest
to good diastereoselectivity. Enantioselectivity, how-
ever, was poor but can be enhanced either via the
reaction of allylic sulfides 1 (R1=bulky group) and
diazoacetates 2 (R2=bulky group)5b,d–f or via double
(1)
In connection with our investigations6 on 1,2-relative
induction3b during bond formation in Rh(II)–carbenoid
mediated transformations, we have investigated the
tandem sulfonium ylide formation-[2,3]-sigmatropic
rearrangement of chiral non-racemic allylic sulfides typ-
ified by 6, 9 and 10. The dioxolanyl moiety has been
found to be effective in providing good to high diastereo-
control in related sigmatropic rearrangements7 and it
was anticipated that similar levels of diastereocontrol
* Corresponding author. E-mail: andrew.wee@uregina.ca
† 2002 NSERC undergraduate summer research assistant.
0957-4166/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0957-4166(03)00079-X

898
A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
should be possible in the [2,3]-sigmatropic rearrange-
ment in 6, 9 and 10. It should be pointed out that the
sulfonium ylide-[2,3]-sigmatropic rearrangement reac-
tion of the allylic sulfide 3 with diazo compounds
catalyzed by Rh2(OAc)4 was briefly investigated by
Takano and co-workers (Eq. (2)). The rearrangement
was found to afford products 4 in modest to good
chemical yields; however, the diastereoselectivity of the
reaction was not reported.8
(2)
We report herein the details of our investigations into
the tandem sulfonium ylide formation-[2,3]-sigmatropic
rearrangement of chiral non-racemic allylic sulfides.
Excellent diastereoselectivity is possible only with the
use of secondary sulfides and the diastereoselectivity is
controlled by the C-5 stereocenter that bears the sulfide
moiety and not the dioxolanyl moiety of the starting
material. The use of chiral Rh(II) catalysts did not
result in any enhancement in the diastereoselectivity of
the rearrangement reaction.
Scheme 1.
investigated first. Benzene was initially chosen as the
solvent because it is the most commonly used solvent
for [2,3]-sigmatropic rearrangements of sulfonium
ylides.2 The reaction gave an improved yield (43%) of
the rearrangement products 7. Maleate and fumarate
side-products as well as varying amounts of 1-carbo-
ethoxy-2,4,6-cycloheptatriene, which had resulted from
the cycloaddition of the Rh(II)–carbenoid to the ben-
zene ring,11 were also obtained. Neither the cyclo-
propane product, which could have arisen from
metallocarbenoid attack of the double bond nor prod-
ucts that could have formed from decomposition of
oxonium ylide intermediates was observed. The use of
chlorobenzene as solvent did result in a slight increase
in the yield (48–52%) of 7, but did not completely
2. Results and discussion
As described above, the reaction depicted in Eq. (2) has
been studied.8 Interestingly, the same reaction of 3 with
ethyl diazoacetate (EDA) was not explored. Our pre-
liminary studies of the reaction of 3 with EDA showed
the reaction to be inefficient; poor yields of the rear-
rangement product 4 (X=CO2Et, Y=H; 11%) were
obtained, and substantial amounts of maleate and
fumarate, by-products formed via metallocarbenoid
dimerization, were also produced. It was reasoned, in
light of Takano’s results, that the poor yield of the
rearrangement product may be due to the inefficient
interception of the Rh(II)–carbenoid intermediate by
the sulfur atom of the phenylsulfide moiety because of
the decreased electrophilicity of the metallocarbenoid
carbon. It was decided to replace the phenylthio unit
with a 4-methoxyphenylthio group because the nucleo-
philicity of the sulfur atom in the latter group should be
enhanced due to the electron-donating effect of the
4-methoxy substituent. Therefore, the allylic sulfides
6a,b were prepared and used in our initial studies.
inhibit
cycloheptatriene
formation.
Therefore,
chlorobenzene was used as the solvent in subsequent
experiments.
The rearrangement products 7 were obtained as an
inseparable mixture of four diastereomers. In the H
1
NMR spectrum of the mixture, the diagnostic H-3
multiplet (ddd) for each of the four diastereomers
appeared at l 2.49, 2.67, 2.69 and 2.85. Due to exten-
sive signal overlap in the 1H NMR spectrum, no
attempts were made to assign the stereochemistry at the
newly created C-2 and C-3 stereocenters in each of the
four diastereomers nor was it possible to ascertain the
ratio of each of the four diastereomers. To facilitate the
determination of the diastereoselectivity during the for-
mation of the C-3 stereocenter, the diastereomeric mix-
ture 7 was subjected to reductive desulfurization using
Bu3SnH.12 This afforded two diastereomers 8a,b having
very similar polarity, which were separated by careful
column chromatography. The less polar component
(Rf=0.32) was assigned structure 8a and the more
polar component (Rf=0.25) was assigned structure 8b
2.1. Preparation of and [2,3]-sigmatropic rearrangement
of allylic sulfides (E)-6a and (Z)-6b
The readily available, known9 (E)- and (Z)-allylic alco-
hols 5a,b were reacted with bis(4-methoxyphenyl)-
disulfide10 and tributylphosphine, under ultrasound
irradiation (1.5 h), to afford high yields (86–92%) of the
corresponding (E)- and (Z)-allylic sulfides 6a,b, respec-
tively (Scheme 1). The [2,3]-sigmatropic rearrangement
reaction of 6a catalyzed by Rh2(OAc)4 (5 mol %) was
on the basis of comparison of their H and 13C NMR
1
data with those reported in the literature.13 For the
determination of the diastereomeric ratio of 8a:8b, the
1H NMR spectrum of 8a,b was recorded in a 1:1.2 v/v
CDCl3:C6D6 solvent mixture.14 This solvent mixture

A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
899
was conducive for the resolution of some of the over-
lapping resonances in the spectrum. The integration of
the two singlets, each of which were due to one of the
methyl groups on the 1,3-dioxolanyl moieties of each of
the diastereomers, at l 1.36 (8a) and l 1.34 (8b), gave
the ratio of 8a:8b. The results are collected in Table 1.
metric induction from the adjacent dioxolanyl stereo-
center during the key [2,3]-sigmatropic rearrangement
step. This lack of diastereoselection led us to investigate
the sulfonium ylide-[2,3]-sigmatropic rearrangement of
chiral non-racemic secondary allylic sulfides 9 and 10 to
ascertain whether higher diastereoselectivity can be
achieved. It is pertinent to note that two previous
studies on the [2,3]-sigmatropic rearrangement of sec-
ondary sulfides have been reported.4e,f However, in
these studies the sulfonium ylides were generated by
base-mediated deprotonation of sulfonium salts. The
[2,3]-sigmatropic rearrangement of sulfonium ylides
generated in situ from the reaction of secondary sulfides
with EDA catalyzed by Rh(II) catalysts has not been
investigated.
The Rh2(OAc)4-catalyzed reaction of (E)-6a showed no
diastereoselectivity during the rearrangement, affording
almost equal amounts of 8a and 8b. It is also interesting
to compare this result to that of the related ortho ester
Claisen rearrangement (MeC(OEt)3, cat. MeCH2-
CO2H)15 of the (E)-allylic alcohol of the type 5, which
was independently studied by Suzuki, Kametani et al.16
and Mulzer et al.13b The former group reported a 3:1
ratio of 8a:8b whereas the latter group obtained a 1:1
ratio of 8a:8b. We therefore studied the reaction of
(Z)-6b because it has been shown7 in other related
reactions that a (Z)-double bond generally provides
enhanced diastereoselectivity. However, in our case, no
diastereoselection was achieved; in fact, the ratio of
8a:8b obtained with (Z)-6b was very similar to that
realized with (E)-6a (entries 1 and 2).
2.2. Preparation of allylic sulfides 9 and 10
The alcohol 12a has been prepared13a by Mulzer and
co-workers via reduction (diacetone glucose/9-BBN ate
complex) of the enone 11a. Since we also required
alcohol 13a, we decided to examine the asymmetric
reduction of the enones 11a and 11b9a using the CBS
method.17 It was of concern to us at the outset that the
presence of the chiral 1,3-dioxolanyl group in 11a and
11b would be detrimental to the diastereoselectivity of
the reduction. We were pleased to find that our con-
cerns were unfounded.
Due to the poor diastereoselectivity observed with
Rh2(OAc)4, we evaluated the use of the enantiomeric
set of Rh2(MEPY)4 to determine whether the
diastereoselectivity of the reaction can be enhanced
through double stereoselection. Disappointingly, both
chiral Rh(II) catalysts gave 8a and 8b in ratios that
were very similar to those obtained with Rh2(OAc)4.
More interestingly, the ratio of 8a:8b obtained for the
enantiomeric pair of catalysts are also very similar, and
show the same preference for the formation of 8a
(entries 3 and 4).
Thus, the asymmetric reduction of 11a was studied first
to develop the optimal reaction conditions for the
reduction. It was found that the best conditions were
(1) a reaction temperature of −20°C, (2) a reaction time
of 30 min, and (3) a mole ratio of 1:0.6:0.6 for the
11a:borane–methyl sulfide complex:(S)-2-methyl-CBS-
oxazaborolidine (2-MeCBS) catalyst and under these
conditions the allylic alcohols 12a and 12b were
obtained in a combined yield of 90%. Both 12a and 12b
have very similar polarity and were not readily sepa-
The low diastereoselectivity of the sulfonium ylide-[2,3]-
sigmatropic rearrangement involving allylic sulfides of
the type 6 suggested that there was minimal 1,2-asym-
Table 1. Diastereoselectivity of [2,3]-sigmatropic rearrangement of 6a,b, 9 and 10a
Entry
Sulfide
Catalyst
Solvent
Temp. (°C)
Yield (%)b
Relative yield (%)
8a:8bc 20a:20bd
1
2
3
4
5
6
7
8
(E)-6a
(Z)-6b
(Z)-6b
(Z)-6b
(E)-9
(E)-9
(E)-9
(E)-9
(E)-9
(E)-9
(E)-9
(E)-9
(Z)-10
(Z)-10
Rh2(OAc)4
Rh2(OAc)4
Rh2(5R-MEPY)4
Rh2(5S-MEPY)4
Rh2(OAc)4
Rh2(OAc)4
Rh2(OAc)4
Rh2(OAc)4
Rh2(OAc)4
Rh2(5R-MEPY)4
Rh2(5S-MEPY)4
Rh2(Cap)4
Rh2(OAc)4
Rh2(Cap)4
PhCl
PhCl
PhCl
PhCl
PhCl
PhCl
PhCl
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
80
80
80
80
80
40
25
40
25
25
25
25
25
25
52
50
48
50
49
50
32
54
51
80
82
89
65
90
59:41
62:38
56:44
63:57
–
–
–
–
–
–
–
–
–
–
–
–
–
–
87:13
90:10
96:4
97:3
98:2
95:5
94:6
95:5
99:1
98:2
9
10
11
12
13
14
a 5 mol % Rh(II) catalyst, 3 mol equiv. EDA in the appropriate dry solvent.
b Combined yield of diastereomeric products 7 or 18.
c Ratio based on 1H NMR integration of the methyl singlets at l 1.36 (8a) and l 1.34 (8b) of the dioxolanyl moiety in each of the diastereomers
in the mixture 8.
d Ratio based on GC analysis of 20.

