Asymmetrie [2,3]-wittig rearrangement of oxygenated allyI benzyl ethers in the presence of a chiral di-tBu-bis(oxazoline) ligand: A novel synthetic approach to THF lignans

Article abstract of DOI:10.1002/chem.200901212

We have accomplished highly enantioselective [2,3]-Wittig rearrangements of functionalized allyl benzyl ethers in the presence of a chiral ditBu-bis(oxazoline) ligand. In various oxygenated benzylic ethers, the reactions proceeded with excellent diastereo

Full text of DOI:10.1002/chem.200901212

FULL PAPER
DOI: 10.1002/chem.200901212
Asymmetric [2,3]-Wittig Rearrangement of Oxygenated Allyl Benzyl Ethers
in the Presence of a Chiral di-tBu-bis(oxazoline) Ligand: A Novel Synthetic
Approach to THF Lignans
Maria Kitamura, Yoshimi Hirokawa, and Naoyoshi Maezaki*[a]
Abstract: We have accomplished highly
enantioselective [2,3]-Wittig rearrange-
ments of functionalized allyl benzyl
ethers in the presence of a chiral di-
tBu-bis(oxazoline) ligand. In various
oxygenated benzylic ethers, the reac-
tions proceeded with excellent diaster-
eo- and enantioselectivities, although
the presence of a methoxy substituent
at the ortho-position on the benzyl
group drastically decreased the enan-
tioselectivity. Conversely, o-ethyl and
o-phenyl substituents had no significant
effect on the selectivity. We found that
the C2-substituent of the allylic moiety
played an important role in producing
high enantioselectivity. Highly enantio-
selective [2,3]-Wittig rearrangement in
the presence of catalytic amounts of
the chiral ligands is also described.
Keywords: Wittig rearrangement ·
asymmetric synthesis
·
external
chiral ligands · lignans · stereoselec-
tivity
Introduction
ties have remained moderate to low, in contrast to the high
enantioselectivities seen with alkenyl propargyl ethers.[4,5] To
obtain an enantioselectivity greater than 85% ee, conversion
of the aryl group into a tricarbonylchromium complex was
required.[6]
Lignans, secondary metabolites arising from phenylpropa-
noid biosynthesis, are natural products widely distributed in
vascular plants. Their interesting biological activities—cyto-
toxic, antiviral, antifungal, immunosuppressive, anti-asth-
matic, and antioxidant—are the main reason they have at-
tracted so much interest.[1] Lignans are also fascinating syn-
thetic targets from the point of view of natural products syn-
thesis, because they possess a variety of frameworks with di-
verse stereochemistry and with many substitution patterns
on aryl groups.
In the course of our synthetic studies on lignans,[2] we en-
visioned chiral benzylic alcohol derivatives such as
2
(Scheme 1) as key intermediates for various lignans. These
alcohols can be synthesized by asymmetric [2,3]-Wittig rear-
rangement[3] of oxygenated benzylic ethers 1 in the presence
of an external chiral ligand. Asymmetric [2,3]-Wittig rear-
rangement is an intriguing method by which to provide
chiral building blocks for synthesis of natural products. For
benzylic ether-type substrates, however, the enantioselectivi-
Scheme 1.
We were interested in Hondaꢀs work relating to the syn-
theses of kallolide A and pinnatin A, in which highly enan-
tioselective [2,3]-Wittig reactions of cyclic furfuryl ethers
were accomplished in the presence of (S,S)-2,2’-(dimethyl-
methylene)- and (S,S)-2,2’-(diethylmethylene)bis(4-tert-
butyl-2-oxazoline) as chiral ligands (91 and 93% ee, respec-
tively), albeit in low yields (16 and 19%, respectively).[7] In
spite of the high selectivity, application of the di-tBu-bis(oxa-
[a] Prof. M. Kitamura, Prof. Dr. Y. Hirokawa, Prof. Dr. N. Maezaki
Faculty of Pharmacy, Osaka Ohtani University
3-11-1 Nishikiori-Kita, Tondabayashi
Osaka 584-8540 (Japan)
Fax : (+81)721-24-9541
E-mail: maezan@osaka-ohtani.ac.jp
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200901212.
Chem. Eur. J. 2009, 15, 9911 – 9917
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9911

zoline) ligands to asymmetric [2,3]-Wittig rearrangement
was limited to this cyclic substrate.
We next examined the [2,3]-Wittig reactions of the ben-
zylic ethers (Scheme 3 and Table 2). Treatment of the unsa-
turated benzyl ether 5a with tBuLi in THF exclusively af-
forded the [2,3]-Wittig-rearranged product 7a (syn isomer)
in 86% yield (entry 1). In the case of 5j the stereoselectivity
was reversed, with the anti isomer 7j being obtained as the
main product (50%) along with 7a (27%).
We therefore examined the asymmetric [2,3]-Wittig rear-
rangement of various oxygenated benzylic ethers in the
presence of di-tBu-bis(oxazoline) ligands. In a previous com-
munication[8] we reported the highly enantioselective [2,3]-
Wittig rearrangement of benzylic ethers and also the effect
of the structure of the benzyl moiety on selectivity
(Scheme 1).
Here we disclose the scope and limitations of the diaster-
eo- and enantioselective asymmetric [2,3]-Wittig rearrange-
ment of oxygenated benzylic ethers in the presence of exter-
nal chiral ligands.
Results and Discussion
ACHTUNGTRENNUNG[2,3]-Wittig rearrangement in the absence of chiral ligand:
Scheme 3. ACTHNUTRGNE[UNG 2,3]-Wittig rearrangements of benzylic ethers 5a–j.
