Hydroxy direction of zinc carbenoid addition to a remote olefin. Analysis and improvement of 2,4-pentanediol tethered Furukawa cyclopropanation

Article abstract of DOI:10.1016/S0040-4020(01)00727-X

Zinc carbenoid addition to an olefin stereocontrolled by a chiral and remote intramolecular hydroxy group was studied in detail. Stereoselectivity as well as chemical yield of the Furukawa cyclopropanation was dramatically improved by modifications of the

Full text of DOI:10.1016/S0040-4020(01)00727-X

TETRAHEDRON
Pergamon
Tetrahedron 57 42001) 7495±7499
Hydroxy direction of zinc carbenoid addition to a remote ole®n.
Analysis and improvement of 2,4-pentanediol tethered
Furukawa cyclopropanation
Takashi Sugimura,p Tohru Futagawa, Masato Yoshikawa, Toshifumi Katagiri,
Ryoso Miyashige, Miki Mizuguchi, Shinya Nagano, Seiji Sugimori, Akira Tai,
Takahiro Tei and Tadashi Okuyama
Faculty of Science, Himeji Institute of Technology, 3-2-1 Kohto, Kanaji, Kamigori, Ako-gun, Hyogo 678-1297, Japan
Received 15 June 2001; accepted 13 July 2001
AbstractÐZinc carbenoid addition to an ole®n stereocontrolled by a chiral and remote intramolecular hydroxy group was studied in detail.
Stereoselectivity as well as chemical yield of the Furukawa cyclopropanation was dramatically improved by modi®cations of the procedure
of addition of the reagents. The structure of an active species to give the hydroxy-directed product is proposed. q 2001 Elsevier Science Ltd.
All rights reserved.
1. Introduction
alcohols can be activated by a Lewis acid, but the effective-
ness of this method for homoallylic or other remote ole®nic
alcohols is unknown. Stereoselective zinc carbenoid addi-
tion to 1 shown in Scheme 1 is an example of the entropi-
cally unfavorable hydroxy-directed reaction, where an
ole®n of the reaction site is separated by ®ve bonds from
the hydroxy group.7 Stereoselectivity of the reaction is
sensitive to the reaction conditions, and the product yield
is easily reduced under inadequate conditions due to conver-
sion of 1 to the corresponding acetal.8 In this study, effects
of reaction conditions on the stereoselectivity and the yield
of the products 2 and 3 in the carbenoid reaction of 1 were
carefully analyzed, and the dramatic improvements could be
achieved based on these analyses.9
Hydroxy group-directed zinc carbenoid addition to an ole®n
is a promising method for constructing a cyclopropane ring
under strict regio- and diastereo-control.1 The reactions can
be performed easily just by adding diiodomethane to a
mixture of the substrate and activated zinc 4Simmons±
Smith reaction2) or diethylzinc instead of activated zinc
4Furukawa procedure3). However, when the hydroxy
group to direct the cyclopropanation exists at the remote
position from the reaction site of an ole®n, the selectivities
depend much on the reaction conditions due to the compe-
titive intermolecular reactions of the uncoordinated carbe-
noids. Since zinc carbenoids produced by those procedures
are mixtures of a variety of species having varying reactiv-
ities4 and the proportion of the active species is changing
during the reaction, controlling of the reaction species is
vital to achieve high selectivity of the reaction. Especially
in the cases of remote hydroxy-directed cyclopropanations
of entropically unfavorable quasi-intramolecular process,
the intermolecular reactions proceed easily and controlling
of the reaction species becomes critical. Although several
modi®ed Furukawa procedures have recently been disclosed
to improve activity of zinc carbenoids to unfunctionalized
ole®ns,5 stereoselective cyclopropanation by the remote hy-
droxy direction has not been studied well. Charret and co-
workers6 reported that the Furukawa procedure for allylic
2. Results and discussions
2.1. Relation between the selectivity and the conversion
Stereoselectivity of cyclopropanation of 1 by the Furukawa
procedure largely depends on the solvent employed.8 The
reactions in some ethers give high diastereomeric excess of
Keywords: asymmetric synthesis; zinc carbenoid; cyclopropanation; stereo-
control.
p
Corresponding author. Tel.: 181-791-58-0168; fax: 181-791-58-0132/
0115; e-mail: sugimura@sci.himeji-tech.ac.jp
Scheme 1.