900
A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
1
sulfides 9 and 10. The ultrasound-mediated Mitsunobu
reaction that worked very well for the conversion of
5a,b“6a,b was found to be ineffective for the prepara-
tion of 9 from 12a as substantial amounts of the
starting 12a remained after reaction times of 1.5 and 22
h. Conducting the same reaction at a temperature of
60°C (ultrasound water bath) was also fruitless. After
many attempts at varying the reaction conditions (e.g.
solvent, phosphine) it was found that a good yield
(81%) of 9 was obtained when freshly prepared bis(4-
methoxyphenyl)disulfide was used with tributylphos-
phine, and toluene (reflux) was employed as the solvent.
Application of the same method to 13a led to a 90%
yield of the sulfide 10. It is noteworthy that SN2%
reaction products were not detected under these reac-
tion conditions.
rated by column chromatography. Furthermore, the H
NMR spectrum of the mixture was not useful in that
only a single set of signals could be observed. In order
to determine the ratio of 12a:12b, the mixture was
1
converted to the benzyl ethers 14, the H NMR spec-
trum of which revealed two sets of signals. The
diastereomeric ratio of 14a:14b was found to be 88:12,
based on integration of the H-1 double doublets cen-
tered at l 3.56 (14a) and at l 3.59 ppm (14b), which in
turn provided the ratio for 12a:12b (Scheme 2).
Since the mixture of 12a and 12b was very difficult to
separate by column chromatography, the mixture was
converted to the crystalline 3,5-dinitrobenzoate deriva-
tives 15. Fractional crystallization of the mixture 15
then gave an 86% yield of the pure diastereomer 15a
and subsequent base hydrolysis of 15a regenerated the
pure 12a {[h]2D2 +24 (c 1.8, CHCl3), lit.18 +28 (c 1.4,
CHCl3)}. A sample of pure 12a was converted to the
2.3. [2,3]-Sigmatropic rearrangement reaction of 9 and
10
1
benzyl ether 14a and the H NMR spectrum showed
The reactions of 9 and 10 with EDA in the presence of
Rh(II) catalysts (5 mol%) in chlorobenzene and
dichloromethane were studied. Unlike the reaction of
the primary allylic sulfides 6a,b the sulfonium ylide-
[2,3]-sigmatropic rearrangement reaction of 9 and 10
led to the formation of a mixture of only two
diastereomers as evidenced by the presence of the char-
acteristic H-3 resonances, one centered at l 2.36 (ddd)
and the other at l 2.70 (ddd), of the diastereomers. For
characterization purposes, the diastereomeric mixture
from the Rh2(OAc)4-catalyzed rearrangement of 9, were
carefully separated by column chromatography to
provide pure 18a (less polar, Rf=0.38) and 18b (more
only one set of signals. The absolute configuration at
C-5 in alcohol 12a was confirmed as R by degradation
to the known19a alcohol 17a, whose specific rotation
value {[h]2D2 −32 (c 0.3, CHCl3), lit.19a −47 (c 1, CHCl3)}
had the same sign and magnitude to that reported in
the literature.
For the preparation of the alcohol 13a, the enone 11b
was reduced (BH3·Me2S) in the presence of the (R)-2-
MeCBS catalyst under the optimal conditions specified
for 11a. Compounds 13a,b were obtained in a com-
bined yield of 85%. GC analysis of the alcohol mixture
showed the ratio of 13a (tR=9.09 min):13b (tR=9.15
min) was 86:14. The mixture of alcohols 13a,b was
amenable to separation by careful column chromatog-
raphy, which afforded pure diastereomer 13a (58%).
The configuration at C-5 in compound 13a was con-
firmed as S by its conversion to the known 17b,19b and
comparison of its specific rotation value {[h]2D2 +44 (c
1.3, CHCl3)} with that reported in the literature {[h]D22
+48 (c 1.0, CHCl3)}.
1
polar, Rf=0.25). Analysis of the H NMR spectra of
18a,b permitted assignment of the H-3 resonance at l
2.36 to 18a and the one at l 2.70 to 18b. Furthermore,
the vicinal coupling constant of the H-2 doublet in 18a
and 18b was found to be 8.8 and 10.3 Hz, respectively.
This suggests that the relative stereochemistry of H-2
and H-3 in 18a is synclinal (q=53o) and in 18b they
have an antiperiplanar (q=175o) relationship.20 Other
1
salient features in the H NMR spectra were the reso-
With the allylic alcohols 12a and 13a in hand, we
investigated the preparation of the chiral non-racemic
nances due to H-4 and the olefinic hydrogens. In 18a
the H-4 signal was deshielded and occurred at l 4.72
Scheme 2.

A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
901
new C-3 stereocenter. The results are summarized in
Table 1.
whereas in 18b, the same proton was shielded and
resonated in the region l 3.91–4.08. The olefinic hydro-
gens in 18a,b exhibited large coupling constants, J=14–
15.7 Hz, which indicated that the double bonds in 18a,b
had E-stereochemistry (Scheme 3).
From Table 1, it is clear that the reaction of secondary
sulfide 9 with EDA, under the optimal reaction condi-
tions (PhCl, 80°C) used for compounds 6, gave signifi-
cantly higher diastereoselectivity (entries 1 and 5).
However, the combined yield of the rearrangement
products 7 and 18 was the same in both cases. It is
noteworthy that in chlorobenzene, the cycloheptatriene
product that would result from the cycloaddition reac-
tion of the metallocarbenoid with chlorobenzene was
not detected. Temperature effects were also evident: At
40°C, an increase in the diastereoselectivity was
achieved, but the chemical yield of the products
remained the same. On the other hand, the room
temperature (25°C) reaction afforded substantially
higher diastereoselectivity, but at the expense of chemi-
cal yield (entries 6 and 7). The increased diastereoselec-
tivity observed with chlorobenzene at 40°C prompted
us to evaluate dichloromethane as a solvent. We were
pleased to find that in dichloromethane, the diastereose-
lectivity of the reaction remained consistently high and
was not influenced by reaction temperature (entries 8
and 9).
It is useful to note that with Rh2(OAc)4 and Rh2(Cap)4
as catalysts, the rearrangement of 9 resulted in the
formation of 18a:18b in a ratio of 1:1.6–2.3. For com-
pound 10, however, its rearrangement catalyzed by
Rh2(OAc)4 gave 18a and 18b in a 1:4 ratio, but this
trend was reversed with Rh2(Cap)4 where the 18a:18b
ratio was 2:1. This latter outcome is intriguing and
suggests that the diastereoselection at C-2 may be influ-
enced somewhat by the Rh2(Cap)4 catalyst. Presently,
we do not have a satisfactory explanation for this
observation.
The products 18 were subjected to reductive desulfur-
1
ization using Bu3SnH. The H NMR spectrum of the
desulfurized product 19 indicated that only one
diastereomer was present. The diastereomer was
1
assigned structure 19a by comparison of its H and 13C
NMR data with literature values.13a This result also
confirmed that the diastereomers 18a,b differed only in
the stereochemistry at C-2 and indicated that there was
excellent and complete transfer of chirality from the
C-5 stereocenter in 9 and 10 to the newly formed C-3
stereocenter in product 18. The diastereocontrol during
the formation of the C-2 stereocenter was modest at
best.
A similar degree of and the same sense of diastereose-
lection was realized with the chiral catalysts, Rh2[(5R)-
MEPY]4 and Rh2[(5S)-MEPY]4 (entries 10 and 11).
These results are in accord with those obtained with 8a
and, collectively, they strongly indicate that the key
[2,3]-sigmatropic rearrangement step involved a free
sulfonium ylide intermediate.5b,e,f This line of reasoning
was confirmed with the use of the achiral catalyst
Rh2(Cap)4, which afforded an almost identical ratio of
20a:20b (entries 10, 11 and 12) as well as the same sense
of diastereoselection.
To facilitate the determination of the diastereoselectiv-
ity of the reaction of 9 and 10, ester 19a was hydro-
genated to obtain 20a, which was subjected to GC–MS
analysis. The chromatogram showed two well resolved
peaks; a major peak at tR=17.3 min was attributed to
20a, whereas a minor peak at tR=17.1 min was
It is also interesting to note that the reaction of 9
catalyzed with Rh2(OAc)4 proceeded with slightly
higher diastereoselectivity than those catalyzed using
1
ascribed to 20b (not detected by H NMR). Each of
these compounds showed m/z 229 (M−15) peaks in
their mass spectra. All the rearrangement products in
subsequent reactions were first transformed to 20a,b,
which were then analyzed by GC to ascertain the
degree of diastereocontrol during the formation of the
either
Rh2[(5R)-MEPY]4,
Rh2[(5S)-MEPY]4
or
Rh2(Cap)4; however, the chemical yields of the products
obtained with Rh2(OAc)4 are lower than those with the
Rh(II) carboximidates (entries 8–12).
Scheme 3.

902
A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
The influence of the double bond geometry in the
secondary allylic sulfide on the diastereoselectivity of
the rearrangement reaction was studied next. Our expe-
rience with the rearrangement of the primary sulfides
indicated that there was no significant improvement in
the diastereoselectivity of the reaction when the (Z)-
allylic sulfide 6b (entries 1 and 4) was used. In the case
of (Z)-10, however, we were pleased to find that high
diastereoselectivity was achieved with both Rh2(OAc)4
and Rh2(Cap)4 (entries 13 and 14).
group occupies the sterically less demanding pseudo-
equatorial position in the five-membered envelope21
TSs. The TSs B, B% and D, D% are higher in energy and
are not involved; B and B% are destabilized by an allylic
A1,3 strain22 between the pseudoaxial methyl group and
the olefinic hydrogen, and D and D% are destabilized by
the steric interaction between the pseudoaxial methyl
group and the dioxolanyl moiety.