The benzylic ethers 5a–i for [2,3]-Wittig rearrangement
were synthesized in 63–85% yields by tBuOK-promoted nu-
cleophilic substitution of various benzylic alcohols on the al-
lylic bromide 4, which was prepared from the known unsatu-
rated ester 3[9] by DIBAL reduction of the ester followed by
triisopropylsilyl (TIPS) protection (Scheme 2 and Table 1).
The benzylic ether 5j (the Z isomer of 5a) was synthesized
from the known allylic alcohol 6[10] by benzylation (NaH,
BnBr, nBu4NI, hexane/THF) in 68% yield.
Table 2.
Entry Ar
[2,3]-Wittig rearrangements of 5a–i.[a]
ACHTUNGTRENNUNG
A
Yield[b]
[%]
[1,2] Yield[b]
ACHTUNGTRENNUNG
(syn)
[%]
1
2
3
4
5
6
7
8
9
phenyl
7a
7b
7c
7d
86
44
64
68
43
93
72
58
79
8a
8b
8c
8d
8e
8 f
8g
8h
8i
0
17
0
0
6
0
0
0
0
4-methoxyphenyl
3-methoxyphenyl
2-methoxyphenyl
3,4-dimethoxyphenyl 7e
3,5-dimethoxyphenyl 7 f
2,3-dimethoxyphenyl 7g
2,5-dimethoxyphenyl 7h
3,4,5-trimethoxyphen- 7i
yl
[a] Reactions were carried out with benzylic ether (1 equiv) and tBuLi
(10 equiv) in THF at À788C for 2 h. [b] Isolated yield.
Substitution on the benzene ring also affected both the re-
activity and the selectivity. The 4-methoxybenzyl derivative
5b afforded the required [2,3]-Wittig product 7b in 44%
yield, but a 17% yield of the [1,2]-Wittig product 8b was
also produced (entry 2). A two-dimensional 1H/13C-hetero-
nuclear multiple bond correlation (HMBC) experiment re-
vealed that the [1,2]-Wittig rearrangement product 8b had
resulted from abstraction of the allylic proton rather than
the benzylic proton. The regioselectivity of the deprotona-
tion was improved when the 3-methoxybenzyl derivative 5c
and the 2-methoxybenzyl derivative 5d were used, giving 7c
and 7d in 64 and 68% yield, respectively, without formation
of the [1,2]-Wittig rearrangement products (entries 3 and 4).
The same reaction conditions were applied to the highly
oxygenated benzylic ethers 5e–i, with diverse substitution
patterns, some of these aryl groups having been found in
natural lignans. In these compounds, only the 3,4-dimethox-
ybenzyl derivative 5e afforded the [1,2]-Wittig rearrange-
ment product 8e, but the proportion of the byproduct was
smaller than that obtained from the 4-methoxybenzyl deriv-
ative 5b.
Scheme 2. Preparation of benzylic ethers 5a–j.
Table 1. Synthesis of benzylic ethers 5a–i.[a]
Entry
Ar
Benzylic ether
Yield [%][b]
1
2
3
4
5
6
7
8
9
phenyl
5a
5b
5c
5d
5e
5 f
5g
5h
5i
75
63
80
85
85
74
85
85
77
4-methoxyphenyl
3-methoxyphenyl
2-methoxyphenyl
3,4-dimethoxyphenyl
3,5-dimethoxyphenyl
2,3-dimethoxyphenyl
2,5-dimethoxyphenyl
3,4,5-trimethoxyphenyl
[a] Reactions were carried out with bromide 4 (1 equiv), benzylic alcohol
(2 equiv), and tBuOK (2.2 equiv) in THF at RT for 16 h. [b] Isolated
yield.
9912
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9911 – 9917

Asymmetric [2,3]-Wittig Rearrangement
FULL PAPER
The assignment of relative stereochemistry was conducted
ligand. Initially, we examined the solvent effect on the asym-
metric [2,3]-Wittig rearrangement of the benzylic ether 5b
in the presence of the absence of the bis(oxazoline) 11a
(Figure 1 and Table 3). The enantiomeric excesses were
strongly affected by the solvent. In THF the reaction pro-
ceeded even in the absence of the chiral ligand, but the re-
action in the presence of the chiral ligand 11a afforded a
racemic [2,3]-Wittig rearranged product 7b (entries 1 and
4). In sharp contrast, in a nonpolar solvent such as hexane
or toluene (entries 2 and 3) the [2,3]-Wittig rearrangement
did not take place in the absence of the chiral ligand. In-
stead, considerable amounts of the tert-butyl adduct 12b—
the result of an SN2ꢀ-type reaction—were produced both in
toluene and in hexane, in 85 and 68% yield, respectively. In
the presence of the chiral ligand 11a, however, the desired
reaction took place and the selectivity was dramatically in-
creased to 96–98% ee (entries 5 and 6). These results indi-
cate that in the nonpolar solvent tBuLi cannot deprotonate
the benzylic proton without forming the tBuLi/ligand com-
plex, and so the product is obtained with high optical purity.
by conversion of product 7d into the known g-lactone syn-
1
9[11] in two steps (Scheme 4). The H and 13C NMR spectral
data were identical to the published data. The anti isomer of
7d was selectively synthesized from 7d through oxidation
followed by diastereoselective reduction with l-Selectride.
