Regioselective SN2 opening of vinylic epoxides with trialkylzincates and trialkylaluminates

Article abstract of DOI:10.1002/ejoc.200300759

The use of trialkylorganozincates and tetraalkylaluminates allows regioselective S_N2 nucleophilic opening of vinylic epoxides. The reaction occurs with an anti-substitution pattern and can be applied to a wide range of substrates. We also show that the solvent and the structure of the epoxide have an influence on the substitution products' ratio. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200300759

FULL PAPER
Regioselective SN2 Opening of Vinylic Epoxides with Trialkylzincates and
Trialkylaluminates
Olivier Equey,[a] Emmanuel Vrancken,[b] and Alexandre Alexakis*[a]
Keywords: Aluminum / Nucleophilic substitution / Regioselectivity / Vinylic epoxides / Zinc
The use of trialkylorganozincates and tetraalkylaluminates
allows regioselective SN2 nucleophilic opening of vinylic ep-
oxides. The reaction occurs with an anti-substitution pattern
and can be applied to a wide range of substrates. We also
show that the solvent and the structure of the epoxide have
an influence on the substitution products’ ratio.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2004)
Introduction
Vinylic epoxides are interesting starting materials that are
commonly used in organic synthesis.[1] The nucleophilic ad-
dition of organometallic reagents to vinylic epoxides pro-
vides 1,2- and 1,4-addition products,[2] with 1,4-addition
predominating with organocopper reagents[3] and copper-
catalysed reactions of organomagnesium[4,5] and organozinc
compounds.[6Ϫ8] Marino et al.[9,10] have shown that the use
of cyanocopper derivatives RCuCNLi affords exclusively
the SN2Ј products. On the other hand, Grignard
reagents,[11Ϫ14] alkyllithium,[15Ϫ18] organoaluminum[19,20]
and organozinc[21] species favor the 1,2-addition products,
although a mixture of regioisomers is generally obtained
(Scheme 1). Some exceptions exist. For example, Maruy-
ama et al.[22] found that allylstannane exclusively gives the
1,2-adduct in the presence of BF3·OEt2 with various vinylic
epoxides. Recently, Zaidlewicz et al.[23] reported that allyl-
boration of 3,4-epoxycycloalkenes of six- to eight-mem-
bered rings with allyldiethylborane and (2-cyclohexenyl)d-
icyclohexylborane favors the cis-1,2-addition. Xue et al.[24]
reported that the use of diethylzinc in the presence of
CF3COOH gives the cis-1,2-addition product.
Scheme 1. Nucleophilic addition of various organometallic re-
agents to 1,3-cyclohexadiene monoepoxide 1
It might be predicted that a hard nucleophile should have
a preference for the SN2 process, whereas a soft nucleophile
(such as a copper reagent) would prefer a softer center, and
therefore an SN2Ј process. Indeed, our group has reported
that an RLi/BF3 combination allows the regioselective SN2
nucleophilic cleavage of a variety of cyclic as well as acyclic
vinylic epoxides.[25,26] A clean anti process is involved and
the reaction works well with a variety of RLi reagents (R ϭ
Me, nBu, Ph, hexynyl, benzyl). However, the reaction must
be performed at low temperatures (Ϫ75 to Ϫ116 °C) and
with cyclopentadiene oxide a mixture of unidentified prod-
ucts was observed. This epoxide seems to be too sensitive
to strongly basic or Lewis acidic conditions. That is the rea-
son why we turned our attention to slightly less basic re-
agents such as zincates and aluminates. Here we want to
report the regioselective SN2 anti addition of trialkylzin-
cates and tetraalkylaluminates to various vinylic epoxides
at room temperature.
[a]
´
´
`
Departement de Chimie Organique, Universite de Geneve,
Sciences II,
`
30 quai Ernest-Ansermet, 1211 Geneve 4, Switzerland
Fax: (internat.) ϩ 41-22-379-3215
E-mail: alexandre.alexakis@chiorg.unige.ch
[b]
´ ´
Laboratoire de Chimie des Organoelements, UMR 7611,
´
Universite Pierre et Marie Curie,
4 place Jussieu, 75252 Paris Cedex 05, France
Eur. J. Org. Chem. 2004, 2151Ϫ2159
DOI: 10.1002/ejoc.200300759
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2151

O. Equey, E. Vrancken, A. Alexakis
FULL PAPER
Table 1. Regioselective opening of 1,3-cyclohexadiene monoepoxide (1)
Entry
Reagent
Additive
Solvent
Temp.
Time
Yield (%)[a]
SN2/SN2Ј[b]
1
2
3
4
5
6
7
8
LiZnBu3
LiZnMe3
LiZnBu3
LiZnBu3
LiZnBu3
LiZnMe3
LiZnMe3
LiZnMe3
LiZnMe3
LiAlBu4
LiAlBu4
AlBu3
BF3·OEt2
Et2O
Et2O
THF
Et2O
hexane
THF
THF
Et2O
hexane
Et2O
hexane
Et2O
THF
hexane
Ϫ90 °C
10 min
10 min
2 h
74
68
86
85
89
0
61
16
52
83
89
0
95:5
96:4
100:0
100:0
97:3
Ϫ
100:0
65:35
94:6
98:2
97:3
Ϫ
BF3·OEt2
Ϫ90 °C
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
room temp.
room temp.
room temp.
room temp.
reflux
room temp.
room temp.
room temp.
room temp.
room temp.
reflux
2 h
2 h
22 h
15 h
18 h
17 h
2 h
9
10
11
12
13
14
2 h
20 h
15 h
17 h
LiAlMe4
LiAlMe4
39
60
100:0
96:4
room temp.
[a]
[b]
1
Isolated yield after flash chromatography (FC). Determined by H NMR spectroscopy from the crude product.
2a and anti-3a with 89% overall yield (Entry 5). In THF,
with the less reactive trimethylzincate species, it was neces-
sary to heat the solution overnight, because coordination
of the zincate species by the solvent attenuated its reactivity.