2.5. Synthesis of (R)-4-(4-chlorophenyl)-2-butyrolactone
2.4. Reaction pathway
The excellent stereocontrol that is possible during bond
formation in the sulfonium ylide rearrangement of the
chiral non-racemic secondary allylic sulfide systems
encouraged us to apply the method to the synthesis of
(R)-4-(4-chlorophenyl)-2-butyrolactone 25, a key inter-
mediate used in the synthesis of the muscle relaxant and
GABAB receptor agonist (R)-balcofen.23
The overall results suggest that the 1,3-dioxolanyl moi-
ety is ineffective for 1,2-relative induction in the sulfo-
nium ylide rearrangement and furthermore, chiral
Rh(II) catalysts have no effect on the diastereoselectiv-
ity of the reaction. The C-5 stereocenter bearing the
sulfide unit in 9 and 10 was found to be a critical and
dominant factor that is essential for high diastereoselec-
tivity during the formation of the new stereocenter at
C-3. These findings are important in the context of the
synthesis, because the stereochemical information is
completely transferred from the C-5 stereocenter of the
sulfide to the new C-3 stereocenter in the product, and
occurred independently of the dioxolanyl stereocenter
located on the carbon directly adjacent to the reacting
prochiral C-3 carbon.
The known24 allylic alcohol 21 was prepared from the
(E)-4-(4-chlorophenyl-3-buten-2-one25 via reduction
with borane-dimethylsulfide catalyzed by (S)-2-MeCBS.
Under the usual optimized reaction conditions (vide
supra), a 78% ee (Chiralpak AD) of the product 21 was
realized. However, upon further investigations it was
found that if the time of addition of the enone to the
borane–catalyst mixture was increased to 1 h, the ee of
the alcohol 21 was increased to 92%. A 98% ee was
achieved when the addition time was increased to 3.5 h.
The results of the rearrangement reactions of the pri-
mary and secondary sulfides, 6, 9 and 10, indicate that
free sulfonium ylide intermediates were involved. The
poor diastereoselectivity of the reaction of the primary
sulfides (E)-6a and (Z)-6b, suggests that the 6a-derived
sulfonium ylide intermediate underwent sigmatropic
rearrangement via all four transition states (TS) A, A%,
B, B% (R=H, Fig. 1E) and the 6b-derived sulfonium
intermediate via TS-C, C%, D, D% (R=H, Fig. 1Z). In
the case of the secondary sulfides (E)-9 and (Z)-10,
excellent diastereocontrol during the rearrangement was
achieved. Furthermore, only two diastereomers were
present in the product 18, which differed only in the
stereochemistry at C-2. These results suggest that for
(E)-9 and (Z)-10, only TS-A and A% (R=Me, Fig. 1E)
and TS-C and C% (R=Me, Fig. 1Z) are involved,
respectively. The minor diastereomer 18a was formed
via TS-A% and C% and the major diastereomer 18b was
generated via TS-A and C. In all these TSs, the methyl
Treatment of the alcohol 21 (98% ee) with bis(p-
methoxyphenyl) disulfide using the above described
procedure (e.g. 12a“9) yielded the secondary sulfide
2226 (81%). The Rh(II)-catalyzed reaction of 22 with
EDA provided a mixture of diastereomers 23 that was
not separated, but was subjected to direct reductive
desulfurization to give 24 along with a very small
amount of an unidentified impurity (1H NMR).
Attempts to further purify 24 by repeated chromatogra-
phy were unsuccessful because the impurity co-eluted
with 24. The double bond in 24 was ozonized and then
reduced with sodium borohydride to give the corre-
sponding primary alcohol, which was subjected to lac-
tonization, catalyzed by p-toluenesulfonic acid (PTSA),
to form the target 2-butyrolactone 25 in 72% yield.
Chiral HPLC analysis of 25, however, indicated an ee
of 63%. The low enantioselectivity was unexpected and
surprising in light of the fact that the starting alcohol
Figure 1.

A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
903
21 had an ee of 98% and the fact that we have shown
that complete transfer of chirality was achieved for the
secondary sulfides 9 and 10 (vide supra). We reasoned
that partial racemization (ꢀ36%) of the C-2 stereocen-
ter must have occurred during the preparation of the
sulfide 22 from the alcohol 21. This may involve the
participation of a solvent caged ion pair 26, under the
Mitsonobu reaction conditions, which would lead to
the formation of 22 (the product of inversion) as the
major product and its enantiomer (the product of reten-
tion) as the minor product. The formation of the ion
pair is facilitated by resonance stabilization of the
carbocation intermediate 27.27 We did not detect any
SN2% product resulting from attack at the benzylic posi-
tion. The primary alcohol 28 was prepared from the
desulfurized product 24 and chiral HPLC analysis (Chi-
ralcel OD) showed its ee was 66%. This result further
supports the notion that partial loss of stereochemical
information at C-2 occurred at the sulfide formation
step (Scheme 4).
enantioselectivity was due to partial racemization dur-
ing the formation of the secondary allylic sulfide 22.
Nevertheless, the method developed here should find
application in synthesis. Other methods, which obviate
partial racemization during sulfide formation, are cur-
rently being investigated. These studies will be reported
in the future.
4. Experimental
4.1. General
Melting points were determined on a Kofler hot-stage
melting point apparatus and were uncorrected. Infrared
spectra were recorded using a Perkin-Elmer 1600FT
infra-red spectrophotometer and only diagnostic signals
are reported. NMR spectra were obtained on a Bruker
AC200 QNP spectrometer; chemical shifts are given in
parts per million (l) relative to the appropriate refer-
ence signal. 1H NMR spectra were obtained at 200
MHz in CDCl3 using tetramethylsilane (lH=0.00) or
CHCl3 (lH=7.26) as reference. Where appropriate,
double irradiation experiments were used for proton
assignments. 13C NMR spectra were obtained at 50.32
MHz in CDCl3 using CDCl3 (lC=77.00) as reference.
Optical rotations were measured on an Optical Activity
(AA-5) polarimeter at the sodium D line (u=589 nm).
High resolution mass spectra and elemental analyses
were performed at the Department of Chemistry, Uni-
versity of Saskatchewan. Gas chromatography (30 m×
0.25 m×0.25 mm SPB™-5 capillary column) was carried
out with a Hewlett–Packard Model 5890 Series II Gas
Chromatography equipped with a flame ionization
detector. GC–MS was carried out with a Hewlett–Pack-
ard Model 5890 Gas Chromatography (30 m×0.25 m×
0.25 mm SPB™-5 capillary column) equipped with a
Hewlett–Packard 5970A Series Mass Selective Detector
(70 eV ionization energy). Reaction progress was moni-
tored by thin-layer chromatography performed on
Merck® silica gel 60F254 precoated (0.25 mm) on alu-
minum-backed sheets. Flash chromatography was per-
formed on Merck® silica gel 60 (230–400 mesh).
Petroleum ether used is the fraction with bp 35–60°C.
Ultrasound-mediated reactions were done in a Sono-
gen® Automatic Cleaner (Model D-100, Branson
Instruments, Inc) with a frequency of 33 kHz and
average power output of 100 W. All air and moisture
sensitive reactions were carried out under a static pres-
sure of argon. All organic extracts were dried over
anhydrous Na2SO4. Acetone was dried by distillation
over KMnO4. CH2Cl2, Cl(CH2)2Cl, CH3CN, toluene,
benzene, pyridine and Et3N were dried by distillation
over CaH2. MeOH was dried by distillation over mag-
nesium methoxide (unless otherwise noted). THF was
dried by distillation over sodium using sodium benzo-
phenone ketyl as an indicator. HPLC was performed
using a Waters 600 pump equipped with a Waters 2487
UV detector (u=254 nm) and interphased to a
Millenium32 PC workstation. Allylic alcohols 6a,b were
prepared according to literature procedures.9 Rh2(Cap)4
[dirhodium(II) tetrakis(caprolactamate)] was prepared
according to literature procedures, (R)- and (S)-
3. Conclusions
The chiral dioxolanyl moiety that has served well as a
diastereocontrol group in other systems was found to
be ineffective in the tandem sulfonium ylide-[2,3]-sigma-
tropic rearrangement of the primary and secondary
allylic sulfides 6, 9 and 10. The rearrangement reaction
was found to involve free sulfonium ylide intermediates
and high diastereoselectivity was observed only for the
chiral non-racemic secondary sulfides 9 and 10. In these
cases, excellent and complete transfer of chirality (C-5-
substrate“C-3-product) was realized. The method was
used in the synthesis of (R)-4-(4-chlorophenyl)-2-buty-
rolactone, which was obtained in 63% ee. The modest
Scheme 4.

904
A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
Rh2(MEPY)4 [dirhodium(II) tetrakis(methyl 2-pyrro-
lidinone-5-carboxylate)] were purchased from Aldrich.
subjected to flash column chromatography (from 5:1
v/v pet. ether:ether to 1:1 v/v pet. ether:ether) to give
the alcohol 12 (896 mg, 90%). IR wmax: 3695–3084 cm−1.
1H NMR (signals of diastereomers were indistinguish-
able) l: 1.24 (d, 3H, J=6.8 Hz, H-6), 1.34 (s, 3H,
O-C-Me), 1.39 (s, 3H, O-C-Me), 1.61 (s, 1H, -OH), 3.54
(dt, 1H, J=8.4, 8.4, 1.0 Hz, H-1), 4.05 (ddd, 1H,
J=8.8, 6.7, 0.8 Hz, H-1), 4.22–4.39 (m, 1H, H-5),
4.47–4.53 (m, 1H, H-2), 5.61 (dd, 1H, J=15.4, 7.4 Hz,
H-3), 5.83 (dd, 1H, J=15.3, 5.8 Hz, H-4). 13C NMR l:
discernible signals for minor diastereomers in brackets,
23.0, (23.2), (25.7), 25.8, (26.5), 26.6, (66.2), 67.8, 69.3,
(68.0), (72.2), 76.4, 112.4, 126.8, (128.3), (136.6), 138.4.