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020401)00727-X

7496
T. Sugimura et al. / Tetrahedron 57 *2001) 7495±7499
unexplainable change of the de that remains low throughout
the reaction.
These results indicate that there are at least two reaction
species to give the opposite stereoselectivity. So, if one
desirable species can be effectively generated, the high de
as well as the high product yield may be achieved.
2.2. Modi®cations of the Furukawa procedure
During the reaction according to the Furukawa procedure,
the reaction species changes in two ways; formation of the
zinc carbenoid and its consumption or decomposition. To
separate these two factors, we ®rst investigated the reaction
with pre-formed carbenoid reagents. By mixing a 1/1 molar
ratio of diethylzinc and diiodomethane, a species expressed
as EtZnCH2I is expected to be produced, and the 1/2 mixture
gives Zn4CH2I)2. The reactions were carried out by adding
the pre-formed carbenoid to a solution of 1. Results of the
reactions in diethyl ether, THF, and DME are summarized
in Table 1.
Figure 1. Diastereomeric excess of 2 and 3 as a function of conversion of 1
during the reaction in DME with the reagents at the ratio 4diethylzinc/
diiodomethane) of 5/10 4W), and in diethyl ether at 5/10 4X), at 5/5 4A),
and at 5/2 4B) according to the Furukawa procedure.
the products 2 and 3; 94% de in THF, 94% de in dioxane,
and 95% de in DME at room temperature ꢀ%de ˆ 100ꢀ2 2
3†=ꢀ2 1 3††: The yields under these conditions are moderate
in a range of 56±65% even with the excess reagents;
5 equiv. of diethylzinc and 10 equiv. of diiodomethane
4abbreviated as 5/10). On the other hand, use of non-ethereal
solvent lowers the de to 41% 4hexane), 49% 4benzene), and
14% 4dichloromethane). Surprisingly, the selectivity in
diethyl ether becomes lowest and even reversed 428% de,
73% yield). The present study started with observation of
this anomalous solvent dependency.
All the values of de given in entries 1±6 are as high as 93±
96%, and the low value of the de in diethyl ether by the
original Furukawa procedure was dramatically improved.
The solvent dependency of the de was thus solved, and
the formation process of the carbenoid was found to be
important. Although the high de was achieved in all ethereal
solvents, the product yields are not satisfactory even with
5 equiv. of the carbenoid species. We assumed that this is
due to the decomposition of the carbenoids after their
formation.10 As the chemical yields are always high with
the 1/1 mixture, EtZnCH2I seems to be more stable than
Zn4CH2I)2. When the pre-formed EtZnCH2I was added
dropwise until the substrate was consumed, the de or the
chemical yield was not improved, but amounts of the
reagents could be reduced in diethyl ether 4entries 7±9).
When the reactions were interrupted before the completion
of reaction, the de was found to change with conversion.
The de vs. conversion relationships shown in Fig. 1 are
different depending on the solvent and amounts of the
reagents. In DME, the de was kept high during the reaction
under the 5/10 conditions, and the de was still higher at
lower conversions; e.g. 99% de at 5% conversion. In diethyl
ether, the de decreases toward zero as the reaction proceeds
under the same 5/10 conditions. This experiment suggests
that the de in diethyl ether is high with a speci®c reaction
species formed in the early stage of reaction. The high de of
over 99% was in fact observed under the 5/5 conditions in
diethyl ether at 21% conversion. In contrast, the use of a less
amount of diiodomethane 45/2 conditions) resulted in an
To achieve the high chemical yield while maintaining the
high de, decomposition of the active carbenoid should be
minimized, because it provides other zinc species, which
may result in a change in the coordination mode of the
zinc carbenoid and in turn decomposition of 1. Our ®nal
modi®cation of the Furukawa procedure is a slow addition
of diiodomethane to a mixture of diethylzinc and 1. The
reaction was monitored by TLC and the addition of diiodo-
methane was stopped at the consumption of 1 when the
Table 1. De and yield of the product in the reaction of pre-formed carbenoids
Entry
Solvent
Et2Zn 4equiv.)
