The unusual 1,4-chelation-controlled nucleophilic addition to aldehydes with high stereoselectivity. A systematic study of stereoselectivity in the addition reaction of carbon nucleophiles to cis-substituted cyclopropanecarbaldehydes

Article abstract of DOI:10.1016/j.tet.2004.05.063

The addition reaction of carbon nucleophiles to cis-substituted cyclopropanecarbaldehydes was systematically investigated. Ab initio calculations of model cyclopropanecarbaldehydes suggested that the bisected s-cis and s-trans conformers are the only two minimum energy conformers, which are stabilized due to the π-donating stereoelectronic effect of the cyclopropane ring. The experimental results of a series of substrates, that is, cyclopropanecarbaldehydes 1-5 bearing a cis-(tert-butyldiphenylsilyloxy)methyl group, a cis-benzyloxymethyl group, a cis-(p-methoxybenzyloxy)methyl group, cis-N,N-diethylcarbamoyl and trans-phenyl groups, and cis-(tert- butyldiphenylsilyloxy)methyl and trans-phenyl groups, respectively, showed that highly anti-selective Grignard additions could be realized. It turned out that it occurred via an unusual 7-membered 1,4-chelation-controlled pathway. Highly stereoselective Grignard addition via the chelation-controlled pathway occurred even in the reaction of the usually non-chelating silyl ether-type substrate 5. The results have great importance because the 1,4-chelation-controlled stereoselective addition reactions can indeed be realized. Under non-chelation conditions, the syn-products were produced with moderate stereoselectivity, which are likely to be formed via the bisected s-cis conformation-like transition state stabilized by the characteristic orbital interaction. These reactions, especially the chelation-controlled reaction, should be useful because of their t stereoselectivity and stereochemical predictability.

Full text of DOI:10.1016/j.tet.2004.05.063

Tetrahedron 60 (2004) 6689–6703
The unusual 1,4-chelation-controlled nucleophilic addition to
aldehydes with high stereoselectivity. A systematic study of
stereoselectivity in the addition reaction of carbon nucleophiles to
cis-substituted cyclopropanecarbaldehydes
Yuji Kazuta,a Hiroshi Abe,a Tamotsu Yamamoto,b Akira Matsudaa and Satoshi Shutoa
,
*
aGraduate School of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
bInstitute for Life Science Research, Asahikasei Pharma Co., Ohito-cho, Shizuoka 410-2321, Japan
Received 12 April 2004; accepted 17 May 2004
Available online 19 June 2004
Abstract—The addition reaction of carbon nucleophiles to cis-substituted cyclopropanecarbaldehydes was systematically investigated. Ab
initio calculations of model cyclopropanecarbaldehydes suggested that the bisected s-cis and s-trans conformers are the only two minimum
energy conformers, which are stabilized due to the p-donating stereoelectronic effect of the cyclopropane ring. The experimental results of a
series of substrates, that is, cyclopropanecarbaldehydes 1–5 bearing a cis-(tert-butyldiphenylsilyloxy)methyl group, a cis-benzyloxymethyl
group, a cis-(p-methoxybenzyloxy)methyl group, cis-N,N-diethylcarbamoyl and trans-phenyl groups, and cis-(tert-butyldiphenylsilyloxy)-
methyl and trans-phenyl groups, respectively, showed that highly anti-selective Grignard additions could be realized. It turned out that it
occurred via an unusual 7-membered 1,4-chelation-controlled pathway. Highly stereoselective Grignard addition via the chelation-controlled
pathway occurred even in the reaction of the usually non-chelating silyl ether-type substrate 5. The results have great importance because the
1,4-chelation-controlled stereoselective addition reactions can indeed be realized. Under non-chelation conditions, the syn-products were
produced with moderate stereoselectivity, which are likely to be formed via the bisected s-cis conformation-like transition state stabilized by
the characteristic orbital interaction. These reactions, especially the chelation-controlled reaction, should be useful because of their t
stereoselectivity and stereochemical predictability.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
Much attention has been focused on the stereoselective
synthesis of cyclopropane derivatives, because cyclo-
propanes are important as key fragments in many natural
products and as synthetic intermediates due to the ease of
their ring-opening.1–4 A cyclopropane ring is also useful for
Figure 1. Bisected s-cis and s-trans conformations of a,b-unsaturated
cyclopropanes.
restricting the conformation of biologically active
compounds in medicinal chemical studies.3
carbaldehydes by carbon nucleophiles8 – 10 have been
Cyclopropyl ketones and cyclopropanecarbaldehydes,
reported,11 and these stereoselectivities are thought to be
which are useful as synthetic intermediate for various
related to the characteristic stereoelectronic effect of the
biologically active compounds having a cyclopropane
cyclopropane ring. We most recently showed that stereo-
backbone, prefer the bisected s-trans and s-cis confor-
selective hydride reduction of cyclopropyl ketones is
mations, as shown in Figure 1, due to the characteristic
explained effectively by the bisected s-cis-transition-state
stereoelectronic effects of the cyclopropane ring.1a,b,5
reaction model.7
Highly stereoselective nucleophilic hydride reduction of
cyclopropyl ketones6–8 and addition to cyclopropane-
In the course of our studies using a cyclopropane ring for
conformational restriction of biologically active
Keywords: Aldehydes; Conformational analysis; Cyclopropanes;
compounds,3,4 we have found that nucleophilic attack of
Grignard reagents on the cyclopropanecarbaldehydes I
occurs highly stereoselectively to give the anti-product III
Diastereoselectivity; Grignard reactions.
*
Corresponding author. Tel.: þ81-11-706-3228; fax: þ81-11-706-4980;
e-mail address: shu@pharm.hokudai.ac.jp
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.05.063