4.2. (2S,3E)-1,2-Dihydroxy-1,2-O-isopropylidene-5-(4-
methoxyphenylthio)-3-pentene, 6a
The alcohol 5a (78 mg, 0.49 mmol) and freshly made
bis(4-methoxyphenyl) disulfide (178.5 mg, 0.64 mmol)
were dissolved in dry MeCN (12 mL). Bu3P (299.3 mg,
1.48 mmol) was added to the above solution and the
reaction mixture was immersed in an ultrasound water
bath. The reaction mixture was sonicated at rt for 1.5 h.
It was then diluted with CH2Cl2 (30 mL), washed with
10% aqueous NaOH solution (2×10 mL) then brine
(2×10 mL), dried, filtered and evaporated. The residue
was subjected to column chromatography (20:1 pet.
ether:ether and then 4:1 v/v pet. ether:ether) to yield the
sulfide 6a (118.7 mg, 86%). [h]2D2 +40.2 (c 3.5, CHCl3).
The diastereomeric mixture of 12a,b was converted
(NaH, PhCH2Br, cat. Bu4NI) to the corresponding
benzyl ether 14. The 1H NMR spectrum of 14a,b
showed two sets of signals. Integration of the H-1
double doublets centered at l 3.56 (14a) and at l 3.59
ppm (14b) gave an 88:12 ratio of 14a:14b. Data for
benzyl ether 14a,b: IR: 3064, 3031, 1496, 1454, 1372
cm−1. 1H NMR l (discernible signals for minor
diastereomers in square brackets): 1.20 (d, 3H, J=6.9
Hz, H-6), [1.21 (d, J=6.9 Hz)], 1.41 (s, 3H, O-C-Me),
1.44 (s, 3H, O-C-Me), 3.56 (dd, 1H, J=8.0, 7.4 Hz,
H-1), [3.59 (dd, J=8.6, 8.1 Hz)], 3.88–4.00 (m, 1H,
H-5), 4.02 (dd, 1H, J=8.6, 7.1 Hz, H-1), 4.31 (d, 1H,
J=11.9 Hz, O-CH-Ph), [4.30 (d, J=12.5 Hz)], 4.40–
4.53 (m, 2H, H-2, O-CH-Ph), 5.49–5.76 (m, 2H, H-3,
H-4), 7.13–7.33 (m, 5H, ArH).
1
IR wmax: 1592 cm−1. H NMR l: 1.36 (s, 3H, O-C-Me),
1.37 (s, 3H, O-C-Me), 3.34–3.44 (m, 3H, H-1, 2H-5),
3.80 (s, 3H, OMe), 3.98 (dd, 1H, J=8.6, 6.9 Hz, H-1),
4.37–4.50 (m, 1H, H-2), 5.27–5.40 (m, 1H, H-3), 5.71–
5.88 (m, 1H, H-4), 6.78–7.37 (m, 4H, ArH). 13C NMR
l: 25.8, 26.6, 38.1, 55.3, 69.3, 76.3, 109.5, 114.4, 129.8,
130.6, 134.5, 159.3. HRMS calcd for C14H18SO3
280.1133, found 280.1129.
4.3. (2S,3Z)-1,2-Dihydroxy-1,2-O-isopropylidene-5-(4-
methoxyphenylthio)-3-pentene, 6b
The sulfide 6b was prepared, as described for 6a, start-
ing from 5b (351.8 mg, 2.22 mmol), freshly made
bis(4-methoxyphenyl) disulfide (742.9 mg, 2.67 mmol),
Bu3P (1.13 g, 5.56 mmol) in MeCN (12 mL). Yield of
6b (20:1 pet. ether:ether and then 4:1 v/v pet.
ether:ether) was 576.8 mg, 92%. [h]2D2: −86.1 (c 0.9,
4.5. (2S,5S,3Z)-1,2-Dihydroxy-1,2-O-isopropylidene-3-
hexen-5-ol, 13a
Compound 13 was prepared, following the procedure
used for the formation of 12, from enone 11b (942 mg,
5.53 mmol), BH3–Me2S (2.0 M solution in THF, 1.66
mL, 3.32 mmol) and (R)-2-methyl-CBS-oxazaboro-
lidine (1 M solution in toluene, 3.32 mL, 3.32 mmol).
Yield of 13 was 801 mg (85%). GC analysis (inlet
temperature: 200°C, detector temperature: 300°C, oven
temperature: 80°C, 2 min; 80–100°C, 1°C/min ramp) of
13 revealed two well resolved peaks at tR=9.09 min
(13a) and at tR=9.15 min (13b). Integration of the GC
peak areas provided the 13a:13b ratio of 86:14. The
diastereomeric mixture was carefully separated using
flash column chromatography (1:1 v/v pet. ether:ether)
to give pure 13a (552 mg, 58%) as a white solid, mp
49–51°C. [h]D22: +18.6 (c 2.6, CHCl3). IR wmax: 3554–
1
CHCl3). IR wmax: 1591 cm−1. H NMR l: 1.31 (s, 3H,
O-C-Me), 1.36 (s, 3H, O-C-Me), 3.27 (‘t’, 1H, J=9.2
Hz, H-1), 3.37 (ddd, 1H, J=15.3, 8.8, 1.5 Hz, H-5),
3.57 (ddd, 1H, J=15.2, 10.6, 1.1 Hz, H-5), 3.62 (dd,
1H, J=9.9, 8.1 Hz, H-1), 3.80 (s, 3H, OMe), 4.40–4.56
(m, 1H, H-2), 5.27–5.52 (m, 1H, H-3), 5.63–5.80 (m,
1H, H-4), 6.72–7.37 (m, 4H, ArH). 13C NMR l: 25.9,
26.7, 33.8, 55.4, 69.1, 71.4, 109.2, 114.5, 125.0, 129.8,
129.8, 134.5, 159.6. HRMS calcd for C14H18SO3
280.1133, found 280.1135.
4.4. (2S,3E)-1,2-Dihydroxy-1,2-O-isopropylidene-3-
hexen-5-ol, 12
1
3248 cm−1. H NMR l: 1.21 (d, 3H, J=6.9 Hz, H-6),
The enone 11a (988 mg, 5.80 mmol, dried by azeotropic
distillation with dry benzene) in THF (10 mL) was
added dropwise via cannula to a solution of BH3–Me2S
(2.0 M solution in THF, 1.75 mL, 3.48 mmol), (S)-2-
methyl-CBS-oxazaborolidine (1 M solution in toluene,
3.50 mL, 3.48 mmol) and THF (60 mL) at −20°C. The
reaction was quenched after 30 min with cold dry
methanol (10 mL) at −20°C. Saturated aqueous
NaHCO3 (20 mL) was added 10 min later and the
mixture turned milky white. The mixture was extracted
with Et2O (20 mL×3). The combined organic phases
were washed with brine (20 mL), dried with anhydrous
MgSO4, filtered and concentrated. The residue was
1.37 (s, 3H, O-C-Me), 1.40 (s, 3H, O-C-Me), 2.50 (s,
1H, -OH), 3.52 (dd, 1H, J=8.7, 8.6 Hz, H-1), 4.04 (dd,
1H, J=9.2, 7.1 Hz, H-1), 4.57–4.73 (m, 1H, H-5),
4.79–4.93 (m, 1H, H-2), 5.42 (dd, 1H, J=11.9, 8.9 Hz,
H-3), 5.63 (dd, 1H, J=12.0, 9.3 Hz, H-4). 13C NMR l:
23.5, 25.9, 26.7, 63.7, 69.4, 71.7, 109.4, 127.5, 138.6.
A small sample of 13a (50 mg, 0.29 mmol) was con-
verted to the benzyl ether 16 (60 mg, 79%). [h]2D2: −19.1
1
(c 2.9, CHCl3). IR wmax: 1454, 1370 cm−1. H NMR l:
1.23 (d, 3H, J=6.8 Hz, ME), 1.38 (s, 3H, O-C-Me),
1.43 (s, 3H, O-C-Me), 3.52 (dd, 1H, J=8.3, 7.8 Hz,
H-1), 3.97 (dd, 1H, J=8.6, 7.0 Hz, H-1), 4.27 (ddd, 1H,

A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
905
J=7.6, 7.1, 1.4 Hz, H-5), 4.38 (d, 1H, J=13.3 Hz,
OCHPh), 4.62 (d, 1H, J=13.6 Hz, OCHPh), 4.65–4.79
(m, 1H, H-2), 5.54–5.70 (m, 2H, H-3, H-4), 7.23–7.39
(m, 5H, ArH). 13C NMR l: 22.0, 25.9, 26.7, 69.3, 69.9,
71.8, 109.4, 127.5, 127.8, 128.3, 129.5, 136.3, 138.5.
HRMS calcd for C16H22O3 262.1569, found 262.1571.
−3.5 (c 9.9, CHCl3). IR wmax: 3104, 1729, 1629, 1547
1
cm−1. H NMR l: 1.37 (s, 3H, O-C-Me), 1.42 (s, 3H,
O-C-Me), 1.51 (d, 3H, J=6.4 Hz, Me), 3.59 (dd, 1H,
J=7.8, 7.6 Hz, H-1), 4.12 (dd, 1H, J=8.2, 6.4 Hz,
H-1), 4.52 (m, 1H, H-5), 5.62–6.02 (m, 3H, H-2, H-3,
H-4), 9.12 (d, 2H, J=2.0 Hz, Ar-H), 9.21 (t, 1H, J=2.0
Hz, Ar-H), 13C NMR l: 19.9, 25.7, 26.5, 29.6, 69.2,
73.0, 75.9, 109.6, 122.3, 129.4, 131.0, 131.6, 134.1,
148.6, 161.6. Anal. calcd for C16H18N2O8: C, 54.46; H,
4.95; N, 7.65. Found: C, 52.50; H, 5.11; N, 7.57.
4.6. Confirmation of absolute configuration of 12a and
13a
Ozone was bubbled into a solution of benzyl ether
14a,b (60 mg, 0.229 mmol) or 16 (52 mg, 0.20 mmol) in
methanol (10 mL) at −78°C for 46 s. The reaction
mixture was warmed to 0°C and NaBH4 (26 mg, 0.686
mmol) was added. After stirring for 1 h, the reaction
mixture was warmed to rt. The reaction was monitored
by TLC and when the reaction was complete glacial
AcOH (2 drops) was added to destroy excess NaBH4.
50% NaHCO3 aqueous solution was used to neutralize
the mixture. After the solvent was evaporated, the
residue was subjected to flash column chromatography
(1:1 v/v pet. ether:ether) to obtain the product: (R)-17a
(30 mg, 80%) or (S)-17b (27.5 mg, 85%).