CH2I2 4equiv.)
Reaction time 4h)a
Addition time 4h)b
de 4%)
Yield 4%)
1
2
3
4
5
6
7
8
9
10
OEt2
OEt2
THF
THF
DME
DME
OEt2
OEt2
OEt2
THF
5
5
5
5
5
5
2
3
2
4
10
5
10
5
10
5
2
3
2
4
1.5
1.0
1.5
3.5
2.0
1.5
±
±
±
±
±
±
±
±
±
95
96
90
96
93
95
96
97
97
96
59.5
62.6
14.5
56.0
18.6
30.7
31.6
53.6
51.3
36.3
±
0.5
1.0
2.0
18.0
a
The pre-formed carbenoid was added within a few minutes and stirring was continued during the reaction time.
The pre-formed carbenoid was added dropwise during the addition time.
b

T. Sugimura et al. / Tetrahedron 57 *2001) 7495±7499
Table 2. The product yields and de's obtained by the slow addition of diiodomethane
7497
Entry
Solvent
Et2Zn 4equiv.)
CH2I2 4equiv.)
Addition time 4h)
de 4%)
Yield 4%)
1
2
3
4
5
6
7
8
9
OEt2
OEt2
OEt2
DME
THF
Hexane
Hexane
Hexane
CH2Cl2
4
2
1
4
2
4
2
1
4
3
1
1±2
3
1±3
3
1
6.5
1.0
1.0±2.0
3.5
1.0±3.0
4.5
1.0
1.0
2.0
96
97
±
94
±
86
81
±
92.4
61.5
0
63.6
0
55.1
75.1
0
1
3
±
0
cycloaddition occurred. The results are summarized in
Table 2.
2.4. Cyclopropanation of hydroxy-protected 1
The reaction selectivity of 1 in diethyl ether under the origi-
nal Furukawa conditions shown in Fig. 1 indicates that, in
the later stage, the selectivity is reversed and 3 is produced
as a major product. In this stage, the quasi-intramolecular
reaction species might be deactivated and intermolecular
reaction to produce 3 may show up. The selectivity of the
intermolecular reaction is studied with two analogous
substrates, 4 and 5, where the hydroxy group of 1 was
protected with acetyl and t-butyldimethylsilyl groups,
respectively. The reaction was carried out under the Furu-
kawa and the Simmons±Smith conditions. The results are
shown in Table 3. The results with dichlorocarbene are also
given for comparison.
Contrasting solvent effects were observed in this procedure.
Although none of the products were obtained in dichloro-
methane or THF, moderate yields of the products were
attained even with minimum amounts of diethylzinc
42 equiv.) and diiodomethane 41 equiv.) in diethyl ether
and hexane 4entries 2 and 7). The best results in terms of
de and yield were achieved by using somewhat more but
still reasonable amounts of the reagents in diethyl ether:
96% de and 92% yield as in the case of entry 1. The solvent
dependency on the yield with the minimum reagents indi-
cates that zinc carbenoid produced by this method reacts
highly selectively but is not active enough in a polar solvent,
THF.