6690
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
in high yield, as shown in Scheme 1A.4a–c,8 We initially
attributed these highly selective stereochemical results to
the bisected-conformational transition state stabilized by the
characteristic electron-donating effect of the cyclopropane
ring.8 However, further studies were needed to confirm the
reaction mechanism. Thus, in order to clarify the mode of
addition of carbon nucleophiles to cis-substituted cyclo-
propanecarbaldehydes in a general way and to develop a
versatile method for the highly stereoselective reactions, we
performed a systematic study of the addition reaction using
a series of cyclopropanecarbaldehyde substrates 1–5, the
structures of which are shown in Figure 2. Conformational
analysis of the substrates and their model compounds was
also performed to clarify the mode of nucleophilic addition
reaction to the cyclopropanecarbaldehydes. These studies
showed that the highly anti-selective Grignard addition to
cis-substituted cyclopropanecarbaldehydes occurs via an
unusual 7-membered 1,4-chelation-controlled pathway.
above, the addition of Grignard reagents to the cyclo-
propanecarbaldehydes I having a phenyl and a carbamoyl
group at the trans and cis positions, respectively, to the
formyl group on the cyclopropane ring, occurs highly
stereoselectively to produce the corresponding anti-alcohol
III as the major product (Scheme 1A).8 The stereochemical
results in the reaction were explained by the nucleophilic
attack occurring from the less hindered face of the substrates
in the bisected s-trans conformation, as shown in
Figure 3A.8 The previous studies showed that cyclo-
propanecarbaldehydes preferentially exist in the bisected
s-trans and the s-cis conformation,5 and the X-ray crystal-
lographic analysis of the aldehyde I (X¼Ph, Y¼Et) actually
Figure 3. Conceivable reaction pathways of the nucleophilic attack on
cyclopropanecarbaldehydes.
showed that it assumes the bisected s-trans conformation in
the solid state.8b
After publication of our results, Reiser and co-workers
reported very efficient reactions that nucleophilic addition
reaction to the 2,3-disubstituted cyclopropanecarbaldehyde
IV occurred highly stereoselectively (Scheme 1B).10
Because the substrate was unstable under basic conditions,
they investigated Lewis acid-promoted addition reactions
with organo-silicon nucleophiles and also nitroaldol
reactions, as summarized in Table 1. Surprisingly, the
stereochemical outcome was opposite to that in our above
described case. For example, when IV was subjected to the
reaction with CH2vCHCH2TMS/BF3·OEt2, the corre-
sponding syn-alcohol was produced almost exclusively.
They concluded that the stereoselectivity could be explained
by the Felkin–Anh model, as shown in Figure 3B. They
summarized the Felkin–Anh-type reactions in an excellent
review and suggested that the anti-selective stereocontrol in
our studies (Scheme 1A) would occur via a chelation-
controlled pathway.10b However, we initially thought this
unlikely since highly stereoselective addition to carbonyls
via such a 7-membered chelation-controlled pathway was
unknown.10b,13–15
Scheme 1. Nucleophilic additions to cis-cyclopropanecarbaldehydes
reported previously.
Figure 2. Structurally simplified cyclopropanecarbaldehydes as reaction
substrates.
2. Results and discussion
2.1. The previous studies
On the other hand, Dauben and co-workers reported that the
aldol reaction with a cyclopropanecarbaldehyde VII was
moderately stereoselective to give a syn/anti-mixture in a
5:1 ratio (Scheme 1C).9
Only limited examples of nucleophilic additions to
cyclopropanecarbaldehydes are known.8–10,12 As described