4.7.2. (2) Base hydrolysis of derivative 15a. The ester 15a
(300 mg, 0.819 mmol) was dissolved in methanol (10
mL) and powdered K2CO3 (170 mg, 1.23 mmol) was
added. The pink solution was stirred at rt for 20 h. The
solvent was then evaporated and the residue was taken
into CH2Cl2 (20 mL), washed with brine (2×5 mL),
dried, concentrated. Purification by flash column chro-
matography (1:1 v/v pet. ether:ether) yielded pure alco-
hol 12a (140 mg, 100%). [h]2D2: +24 (c 1.8, CHCl3) {lit.18
+28 (c 1.4)}. IR wmax: 3695–3085 cm−1. 1H NMR l: 1.24
(d, 3H, J=6.6 Hz, Me), 1.35 (s, 3H, O-C-Me), 1.39 (s,
3H, O-C-Me), 1.97 (s, 1H, -OH), 3.55 (dd, 1H, J=8.1,
8.0 Hz, H-1), 4.05 (dd, 1H, J=8.1, 6.6 Hz, H-1),
4.22–4.37 (m, 1H, H-5), 4.40–4.53 (m, 1H, H-2), 5.61
(dd, 1H, J=15.4, 7.8 Hz, H-3), 5.83 (dd, 1H, J=15.6,
5.5 Hz, H-4). 13C NMR l: 23.0, 25.8, 26.6, 67.7, 69.3,
76.4, 109.0, 126.9, 138.4. A small sample of pure 12a
was converted to the benzyl ether 14a for further
characterization. Its NMR data showed only one set of
signals. [h]2D2: +59.4 (c 0.8, CHCl3). IR wmax: 1454, 1370
4.6.1. (R)-2-Benzyloxy-1-propanol, 17a. [h]2D2: −32 (c 0.3,
CHCl3) {lit.19a −47 (c 1.0, CHCl3)}. IRwmax: 3601–3143,
1
3063, 3031, 1453, 1375 cm−1. H NMR l: 1.16 (d, 3H,
J=5.8 Hz, -Me), 2.20 (s, 1H, -OH), 3.45–3.76 (m, 3H,
-CH2-CH-), 4.50 (d, 1H, J=11.4 Hz, -OCHPh), 4.66 (d,
1H, J=11.4 Hz, -OCHPh), 7.22–7.40 (m, 5H, ArH).
13C NMR l: 15.8, 66.3, 70.7, 75.5, 127.7, 127.7, 128.4,
128.4, 138.4. HRMS calcd for C10H14O2 166.0994,
found 166.0996.
1
cm−1. H NMR l: 1.20 (d, 3H, J=6.8 Hz, Me), 1.32 (s,
3H, O-C-Me), 1.37 (s, 3H, , O-C-Me), 3.49 (‘t’, 1H,
J=8.6 Hz, H-1), 3.82–3.97 (m, 1H, H-5), 4.02 (dd, 1H,
J=9.0, 6.9 Hz, H-1), 4.31 (d, 1H, J=13.0 Hz, OCH-
Ph), 4.39–4.53 (m, 2H, H-5, O-CH-Ph), 5.50–5.77 (m,
2H, H-3, H-4), 7.15–7.28 (m, 5H, ArH). 13C NMR l:
21.4, 25.9, 26.7, 69.5, 70.1, 74.8, 76.5, 109.4, 127.4,
127.5, 127.6, 128.3, 129.3, 136.0, 138.6. HRMS calcd
for C16H22O3 262.1569, found 262.1559.
4.6.2. (2S)-2-Benzyloxy-1-propanol, 17b. [h]2D2: +44 (c
1.3, CHCl3) {lit.19b +48.0 (c 6.4, CHCl3)}. IR wmax
:
1
3601–3143, 3063, 3031, 1453, 1375 cm−1. H NMR l:
1.18 (d, 3H, J=5.9 Hz, -Me), 2.41 (bs, 1H, -OH),
3.45–3.77 (m, 3H, -CH2-CH-), 4.49 (d, 1H, J=11.5 Hz,
-OCHPh), 4.67 (d, 1H, J=11.5 Hz, -OCHPh), 7.20–
7.45 (m, 5H, ArH). 13C NMR l: 15.8, 66.3, 70.7, 75.5,
127.7, 127.7, 128.4, 138.0.
4.8. (2S,5S,3E)-1,2-dihydroxy-1,2-O-isopropylidene-5-
(4-methoxyphenylthio)-3-hexene, 9
4.7. Obtaining pure (2S,5R,3E)-1,2-dihydroxy-1,2-O-
isopropylidene-3-hexen-5-ol, 12a
The alcohol 12a (750 mg, 4.36 mmol) and freshly made
bis(4-methoxyphenyl) disulfide (2.43 g, 8.71 mmol) was
dissolved in dry toluene (100 mL). Bu3P (3.53 g, 17.42
mmol) was added and the mixture was heated at reflux
for 3 h. Toluene was evaporated and the residue was
diluted with CH2Cl2 (30 mL) and then washed with
10% NaOH (10 mL×2). The aqueous phase was re-
extracted with CH2Cl2 (10 mL); the combined CH2Cl2
layers were washed with brine (10 mL×2), dried, con-
centrated and subjected to flash column chromatogra-
phy (from 20:1 pet. ether:ether to 1:1 v/v pet.
ether:ether) to give the sulfide 9 (989 mg, 81%). GC
analysis (inlet temperature: 200°C, detector tempera-
ture: 300°C, oven temperature: 100°C (2 min), 100–
200°C, 10°C/min) showed a single peak at tR=18.2
min. [h]2D1: −1.9 (c 6.4, CHCl3). IR wmax: 1592, 1494
4.7.1. (1) Preparation of 3,5-dinitrobenzoate derivative
15a,b. The alcohol 12a,b (900 mg, 5.23 mmol) and
DMAP (63.9 mg, 0.523 mmol) were dissolved in
CH2Cl2 (20 mL) and were cooled to 0°C. Et3N (5 mL)
was added to the solution and it was stirred for 10 min.
3,5-Dinitrobenzoyl chloride (1.57 g, 6.79 mmol) in
CH2Cl2 (10 mL) was added via cannula to the above
mixture. The solution turned initially orange then
brown. The reaction mixture was slowly warmed to rt
and then stirred overnight. The reaction mixture was
washed with saturated NaHCO3, brine, dried with
anhydrous Na2SO4, filtered, evaporated and subjected
to flash column chromatography (2:1 pet. ether:ether)
to give a light yellow solid (1.38 g, 72%). The
diastereomeric mixture was subjected to fractional
recrystallization (3:1 v/v pet. ether:EtOAc) to obtain
the major (5R)-15a (1.19 g, 86% ). mp: 88–89.5°C. [h]2D1:
1
cm−1. H NMR l: 1.32 (d, 3H, J=7.5 Hz, Me), 1.32 (s,
3H, O-C-Me), 1.33 (s, 3H, O-C-Me), 3.19 (dd, 1H,
J=8.2, 7.7 Hz, H-1), 3.42–3.62 (m, 1H, H-5), 3.76 (s,

906
A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
3H, -OMe), 3.87 (dd, 1H, J=7.7, 6.0 Hz, H-1), 4.28–
4.42 (m, 1H, H-2), 5.02 (dd, 1H, J=15.0, 7.9 Hz, H-3),
5.60 (dd, 1H, J=15.0, 8.8 Hz, H-4), 6.65–7.35 (m, 4H,
ArH). 13C NMR l: 19.9, 25.8, 26.6, 47.0, 55.3, 69.2,
76.6, 109.2, 114.2, 115.0, 124.7, 127.7, 136.3, 136.6.
HRMS calcd for C16H22O3S 294.1290, found 294.1290.
9.6, 2.8 Hz, H-3), 3.69 (dd, 1H, J=7.9, 7.2 Hz, H-5),
3.75 (d, 1H, J=10.5 Hz, H-2), 3.80 (s, 3H, OMe), 3.95
(dd, 1H, J=8.4, 6.9 Hz, H-5), 3.98–4.26 (m, 3H, H-4,
-CO2CH2Me), 5.20 (dd, 1H, J=17.5, 2.3 Hz, ꢁCH),
5.38 (dd, 1H, J=10.7, 2.4 Hz, ꢁCH), 5.67–5.88 (m, 1H,
-CHꢁ), 6.77–7.48 (m, 4H, ArH). HRMS (total mixture)
calcd for C19H26SO5 366.1501, found 366.1495.
4.9. (2S,5R,3Z)-1,2-dihydroxy-1,2-O-isopropylidene-5-
(4-methoxyphenylthio)-3-hexene, 10
Rearrangement products 18 were separated into two
components by careful column chromatography (20:1
v/v pet. ether:ether to 2:1 v/v pet. ether:ether): The less
polar component 18a, (Rf=0.38, 5:1 pet. ether:Et2O):
Compound 10 was prepared in the same way as allylic
sulfide 9, starting from 13a (390 mg, 2.26 mmol),
bis(4-methoxyphenyl) disulfide (1.58 g, 5.66 mmol) and
Bu3P (1.83 g, 9.06 mmol) to give the sulfide 10 (514 mg,
90%). GC analysis (inlet temperature: 200°C, detector
temperature: 300°C, oven temperature: 100°C (2 min),
100–200°C, 10°C/min) showed a single peak at tR=15.8
min. [h]2D3: −150 (c 0.8, CH3OH) . IR wmax: 1591, 1493
1
IR wmax: 1732, 1593, 1495 cm−1. H NMR l: 1.08 (t,
3H, J=7.3 Hz, -CO2CH2Me), 1.26 (s, 3H, O-C-Me),
1.31 (s, 3H, O-C-Me), 1.62 (dd, 3H, J=7.2, 1.6 Hz,
ꢁCHMe), 2.36 (ddd, 1H, J=11.0, 8.7, 1.8 Hz, H-3),
3.56 (t, 1H, J=7.5 Hz, H-5), 3.66 (d, 1H, J=8.8 Hz,
H-2), 3.73 (s, 3H, OMe), 3.86–4.06 (m, 3H, H-5%,
-CO2CH2Me), 4.72 (ddd, 1H, J=7.5, 7.5, 2.5 Hz, H-4),
5.30 (ddd, 1H, J=15.7, 8.8, 1.6 Hz, -CHꢁ), 5.45 (ddd,
1H, J=15.7, 6.3, 6.3 Hz, ꢁCHMe), 6.75 (d, 2H, J=8.2
Hz, ArH), 7.37 (d, 2H, J=8.0 Hz, ArH). 13C NMR l:
14.3, 18.1, 25.2, 26.1, 46.3, 54.9, 55.3, 60.9, 66.9, 73.9,
108.8, 114.4, 124.2, 125.2, 131.4, 135.7, 160.0. The more
polar component, 18b (Rf=0.25, 5:1 pet. ether:Et2O):
1
cm−1. H NMR l: 1.33 (d, 3H, J=7.3 Hz, Me), 1.28 (s,
3H, O-C-Me), 1.31 (s, 3H, O-C-Me), 3.02 (dd, 1H,
J=9.0, 8.6 Hz, H-1), 3.19 (dd, 1H, J=9.0, 7.1 Hz,
H-1), 3.77 (s, 3H, OMe), 3.82–4.00 (m, 1H, H-5),
4.26–4.40 (m, 1H, H-2), 5.26 (dd, 1H, J=11.4, 10.0 Hz,
H-3), 5.47 (dd, 1H, J=11.6, 11.4 Hz, H-4), 6.77–7.39
(m, 4H, ArH). 13C NMR l: 20.6, 25.8, 42.5, 55.3, 68.8,
71.5, 108.5, 114.3, 124.5, 127.2, 136.5, 137.3, 159.8.