The reactions of the acetyl-protected substrate 4 4entries 1±
6) again showed a characteristic change among the solvents
employed. Higher values of the de in non-polar solvents
may suggest occurrence of a weak interaction between the
acetyl group and zinc carbenoid. The reversal of the de was
observed in THF 4entries 4 and 6), but the degree is lower
than values expected from the results shown in Fig. 1. The
high de in the opposite direction was observed with the
TBS-protected 5 up to 277% 4in hexane, entry 7). With
this substrate, the reaction did not proceed at all in ethereal
solvents by either of the methods, and the reactant 5 was
recovered quantitatively. The stereochemical switching by
the TBS protection was also observed in the reaction with
dichlorocarbene, which is expected to be weakly directed by
the hydroxy group.12
2.3. Active species for the hydroxy-directed
cyclopropanation
The success in the minimization of the reagents for the
stereoselective addition prompted us to characterize the
reaction species in the quasi-intramolecular addition. A
minimum unit of the species is expected to consist of two
organozinc and one methylene group. When 1 was added to
a 1/1 mixture of diethylzinc and diiodomethane in diethyl
ether, the reaction did not occur at all but addition of another
equivalent of diethylzinc to this mixture promoted the reac-
tion to give 41.4% yield of the product in 96% de. Another
experiment was carried out by starting with reaction of 1
with sodium hydride and then adding zinc iodide to give a
species like ROZnI. To this mixture, 1 equiv. each of
diethylzinc and diiodomethane were added to result in
45.6% yield of the product of 96% de. These experiments
support that the above minimum required unit is essential of
the quasi-intramolecular addition. The active species in the
present system can be represented as shown in Fig. 2.11
The selectivity switching due to the hydroxy group protec-
tion was experimentally demonstrated with the Furukawa
procedure. However, the de dependency in the reaction of 5
on the solvent is not the same as that in the later stage of the
reaction of 1. This may exhibit an additional role of zinc
incorporated in 1: a deactivated zinc species placed at the
hydroxy group has an ability to coordinate to the vinylic
ether oxygen to ®x the conformation of the pentanediol
tether.
3. Conclusion
The slight modi®cation of the Furukawa procedure
improved cyclopropanation of 1 to give the high de irrespec-
tive of the solvent used. This improvement makes possible
to apply the present stereoselective reaction to syntheses of
optically active compounds irrespective of the solubility and
Figure 2. Possible structures of active species for the intramolecular carbe-
noid reaction.

7498
T. Sugimura et al. / Tetrahedron 57 *2001) 7495±7499
Table 3. Reactions of hydroxy-protected substrates 4 and 5
Entry
Substrate
Conditionsa
Solvent
Temperature 48C)
Time 4h)
de 4%)
Yield 4%)
1
2
3
4
5
6
7
8
4
4
4
4
4
4
5
5
5
5
5
5
1
5
5
F
F
F
F
SS
SS
F
F
F
Hexane
CH2Cl2
OEt2
THF
OEt2
0
0
0
1.5
1
4
75
86
47
218
36
211
277
235
±
±
±
225
16
246
235
80.1
89.8
21.0
29.8
38.4
18.1
20.7
69.5
0
0
0
12.0
43.0
97.0
34.2
20
40
60
20
0
20
20
40
60
0
65
2.5
4.5
20
1
15
15
7.5
50
1
THF
Hexane
CH2Cl2
OEt2
THF
OEt2
9
10
11
12
13
14
15
F
SS
SS
:CCl2
:CCl2
:CCl2
THF
CHCl3
CHCl3
Hexaneb
0
0
1
2
a
F: The Furukawa procedure with 5 equiv. of Et2Zn and 10 equiv. of CH2I2. SS: The Simmons±Smith reaction with 3 equiv. each of Zn/Cu and CH2I2. :CCl2:
Reaction of dichlorocarbene generated from CHCl3 and NaOH/Bu4NCl in CHCl3.
10% 4w/w) of CHCl3 in hexane.
b
reactivity of the substrate as well as incorporation of addi-
ethyl acetate in hexane) gave a colorless oil 4168 mg, 92.4%
yield).
tional ethereal functionalities in the substrate. We also
found the minimum unit of the hydroxy-directed species,
which was not found during study of the common entropi-
cally favorable systems. In those cases, formation of various
carbenoids does not affect much the results due to the quick
addition of the hydroxy-directed species. Coordination or
reaction of the hydroxy group donates electrons to the
zinc atom to reduce the electrophilicity of the zinc carbe-
noid, which retards the cyclopropanation. The present
results reasonably explain how to re-activate the carbenoid
to minimize the intermolecular side reaction. The present
study also indicates that survey of the reaction solvent is
important to ®nd optimized conditions when a hydroxy
group to direct zinc carbenoid is at a remote position of
the reaction site.