Its stereochemistry was also confirmed by conversion into
the known g-lactone anti-9[11] in a manner similar to that de-
scribed for syn-9. The relative stereochemistries of other
[2,3]-Wittig rearrangement products were confirmed by
comparison of the coupling constants between the C1- and
1
C2-methine protons in their H NMR data with those for 7d
and its anti isomer. That is, the coupling constants for prod-
ucts 7a–i ranged from 3.4 to 4.4 Hz and are close to the J=
4.0 Hz seen for 7d (syn). On the other hand, the anti isomer
7j shows a large J value (8.8 Hz) and is well matched to the
anti isomer of 7d (J=8.3 Hz). A similar empirical rule was
reported by Nakai.[12]
Figure 1. External chiral ligands (ECLs).
Scheme 4.
Table 3. Solvent effect on asymmetric [2,3]-Wittig rearrangement.[a]
The effect of olefin geometry on diastereoselectivity was
consistent with the general rule. That is, E and Z isomers ex-
hibit syn and anti selectivity, respectively. The selectivity in
the case of the Z isomer 5j is moderate because of gauche
interaction between the phenyl group and the methyl group
on the olefin in the transition state leading to the anti
isomer 7j.[3d] The effect of the aryl substituent on the regio-
selectivity of the deprotonation (the selectivity with regard
to [2,3]- and [1,2]-Wittig rearrangement) was interpreted as
follows. The decreases in regioselectivity seen with the ben-
zylic ethers 5b and 5e were presumably caused by the weak
acidity at their benzylic positions resulting from the elec-
tron-donating resonance effects of their para-methoxy
groups. The corresponding meta-isomers 5c and 5 f exhibited
high regioselectivity. On the other hand, the high regioselec-
tivity observed in the substrates 5d, 5g, and 5h, each con-
taining a 2-methoxy substituent, were presumably due to the
lateral metallation effect through coordination to the coun-
ter-cation of the base.
Entry
ECL
Solvent
Yield of 7b[b] [%]
ee[c] [%]
1
2
3
4
5
6
none
none
none
11a
11a
11a
THF
44
0[d]
0[e]
35
34
34
–
–
–
0
96
98
toluene
hexane
THF
toluene
hexane
[a] Reactions were carried out with benzylic ether 5b (1 equiv), tBuLi
(10 equiv), and bis(oxazoline) 11a (5 equiv) in solvent at À788C for 2 h.
[b] Isolated yield. [c] Determined by chiral HPLC analysis. [d] An 85%
yield of 12b was produced. [e] A 68% yield of 12b was produced and a
21% yield of 5b was recovered.
We also examined the effect of other external chiral li-
gands (ECLs) with 5a (Figure 1 and Table 4). The widely
utilized (À)-sparteine (10) afforded the adduct arising from
the SN2ꢀ reaction of tBuLi, producing 12a[13] in 56% yield
(entry 1). The bis(oxazoline) 11b afforded a yield and an
enantioselectivity comparable to those obtained with bis(ox-
azoline) 11a (entries 2–3), but the product 7a contained an
Asymmetric [2,3]-Wittig rearrangement in the presence of a
chiral ligand: We next turned our attention to the asymmet-
ric [2,3]-Wittig rearrangement in the presence of a chiral
Chem. Eur. J. 2009, 15, 9911 – 9917
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9913

N. Maezaki et al.
Table 5. Asymmetric [2,3]-Wittig rearrangements of benzylic ethers
inseparable impurity, presumably originating from the chiral
ligand. The bis(oxazoline) 11c seems to decompose under
the reaction conditions, and the SN2ꢀ product 12a was ob-
tained in 51% yield. The chiral ligand 11d, the diethyl-
methylene equivalent of 11a, provided the product 7a in
73% yield, but the enantioselectivity was significantly re-
duced to 46% ee. The SN2ꢀ reaction of tBuLi also predomi-
nated when the 5,5-dimethyloxazoline-type chiral ligand 11e
was employed. The bis(oxazoline) 11a was therefore the
best external chiral ligand out of those we tested.
5a–i.[a]
Entry
Ar
Product
Yield[b] [%]
ee[c] [%]
1
phenyl
7a
7b
7c
7d
7e
7 f
7g
7h
7i
65
34
75
73
32
72
58
58
76
98
98
95
8
87
85[e]
17
22
85
2[d]
3
4-methoxyphenyl
3-methoxyphenyl
2-methoxyphenyl
3,4-dimethoxyphenyl
3,5-dimethoxyphenyl
2,3-dimethoxyphenyl
2,5-dimethoxyphenyl
3,4,5-trimethoxyphenyl
4
5
6
7
8
9
Table 4. Effect of chiral ligands on the asymmetric [2,3]-Wittig rearran-
gement.[a]
[a] Reactions were carried out with the benzylic ether (1 equiv), tBuLi
(10 equiv), and the bis(oxazoline) 11a (5 equiv) in hexane at À788C for
2 h. [b] Isolated yield. [c] Determined by chiral HPLC analysis. [d] A
trace of [1,2]-Wittig rearrangement product was produced as a by-prod-
uct. [e] The ee value reported in our communication (ref. [8]) was cor-
rected as a result of the retest.
Entry
External chiral ligand
Yield of 7a[b] [%]
ee[c] [%]
1
2
3
4
5
6
(À)-sparteine (10)
0[e]
71
–
11a
11b
11c
11d
11e
98
99
–
46
–
75[d]
0[f]
73
ported[18] [a]2D3 =+57.1 (c=0.2, CHCl3)}. Because the g-lac-
tone 13 is a synthetic intermediate of 3-epi-eupomatilone 6
(cf. Scheme 1), we have achieved a formal synthesis of this
compound.[18]
0[g]
[a] Reactions were carried out with the benzylic ether 5a (1 equiv),
tBuLi (10 equiv), and the external chiral ligand (1 equiv) in hexane at
À788C for 2 h. [b] Isolated yield. [c] Determined by chiral HPLC analy-
sis. [d] Containing a small amount of by-product. [e] A 56% yield of 12a
was produced. [f] A 51% yield of 12a was produced. [g] A 73% yield of
12a was produced.