Indeed, we simply recovered the starting material when the
reaction was performed at room temp. (Entry 6). After 15 h
at reflux, we obtained 61% of the SN2 product anti-2b only
(Entry 7). On the other hand, in Et2O the reaction could
be done at room temp. Surprisingly, we obtained only 16%
of a 65:35 mixture of substitution products anti-2b and anti-
3b (Entry 8). In view of this bad result, we decided to avoid
the use of Et2O with trialkylzincates for the rest of our
study. In hexane, we obtained a 94:6 mixture of SN2 and
SN2Ј products anti-2b and anti-3b, respectively, in 52% over-
all yield (Entry 9). These results were quite surprising be-
cause organozincates are known to undergo a 1,4-conju-
gated addition reaction with α,β-unsaturated carbonyl
compounds.[27Ϫ31]
Results and Discussion
Cyclic Vinylic Epoxides
We began our investigation with 1,3-cyclohexadiene
monoepoxide (1) as a substrate (Table 1). The organozinc-
ate species can be generated by reaction of an equimolar
amount of dialkylzinc and alkyllithium. When cyclohexadi-
ene monoepoxide (1) was treated with 1.5 equiv. of tributyl-
or trimethylzincate and 1.5 equiv. of BF3·OEt2 at Ϫ90 °C,
after 10 min we obtained a 95:5 mixture of the SN2 and
SN2Ј products anti-2a and anti-3a, respectively, in 74% over-
all yield, for tributlyzincate and a 96:4 mixture of SN2 and
SN2Ј products anti-2b and anti-3b, respectively, in 68% over-
all yield, for trimethylzincate (Entries 1 and 2). Hydrogen-
ation of the double bond and comparison with the reported
1H NMR spectrum of trans-2-butylcyclohexanol (4) al-
lowed us to confirm that the addition product arises from
a clean anti process (Scheme 2).
We thought that it should be possible to use tetraalkylal-
uminate to open vinylic epoxides regioselectively. Indeed,
we observed the same tendency as with triorganozincates.
The reaction of tetrabutylaluminate with 1 afforded, after
2 h at room temp., a 98:2 mixture of anti-2b and anti-3b
with 83% yield in Et2O and a 97:3 mixture of anti-2b and
anti-3b with 89% yield in hexane (Entries 10 and 11). When
1 was treated with 1.5 equiv. of AlBu3 in Et2O at room
Scheme 2. Hydrogenation of the double bond of 2a
Interestingly, the reaction could be performed at room temp., only degradation products were observed (Entry 12).
temperature without the need for a Lewis acid. After 2 h at Tetrabutylaluminate can be readily obtained by treating n-
room temperature in THF, we obtained 86% of the pure butyllithium with aluminum chloride in a 4:1 ratio in Et2O
SN2 product anti-2a (Entry 3). The reaction could also be or in a hydrocarbon solvent. To avoid the tedious removal
carried out in Et2O; we obtained 85% of pure SN2 product of the solvent, we decided to choose Et2O as polar solvent
anti-2a (Entry 4). The use of a nonpolar solvent such as with tetrabutylaluminate for our study instead of THF. The
hexane gave a 97:3 mixture of SN2 and SN2Ј products anti- reaction carried out with tetramethylaluminate gave similar
2152
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2151Ϫ2159

Regioselective SN2 Opening of Vinylic Epoxides
FULL PAPER
results. In THF, after 15 h at reflux, we obtained 39% of product anti-6b in 33% yield (Entry 5). In hexane, we ob-
anti-2b only and in hexane, after 17 h at room temp., a 96:4 tained an 83:17 mixture of anti-6b and anti-7b with 34%
mixture of anti-2b and anti-3b in 60% overall yield (Entries overall yield after 15 h at room temp. (Entry 6). With Li-
13 and 14). The tetramethylaluminate species can be pre- AlMe4, the yields were still lower. In THF, we did not iso-
pared by mixing 1 equiv. of trimethylaluminum with 1 late any product after flash chromatography (Entry 7), and
equiv. of methyllithium [Scheme 3, Equation (1)]. Alterna- in hexane we obtained a 95:5 mixture of anti-6b and anti-
tively, the reaction can be performed with the tetramethylal- 7b with only 18% overall yield (Entry 8).
uminate species prepared by adding 4 equiv. of methyllith-
Compared to monoepoxides 1 and 5, 1,3-cyclooctadiene
ium to aluminum chloride [Scheme 3, Equation (2)]. In monoepoxide (8) is much less reactive, so that even with
both cases, we observed exactly the same regioselectivities.
LiZnBu3 and LiAlBu4 the reaction required prolonged
heating to go to completion. Moreover, we observed a great
discrepancy between LiZnBu3 and LiZnMe3. When the re-
action was carried out in THF, with LiZnBu3, we obtained
54% of a 97:3 mixture of ring-opening products in favor of
the SN2 product anti-9a, while with LiZnMe3 we obtained
72% of a 1:1 mixture of anti-9b and anti-10b (Entry 9 vs.
Entry 12). This result was quite surprising since we nor-
mally obtained mainly the product of SN2 substitution in
polar solvent. The use of hexane, with LiZnBu3 and Li-
AlBu4, gave 85:15 and a 70:30 mixtures of anti-9a and anti-
10a in favor of the SN2 product, with 46% and 58% overall
yield respectively (Entries 10 and 11).
Scheme 3. Preparation of the tetraaluminate species
Encouraged by these initial results, we decided to expand
the scope of the method and the reaction conditions were
examined with various simple vinylic epoxides. We started
with cyclopentadiene monoepoxide (5) (Table 2). The reac-
tion of 5 with 1.5 equiv. of LiZnBu3 in THF afforded, after
3 h at room temperature, 75% of the SN2 product anti-6a
(Entry 1). In hexane, we observed the formation of more
SN2Ј product anti-7a than with cyclohexadiene monoepox-
Acyclic Vinylic Epoxides
We extended our study to acyclic vinylic epoxides. First,
ide (1). We obtained an 84:16 mixture of anti-6a and anti- the reaction was examined with very simple substrates such
7a with 66% overall yield (Entry 2). With LiAlBu4, the reac- as the monoepoxides 11 and 12 of 1,3-butadiene and iso-
tion gave a mixture of anti-6a and anti-7a, either in Et2O prene, respectively (Table 3). In every case we observed the
or in hexane, but largely in favor of the product arising formation of almost all the regio- and stereoisomers pos-
from the SN2 substitution (Entries 3 and 4). The yields ob- sible, the ratio of these isomers depending strongly on the
tained for the transfer of a methyl group in the C-2 position solvent used. When a solution of LiZnBu3 and 11 in THF
of monoepoxide 5 were moderate. With LiZnMe3, the reac- was stirred at room temperature for 4 h, we observed the
tion afforded, after 14 h at reflux in THF, only the SN2 formation of a 92:8 mixture of SN2 and SN2Ј products (En-
Table 2. Regioselective opening of cyclopentadiene monoepoxide (5) and 1,3-cyclooctadiene monoepoxide (8)
Entry
Substrate
Reagent
Solvent
Temp.