4.2.1. Preparation of 4. A mixture of 1 4498 mg,
2.71 mmol) and acetic anhydride 40.8 ml, 8.48 mmol) in
anhydrous pyridine 42 ml) was stirred for 3 h at room
temperature. After addition of water, the mixture was
extracted with diethyl ether 420 ml£2) and washed with
saturated aqueous copper sulfate 420 ml£4). The organic
layer was dried over magnesium sulfate, concentrated, and
puri®ed by silica gel column chromatography 4elution with
20% ethyl acetate in hexane) to give 390 mg of a colorless
oil 463.7% yield). [a]D ˆ242.8 4c 1.1, methanol), IR 4neat,
20
KCl) 2980, 2940, 2850, 1740, 1670, 1380, 1240,
1
1180 cm21; H NMR 4CDCl3) 5.12 4m, 1H), 4.59 4m, 1H),
4.12 4m, 1H), 2.06±1.92 4m, 4H), 2.00 4s, 3H), 1.79±1.48
4m, 6H), 1.23 4d, Jˆ6.4 Hz, 3H), 1.19 4d, Jˆ6.4 Hz, 3H);
Anal. Found: C, 69.04; H, 9.51. Calcd for C13H22O3: C,
69.30; H, 9.40.
4. Experimental
4.1. General
4.2.2. Preparation of 5. To a mixture of t-butyldimethyl-
silyl chloride 42.40 g, 15.9 mmol), imidazole 42.44 g,
35.8 mmol), and a catalytic amount of 4-dimethylaminopyr-
idine in DMF 420 ml), 1 41.65 g, 8.97 mmol) was added in
5 min at room temperature. After stirring for 30 min, the
mixture was poured into a saturated aqueous sodium bicar-
bonate solution 450 ml) and extracted with diethyl ether
480 ml£3). The organic layer was dried over sodium sulfate
and concentrated to give a yellow oil 44.21 g), which was
puri®ed by silica gel column chromatography 4elution with
2% ethyl acetate in hexane) to afford 2.58 g of a colorless oil
All temperatures are uncorrected. 1H NMR 4400 MHz)
spectra were recorded on a JEOL GX-400 spectrometer in
CDCl3 or C6D6 using residual solvent peak as an internal
standard 47.24 or 7.10 ppm). Analytical GLC of products 2
and 3 was conducted by the reported method8 with a capil-
lary column 4PEG-20M, 0.25 mm i.d. 50 m). Diethylzinc
was purchased from Kanto Kagaku Co. as a hexane solution
41.04 M). All solvents were puri®ed by distillation in the
presence of a proper dehydrating agent. All reactions were
carried out under a dry nitrogen atmosphere.
496.0% yield). [a]D ˆ259.8 4c 0.9, methanol), IR 4neat,
20
KCl) 2925, 2850, 1670, 1370, 1250, 1180, 1060, 840,
1
770 cm21; H NMR 4C6D6) 4.72 4m, 1H), 4.41 4m, 1H),
4.2. Cyclopropanation of 1 resulting in high yield and
high de -entry 1 in Table 2)
4.36 4m, 1H), 4.20 4m, 1H), 2.20 4m, 1H), 2.11 4m, 1H),
1.75±1.52 4m, 6H), 1.20 4d, Jˆ6.1 Hz, 3H), 1.13 4d,
Jˆ6.1 Hz, 3H), 1.06 4s, 9H), 1.00±0.98 4m, 2H), 0.17±
0.14 4m, 6H); Anal. Found: C, 68.29; H, 11.38. Calcd for
C17H34O2Si: C, 68.39; H, 11.48.