With the tBuLi/11a complex,[14] asymmetric [2,3]-Wittig
rearrangements were conducted on the various oxygenated
benzylic ethers 5a–i in hexane. The results are summarized
in Table 5. High enantioselectivities were observed with the
unsubstituted benzyl ether 5a, the 4-methoxybenzyl ether
5b, and the 3-methoxybenzyl ether 5c (entries 1–3). Inter-
estingly, the enantioselectivity was greatly decreased in the
case of the 2-methoxybenzyl ether 5d (entry 4). A similar
tendency was observed with the other ethers 5e–i. Whereas
the 3,4- and 3,5-dimethoxybenzyl ethers 5e and 5 f showed
comparatively high enantioselectivities (entries 5 and 6), the
2,3- and 2,5-dimethoxybenzyl ethers 5g and 5h, each con-
taining an ortho-methoxy substituent, afforded products
with low enantiomeric excesses (entries 7 and 8). The 3,4,5-
trimethoxybenzyl ether 5i, bearing no methoxy group at its
ortho position, afforded a product with high enantiomeric
excess (entry 9).
The absolute configurations of the products were assigned
as (1R,2S) by a modified Mosher method[15,16] after conver-
sion of the secondary alcohols 7b–d into the corresponding
(+)- and (À)-a-methoxy-a-(trifluoromethyl)phenylacetic
acid (MTPA) esters.[17] The absolute configurations of the
other products were assigned analogously. The assignment
was also confirmed by comparison of the specific rotation of
g-lactone 13 (Scheme 5) derived from 7i (85% ee) with the
reported value {[a]2D1 =+67.3 (c=0.34, CHCl3) versus the re-
Scheme 5.
The effect of the 2-methoxy group on enantioselectivity
cannot be explained clearly at present, but oxygen appears
to play an important role. To confirm the role of the 2-me-
thoxy group, we synthesized the 2-ethylbenzyl and 2-phenyl-
benzyl ethers 5k and 5l in 63 and 62% yield, respectively,
by the procedure shown in Scheme 2.
The asymmetric [2,3]-Wittig reaction of 5k afforded
(1R,2S)-7k in 69% yield and with excellent enantioselectivi-
ty (>99% ee, Scheme 6), in contrast to that of the 2-me-
thoxybenzyl ether 5d (8% ee).[19] Even when the o-substitu-
ent was a large phenyl group (substrate 5l), the enantiose-
lectivity decreased to 92% ee, but was still far higher than
that seen in the case of the methoxy substituent. We pre-
sume that this decline in the enantioselectivity was the
result of coordination of the methoxy group to the lithium
cation, rather than steric effects.
In contrast with the high diastereo- and enantioselectivi-
ties seen with the E isomers, an attempt with the Z isomer
5j was fruitless. Not only were the yields low, but the dia-
stereo- and enantioselectivities were also poor, with
9914
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9911 – 9917

Asymmetric [2,3]-Wittig Rearrangement
FULL PAPER
ethers—possess no C2-substituents on their allyl ether moi-
eties, and these substrates exhibited moderate to low enan-
tioselectivities.[5] We suspect that the presence of a large C2-
substituent on the allylic moiety plays some important role
in producing high enantioselectivity.
AHCTUNGTREG[NNNU 2,3]-Wittig rearrangements in the presence of catalytic
amounts of chiral ligand are also of much interest. Kimatch
reported that enantioselectivity was partially retained when
(À)-sparteine (0.2 equiv) was used (2.2 equiv, 60% ee;
0.2 equiv, 48% ee).[5b] Similar results were reported by
Nakai et al.,[4a] who found that (S,S)-2,2’-(diethylmethylene)-
bis(4-isopropyl-2-oxazoline) (0.2 equiv) showed the same
enantioselectivity as when 2.0 equiv of the chiral ligand was
used in ether (34% ee). We were interested in exploring
whether or not the selectivity at higher levels (>95% ee)
would be maintained, and the results are summarized in
Table 7. We found that the yields were reduced with de-
creasing quantities of chiral ligand, but the selectivity re-
mained high at 98% ee (entries 1–5). The yields were slight-
ly improved by decreasing the amount of base (entries 5–7),
as a result of a reduction in by-products including the SN2ꢀ
product 12a. To the best of our knowledge, this is the first
example of excellent enantioselectivity being achieved in
the presence of a catalytic amount of chiral ligand in the
[2,3]-Wittig rearrangement of benzylic ethers.
Scheme 6.
(1R,2S)-7a and the anti isomer 7j being obtained in 9%
yield (10% ee) and 11% yield (39% ee), respectively
(Scheme 7).[20]
Scheme 7.
We next examined [2,3]-Wittig rearrangements of sub-
strates with different C2-substituents (Table 6). Compound
5m, containing a TBS ether, gave a low yield, but a compa-
ratively high enantioselectivity. For compound 5n, with a
Table 7. Catalytic loading of the bis(oxazoline) 11a for the rearrange-
ment of 5a.