Time
Yield (%)[a]
SN2/SN2Ј[b]
1
2
3
4
5
6
7
8
9
10
11
12
5
5
5
5
5
5
5
5
8
8
8
8
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
LiZnMe3
LiZnMe3
LiAlMe4
LiAlMe4
LiZnBu3
LiZnBu3
LiAlBu4
LiZnMe3
THF
hexane
Et2O
hexane
THF
hexane
THF
hexane
THF
hexane
hexane
THF
room temp.
room temp.
room temp.
room temp.
reflux
room temp.
reflux
room temp.
reflux
reflux
3 h
3 h
3 h
75
68
86
89
33
34
0
18
54
46
58
72
100:0
84:16
96:4
3 h
92:8
14 h
15 h
14 h
15 h
16 h
19 h
22 h
36 h
100:0
83:17
Ϫ
95:5
97:3
85:15
70:30
49:51
reflux
reflux
[a]
[b]
1
Isolated yield after FC. Determined by H NMR spectroscopy from the crude product.
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2151Ϫ2159
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2153

O. Equey, E. Vrancken, A. Alexakis
FULL PAPER
Table 3. Regioselective opening of 1,3-butadiene monoepoxide (11) and isoprene monoepoxide (12)
Entry
Substrate
Reagent
Solvent
Temp.
Time
Yield
(%)[b]
13:14/15:16 or
17:18/19:20[a]
SN2/SN2Ј[a]
1
2
3
4
5
6
7
8
11
11
11
11
12
12
12
12
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
THF
hexane
Et2O
hexane
THF
hexane
Et2O
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
4 h
3 h
3 h
4 h
17 h
5 h
5 h
4 h
55
80
56
70
70
89
85
79
63:29/6:2
83:0/16:1
12:1/42:45
55:1/39:5
35:50/10:5
33:0/55:12
17:0/14:69
43:0/36:21
92:8
83:17
13:87
56:44
85:15
33:77
17:83
43:57
hexane
[a]
1
[b]
Determined by H NMR spectroscopy from the crude product. Isolated yield after FC.
try 1). SN2 substitution was favored, but we isolated a mix- formed. With LiAlBu4 in Et2O, the reaction afforded
ture of products of addition to both carbon atoms bearing mainly both SN2Ј products 15 and 16, in a 1:1 ratio between
the epoxide (13 and 14). In hexane, the reaction gave an (E) and (Z) isomers, while in hexane the reaction provided
83:17 mixture of SN2 and SN2Ј products. In this case, we a 56:44 mixture of SN2 and SN2Ј products (Entries 3 and 4).
observed only the SN2 substitution product 13 arising from
We thought that the presence of the methyl group would
a nucleophilic attack at the carbon atom of the epoxide next favor attack at the terminal position of the monoepoxide
to the double bond (Entry 2). The product 14 was not 12, whereas in fact the reaction was shown to favor SN2Ј
Table 4. Regioselective opening of (2E)-2-styryloxirane (21) and (E)-2-(hex-1-enyl)oxirane (25)
Entry
Substrate
Reagent
Solvent
Temp.
Time
Yield
(%)[b]
22:23:24 or
26:27:28[a]
SN2/SN2Ј[a]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
21
21
21
21
21
21
21
21
21
25
25
25
25
25
25
25
25
25
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
LiZnMe3
LiZnMe3
LiAlMe4
LiAlMe4
LiAlMe4
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
LiZnMe3
LiZnMe3
LiAlMe4
LiAlMe4
LiAlMe4
THF
hexane
Et2O
hexane
THF
hexane
THF
Et2O
hexane
THF
hexane
Et2O
hexane
THF
hexane
THF
Et2O
hexane
room temp.
room temp.
room temp.
room temp.
reflux
room temp.
reflux
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
2 h
2 h
2 h
84
83
65
81
77
94
0
91:3:6
81:0:19
82:0:18
81:0:19
94:5:1
91:0:9
Ϫ
92:0:8
76:0:24
90:8:2
99:0:1
92:4:4
97:0:3
96:1:3
95:0:5
Ϫ
94:6
81:19
82:18
81:19
99:1
91:9
Ϫ
92:8
76:24
98:2
99:1
96:4
97:3
97:3
95:5
Ϫ
2 h
22 h
22 h
5 d
16 h
21 h
4 h
2 h
2 h
2 h
23 h
20 h
5 d
41 h
20 h
74
65
65
69
63
66
53
66
0
48[c]
62
99:0:1
98:0:2
99:1
98:2
[a]
1
[b]
[c]
Determined by H NMR spectroscopy from the crude product. Isolated yield after FC. Conversion: 75%.
2154
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2151Ϫ2159

Regioselective SN2 Opening of Vinylic Epoxides
FULL PAPER
substitution. When the reaction was carried out with The results obtained with LiZnMe3 and LiAlMe4 were
LiZnBu3 in THF, we obtained an 85:15 mixture of SN2 and quite similar to those obtained for the transfer of a butyl
SN2Ј products with 70% overall yield; the product 18 arising group. The best regioselectivity was obtained in THF with
from an SN2 substitution at the terminal carbon atom of LiZnMe3 (Entry 5). After 22 h at room temperature, we ob-
the epoxide was obtained to a much greater extent than tained a 99:1 mixture of SN2 and SN2Ј products with 77%
with monoepoxide 11 (Entry 5 vs. Entry 1). In hexane, the overall yield. In hexane, the ratio between SN2 and SN2Ј
reaction afforded a 33:77 mixture favoring the (E)-SN2Ј products dropped to 91:9 (Entry 6). With LiAlMe4, in
product 19, with 89% overall yield (Entry 6). Again, in Et2O, we isolated a 92:8 mixture of SN2 and SN2Ј products
Et2O LiAlBu4 favored formation of the SN2Ј products, but with 74% overall yield (Entry 8). In hexane, this ratio
with a preference for the (Z) isomer 20 (Entry 7). In hexane, dropped to 76:24 (Entry 9). In THF, after 5 d at room tem-
we obtained a 43:57 mixture of SN2 and SN2Ј products with perature, we only recovered the starting material (Entry 7).
79% overall yield (Entry 8).