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6691
Table 1. Nucleophilic additions to the 2,3-di-substituted cyclopropanecarbaldehydes IV reported by Reiser and co-workers10
Substrate
Nucleophile
Promoter
Solvent
Temp (8C)
Yield (%)
Syn/anti
IV
IV
IV
IV
IV
TMSCH2CH2vCH
CH2vCH(Ph)OTMS
TMSCN
TMSCN
i-PrNO2 (solvent)
BF3·OEt2
BF3·OEt2
BF3·OEt2
none
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
—
278
278
278
25
92
71
99:1
99:1
9:1
10:1
99:1
94
100a
91b
Et3N
0
a
The corresponding O-TMS ether products were obtained.
The N-formyl group migrated to the hydroxyl.
b
As stated above, the stereochemical outcome of the
nucleophilic additions to the cis-substituted cyclopropane-
carbaldehydes is significantly different depending on the
structure of substrates and the reaction conditions used, and
therefore the reaction mechanisms have not been clearly
understood.
active form, since the stereochemistries of the resulting
secondary alcoholic products should be easily determined
by the modified Mosher’s method.16 The substrates 1 and 2
were synthesized from chiral epichlorohydrins according to
the method recently developed in our laboratory.3,7 The
O-PMB-protected substrate 3 was synthesized from
(S)-epichlorohydrin via the key intermediate 6, as shown
in Scheme 2.
2.2. Design and synthesis of the structurally simplified
substrates
The substrates used in the previous studies possessed
different functional groups, which could affect the stereo-
chemical results, at least to some extent. Furthermore, the
nucleophiles and the reaction conditions employed in these
studies were quite different. We planned to perform a
systematic study of the reactions using structurally simpli-
fied substrates to clarify the mode of nucleophilic addition
to the cyclopropanecarbaldehydes.
Thus, we designed the structurally simplified substrates,
the cyclopropanecarbaldehyde bearing a cis-(tert-butyl-
diphenylsilyl (TBDPS) oxy)methyl group (1) and its
O-benzyl- and O-p-methoxybenzyl (PMB)-protected
congeners 2 and 3, for investigating the reactions under
various conditions. We used these substrates in an optically
Scheme 2. Preparation of the O-PMB-protected substrate 3.
2.3. Conformational analysis of the substrates
The X-ray crystallographic structure of the O-TBDPS-
protected substrate 117 shows that 1 exists in the bisected
Figure 4. X-ray crystallographic structure of 1.