Table 6. Effect of the allylic ether C2-substituent on the asymmetric
[2,3]-Wittig reaction.[a]
Entry tBuLi
Bis(oxazoline) 11a
(equiv)
Yield of 7a
ee
(equiv)
[%][b]
[%][c]
1
2
3
4
5
6
7
10
10
10
10
10
6
5
1
65
71
52
45
23
37
40
98
98
98
98
98
96
96
0.5
0.2
0.1
0.1
0.1
Entry Substrate
R
tBuLi
(equiv)
Product Yield
[%][b]
ee
[%][c]
1
2
3
4
5
6
5a
5a
5m
5n
5o
5p
CH2OTIPS 10
7a
7a
7m
7n
7o
7p
71
64
18
11
76
63
98
98
91
29
63
38
3
CH2OTIPS
CH2OTBS
CH2OMe
Me
6
6
6
6
[a] Reactions were carried out with the benzylic ether 5a (1 equiv),
tBuLi, and 11a in hexane at À788C for 2 h. [b] Isolated yield. [c] Deter-
mined by chiral HPLC analysis.
H
10
[a] Reactions were carried out with the benzylic ether (1 equiv), tBuLi,
and 11a (1 equiv) in hexane at À788C over 2 h. [b] Isolated yield. [c] De-
termined by chiral HPLC analysis.
Conclusions
We have found that the [2,3]-Wittig rearrangement of vari-
ous benzylic ethers proceeds with very high diastereo- and
enantioselectivities. The choice of chiral ligand and substitu-
ents on the substrates are important for achievement of high
selectivity. Interestingly, high enantioselectivities were re-
tained when the chiral ligand was used in catalytic amounts.
The products may become versatile intermediates for the
synthesis of lignans.
methyl ether, however, both the yield and enantioselectivity
were reduced. Compound 5o, with a methyl substituent, ex-
hibited moderate yield and selectivity. The (Z)-crotyl ether
5p showed enantioselectivity comparative to that described
in the previous report.[4a,b] The allyl benzyl ethers used in
the previous reports—such as allyl, crotyl, and prenyl
Chem. Eur. J. 2009, 15, 9911 – 9917
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9915

N. Maezaki et al.
and solvent evaporation. Purification by flash column chromatography
on silica gel with hexane/Et2O 10:1 yielded 7a (38.6 mg, 86%) as a color-
less oil.
Experimental Section
Optical rotations were measured with a JASCO P-1020 digital polarime-
ter. H and 13C NMR spectra were recorded in CDCl3 solution at 400 and
1
General procedure for asymmetric [2,3]-Wittig rearrangement of allyl
benzyl ethers in the presence of a chiral ligand: tBuLi (1.58m in pentane,
0.72 mL, 1.14 mmol) was added with stirring at À788C under Ar to a so-
lution of allyl benzyl ether 5a (39.8 mg, 0.114 mmol) and bis(oxazoline)
11a (33.6 mg, 0.114 mmol) in dry hexane (0.57 mL). The stirring was con-
tinued for 2 h at this temperature. The reaction mixture was quenched
with saturated aqueous NH4Cl and extracted with EtOAc. The combined
extracts were washed with saturated aqueous NH4Cl, water, and brine
prior to drying and solvent evaporation. The crude product was purified
by PTLC (SiO2) with hexane/Et2O 6:1 to give (1R,2S)-7a (28.2 mg, 71%)
as a colorless oil.
100 MHz, respectively, with a JEOL JNM-AL-400 spectrometer. Chemi-
cal shifts of H NMR are expressed in ppm downfield from tetramethylsi-
1
lane as an internal standard (d=0 ppm). Chemical shifts of 13C NMR are
expressed as ppm in CDCl3 as an internal standard (d=77 ppm). The fol-
lowing abbreviations are used: broad=br, singlet=s, doublet=d, trip-
let=t, quartet=q, quintet=qn and multiplet=m. IR absorption spectra
(FT: diffuse reflectance spectroscopy) were recorded in KBr powder with
a JASCO FT-6300 IR spectrophotometer, and only noteworthy absorp-
tions (cmÀ1) are listed. Mass spectra were obtained with a JEOL GC-
Mate II mass spectrometer. Purification of the crude products was carried
out by flash column chromatography or preparative TLC (PTLC). Fuji
Silysia Silica Gel BW-300 was used as an adsorbent for column chroma-
tography. For PTLC, silica gel 60 F254 (Merck) was used. All air- or mois-
ture-sensitive reactions were carried out in flame-dried glassware under
Ar or N2. All organic extracts were dried over anhydrous MgSO4 or
Na2SO4, filtered, and concentrated under reduced pressure with a rotary
evaporator.
AHCTUNGTERG(NNUN 1R,2S)-2-Methyl-1-phenyl-3-(triisopropylsilyloxymethyl)but-3-en-1-ol
(7a): [a]2D5 =À12.2 (c=1.24 in CHCl3); 1H NMR: d=0.99 (d, J=7.1 Hz,
3H), 1.06–1.18 (m, 21H), 2.60 (qd, J=7.1, 4.0 Hz, 1H), 3.05 (d, J=
2.4 Hz, 1H), 4.08 (d, J=12.9 Hz, 1H), 4.22 (d, J=12.9 Hz, 1H), 4.81 (dd,
J=4.0, 2.4 Hz, 1H), 4.98 (brs, 1H), 5.20 (brs, 1H), 7.21–7.36 ppm (m,
5H); 13C NMR: d=12.0 (3C), 12.6, 18.0 (6C), 44.9, 66.0, 76.0, 112.5,
126.2 (2C), 126.9, 127.9 (2C), 143.0, 150.5 ppm; IR (KBr): n˜ =3409, 1650,
1603, 1494 cmÀ1; MS (FAB): m/z: 349 [M+H]+; HRMS (FAB): m/z:
calcd for C21H37O2Si: 349.2563 [M+H]+; found: 349.2562.