HRMS calcd for C16H22SO3 (M+−1) 294.1290, found
294.1285.
1
IR wmax: 1727, 1593, 1494 cm−1. H NMR l: 1.05 (t,
3H, J=6.5 Hz, -CO2CH2Me), 1.22 (s, 3H, O-C-Me),
1.30 (s, 3H, O-C-Me), 1.71 (dd, 3H, J=7.0, 1.2 Hz,
ꢁCHMe), 2.70 (ddd, 1H, J=9.4, 9.4, 3.0 Hz, H-3), 3.60
(t, 1H, J=7.5 Hz, H-5), 3.68 (d, 1H, J=10.3 Hz, H-2),
3.72 (s, 3H, OMe), 3.85 (dd, 1H, J=7.5, 6.5 Hz, H-5%),
3.91–4.08 (m, 3H, H-4, MeCH2O), 5.28 (ddd, 1H,
J=14.0 9.5, 1.1 Hz, CHꢁ) 5.53 (ddd, 1H, J=13.7, 6.3,
6.3 Hz, ꢁCHMe), 6.75 (d, 2H, J=8.2 Hz, ArH), 7.33
(d, 2H, J=8.2 Hz, ArH). 13C NMR l: 14.0, 18.1, 25.1,
26.0, 46.5, 53.6, 55.3, 60.9, 66.8, 76.1, 109.1, 114.3,
123.3, 125.7, 131.2, 135.5, 136.4, 171.9. Anal. calcd for
C20H28O5S (total mixture): C, 63.13; H, 7.42. Found: C,
63.15; H, 7.35.
4.10. General procedure for the sulfonium ylide forma-
tion-[2,3]-sigmatropic rearrangement reaction of 6a,b, 9
and 10
The primary or secondary allylic sulfide 6a, 6b, 9 or 10
(1.0 mmol) was dissolved in dry chlorobenzene or
dichloromethane (10 mL) under Ar. The Rh (II) cata-
lyst (5 mol %, predried at 100°C at 0.5 mmHg for 1 h)
was added and the mixture was heated at the appropri-
ate temperature (Table 1) or was stirred at rt. A solu-
tion of EDA (3.0 mmol) in the same solvent was added
via syringe pump over a period of 30 min. After the
addition was complete, the mixture was stirred for an
additional 30 min. The solvent was then evaporated
and the residue was purified by flash column
chromatography.
4.11. General procedure for the reduction of rearrange-
ment products 7 and 18
Bu3SnH (0.22 mmol) was syringed to a 25 mL flask
under argon. Toluene (5 mL) was added to the above
flask and the solution was heated to 80°C. To the above
solution was added sulfide 7 or 18 (0.1 mmol) and
AIBN (0.02 mmol) in toluene (5 mL) via syringe pump
over a period of 1 h. The reaction was kept at 80°C for
2 h. The solvent was evaporated and the residue was
added 10% KF aqueous solution (5 mL) and ether (5
mL). The resulting mixture was vigorously stirred at rt
for 1 h. The organic phase was collected and the
inorganic phase was extracted with ether (2×5 mL). The
combined organic phases were evaporated and the
residue was purified by flash column chromatography.
Rearrangement products 7 were separated by careful
column chromatography (20:1 v/v pet. ether–ether and
then 3:1 v/v pet. ether–ether) into two diastereomeri-
cally enriched components: The more polar component
consisted of mainly one diastereomer whereas the less
polar component consisted of at least three
diastereomers. The less polar component: IRwmax: 1732,
1
1593 cm−1. H NMR l: [1.16 (t, J=7.2 Hz), 1.17 (t,
J=7.5 Hz), 1.19 (t, J=7.5 Hz)] (3H, -CO2CH2Me),
[1.32 (s), 1.33 (s), 1.37 (s)] (6H, O-C-Me), [2.49 (ddd,
J=10.8, 10.0, 3.0 Hz), 2.69 (ddd, J=9.2, 8.8, 5.0 Hz),
2.85 (ddd, J=9.4, 9.0, 6.0 Hz)] (1H, H-3), 3.61–3.83 (m,
2H), 3.80 (s, 3H, OMe), 3.89–4.33 (m, 4H), 5.07–5.98
(m, 3H, -CHꢁCH2), 6.71–7.51 (m, 4H, ArH). The more
polar component: IR: 1730, 1592 cm−1. 1H NMR l: 1.13
(t, 3H, J=7.4 Hz, -CO2CH2Me), 1.29 (s, 3H, O-C-
CH3), 1.38 (s, 3H, O-C-Me), 2.67 (ddd, 1H, J=10.0,
4.12. Rearrangement products 8a,b: yield: 66–73%
4.12.1. Less polar diastereomer 8a. (Rf=0.32, 5:1 pet.
ether:ether): [h]D22: +23.2 (c 1.0, CHCl3) {lit.13b enan-
tiomer of 8a: −14.5 (c 2.0)}. IRwmax: 1735 cm−1. 1H
NMR l: 1.18 (t, 3H, J=7.3 Hz, -CO2CH2Me), 1.27 (s,

A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
907
3H, O-C-Me), 1.33 (s, 3H, O-C-Me), 2.27 (dd, 1H,
J=16.0 10.5 Hz, H-2), 2.58–2.69 (m, 2H, H-2, H-3),
3.56–3.67 (m, 1H, H-5), 3.83–3.96 (m, 2H, H-4, H-5),
4.07 (q, 2H, J=7.2 Hz, -CO2CH2Me), 5.00–5.66 (m,
3H, -CHꢁCH2). 13C NMR l: 14.3, 25.2, 26.2, 35.9,
42.8, 60.5, 66.6, 77.6, 109.1, 117.8, 135.9, 172.1.
17.0, 9.0 Hz, H-2), 3.59 (t, 1H, J=7.8 Hz, H-5), 3.96
(dd, 1H, J=7.8, 6.3 Hz, H-5), 4.12 (q, 2H, J=7.2 Hz,
-CO2CH2Me), 4.00–4.13 (m, 1H, H-4). 13C NMR l:
14.1, 14.2, 20.0, 25.2, 26.4, 32.9, 34.9, 37.3, 60.3, 66.4,
77.6, 108.0, 173.0. GC–MS: 20a, tR=10.8 min, m/z (rel.
intensity): 229 (M+, 68%), 187 (M−MeCHꢁCH2, 14%),
141 (M−CH3CO2Et, 100%), 101 (2,2-dimethyldiox-
olanyl cation, 93%); 20b, tR=10.7 min, m/z (rel. inten-
sity): 229 (M+, 78%), 141 (M−CH3CO2Et), 101
(2,2-dimethyldioxolanyl cation, 100%) HRMS: calcd
for C12H21O4 (M+−15) 229.1440, found 229.1444.
4.12.2. More polar diastereomer 8b13a. (Rf=0.25, 5:1
pet. ether:ether): [h]2D2: +15.7 (c 1.6, CHCl3) {lit.13b
1
enantiomer of 8b: −17.2 (c 1.8)}. IRwmax: 1736 cm−1. H
NMR l: 1.18 (t, 3H, J=7.4 Hz, -CO2CH2Me), 1.27 (s,
3H, O-C-Me), 1.34 (s, 3H, O-C-Me), 2.31 (dd, 1H,
J=15.0, 8.6 Hz, H-2), 2.46 (dd, 1H, J=15.2, 5.6 Hz,
H-2), 2.64–2.78 (m, 1H, H-3), 3.59 (dd, 1H, J=8.0, 6.7
Hz, H-5), 3.92 (dd, 1H, J=8.5, 7.0 Hz, H-5), 4.02–4.11
(m, 1H, H-4), 4.05 (q, 2H, J=7.2 Hz, -CO2CH2Me),
5.02–5.77 (m, 3H, -CHꢁCH2). 13C NMR l: 14.2, 25.5,
26.7, 36.6, 44.9, 60.3, 67.9, 77.5, 109.9, 117.8, 136.4,
172.1.
4.15. (2R,3E)-4-(4-Chlorophenyl)-3-buten-2-ol, 21
(3E)-4-(4-Chlorophenyl)-3-buten-2-one (336 mg, 1.86
mmol) was dissolved in THF (8 mL) and the solution
was added over a period of 3.5 h, via a syringe pump,
to a mixture of BH3·Me2S (2.0 M in THF, 1.11 mmol,
0.57 mL) and (S)-2-methyl-CBS-oxazaborolidine (1 M
in toluene, 1.11 mmol, 1.11 mL) in THF (10 mL) at −10
to −15°C. After 40 min, the reaction mixture was
quenched with dry cold MeOH (5 mL) at −15°C. Then
saturated aqueous NaHCO3 solution (20 mL) was
added and after 5 min the mixture turned milky white.