The vinylic epoxide 25 possesses an alkyl group instead
At this stage, we thought that the absence of any substitu- of a phenyl group at the terminal olefinic carbon atom. The
ent at the terminal olefinic carbon atom could be respon- regioselectivities observed in apolar solvents were better
sible for the bad regioselectvities observed with monoepox- with epoxide 25 than with epoxide 21. With LiZnBu3, in
ides 11 and 12. In order to check this assumption, we pre- THF, we obtained a 98:2 mixture of SN2 and SN2Ј products
pared two vinylic epoxides through the two-step sequence (Entry 10). However, both SN2 substitution products 26a
described by Matteson et al.[32] Monoepoxide 21 was syn- and 27a were formed. In hexane, the ratio was 99:1 in favor
thesized starting from cinnamaldehyde. LiZnBu3 was used of the SN2 substitution product 26a (Entry 11). Interest-
first. In THF, after 2 h at room temperature, the reaction ingly, we observed only one product arising from an SN2
afforded mainly the SN2 substitution products 22a, with a substitution. With LiAlBu4 in Et2O, we obtained a 96:4
good overall yield (84%) (Table 4, Entry 1). In hexane, the mixture of SN2 and SN2Ј products and we observed both
reaction gave a greater amount of SN2Ј product 24a, but SN2 substitution products 26a and 27a (Entry 12). In hex-
the product 23a arising from an SN2 substitution on the ane, the ratio was 97:3 in favor of the SN2 substitution
terminal carbon atom of the epoxide was not formed (Entry product 26a (Entry 13). The reaction was performed with
2). With LiAlBu4, we observed no difference of regioselec- LiZnMe3 and LiAlMe4. With LiZnMe3, after 23 h at room
tivity between polar and apolar solvents (Entries 3 and 4). temperature in THF, we obtained a 97:3 mixture of SN2
Table 5. Regioselective opening of (5Z,7Z)-5,7-dodecadiene monoepoxide (29) and (6E,8E)-6,8-tetradecadiene monoepoxide (30)
Entry
Substrate
Reagent
Solvent
Temp.
Time
Yield (%)[a]
SN2/SN2Ј[b]
1
2
29
29
29
29
29
29
29
29
30
30
30
30
30
30
30
LiZnBu3
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
LiZnMe3
LiZnMe3
LiZnMe3
LiZnBu3
LiZnBu3
LiAlBu4
LiAlBu4
LiZnMe3
LiZnMe3
LiZnMe3
THF
THF
hexane
Et2O
hexane
THF
THF
hexane
THF
hexane
Et2O
hexane
THF
room temp.
reflux
room temp.
room temp.
room temp.
room temp.
reflux
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
room temp.
6 d
16 h
2 h
16 h
2 h
54[c]
96
94
8
89
26[d]
77
75
43[e]
77
75
83
0
95:5
95:5
76:24
88:12
71:29
99:1
3
4
5
6
7
8
9
10
11
12
13
14
15
7 d
18 h
64 h
5 d
3 h
24 h
4 h
6 d
12 d
6 d
99:1
65:35
87:13
73:27
91:9
41:59
Ϫ
79:21
39:61
[f]
THF
hexane
84
88
[a]
[b]
1
[c]
[e]
Isolated yield after FC. Determined by H NMR spectroscopy from the crude product. Conversion: 70%. [d] Conversion: 45%.
[f]
Conversion: 62%. The reaction was carried out with 5 equiv. of LiZnMe3.
Eur. J. Org. Chem. 2004, 2151Ϫ2159
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2155

O. Equey, E. Vrancken, A. Alexakis
FULL PAPER
and SN2Ј products with 53% overall yield, while in hexane
The reaction was also studied with LiZnMe3. In THF,
we obtained a 95:5 mixture of SN2 and SN2Ј products with after 7 d at room temperature, the ratio was excellent (99:1
66% overall yield (Entries 14 and 15). With LiAlMe4, after in favor of anti-31b) but the conversion was not complete
5 d in THF, we simply recovered the starting material (En- (45%) (Entry 6). Again, when the reaction mixture was
try 16). In Et2O, after 41 h at room temperature, the conver- heated to reflux, after 18 h we obtained a better yield (77%)
sion was not complete, but we obtained 48% of a 99:1 mix- and the same regioselectivity (Entry 7). In hexane, after
ture of SN2 and SN2Ј products (Entry 17). In hexane, after 64 h at room temperature, we obtained a 65:35 mixture of
20 h at room temperature, the reaction gave a 98:2 mixture SN2 and SN2Ј products with 75% overall yield (Entry 8).
of SN2 and SN2Ј products with 62% overall yield (Entry 18). These results confirmed that SN2 substitution is preferred
Finally, the reaction was extended to acyclic vinylic epox- in polar solvents such as THF.
ides having the same substitution pattern at C-1 and C-4
The reaction was then carried out with monoepoxide 30.
(Table 5). We thought that with an unbiased system we This epoxide possesses a trans configuration. In THF with
would no longer observe products arising from an SN2 sub- LiZnBu3, we obtained an 87:13 mixture of SN2 and SN2Ј
stitution at the terminal carbon atom of the epoxide. products 34a and 35a, respectively, with 43% overall yield
(5Z,7Z)-5,7-Dodecadiene monoepoxide (29) and (6E,8E)- (conv. 62%; Entry 9). Here also, the double bond retained
6,8-tetradecadiene monoepoxide (30) were selected as rep- its configuration and only anti-SN2 substitution products
resentative substrates. We observed a slight difference of re- were detected. In hexane, after 3 h at room temperature, we
activity between both substrates. The monoepoxide 29 re- obtained a 73:27 mixture of substitution products in favor
acted faster than the monoepoxide 30. This difference of anti-34a (Entry 10). With LiAlBu4 in Et2O, we obtained
should arise from the fact that both substituents on the 75% of a 91:9 mixture of substitution products in favor of
epoxide 29 are on the same side, so that the carbon atom anti-34a (Entry 11). In hexane, after 4 h at room tempera-
of the epoxide next to the double bond is sterically less hin- ture, more SN2Ј substitution products were formed (41:59)
dered.[3,26] Moreover, the regioselectivities observed with but the overall yield was better (83%; Entry 12). The reac-
monoepoxide 29 were slightly better. When the reaction was tion was then performed with LiZnMe3. It was very slow.
carried out with LiZnBu3 in THF, after 6 d at room tem- After 6 d at room temperature we just recovered the starting
perature we isolated 54% of a 95:5 mixture of SN2 and SN2Ј material (Entry 13). With 5 equiv. of LiZnMe3, after 12 d
products in favor of anti-31a (Entry 1). It is noteworthy that at room temperature we obtained a 79:21 mixture of SN2
the (Z) double bond retained its configuration and that no and SN2Ј products in favor of anti-34b with an overall yield
cis-1,2-addition products were formed. The conversion was of 84% (Entry 14). In hexane, after 6 d at room tempera-
not complete (70%), but when the solution was heated at ture, we obtained a 39:61 mixture of ring-opening products
reflux, after 16 h the reaction was complete. We isolated in favor of the SN2Ј substitution product anti-35b with a
96% of products of substitution with exactly the same ratio good overall yield (88%; Entry 15).