(Z)-1-Bromo-2-(triisopropylsilyloxymethyl)but-2-ene (4): DIBAL (1.02m
in n-hexane, 58 mL, 57.0 mmol) was added with stirring at À208C under
Ar to a solution of ester 3 (5.0 g, 25.9 mmol) in CH2Cl2 (50 mL). After
stirring at the temperature for 1.2 h, the reaction mixture was quenched
with saturated aqueous Rochelle salt. The pH was then adjusted to 4
with HCl (2m) and the mixture was extracted with CH2Cl2. The com-
bined extracts were washed with water and brine prior to drying and sol-
vent evaporation.
AHCTUNGERTG(NNUN 4RS,5SR)-5-(2-Methoxyphenyl)-4-methyl-3-methylenedihydrofuran-
2(3H)-one (syn-9): Tetrabutylammonium fluoride (1m solution in THF,
0.21 mL, 0.21 mmol) was added at room temperature under Ar to a solu-
tion of 7d (syn) (25.9 mg, 0.068 mmol) in THF (0.68 mL) and the stirring
was continued for 2 h. The reaction mixture was quenched with saturated
aqueous NH4Cl, and the resulting mixture was extracted with EtOAc.
The combined extracts were washed with saturated aqueous NH4Cl,
water, and brine prior to drying and solvent evaporation. The crude
product was purified by PTLC (SiO2) with n-hexane/EtOAc 1:1 to give
the alcohol (15.2 mg, quant.) as a colorless oil.
The crude residue was used in the next step without purification. 2,6-Lu-
tidine (3.05 g, 28.5 mmol) and TIPSOTf (7.81 g, 25.9 mmol) were added
with stirring at 08C to a solution of the crude residue in CH2Cl2 (50 mL),
and the stirring was continued at room temperature for 15 h. The reac-
tion mixture was quenched with saturated aqueous NH4Cl and extracted
with CH2Cl2. The combined extracts were washed with water and brine
prior to drying and solvent evaporation. The crude residue was purified
by flash column chromatography on silica gel with n-hexane/EtOAc 95:5
Catalytic
TEMPO
(1.1 mg,
0.0068 mmol),
nBu4NI
(2.53 mg,
0.0068 mmol), and N-chlorosuccinimide (27.4 mg, 0.21 mmol) were added
with stirring at room temperature to a mixture of the alcohol (15.2 mg,
0.068 mmol) in CH2Cl2 (0.3 mL), aqueous NaHCO3 (0.5m, 0.15 mL), and
aqueous K2CO3 (0.05m, 0.15 mL). The mixture was vigorously stirred at
the same temperature for 1.5 h. The organic layer was then separated
and the aqueous phase was extracted with CH2Cl2. The combined ex-
tracts were washed with brine prior to drying and solvent evaporation.
The crude product was purified by PTLC (SiO2) with n-hexane/EtOAc
1
to give 4 (6.95 g, 84% in two steps) as a colorless oil. H NMR: d=1.05–
1.18 (m, 21H), 1.73 (dt, J=7.0, 1.3 Hz, 3H), 4.08 (s, 1H), 4.50 (qn, J=
1.3 Hz, 1H), 5.75 ppm (qt, J=7.0, 1.3 Hz, 1H); 13C NMR: d=12.0 (3C),
13.1, 18.0 (6C), 27.4, 65.0, 125.7, 136.3 ppm; IR (KBr): n˜ =1667 cmÀ1; MS
(CI): m/z: 321 [M+H]+; HRMS (CI): m/z: calcd for C14H30BrOSi:
321.1249 [M+H]+; found: 321.1242.
1
3:1 to give syn-9 (7.5 mg, 50%) as a colorless oil. The H NMR spectrum
was identical with that reported.[11]
General procedure for the synthesis of benzylic ethers: tBuOK (0.35 g,
3.08 mmol) was added with stirring at room temperature under Ar to a
solution of 4 (0.45 g, 1.40 mmol) and benzyl alcohol (0.3 g, 2.8 mmol) in
THF (10 mL). After the stirring had been continued for 15 h, the reac-
tion mixture was quenched with saturated aqueous NH4Cl and the mix-
ture was extracted with CH2Cl2. The combined extracts were washed
with water and brine prior to drying and solvent evaporation. The crude
residue was purified by flash column chromatography on silica gel with
n-hexane/EtOAc 9:1 to give 5a (0.36 g, 75%) as a colorless oil.
(E)-1-Benzyloxy-2-(triisopropylsilyloxymethyl)but-2-ene (5a): 1H NMR:
d=1.06–1.15 (m, 21H), 1.69 (brd, J=6.9 Hz, 3H), 4.09 (s, 2H), 4.25 (brs,
2H), 4.47 (s, 2H), 5.80 (brq, J=6.9 Hz, 1H), 7.25–7.34 ppm (m, 5H);
13C NMR: d=12.0 (3C), 13.0, 18.0 (6C), 65.0, 65.5, 71.9, 123.7, 127.5,
127.7 (2C), 128.3 (2C), 135.9, 138.6 ppm; IR (KBr): n˜ =1495 cmÀ1; MS
(FAB): m/z: 349 [M+H]+; HRMS (FAB): m/z: calcd for C21H37O2Si:
349.2563 [M+H]+; found: 349.2556.