The mixture was extracted thoroughly with ether (3×20
mL). The combined organic layer was washed with
brine (20 mL), dried, filtered and concentrated. The
residue was purified by column chromatography (pet.
ether–EtOAc 6:1 v/v) to give the allylic alcohol 21 (321
mg, 95%) as a white solid, mp: 61–61°C. [h]2D2: +27.3 (c
0.6, CHCl3), [h]D20 +17.7 (c 1.5, CH2Cl2). % ee: 98%
(Chiralpack AD, 98:2 v/v hexane:2-propanol, 1.0 mL/
min): (S)-21b; tR=14.4 min, (R)-21b; tR=15.0 min). IR
4.13. Rearrangement product 19
Compound 19 (5:1 pet. ether:ether), yield was 76–88%.
1
IR wmax: 1738, 1449 cm−1. H NMR l (only resonances
for 19a observed): 1.21 (t, 3H, J=7.8 Hz,
-CO2CH2Me), 1.31 (s, 3H, O-C-Me), 1.36 (s, 3H, O-C-
Me), 1.64 (dd, 3H, J=6.2, 1.1 Hz, Me), 2.29 (dd, 1H,
J=15.1, 9.2 Hz, H-2), 2.45 (dd, 1H, J=14.9, 5.4 Hz,
H-2%), 2.66–2.78 (m, 1H, H-3), 3.61 (dd, 1H, J=8.0, 7.4
Hz, H-5), 3.92 (dd, 1H, J=8.0, 6.3 Hz, H-5%), 4.01–4.15
(m, 3H, H-4, MeCH2O), 5.28 (ddq, J=15.4 8.6, 1.1 Hz,
CHꢁ), 5.44–5.67 (m, 1H, ꢁCHMe). 13C NMR l: 14.2,
18.0, 25.1, 26.1, 37.1, 41.6, 60.2, 66.5, 77.5, 108.7, 128.5,
172.2. HRMS (19 from rearrangement of 9) calcd for
C12H19O4 (M+−15) 227.1283, found 227.1285. HRMS
(20 from rearrangement of 10) calcd for C12H19O4
(M+−15) 227.1283, found 227.1279.
1
wmax (neat): 3354, 3050, 1652, cm−1. H NMR l: 1.38 (
d, J=6.38 Hz, 3H, CH3), 1.61 (s, 1H, OH), 4.46–4.53
(m, 1H, H-2), 6.24 (dd, J=6.2, 15.9 Hz, 1H, H-3), 6.53
(d, J=15.9 Hz, 1H, H-4), 7.29–7.34 (m, 4H, Ar-H). 13C
NMR l: 23.4, 68.7, 103.6, 127.7, 127.9, 128.7, 134.2,
135.2.
4.14. Ethyl (4S)-4,5-dihydroxy-4,5-O-isopropylidene-3-
(propyl)pentanoate, 20
4.16. (S)-4-(4-Chlorophenyl)-2-(4-methoxyphenylthio)-3-
butene, 22
To a Parr flask was added 19 (12.1 mg, 0.05 mmol),
10% palladized charcoal (70% weight by weight of
catalyst to substrate) and dry MeOH (15 mL). The
flask was then connected to a Parr apparatus. The flask
was evacuated and flushed with H2 gas and this process
was repeated three times. The flask was filled with H2
gas to the pressure of 35 psi. and it was shaken for 1.5
h. Then the flask was evacuated, equilibrated to atmo-
spheric pressure and removed. The reaction mixture
was filtered through Celite and the filter cake was
washed with CH2Cl2 (3×5 mL). The solvent was evapo-
rated and the residue was filtered through a short
silica-gel pad to give the hydrogenated compounds
(11.5 mg, 94%). GC–MS analysis (inlet temperature:
200°C, detector temperature: 300°C, oven temperature:
100°C (2 min), 100–150°C, 3°C/min) of the mixture
provided the ratio of 20a (tR=17.3 min) and 20b (tR=
The sulfide 22 was prepared, as described for com-
pound 9, starting from the alcohol 21 (234 mg, 1.28
mmol) and freshly prepared bis(4-methoxyphenyl)-
disulfide (714 mg, 2.56 mmol) and Bu3P (1.04 g, 5.12
mmol) in toluene (25 mL). After the mixture was
refluxed (20 h) the solvent was evaporated and the
residue was taken into CH2Cl2 (10 mL), washed with
10% aqueous NaOH (10 mL), and the aqueous phase
was reextracted with CH2Cl2 (10 mL), the combined
CH2Cl2 layers were washed with brine (20 mL), dried,
concentrated and purified by column chromatography
(pet. ether:ether 15:1v/v) to give the product 22 (316
mg, 81%), mp: 41–42°C, [h]2D2=+219.0 (c 1.0, CHCl3).
1
IR wmax (neat): 3002, 1591, 1090 cm−1. H NMR l: 1.43
(d, J=6.8 Hz, 3H, CH3), 3.67–3.75 (m, 1H, H-3), 3.78
(s, 3H, OCH3), 5.97–6.19 (m, 2H, H-1. H-2), 6.79–6.83
(m, 2H, Ar-H), 7.16–7.39 (m, 6H, Ar-H). 13C NMR l:
20.2, 47.6, 55.3, 114.1, 127.4, 127.6, 128.5, 128.6, 128.7,
129.2, 132.4, 136.7, 159.7. HRMS calcd for C17H17ClOS
304.0689, found, 304.0693.
1
17.1 min). IR wmax: 1736 cm−1. H NMR l: 0.90 (t, 3H,
J=6.7 Hz, -(CH2)2Me), 1.28–1.42 (m, 4H, -CH2CH2-),
1.26 (t, 3H, J=7.2 Hz, MeCH2O), 1.33 (s, 3H, O-C-
Me), 1.39 (s, 3H, O-C-Me), 2.11 (dd, 1H, J=17.0, 6.8
Hz, H-2), 2.11–2.18 (m, 1H, H-3), 2.32 (dd, 1H, J=

908
A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
4.17. (R)-4-(4-Chlorophenyl)-2-butyrolactone, 25
The primary alcohol (30 mg, 0.12 mmol) was dissolved
in benzene (5 mL) and PTSA (9.1 mg, 0.05 mmol, 40%
mol) was added. The mixture was heated at 80°C for 6
h and then benzene was distilled off over a period of 2
h. Et3N (1 mL) and CH2Cl2 (5 mL) was added to the
residue and the mixture was washed with saturated
NaCl solution, dried, filtered and the crude product
purified by column chromatography (pet. ether–ether,
3:1 v/v) to yield (R)-25 (24.1 mg, 72%). [h]2D0=−29.2
(c 1.2, CHCl3) {lit.19b [h]D20=−48.5 (c 0.5, CHCl3), lit.23
[h]2D6 −50.9 (c 0.7, CHCl3)}. %e.e: 63% (Chiralpak AD,
1.0 mL/min, 96:4 v/v hexane:2-propanol): (R)-25; tR=
31.0 min, (S)-25; tR=29.5 min. IR wmax (neat): 3060,
The sulfide 22 (189 mg, 0.62 mmol) and Rh2(OAc)4
(13.72 mg, 0.031 mmol, 5 mol %, predried at 80°C at
0.5 mmHg for 2 h) were dissolved in CH2Cl2 (6 mL).
A solution of EDA (1.86 mmol) in 5 mL CH2Cl2 was
added via syringe pump over a period of 4 h at 45°C.
The reaction was conducted at 50°C for 4 h. The
solvent was removed under reduced pressure and the
residue was purified by column chromatography (pet.
ether-ether, 30:1 v/v) to obtain 23 as a mixture of
diastereomers (184 mg, 76%). IR wmax (neat): 3013,
3036, 1731, 1638, 1592 cm−1. 1H NMR l: [1.18 (t,
J=7.0 Hz] and 1.17 (t, J=7.0 Hz) (3H,
-CO2CH2CH3) 1.57–1.61 (m, 3H, H-6), 3.83 (s, 3H,
OCH3), 3.74–3.86 (m, 2H, H-2, H-3), 4.03–4.15 (m,
2H, -OCH2CH3), 5.51–5.63 (m, 2H, -CHꢁCH-), 6.73–
6.86 (m, 2H, Ar-H), 7.08–7.45 (m, 6H, Ar-H). HRMS
calcd for C21H23ClO3S 390.1056, found, 390.1061.
1
1781, 1023 cm−1. H NMR l: 2.62 (dd, J=17.4 8.8 Hz,
1H, H-3), 2.93 (dd, J=17.4, 8.7 Hz, 1H, H-3), 3.79
(quint., J=8.2 Hz, 1H, H-4), 4.23 (dd, J=8.8, 7.9 Hz,
1H, H-5), 4.66 (dd, J=8.8, 8.0 Hz, 1H, H-5), 7.17 (d,
J=8.4 Hz, 2H, Ar-H), 7.34 (d, J=8.4 Hz, 2H, Ar-H).
13C NMR l: 35.6, 40.4, 73.7, 128.0, 128.5, 129.2, 137.9,
175.9.
A solution of compound 23 (137 mg, 0.35 mmol) and
AIBN (11.49 mg, 0.082 mmol) in 10 mL benzene (10
mL) was added via syringe (syringe pump) over a
period of 2 h to a solution of Bu3SnH (224 mg, 0.77
mmol) in benzene (10 mL) at 80°C. The reaction was
kept at 80°C for 2 h and benzene was evaporated. The
crude product then dissolved in ether (10 mL) and
10% aqueous KF solution (10 mL) was added. The
mixture was vigorously stirred at rt for 1 h and the
organic phase was separated. The aqueous phase was
reextracted with ether (2×10 mL). The combined
organic phases were dried, filtered and evaporated,
and the residue was purified by column chromatogra-
phy (pet. ether–ether, 20:1 v/v). A small amount of an
unidentified impurity co-eluted with the ester 24 (48
4.18. (R)-3-(4-Chlorophenyl)-1-hexanol, 28
Alkene 24 (130 mg, 0.52 mmol) was hydrogenated
over 10% palladized charcoal using a Parr hydrogena-
tor to provide ethyl 3-(4-chlorophenyl)hexanoate (94
mg, 78%). This compound was used immediately in
the next step.