6692
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
s-trans conformation in the solid state (Fig. 4). The
conformation of 1 in solution was also investigated by
NOE experiments in CD2Cl2 at 278 8C. When the formyl
proton was irradiated, NOEs were observed at H-4 (2.7%),
attached to the position cis to the formyl moiety on the
cyclopropane ring, and at the silyloxymethylene proton
(2.2%) oriented on the concave side, as shown in
Figure 5A. An NOE (5.4%) was also observed at the formyl
proton by the irradiation of the concave proton in the
silyloxymethylene group (Fig. 5B). Thus, there is no
discrepancy between the NOE data and the results of the
X-ray analysis which would indicate that the cyclo-
propanecarboxaldehyde 1 prefers the bisected s-trans
conformation.
Figure 5. NOE experimental data of 1 (500 MHz, CD2Cl2, 278 8C).
Moreover, the conformation of the cyclopropanecarb-
aldehydes was studied using model compounds, that is,
cyclopropanecarbaldehyde (i) and cis23-methylcyclo-
propanecarbaldehyde (ii) (Fig. 6), by ab initio calculations
based on the density functional theory (DFT) using the
GAUSSIAN98 program.18 We calculated the rotational
barrier around the O1–C1–C2–H2 dihedral angle in the
model compounds. The dihedral angle was rotated from 0 to
3608 at 208 intervals, and the conformations were optimized
at RHF/3-21G(d). The single point energies of each
optimized conformer were calculated at RB3LYP/
6-231G(d). The results are shown in Figure 7. The minimum
energy values were observed around 08 (3608) and 1808 for i,
and around 08 (3608) and 1908 for ii, where they assume the
bisected s-trans and s-cis (or s-cis-like) conformations,
respectively. The maximum values are observed around the
angles at 90 and 2708 for both, where they are in
perpendicular conformations. The bisected conformers at
0 or 1808 (or 1908) are 6–7 kcal/mol more stable than the
perpendicular conformers at 90 or 2708 in the model
compounds.
Figure 7. Rotational barrier energies around the O1–C1–C2–H2 dihedral
angle of the model compounds i and ii.
corresponding s-trans conformer in cyclopropanecarbalde-
hyde (i), while the energy difference is insignificant
(D 0.42 kcal/mol). When a methyl group is introduced
into the position cis to the formyl group of i, that is, ii, the
s-cis-conformation becomes somewhat more stable
(D 1.25 kcal/mol). However, the energy difference between
the two bisected s-cis and s-trans conformers is small in
both i and ii, which is in accord with the previous
experimental and theoretical calculation results of the
cyclopropanecarbaldehydes.5h
2.4. Grignard reactions of the structurally simplified
substrates19
Figure 6. Cyclopropanecarbaldehydes as the model compounds for the
calculation studies.
Using the O-TBDPS-protected substrate 1, the Grignard
reactions were examined with Me–, Et–, i-Bu–, or
CH2vCHCH2MgBr in CH2Cl2 or THF at 278 8C
(Scheme 3), and the results are summarized in Table 2.
These reactions gave a mixture of the addition products
9a–d and 10a–d20,21 with small to moderate stereo-
selectivities. These data were clearly different from both
Next, the bisected s-cis and trans conformers of i and ii
were fully optimized at a higher level of theory using
RB3LYP/6-31G(d), and the obtained stable conformations
and their relative energies calculated by RB3LYP/6-31G(d)
are shown in Figure 8. The calculations showed that
while the s-cis conformer is more stable than the