To a solution of 1 4169.1 mg) in dry diethyl ether 430 ml), a
hexane solution of diethyl zinc 43.7 ml) was added at room
temperature 4208C). To this mixture, a solution of diiodo-
methane 40.24 ml) in diethyl ether 460 ml) was added drop-
wise in 6.5 h. The resulting mixture was poured into a
saturated aqueous solution of ammonium chloride. A regu-
lar workup and silica gel chromatography 4elution with 20%
4.3. Dichlorocarbene addition to 1
To a solution of 1 493 mg, 0.51 mmol) and triethylbenzyl-

T. Sugimura et al. / Tetrahedron 57 *2001) 7495±7499
7499
ammonium chloride 436 mg, 0.15 mmol) in chloroform
45 ml) was added an aqueous solution of sodium hydroxide
450%, 4 g) in 15 min at 08C. The mixture was allowed to
stand for 30 min at the same temperature, and was extracted
with dichloromethane 410 ml£4) after addition of water
410 ml). The organic layer was dried over sodium sulfate,
and then concentrated to give 161 mg of a yellow oil. This
was puri®ed by silica gel column chromatography 4elution
with 10% ethyl ether in benzene) to give 57.5 mg of a color-
less oil 4a mixture of the diastereomers, 42.7% yield). The
isomeric ratio of this mixture was determined by the capil-
lary GLC analysis after dehalogenation. Dehalogenation
was carried out with sodium in a mixture of diethyl ether,
methanol, and water at 08C to give a mixture of 2 and 3.
4dd, Jˆ6.3, 5.4 Hz, 1H); MS Found: m/z 198.1611. Calcd
for C12H22O2: 198.1614. The diastereomeric excess of the
product from the dihalocarbene addition to 5 was deter-
mined by the GLC analysis after converting the adduct to
a mixture of 2 and 3 by the same procedure without puri®-
cation.
References
1. Selected examples: 4a) Dauben, E. G.; Berezin, G. H. J. Am.
Chem. Soc. 1963, 85, 468. 4b) Sawada, S.; Takehana, K.;
Inouye, Y. J. Org. Chem. 1968, 33, 1767. 4c) Staroscik,
J. A.; Rickborn, B. J. Org. Chem. 1972, 37, 738. 4d) Kawabata,
N.; Nakagawa, T.; Nakao, T.; Yamashita, S. J. Org. Chem.
1977, 42, 3031. 4e) Charette, A. B.; Marcoux, J. F. J. Am.
Chem. Soc. 1996, 118, 4539. 4f) Baba, Y.; Saha, G.; Nakao,
S.; Iwata, C.; Tanaka, T.; Ibuka, T.; Ohashi, H.; Takemoto, Y.
J. Org. Chem. 2001, 66, 81. 4g) Hoveyda, A. H.; Evans, D. A.;
Fu, G. C. Chem. Rev. 1993, 93, 1307.
4.3.1. Dichlorocarbene addition to 5. The substrate 5 was
treated in the same way as 1. In this case, the two isomers
were separated by silica gel column chromatography
4elution with 1% ethyl acetate in hexane). The minor isomer
4the same stereochemistry as 2) was obtained in 19.3%
yield: [a]D ˆ238.9 4c 1.0, methanol); IR 4neat, KCl)
2. 4a) Simmons, H. E.; Smith, R. D. J. Am. Chem. Soc. 1958, 80,
5323. 4b) Simmons, H. E.; Smith, R. D. J. Am. Chem. Soc.
1959, 81, 4256.
20
2950, 2850, 1200, 840, 780 cm21 1H NMR 4CDCl3)
;
4.02±3.91 4m, 2H), 2.18±1.98 4m, 3H), 1.68 4m, 1H), 1.62
4m, 1H), 1,55±1.49 4m, 2H), 1.46±1.32 4m, 4H), 1.25 4d,
Jˆ6.3 Hz, 3H), 1.12 4d, Jˆ6.1 Hz, 3H), 0.89 4s, 9H), 0.06
4s, 6H). MS Found: m/z 365.1448. Calcd for C17H31Cl2O2Si
4M12CH3): 365.1470. The major isomer 4the same stereo-
3. 4a) Furukawa, J.; Kawabata, N.; Nishimura, J. Tetrahedron
Lett. 1966, 3353. 4b) Furukawa, J.; Kawabata, N.; Nishimura,
J. Tetrahedron 1968, 24, 53.
4. For reviews, see: 4a) Simmons, H. E.; Cairns, T. L.; Vladu-
chick, S. A.; Hoiness, C. M. Org. React. *N.Y.) 1973, 20, 1.