in favor of the SN2 substitution product 31a (Entry 2). In
hexane, the reaction was much faster, as after 2 h at room
temperature we obtained 94% of ring-opening products
(Entry 3). However, more SN2Ј product was formed (76:24
in favor of anti-31a). The reaction has also been run with
LiAlBu4. In Et2O, only 8% of an 88:12 mixture of addition
Influence of the Solvent
The influence of the solvent on the substitution products
ratio can be explained with the following model
(Scheme 5).[33] In a non-coordinating solvent, the lithium
salts of the ‘‘ate’’ complexes must exist essentially in the
form of contact ion pairs. One can then imagine the forma-
products was formed (Entry 4). We isolated mainly the
tion of complex 36 between the organometallic species and
homoallyl ketone 33 in 75% yield (Scheme 4). In hexane, as
the vinylic epoxide. Such a complex activates not only the
with the reaction with LiZnBu3, we obtained a 71:29
carbon atom of the epoxide next to the double bond but
mixture of SN2 and SN2Ј products in favor of anti-31a
also the terminal carbon atom of the double bond, because
(Entry 5).
the positive charge can be stabilized by the mesomeric ef-
fect.[22] In polar solvents such as THF, the ‘‘ate’’ complexes
exist in the form of ion pairs separated by solvents. There-
fore a complex between the oxygen atom of the epoxide
bridge and the organometallic species cannot be formed.
Such complex formation explains also the difference in re-
Scheme 4. Homoallyl ketone 33
activity observed in THF and in hexane.
Scheme 5. Model for the influence of the solvent on regioselectivity
2156
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2151Ϫ2159

Regioselective SN2 Opening of Vinylic Epoxides
FULL PAPER
 solution of nBuLi in hexane was added dropwise. The solution
was stirred for 30 min at room temp. It was then cooled to Ϫ10 °C
before addition of a solution of epoxide (2 mmol) in the required
solvent (3 mL). The solution was then stirred at room temp. until
all the starting material had been consumed (the reaction was
monitored by TLC). The reaction mixture was cooled to 0 °C,
quenched with a 3:1 solution of NH4Cl/NH4OH (5 mL) and stirred
for 30 min at room temp. before being transferred into a separating
funnel. The aqueous phase was extracted three times with Et2O.
The combined organic phases were washed once with brine, dried
with MgSO4, filtered and the solvents were removed under reduced
pressure. The crude product was then subjected to FC on SiO2.
The nature of the organometallic species also plays a role
as ‘‘ate’’ complexes are not as hard as an alkyllithium spec-
ies, for example, but are harder than an organocopper spec-
ies. This intermediate softness can also explain why we ob-
serve more SN2Ј substitution in non-polar solvents.
Conclusion
We have demonstrated with several examples that trialky-
lorganozincates and tetraalkylaluminates provide a con-
venient way to introduce alkyl substituents at the allylic po-
sition of 1,3-diene monoepoxides regio- and diastereoselec-
tively. We have also shown that the solvent has an influence
on the substitution products’ ratio. A polar solvent such as
THF favors addition at the allylic position. We observed
more SN2Ј substitution with non-coordinating solvents such
as hydrocarbons. We have also observed that polar solvents
slow down the reaction rate and that organoaluminates are
more sensitive than organozincates to this effect. Moreover,
the presence of a substituent at the double bond is essential,
otherwise we observe formation of a large amount of SN2Ј
substitution product. The presence of a Lewis acid is not
necessary with our method and the reaction can be carried
out at room temperature. Heating of the solution in some
cases shortens the reaction time, but without having any
influence on the substitution products ratio. For these re-
asons, our method is a good alternative to the already exist-
ing methods.
General Procedure for the Nucleophilic Opening of Epoxides with
Lithium Trimethylzincate: In a 25-mL, two-necked flask equipped
with a thermometer was placed 1.5 mL (3 mmol) of a 2  solution
of ZnMe2 in toluene in the required solvent (6 mL) under nitrogen.
The solution was cooled to Ϫ20 °C and 1.9 mL (3 mmol) of a 1.6
 solution of MeLi in Et2O was added dropwise. The solution was
stirred for 30 min at room temp. It was then cooled to Ϫ10 °C
before addition of a solution of the epoxide (2 mmol) in the re-
quired solvent (3 mL). The solution was then stirred at room temp.
until all the starting material had been consumed (the reaction was
monitored by TLC). The reaction mixture was cooled to 0 °C,
quenched with a 3:1 solution of NH4Cl/NH4OH (5 mL) and stirred
for 30 min at room temp. before being transferred into a separating
funnel. The aqueous phase was extracted three times with Et2O.
The combined organic phases were washed once with brine, dried
with MgSO4, filtered and the solvents were removed under reduced
pressure. The crude product was then subjected to FC on SiO2.
General Procedure for the Nucleophilic Opening of Epoxides with
Lithium Tetrabutylaluminate: In
a 25-mL, two-necked flask
equipped with a thermometer was placed 406 mg (3 mmol) of
AlCl3 in the required solvent (9 mL) under nitrogen. The solution
was cooled to Ϫ20 °C and 6.3 mL (12 mmol) of a 1.9  solution
of nBuLi in hexane was added dropwise. The solution was stirred
for 1 h at room temp. and the formation of a white suspension was
observed. The suspension was then cooled to Ϫ10 °C before ad-
dition of a solution of the epoxide (2 mmol) in the required solvent
(3 mL). The solution was then stirred at room temp. until all the
starting material had been consumed (the reaction was monitored
by TLC). The reaction mixture was cooled to 0 °C, quenched with
a 1  solution of HCl (5 mL) and stirred for 30 min at room temp.
before being transferred into a separating funnel. The aqueous
phase was extracted three times with Et2O. The combined organic
phases were washed once with brine, dried with MgSO4, filtered
and the solvents were removed under reduced pressure. The crude
product was then subjected to FC on SiO2.