Ethyl 3-(4-chlorophenyl)hexanoate (88 mg, 0.35
mmol) was dissolved in dry THF (20 mL). LiAlH4
(40 mg, 1.1 mmol) was added to the solution and the
reaction mixture was refluxed for 5 h. The reaction
mixture was cooled to 0°C, and water (2 mL) was
added. The mixture was stirred for 10 min, 10%
aqueous NaOH (3 mL) was added and the resulting
mixture was stirred for an additional 20 min. The
mixture was extracted with ether (2×10 mL) and the
combined organic phases were dried, evaporated and
the residue was purified by column chromatography
to yield 28 (56 mg, 76%). [h]2D2=−3.3 (c 1.5, CHCl3),
% ee: 66% (Chiralcel OD, 1.0 mL/min, 98:2 v/v hex-
ane:2-propanol): (R)-28; tR=11.2 min, (S)-28; tR=
mg, 54%). IR wmax (neat): 3025, 1732, 1092 cm−1. H
1
NMR l: 1.17 (t, J=7.2 Hz, 3H, -CO2CH2CH3), 1.67
(dd, J=1.4, 4.6 Hz, 3H, H-6), 2.65 (dd, J=7.5, 2.2
Hz, 2H, H-2), 3.76–3.84 (m, 1H, H-3), 4.11 (q, J=7.2
Hz, 2H), 5.49–5.55 (m, 2H, H-4, H-5), 7.12–7.38 (m,
4H, Ar-H). 13C NMR l: 14.2, 17.9, 40.8, 44.3, 60.4,
126.0, 127.3, 128.6, 128.8, 132.7, 141.8, 171.7.
The above ester 24 was treated with ozone in dry
MeOH (10 mL) at −78°C for 50 s. Excess ozone was
then removed and the reaction mixture was warmed
to 0°C. Sodium borohydride (1.0 mmol) was added
and the mixture was allowed to warm slowly to rt (3
h). Glacial AcOH (10 drops) was then added to
destroy excess NaBH4, and the mixture was extracted
with CH2Cl2, the combined organic phases washed
with saturated NaCl solution, and then dried. The
filtered solution was evaporated and the crude oil was
purified by column chromatography (pet. ether–ether,
3:1 v/v) to furnish the primary alcohol (33 mg, 72%).
[h]2D0=−15.6 (c 1.6, CH2Cl2). IR wmax (neat): 3458, 1731
cm−1. 1H NMR l: 1.16 (t, J=7.2 Hz, 3H,
-CO2CH2CH3), 2.27 (s, 1H, OH), 2.57 (dd, J=15.7,
7.9 Hz, 1H, H-2), 2.79 (dd, J=15.6, 6.9 Hz, 1H,
H-2), 3.26–3.34 (m, 1H, H-3), 3.71 (dd, 2H, J=6.3,
3.9 Hz, H-4), 4.05 (q, 2H, J=7.2 Hz, -CO2CH2CH3),
7.13–7.30 (m, 4H, Ar-H). 13C NMR l: 14.0, 37.1,
43.8, 60.6, 66.4, 128.7, 129.1, 132.7, 139.5, 172.3.
1
12.7 min. IR wmax (neat): 3342, 3026, 1493, 1042. H
NMR l: 0.86 (t, 3H, J=7.1 Hz, CH3), 1.10–1.27 (m,
2H, H-5), 1.49–1.99 (m, 5H, OH, H-2, H-4), 2.63–
2.78 (m, 1H, H-3), 3.44–3.52 (m, 2H, HOCH2-), 7.16–
7.34 (m, 4H, Ar-H). 13C NMR l: 14.1, 20.6, 39.2,
39.6, 42.2, 61.2, 126.1, 127.6, 128.4, 145.2.
Acknowledgements
We thank the Natural Sciences and Engineering
Research Council (NSERC), Canada and the Univer-
sity of Regina for support. We thank Mr. Ken
Thoms, Department of Chemistry, University of
Saskatchewan for performing the elemental and high
resolution mass spectral analyses.

A. G. H. Wee et al. / Tetrahedron: Asymmetry 14 (2003) 897–909
909
References
Ann. Chem. 1995, 593; (b) Mulzer, J.; Klaus-Deiter, G.;
Burkhard, K. Liebigs Ann. Chem. 1988, 891.
14. Resolution of some of the overlapping signals in the H
1
1. Mikami, K.; Nakai, T. Synthesis 1991, 594.
2. (a) Lakeev, S. N.; Maydanova, I. O.; Galin, F. Z.;
Tolsikov, G. A. Russ. Chem. Rev. 2001, 70, 655; (b) Li,
A. H.; Dai, L.-X.; Aggarval, V. K. Chem. Rev. 1997, 97,
2341; (c) Padwa, A.; Hornbuckle, S. F. Chem. Rev. 1991,
91, 263; (d) Ando, W. Acc. Chem. Res. 1977, 10, 179.
3. (a) Masamune, S.; Chow, W.; Petersen, J. S.; Sita, L. R.
Angew. Chem., Int. Ed. 1985, 24, 1; (b) Bartlett, P. A.
Tetrahedron 1980, 36, 3.
4. (a) Meyer, O.; Cagle, C. P.; Weickhardt, K.; Dominique,
V.; Gladyz, J. A. Pure Appl. Chem. 1996, 68, 79; (b)
Kurth, M. J.; Tahir, S. H.; Olmstead, M. M. J. Org.
Chem. 1990, 55, 2286; (c) Yoshimoto, M.; Ishihara, S.;
Nakayama, E.; Soma, N. Tetrahedron Lett. 1972, 2923;
(d) Trost, B. M.; Hammen, R. F. J. Am. Chem. Soc.
1973, 95, 962; (e) Garipati, R. S.; Cordova, R.; Parvez,
M.; Weinreb, S. M. Tetrahedron 1986, 42, 2979; (f)
Hartley, R. C.; Richards, I. C.; Warren, S. J. Chem. Soc.,
Perkin Trans. 1 1995, 359.
5. (a) Nishibayashi, Y.; Ohe, K.; Uemura, S. Chem. Com-
mun. 1995, 1245; (b) Fukuda, T.; Irie, R.; Katsuki, T.
Tetrahedron 1999, 55, 649; (c) Aggarwal, V. K.; Ferrara,
M.; Hainz, R.; Spey, S. E. Tetrahedron Lett. 1999, 40,
8923; (d) McMillen, D. W.; Varga, N.; Reed, A. B.; King,
C. J. Org. Chem. 2000, 65, 2532; (e) Kitagaki, S.;
Yamamoto, Y.; Okubo, H.; Nakajima, M.; Hashimoto,
S.-I. Heterocycles 2001, 54, 623; (f) Zhang, X.; Qu, X.;
Ma, Z.; Shi, W.; Jin, X.; Wang, J. J. Org. Chem. 2002,
67, 5621.
NMR was achieved using aromatic solvent induced shift
(ASIS) technique; see (a) Dean, F. M.; Varma, R. S.
Tetrahedron 1982, 38, 2069; (b) Engler, E. M.; Lazlo, P.
J. Am. Chem. Soc. 1971, 93, 1317.
15. Johnson, W. S.; Brockson, T. J.; Loew, P.; Rich, D. H.;
Werthemann, R. A.; Arnold, R. A.; Li, T.; Faulkner, D.
J. J. Am. Chem. Soc. 1970, 92, 4463.
16. Suzuki, T.; Sato, E.; Kamada, S.; Tada, H.; Unno, K.;
Kametani, T. J. Chem. Soc., Perkin Trans. 1 1986, 387.
17. (a) Corey, E. J.; Bakshi, R.; Shibata, S.; Chem, C. P.;
Shing, V. K. J. Am. Chem. Soc. 1987, 109, 7925. Reviews:
(b) Walbaum, S.; Martens, J. Tetrahedron: Asymmetry
1992, 3, 1475; (c) Singh, V. K. Synthesis 1992, 605.
18. Cubero, I. I.; Lopez-Espinosa, M. T. P. Carbohydr. Res.
1986, 148, 209.
19. (a) Fuganti, C.; Grasselli, P.; Spreafico, R.; Zirotti, C. J.
Org. Chem. 1984, 49, 543; (b) Brenna, E.; Caraccia, N.;
Fuganti, C.; Fuganti, D.; Grasselli, P. Tetrahedron:
Asymmetry 1997, 8, 3801.
20. Molecular modeling performed using PC Spartan Pro v.
1.0. 5 ((MMFF v. 6. 0. 6) calculations involve conforma-
tional optimization on 18a,b followed by conformer dis-
tribution calculations.
21. Houk, K. N.; Wu, Y. D. J. Org. Chem. 1991, 56, 5657.
22. (a) Johnson, F. Chem. Rev. 1968, 68, 375; (b) Hoffman,
R. W. Chem. Rev. 1989, 89, 1841.
23. For a lead reference see: Doyle, M. P.; Hu, W. Chirality
2002, 14, 169 and references cited therein.
24. E.g., see: (a) Vedejs, E.; Mackay, J. A. Org. Lett. 2001, 3,
535; (b) Choi, Y. K.; Suh, J. H.; Lee, D.; Lim, I. T.; Jung,
J. Y.; Kim, M. J. J. Org. Chem. 1999, 64, 8423.
25. Prepared in 98% yield (1H NMR >99% E) by Wittig
6. Wee, A. G. H.; McLeod, D. D. Heterocycles 2000, 53,
637.
7. Mengel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191.
8. Takano, S.; Tomita, S.-i.; Takahashi, M.; Ogasawara, K.
Chem. Lett. 1987, 1569.
reaction
of
4-chlorobenzaldehyde
with
Ph3PꢁCHC(ꢁO)Me in the presence of 10 mol % of ben-
zoic acid, cf. Le Boulaire, V.; Gree, R. Chem. Commun.
2000, 2195.
9. (a) Jurczak, J.; Pikul, S.; Bauer, T. Tetrahedron 1986, 42,
447; (b) McGarvey, G. J.; Kimura, M.; Oh, T.; Williams,
J. M. J. Carbohydr. Chem. 1984, 3, 125.
26. Attempted chiral HPLC analysis (98:2 hexane–2-
propanol, 1.0 mL/min, u=254 nm) of 22 using Chiralpak
AD, Chiralcel OB and OD columns were unsuccessful.
27. We also prepared (R)-4-phenyl-2-butyrolactone in 59%
ee, via the reaction (S)-2-(4-methoxyphenylthio)-4-
phenyl-3-butene with EDA, catalyzed by Rh2(OAc)4.
10. Campaigne, J.; Tsurugi, J.; Meyer, W. W. J. Org. Chem.
1961, 26, 2486.
11. Anciaux, A. J.; Demonceau, A.; Noels, A. F.; Hubert, A.
J.; Warin, R.; Teyssie, P. J. Org. Chem. 1981, 46, 873.
12. Neumann, W. P. Synthesis 1987, 665.
13. (a) Mulzer, J.; Klaus-Deiter, G.; Shanyoor, M. Liebigs