AHCTUNGTERG(NNUN 1RS,2RS)-1-(2-Methoxyphenyl)-2-methyl-3-(triisopropylsilyloxymethyl)-
but-3-en-1-ol (anti isomer of 7d): A catalytic amount of TPAP (4.8 mg,
0.0136 mmol) was added with stirring at room temperature under Ar to a
mixture of the alcohol 7d (syn) (51.4 mg, 0.136 mmol), 4-methylmorpho-
line N-oxide (79.5 mg, 0.679 mmol), and molecular sieves (4 ꢂ) in
CH2Cl2/MeCN 9:1 (0.70 mL). The stirring was continued for 1.5 h. After
dilution with CH2Cl2, the mixture was filtered through a pad of Celite
and the filtrate was concentrated under reduced pressure. The crude resi-
due was purified by flash column chromatography on silica gel with n-
hexane/EtOAc 8:1 to give the intermediate ketone (50.0 mg, 98%) as a
colorless oil. 1H NMR: d=0.97–1.11 (m, 21H), 1.33 (d, J=6.8 Hz, 3H),
3.86 (s, 3H), 4.15 (q, J=6.8 Hz, 1H), 4.16 (brs, 2H), 4.93 (s, 1H), 5.22
(brs, 1H), 6.91 (d, J=8.3 Hz, 1H), 6.95 (td, J=7.6, 1.0 Hz, 1H), 7.40
(ddd, J=8.7, 7.0, 1.4 Hz, 1H), 7.50 ppm (dd, J=7.6, 1.7 Hz, 1H);
13C NMR: d=12.0 (3C), 16.2, 18.0 (6C), 48.3, 55.4, 65.2, 110.5, 111.3,
120.6, 129.0, 130.4, 132.6, 148.0, 157.6, 203.8 ppm; IR (KBr): n˜ =1679,
1598, 1486 cmÀ1; MS (FAB): m/z: 377 [M+H]+; HRMS (FAB): m/z:
calcd for C22H37O3Si: 377.2512 [M+H]+; found: 377.2505.
General procedure for [2,3]-Wittig rearrangement of allyl benzyl ethers
in the absence of chiral ligand: tBuLi (1.58m in pentane, 0.82 mL,
1.29 mmol) was added dropwise with stirring at À788C under Ar to a so-
lution of allyl benzyl ether 5a (45.0 mg, 0.129 mmol) in dry THF
(0.65 mL). The stirring was continued for 2 h at this temperature. The re-
action mixture was quenched with saturated aqueous NH4Cl, and the re-
sulting mixture was extracted with EtOAc. The combined extracts were
washed with saturated aqueous NH4Cl, water, and brine prior to drying
l-Selectride (1.0m in THF, 0.144 mL, 0.144 mmol) was added at À788C
with stirring under Ar to a solution of the ketone (45.1 mg, 0.120 mmol)
in THF (1.2 mL). After 2.5 h, H2O2 (30% solution, 0.5 mL) and acetone
(0.25 mL) were added to the mixture, which was then allowed to warm to
room temperature. After 1 h, the mixture was concentrated under re-
duced pressure and the residue was extracted with Et2O. The combined
9916
www.chemeurj.org
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 9911 – 9917

Asymmetric [2,3]-Wittig Rearrangement
FULL PAPER
extracts were washed with brine prior to drying and solvent evaporation.
The crude residue was purified by flash column chromatography on silica
gel with n-hexane/EtOAc 8:1 to give the anti isomer of 7d (41.4 mg,
91%) as a colorless oil. 1H NMR: d=0.95 (d, J=7.1 Hz, 3H), 1.08–1.20
(m, 21H), 2.66 (qn, J=7.6 Hz, 1H), 3.21 (d, J=5.6 Hz, 1H), 3.83 (s, 3H),
4.12 (d, J=13.1 Hz, 1H), 4.24 (d, J=13.1 Hz, 1H), 4.96 (dd, J=8.3,
5.6 Hz, 1H), 5.01 (brs, 1H), 5.24 (brs, 1H), 6.86 (d, J=8.3 Hz, 1H), 6.95
(td, J=7.4, 0.7 Hz, 1H), 7.22 (td, J=7.8, 1.7 Hz, 1H), 7.35 ppm (dd, J=
7.6, 1.7 Hz, 1H); 13C NMR: d=12.0 (3C), 17.3, 18.0 (6C), 45.2, 55.3,
65.7, 72.2, 110.4, 112.4, 120.7, 127.6, 128.1, 131.6, 150.5, 156.8 ppm; IR
prehensive Organic Synthesis (Eds.: B. M. Trost, I. Fleming), Perga-
mon, New York, 1991, Vol. 3, pp. 975–1014.
[4] For propargyl ether-type substrates, see: a) K. Tomooka, N. Komine,
T. Nakai, Chirality 2000, 12, 505–509; b) K. Tomooka, N. Komine,
T. Nakai, Tetrahedron Lett. 1998, 39, 5513–5516; c) S. Manabe,
Chem. Pharm. Bull. 1998, 46, 335–336; d) S. Manabe, Chem.
Commun. 1997, 737–738; e) J. Kang, W. O. Cho, H. G. Cho, H. J.
Oh, Bull. Korean Chem. Soc. 1994, 15, 732–739.
[5] For benzyl ether-type substrates, see: a) I. M. Barrett, S. W. Breeden,
Tetrahedron: Asymmetry 2004, 15, 3015–3017; b) T. Kawasaki, T.
Kimachi, Tetrahedron 1999, 55, 6847–6862; c) T. Kawasaki, T. Kima-
chi, Synlett 1998, 1429–1431 and refs. [4a–d].
[6] S. E. Gibson, P. Ham, G. R. Jefferson, Chem. Commun. 1998, 123–
124.
[7] M. Tsubuki, K. Takahashi, T. Honda, J. Org. Chem. 2003, 68,
10183–10186.
[8] Y. Hirokawa, M. Kitamura, N. Maezaki, Tetrahedron: Asymmetry
2008, 19, 1167–1170.