Experimental Section
General Remarks: THF and Et2O were freshly distilled from so-
dium/benzophenone. Other solvents were obtained from commer-
cial sources and used as received. Flash column chromatography
was carried out on Baker silica gel (0.03Ϫ0.06 mm). TLC was car-
ried out on MachereyϪNagel plastic sheets of silica gel 60 F254. A
solution of 25 g of phosphomolybdic acid, 10 g of Ce(SO4)2·4H2O,
60 mL of concd. H2SO4 and 940 mL of H2O was used to reveal the
1
TLC spots. H and 13C NMR spectra were recorded either with a
Varian Gemini 200 or with a Bruker AC-400 spectrometer. Chemi-
cal shifts are expressed in ppm relative to tetramethylsilane (δ ϭ
0 ppm) and coupling constants are reported in Hz. Deuterated
chloroform was used as solvent. Spin multiplicities are described
with the following symbols: s (singlet), d (doublet), t (triplet), q
(quadruplet), m (multiplet), br. (broad). Mass spectra were ob-
tained with a Finnigan 4023 instrument and IR spectra were re-
corded with a PerkinϪElmer 1420 spectrometer.
General Procedure for the Nucleophilic Opening of Epoxides with
Lithium Tetramethylaluminate: In a 25-mL, two-necked flask
equipped with a thermometer was placed 1.5 mL (3 mmol) of a 2 
solution of AlMe3 in heptane in the required solvent (6 mL) under
nitrogen. The solution was cooled to Ϫ20 °C and 1.9 mL (3 mmol)
of a 1.6  solution of MeLi in Et2O was added dropwise. The
solution was stirred for 30 min at room temp. until it became
cloudy. It was then cooled to Ϫ10 °C before addition of a solution
of the epoxide (2 mmol) in the required solvent (3 mL). The solu-
tion was then stirred at room temp. until all the starting material
had been consumed (the reaction was monitored by TLC). The
reaction mixture was cooled to 0 °C, quenched with a 1  solution
of HCl (5 mL) and stirred for 30 min at room temp. before being
transferred into a separating funnel. The aqueous phase was ex-
tracted three times with Et2O. The combined organic phases were
Preparation of the Epoxides: All the epoxides have already been
described. They were prepared according to the procedure of Cran-
dall et al.,[34] except for the oxiranes 21 and 25, which were pre-
pared according to the two-step procedure described by Matteson
et al.[32] and Lautens et al.[35]
General Procedure for the Nucleophilic Opening of Epoxides with
Lithium Tributylzincate: In a 25-mL, two-necked flask equipped
with a thermometer was placed 3 mL (3 mmol) of a 1  solution
of ZnBu2 in hexane in the required solvent (7 mL) under nitrogen.
The solution was cooled to Ϫ20 °C and 1.6 mL (3 mmol) of a 1.9
Eur. J. Org. Chem. 2004, 2151Ϫ2159
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2157

O. Equey, E. Vrancken, A. Alexakis
FULL PAPER
washed once with brine, dried with MgSO4, filtered and the sol-
vents were removed under reduced pressure. The crude product was
then subjected to FC on SiO2.
3.44Ϫ3.42 (m, 1 H), 3.36Ϫ3.31 (m, 1 H), 2.31Ϫ2.24 (m, 1 H),
2.01Ϫ1.98 (m, 2 H), 1.83 (br. s, 1 H), 1.34Ϫ1.20 (m, 5 H), 0.96 (d,
J ϭ 6.8 Hz, 3 H), 0.89Ϫ0.86 (m, 3 H) ppm. 13C NMR (100 MHz,
CDCl3): δ ϭ 132.2 (d), 132.2 (d), 67.3 (t), 39.7 (d), 32.3 (t), 31.7
(t), 22.2 (t), 22.2 (t), 16.6 (q), 13.9 (q) ppm. The IR spectrum[40]
and the elemental analysis[41] have already been reported.
trans-2-Butyl-3-cyclohexen-1-ol (2a): The product was isolated by
FC (cyclohexane/EtOAc, 9:1) as a colorless oil. The NMR spectra
of 2a are identical to those reported previously.[26]
trans-(7Z)-6-Butyl-7-dodecen-5-ol (31a): The product was isolated
by FC (cyclohexane/EtOAc, 29:1) as a colorless oil. The NMR
spectra of 31a are identical with those reported.[26]
trans-2-Methyl-3-cyclohexen-1-ol (2b): The product was isolated by
FC (cyclohexane/EtOAc, 9:1) as a colorless oil. The NMR spectra
of 2b are identical to those reported previously.[36]
trans-(7Z)-6-Methyl-7-dodecen-5-ol (31b): The product was isolated
by FC (cyclohexane/EtOAc, 29:1) as a slightly yellow oil. 1H NMR
(400 MHz, CDCl3): δ ϭ 5.40 (dt, J ϭ 11.1, 7.8 Hz, 1 H), 5.20 (t,
J ϭ 9.9 Hz, 1 H), 3.39Ϫ3.32 (m, 1 H), 2.57Ϫ2.48 (m, 1 H),
2.10Ϫ1.99 (m, 2 H), 1.57Ϫ1.42 (m, 3 H), 1.38Ϫ1.26 (m, 8 H), 0.98
(d, J ϭ 6.8 Hz, 3 H), 0.91Ϫ0.88 (m, 6 H) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 132.1 (d), 130.6 (d), 75.9 (d), 38.0 (d), 34.1
(t), 32.0 (t), 28.3 (t), 27.4 (t), 22.8 (t), 22.4 (t), 16.6 (q), 14.1 (q),
14.0 (q) ppm. IR: ν˜ ϭ 3410, 3015, 2910, 1020, 740 cmϪ1. C13H26O
(198.34): calcd. C 78.72, H 13.21; found C 78.65, H 13.26.
trans-2-Butyl-3-cyclopenten-1-ol (6a): The product was isolated by
FC (cyclohexane/EtOAc, 9:1) as a pale yellow oil. The NMR spec-
tra of 6a are identical to those reported previously.[37]
trans-2-Methyl-3-cyclopenten-1-ol (6b):The product was isolated by
FC (pentane/Et2O, 3:1) as a slightly yellow oil. The NMR spectra
of 6b are identical to those reported previously.[38]
trans-2-Butyl-3-cycloocten-1-ol (9a): The product was isolated by
FC (cyclohexane/EtOAc, 19:1) as an oil. 1H NMR (400 MHz,
CDCl3): δ ϭ 5.71 (q, J ϭ 8.1 Hz, 1 H), 5.23 (t, J ϭ 9.8 Hz, 1 H),
3.50 (dt, J ϭ 10.1, 3.6 Hz, 1 H), 2.45Ϫ2.40 (m, 1 H), 2.25Ϫ2.18
(m, 1 H), 2.11Ϫ2.08 (m, 1 H), 1.82Ϫ1.60 (m, 6 H), 1.47 (br. s, 1
H), 1.34Ϫ1.19 (m, 6 H), 0.89 (t, J ϭ 7 Hz, 3 H) ppm. 13C NMR
(100 MHz, CDCl3): δ ϭ 133.9 (d), 131.4 (d), 75.9 (d), 42.5 (d), 34.8
(t), 32.1 (t), 30.0 (t), 29.2 (t), 27.6 (t), 22.9 (t), 22.0 (t), 14.2 (q)
ppm. IR: ν˜ ϭ 3400, 2980, 1050, 730 cmϪ1. C12H22O (182.30): calcd.