4b) Charette, A. B. A practical approach. In Organozinc
Reagents, Knochel, P., Jones, P., Eds.; Oxford University:
Oxford, 1999; pp. 263±283.
chemistry as 3) was obtained in 44.3% yield: [a]D ˆ25.35
20
4c 1.1, methanol); IR 4neat, KCl) 2950, 2850, 1200, 840,
780 cm21; 1H NMR 4CDCl3) 3.97±3.87 4m, 2H), 2.13±2.09
4m, 2H), 2.03 4m, 1H), 1.82±1.76 4m, 2H), 1.69±1.58 4m,
2H), 1.47±1.36 4m, 3H), 1.27 4m, 1H), 1.20 4d, Jˆ6.1 Hz,
3H), 1.15 4d, Jˆ6.1 Hz, 3H), 0.89 4s, 9H), 0.07 4s, 6H). MS
Found: m/z 365.1458. Calcd for C17H31Cl2O2Si 4M12Me):
365.1470. To a solution of the major isomer 4227 mg,
0.71 mmol) in diethyl ether 428 ml) was added a small
piece of sodium 4ca. 100 mg) at 08C. To this solution, a
mixture of methanol and water 4100/3.3) was added drop-
wise. The additions of sodium and methanol±water were
repeated until the reactant was not detected on a TLC analy-
sis. The mixture was concentrated in vacuo, extracted with
diethyl ether 4100 ml£3) after addition of water 4100 ml),
washed with a saturated aqueous sodium chloride solution,
and dried over sodium sulfate. Desilylation of the obtained
mixture by the treatment with tetrabutylammonium ¯uoride
in a mixture of THF and triethylamine 42/1) and puri®cation
by silica gel column chromatography 4elution with 15%
ethyl acetate in hexane) afforded diastereomerically pure 3
5. 4a) Denmark, S. E.; Edwards, J. P. J. Org. Chem. 1991, 56,
6974. 4b) Yang, Z. Q.; Lorenz, J. C.; Shi, Y. Tetrahedron Lett.
1998, 39, 8621. 4c) Charette, A. B.; Francoeur, S.; Martel, S.;
Wilb, N. Angew. Chem., Int. Ed. Engl. 2000, 112, 4539.
6. Charette, A. B.; Brochu, C. J. Am. Chem. Soc. 1995, 117,
11367.
7. For other remote hydroxy direction system, see: Charette,
A. B.; Marcoux, J. F. Tetrahedron Lett. 1993, 34, 7157.
8. 4a) Sugimura, T.; Futagawa, T.; Tai, A. Tetrahedron Lett.
1988, 29, 5775. 4b) Sugimura, T.; Yoshikawa, M.; Futagawa,
T.; Tai, A. Tetrahedron 1990, 46, 5955.
9. Preliminary report: Sugimura, T.; Yoshikawa, M.; Mizuguchi,
M.; Tai, A. Chem. Lett. 1999, 831.
10. Reviews for the reactivity of zinc carbenoid other than
cycloaddition to a ole®n: 4a) Furukawa, J.; Kawabata, N.
Advances in Organometallic Chemistry; Stone, F. G. A.,
West, R., Eds.; Academic: New York, 1974; Vol. 12,
pp. 83±134. 4b) Boersma, J. In Comprehensive Organometal-
lic Chemistry, Wilkinson, G., Ed.; Pergamon: New York,
1982; pp. 823±862.
4112 mg, 79.9%). Compound 3: [a]D ˆ269.0 4c 1.0,
20
methanol); IR 4neat, KCl) 3450, 2975, 2950, 2875 cm21
;
1H NMR 4CDCl3) 4.03±3.96 4m, 2H), 2.85 4broad s, 1H),
2.10±1.91 4m, 3H), 1.64 4broad s, 1H), 1.60±1.39 4m, 5H),
1.23±1.22 4d, Jˆ6.4 Hz, 3H), 1.16±1.14 4d, Jˆ6.4 Hz, 3H),
1.26±1.15 4m, 2H), 0.92 4ddm, Jˆ10.7, 5.4 Hz, 1H), 0.26
11. Nakamura, E.; Hirai, A.; Nakamura, M. J. Am. Chem. Soc.
1998, 120, 5844.
12. Fariborz, M.; Still, W. C. Tetrahedron Lett. 1986, 27, 893.