(KBr): n˜ =3413, 1649, 1602, 1588, 1492 cmÀ1
; MS (FAB): m/z: 361
[M+HÀ18]+; HRMS (FAB): m/z: calcd for C22H37O2Si: 361.2563
[M+HÀ18]+; found: 361.2579.
ACHTUNGTRENNUNG(4RS,5RS)-5-(2-Methoxyphenyl)-4-methyl-3-methylenedihydrofuran-
2(3H)-one (anti-9): The anti isomer of 7d was converted into the g-lac-
tone anti-9 (13.7 mg, 86% in two steps) as a colorless oil in the same
manner as described for the procedure of the compound syn-9. Its
1H NMR spectrum was identical with that reported.[11]
[9] a) M. M. Sꢃ, M. D. Ramos, L. Fernandes, Tetrahedron 2006, 62,
11652–11656; b) L. Fernandes, A. J. Bortoluzzi, M. M. Sꢃ, Tetrahe-
dron 2004, 60, 9983–9989.
[10] J. C. Gilbert, J. Yin, Tetrahedron 2008, 64, 5482–5490.
[11] R. Csuk, C. Schrçder, S. Hutter, K. Mohr, Tetrahedron: Asymmetry
1997, 8, 1411–1429.
[12] K. Mikami, Y. Kimura, N. Kishi, T. Nakai, J. Org. Chem. 1983, 48,
279–281.
[13] The optical purities of 12a and 12b were not determined.
[14] a) K. Ishihara, A. Sakakura, Jpn. Kokai Tokkyo Koho
JP2005320304, 2005; b) D. A. Evans, K. A. Woerpel, M. M. Hinman,
M. M. Faul, J. Am. Chem. Soc. 1991, 113, 726–728.
ACHTUNGTRENNUNG(3R,4S,5R)-3,4-Dimethyl-5-(3,4,5-trimethoxyphenyl)dihydrofuran-2(3H)-
one 13: Compound (1R,2S)-7i was converted into the unsaturated g-lac-
tone (13.7 mg, 66% in two steps) as a colorless oil in the same manner as
described in the procedure for compound syn-9. A solution of the unsatu-
rated g-lactone (13.7 mg, 0.049 mmol) in toluene (0.50 mL) was hydro-
genated in the presence of RhACTHNUGTRNEUGN(PPh3)3Cl (45.5 mg, 0.049 mmol) under hy-
drogen (1 atm) at room temperature. The mixture was stirred for 1 d and
the solvent was evaporated. The residue was purified by PTLC (SiO2)
with n-hexane/EtOAc 2:1 to give 13 (8.0 mg, 58%) as a colorless oil.
[a]2D1 =+67.3 (c=0.34 in CHCl3). Its 1H NMR spectrum was identical
with that reported.[18]
[15] I. Ohtani, K. Kusumi, Y. Kashman, H. Kakisawa, J. Am. Chem. Soc.
1991, 113, 4092–4096.
[16] Isobe and co-workers reported that some a-aromatic secondary al-
cohols show irregular Dd values for the b-protons in the modified
Mosher method, presumably due to the anisotropic effects of the a-
aromatic groups. They also reported that these irregular data can be
neglected for determination of the absolute configuration by the
modified Mosher method. Some of our [2,3]-Wittig products also ex-
hibited the same irregular Dd values for the b-protons and the abso-
lute configurations were assigned according to Isobeꢀs report; see I.
Ohtani, K. Hotta, Y. Ichikawa, M. Isobe, Chem. Lett. 1995, 513–
514.
Acknowledgements
We acknowledge financial support in the form of a Grant-in-Aid for Sci-
entific Research (C) from the Japan Society for the Promotion of Science
(No. 18590097) and Osaka Ohtani University Research Fund (Pharma-
ceutical Sciences). We also appreciate financial support provided in the
form of a Grant-in-Aid for Scientific Research (C) from the Japan Soci-
ety for the Promotion of Science (No. 21590032).
[17] In dilute conditions, the ee of the alcohol 7d was improved to
59% ee but with a reduction in the yield to 19%. The Mosher ester
was prepared from 7d with moderate optical purity.
[18] J. B. Johnson, E. A. Bercot, C. M. Williams, T. Rovis, Angew. Chem.
2007, 119, 4598–4602; Angew. Chem. Int. Ed. 2007, 46, 4514–4518.
[19] The absolute configuration of 7k was determined by the modified
Mosher method.
[1] For reviews of lignans, see: a) L. B. Davin, N. G. Lewis, Phytochem.
Rev. 2003, 2, 257–288; b) T. Umezawa, Phytochem. Rev. 2003, 2,
371–390; c) R. S. Ward, Nat. Prod. Rep. 1999, 16, 75–96; d) R. S.
Ward, Nat. Prod. Rep. 1997, 14, 43–74.
[2] M. Kitamura, H. Hayashi, M. Yano, T. Tanaka, N. Maezaki, Hetero-
cycles 2007, 71, 2669–2680.
[3] a) M. Hiersemann, L. Abraham, A. Pollex, Synlett 2003, 1088–1095;
b) G. McGowan, Aust. J. Chem. 2002, 55, 799; c) T. Nakai, K. To-
mooka, Pure Appl. Chem. 1997, 69, 595–600; d) T. Nakai, K.
Mikami, Org. React. 1994, 46, 105–209; e) J. A. Marshall, in Com-
[20] The absolute configuration of the major enantiomer of 7a was con-
firmed by chiral HPLC; that of 7j was not determined.
Received: May 7, 2009
Published online: August 20, 2009
Chem. Eur. J. 2009, 15, 9911 – 9917
ꢁ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9917