C 79.06, H 12.16; found C 79.04, H 12.12.
trans-(8E)-7-Butyl-8-tetradecen-6-ol (34a): The product was iso-
1
lated by FC (cyclohexane/EtOAc, 9:1) as a slightly yellow oil. H
NMR (400 MHz, CDCl3): δ ϭ 5.48 (dt, J ϭ 15.3, 6.8 Hz, 1 H),
5.18 (dd, J ϭ 15.3, 9.3 Hz, 1 H), 3.37 (br. s, 1 H), 2.06Ϫ2.00 (m, 2
H), 1.94Ϫ1.88 (m, 1 H), 1.57Ϫ1.23 (m, 21 H), 0.90Ϫ0.86 (m, 9 H)
ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 134.5 (d), 130.3 (d), 73.8
(d), 49.2 (d), 34.6 (t), 32.7 (t), 32.0 (t), 31.3 (t), 30.8 (t), 29.7 (t),
29.2 (t), 25.4 (t), 22.7 (t), 22.7 (t), 22.5 (t), 14.1 (q), 14.1 (q) ppm.
IR: ν˜ ϭ 3320, 2910, 1430, 1030, 960 cmϪ1. C20H40O (296.53): calcd.
C 81.01, H 13.60; found C 81.92, H 13.54.
2-Butyl-4-phenyl-3-buten-1-ol (22a): The product was isolated by
FC (cyclohexane/EtOAc, 9:1) as a slightly yellow oil. 1H NMR
(400 MHz, CDCl3): δ ϭ 7.41Ϫ7.21 (m, 5 H), 6.51 (d, J ϭ 15.9 Hz,
1 H), 6.01 (dd, J ϭ 15.9, 9.0 Hz, 1 H), 3.65 (dd, J ϭ 10.6, 5.2 Hz,
1 H), 3.52 (dd, J ϭ 10.6, 8.1 Hz, 1 H), 2.42Ϫ2.37 (m, 1 H),
1.74Ϫ1.69 (m, 1 H), 1.53Ϫ1.50 (m, 1 H), 1.40Ϫ1.31 (m, 5 H), 0.92
(t, J ϭ 6.0 Hz, 3 H) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ 137.2
(s), 132.4 (d), 131.7 (d), 128.6 (d), 127.3 (d), 126.2 (d), 66.1 (t), 46.4
(d), 30.9 (t), 29.4 (t), 22.8 (t), 14.1 (q) ppm. IR: ν˜ ϭ 3420, 3025,
2960, 1640, 1330, 1070, 970, 690 cmϪ1. C14H20O (204.31): calcd. C
82.30, H 9.87; found C 82.25, H 9.91.
trans-(8E)-7-Methyl-8-tetradecen-6-ol (34b): The product was iso-
1
lated by FC (cyclohexane/EtOAc, 9:1) as a slightly yellow oil. H
NMR (400 MHz, CDCl3): δ ϭ 5.51 (dt, J ϭ 15.3, 6.7 Hz, 1 H),
5.30 (dd, J ϭ 15.4, 8.5 Hz, 1 H), 3.31 (br. s, 1 H), 2.15Ϫ2.07 (m, 1
H), 2.05Ϫ1.99 (m, 2 H), 1.52Ϫ1.26 (m, 15 H), 1.00 (d, J ϭ 6.8 Hz,
3 H), 0.91Ϫ0.87 (m, 6 H) ppm. 13C NMR (100 MHz, CDCl3): δ ϭ
133.0 (d), 131.7 (d), 74.8 (d), 43.2 (d), 34.3 (t), 32.3 (t), 32.0 (t),
31.3 (t), 28.6 (t), 25.6 (t), 22.6 (t), 22.5 (t), 16.9 (q), 14.1 (q), 14.1
(q) ppm. IR: ν˜ ϭ 3390, 2970, 1470, 1060, 780 cmϪ1. C17H34O
(254.45): calcd. C 80.24, H 13.47; found C 80.17, H 13.51.
2-Methyl-4-phenyl-3-buten-1-ol (22b):[39] The product was isolated
by FC (cyclohexane/EtOAc, 9:1) as a yellow oil. 1H NMR
(400 MHz, CDCl3): δ ϭ 7.40Ϫ7.22 (m, 5 H), 6.50 (d, J ϭ 16 Hz,
1 H); 6.01 (dd, J ϭ 15.9, 7.9 Hz, 1 H), 3.61 (dd, J ϭ 10.5, 5.7 Hz,
1 H), 3.53 (dd, J ϭ 10.5, 7.5 Hz, 1 H), 2.58Ϫ2.53 (m, 1 H), 1.66
(br. s, 1 H), 1.13 (d, J ϭ 6.8 Hz, 3 H) ppm. 13C NMR (100 MHz,
CDCl3): δ ϭ 137.3 (s), 132.6 (d), 130.9 (d), 128.6 (d), 127.2 (d),
126.2 (d), 67.4 (t), 40.2 (d), 16.5 (q) ppm.
Acknowledgments
We thank the Swiss National Science Foundation (No. 20-
068095.02) for financial support.
[1]
Comprehensive Organic Synthesis (Eds.: B. M. Trost, I. Flem-
2-Butyl-3-octen-1-ol (26a): The product was isolated by FC (cyclo-
ing), Pergamon Press, New York, 1991, vol. 6.
1
[2]
hexane/EtOAc, 19:1) as a slightly yellow oil. H NMR (400 MHz,
J. G. Smith, Synthesis 1984, 629Ϫ656.
[3]
J. A. Marshall, Chem. Rev. 1989, 89, 1503Ϫ1511.
E. R. Lee, I. Lacomy, P. Bigler, R. Scheffold, Helv. Chim. Acta
CDCl3): δ ϭ 5.52 (dt, J ϭ 15.3, 6.8 Hz, 1 H), 5.12 (ddt, J ϭ 15.3,
8.9, 1.3 Hz, 1 H), 3.50 (dd, J ϭ 10.4, 5.1 Hz, 1 H), 3.31 (dd, J ϭ
10.4, 8.5 Hz, 1 H), 2.16Ϫ2.10 (m, 1 H), 2.03Ϫ1.99 (m, 2 H), 1.64
(br. s, 1 H), 1.35Ϫ1.19 (m, 10 H), 0.90Ϫ0.85 (m, 6 H) ppm. 13C
NMR (100 MHz, CDCl3): δ ϭ 133.9 (d), 131.3 (d), 66.0 (t), 45.9
(d), 32.4 (t), 31.7 (t), 30.8 (t), 29.3 (t), 22.7 (t), 22.2 (t), 14.0 (q),
13.9 (q) ppm. IR: ν˜ ϭ 3370, 3080, 2960, 1640, 1435, 1050, 990, 910
cmϪ1. C12H24O (184.32) calcd. C 78.20, H 13.12; found C 78.11,
H 13.18.
[4]
1991, 74, 146.
[5]
M. E. Jung, D. C. D’Amico, J. Am. Chem. Soc. 1995, 117, 7379.
[6]
B. H. Lipshutz, K. Woo, T. Gross, D. J. Buzard, R. Tirado,
Synlett 1997, 477Ϫ478.
F. Badalassi, P. Crotti, F. Macchia, M. Pineschi, A. Arnold, B.
[7]
L. Feringa, Tetrahedron Lett. 1998, 39, 7795Ϫ7798.
F. Bertozzi, P. Crotti, F. Macchia, M. Pineschi, B. L. Feringa,
[8]
Angew. Chem. Int. Ed. 2001, 40, 930Ϫ932.
[9]
J. P. Marino, D. P. Floyd, Tetrahedron Lett. 1979, 20, 675Ϫ678.
[10]
2-Methyl-3-octen-1-ol (26b): The product was isolated by FC
(cyclohexane/EtOAc, 9:1) as a slightly yellow oil. 1H NMR
(400 MHz, CDCl3): δ ϭ 5.55Ϫ5.47 (m, 1 H), 5.27Ϫ5.20 (m, 1 H),
J. P. Marino, R. Fernandez de la Pradilla, E. Laborde, J. Org.
Chem. 1987, 52, 4898Ϫ4913.
R. J. Anderson, J. Am. Chem. Soc. 1970, 92, 4978Ϫ4979.
[11]
2158
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 2151Ϫ2159

Regioselective SN2 Opening of Vinylic Epoxides
FULL PAPER
M. Isobe, S. Kondo, N. Nagasawa, T. Goto, Chem. Lett.
1977, 679Ϫ682.
W. T ückmantel, K. Oshima, H. Nozaki, Chem. Ber. 1986,
119, 1581Ϫ1593.
C. A. Musser, H. G. Richey Jr., J. Org. Chem. 2000, 65,
[12]
[27]
[28]
[29]
[30]
[31]
C. B. Rose, S. K. Taylor, J. Org. Chem. 1974, 39, 578.
A. J. Bloodworth, R. J. Curtis, M. D. T. Spencer, N. A. Tallant,
[13]
Tetrahedron 1993, 49, 2729.
B. C. Söderberg, L. R. Austin, C. A. Davis, J. E. Nyström, J.
[14]
O. Vagborg, Tetrahedron 1994, 50, 61.
R. W. Herr, C. R. Johnson, J. Am. Chem. Soc. 1970, 92,
4979Ϫ4981.
J. Staroscik, B. Rickborn, J. Am. Chem. Soc. 1971, 93,
3046Ϫ3047.
D. M. Wieland, C. R. Johnson, J. Am. Chem. Soc. 1971, 93,
3047Ϫ3049.
G. Stork, C. Kowalski, G. Garcia, J. Am. Chem. Soc. 1975,
97, 3258Ϫ3260.
G. A. Crosby, R. A. Stephenson, J. Chem. Soc., Chem. Com-
7750Ϫ7756.
[15]
D. Seebach, H.-O. Kalinowski, W. Langer, C. Gerhard, E.-M.
Wilka, Org. Synth. 1983, 61, 24Ϫ34.
B. L. Feringa, A. H. M. de Vries, Adv. Catal. Processes 1995,
1, 151Ϫ192.
K. M. Sadhu, D. S. Matteson, Tetrahedron Lett. 1986, 27, 795.
G. Boireau, D. Abenhaim, E. Henry-Basch, Tetrahedron 1980,
36, 3061Ϫ3070.
J. K. Crandall, D. B. Banks, R. A. Colyer, R. J. Watkins, J. P.
Arrington, J. Org. Chem. 1968, 33, 423Ϫ425.
M. Lautens, M. L. Maddess, E. L. O. Sauer, S. G. Ouellet, Org.
Lett. 2002, 4, 83Ϫ86.
R. L. Danheiser, C. Martinez-Davila, H. Sard, Tetrahedron
1981, 37, 3943Ϫ3950.
M. Ito, M. Matsuumi, M. G. Murugesh, Y. Kobayashi, J. Org.
Chem. 2001, 66, 5881Ϫ5889.
J. J. Partridge, N. K. Chadha, M. R. Uskovic, J. Am. Chem.
Soc. 1973, 95, 532Ϫ540.
Y. Ohfune, M. Tomita, J. Am. Chem. Soc. 1982, 104,
3511Ϫ3513.
[16]
[17]
[32]
[33]
[18]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[19]
mun. 1975, 287Ϫ288.
A. J. Briggs, K. A. M. Walker, J. Org. Chem. 1990, 55,
2962Ϫ2964.
[20]
[21]
J. J. Eshelby, P. J. Parsons, P. J. Crowley, J. Chem. Soc., Perkin
Trans. 1 1996, 191.
Y. Naruta, K. Maruyama, Chem. Lett. 1987, 963Ϫ966.
M. Zaidlewicz, M. P. Krzeminski, Org. Lett. 2000, 2,
3897Ϫ3899.
[22]
[23]
[24]
S. Xue, Y. Li, K. Han, W. Yin, M. Wang, Q. Guo, Org. Lett.
2002, 4, 905Ϫ907.
N. Bernard, F. Chemla, J. F. Normant, Tetrahedron Lett. 1998,
39, 6715.
C. Hirai, Y. Sakai, T. Matsumoto, Yakagatu 1966, 15,
[25]
306Ϫ310.
W. E. Parham, H. E. Holmquist, J. Am. Chem. Soc. 1954, 76,
1173Ϫ1175.
[26]
A. Alexakis, E. Vrancken, P. Mangeney, F. Chemla, J. Chem.
Soc., Perkin Trans. 1 2000, 3352Ϫ3353.
Received December 3, 2003
Eur. J. Org. Chem. 2004, 2151Ϫ2159
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2159