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6693
the O-TBDPS-protected substrate 1 with MeMgBr were first
carried out in the presence of a variety of additives (entries
1–5). The reactions with BF3·OEt2, AlCl3, or HMPA as the
additive were almost non-stereoselective (entries 1, 2, and
5), similar to the above mentioned reaction without an
additive (Table 2, entry 1). However, addition of ZnBr2 or
TiCl4, which is known to form a chelate effectively with
oxygen atoms,13 to the Grignard reaction system produced
the anti-alcohol 10a as the major product with moderate
selectivity (entries 3 and 4, syn/anti¼1:3.1 and 1:2.8). When
the O-benzyl-protected substrate 2 was used, the reaction
selectively gave the anti-alcohol 12a20 without any
additives (entry 6, syn/anti¼1:4.3). By employing the
O-benzyl substrate 2 the effect of ZnBr2 as the additive
was further enhanced to improve the stereoselectivity (entry
7, syn/anti¼1:7.2). The anti-alcohol 12a seemed to be
produced via the chelation-controlled pathway, since the
anti-selectivity disappeared in the presence of HMPA (entry
8, syn/anti¼1.3:1). The chelation-controlled pathway was
supported by the results with the O-PMB-protected
substrate 3, where the anti-selectivity was further improved
(entry 9, without additive, syn/anti¼1:5.7; entry 10, with
ZnBr2, syn/anti¼1:9.3)22 compared with those of the
O-benzyl-substrate 2 (entries 6 and 7), probably due to
the effective chelate formation for the electron-donating
p-methoxy group. The reactions with EtMgBr or i-BuMgBr
instead of MeMgBr under the conditions using ZnBr2 as the
additive also stereoselectively produced the corresponding
anti-alcohols as the major product (entries 11–14); the
O-Bn-protected substrate 2 showed high stereoselectivity
(entries 12 and 14, syn/anti¼1:10 and 1:20),22 while the
selectivity in the O-TBDPS-protected substrate 1 was
moderate (entries 11 and 13, syn/anti¼1:3.0 and 1:5.9).
Figure 8. Structures of the minimum energy s-cis- and s-trans-conformers
and their relative energies obtained by the ab initio calculations of the
model compound i (A) and ii (B).
2.5. Lewis acid-promoted addition and nitroaldol
reaction of the structurally simplified substrates
Scheme 3. Nucleophilic additions to the cis-cyclopropanecarbaldehydes
1–3.
The Lewis acid-promoted addition and nitroaldol reaction,
which were employed in the previous studies by Reiser and
co-workers,10 were next examined. The reactions were
carried out using the O-TBDPS-protected substrate 1, and
the results are shown in Table 4. Treatment of substrate 1
with CH2vCHCH2TMS in the presence of BF3·OEt2 in
CH2Cl2 at 278 8C gave a mixture of the corresponding syn
and anti-products 9d and 10d in a 7.3:1 ratio (entry 1).
Similar treatment of 1 with CH2vC(Ph)OTMS as the
nucleophile was non-stereoselective (entry 2).21 These
results were in contrast to the highly syn-selective results
of the highly anti-selective additions in our previous results8
and the highly syn-selective ones of Reiser.10 It may be
important that the stereoselectivity changed depending on
the reaction solvent; slightly anti-selective in CH2Cl2
(entries 1–4, syn/anti¼1:1.3–1:2.0), but moderately syn-
selective in THF (entries 5–8, syn/anti¼2.3:1–4.2:1).
The effect of additives on the Grignard reaction was next
investigated in CH2Cl2 (Table 3). The Grignard reactions of
Table 2. Addition reactions of Grignard reagents to the cyclopropanecarbaldehyde 1: effect of solventa
Entry
Substrate
Pg
Nucleophile
Solvent
Time (h)
Products
Yield (%)
Syn/antib
1
2
3
4
5
6
7
8
1
1
1
1
1
1
1
1
TBDPS
TBDPS
TBDPS
TBDPS
TBDPS
TBDPS
TBDPS
TBDPS
MeMgBr
EtMgBr
i-BuMgBr
CH2vCHCHMgBr
MeMgBr
EtMgBr
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
THF
THF
THF
THF
5
5
8
8
3
3
8
8
9a, 10a
9b, 10b
9c, 10c
9d, 10d
9a, 10a
9b, 10b
9c, 10c
9d, 10d
95
77 (23)c
54 (42)c
81
94
84
1:1.3
1:1.5
1:2.0
1:1.7
2.3:1
2.8:1
4.2:1
2.5:1
i-BuMgBr
CH2vCHCHMgBr
99
87
a
Reaction was performed with 2 equiv. of Grignard reagent at 278 8C. A trace of other solvent, toluene/THF in entry 1, Et2O in entries 2, 4, 6 and 8, and THF
in entries 3 and 8, was included in the reaction system: see Section 3.
Determined by 1H NMR.
b
c
Yield of the substrate recovered is given in parenthesis.

6694
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
Table 3. Addition reactions of Grignard reagents to the cyclopropanecarbaldehydes 1–3: effect of additivesa
Entry
Substrate
Pg
Nucleophile
Additive
Time (h)
Products
Yield (%)
Syn/antib
1
2
3
4
5
6
7
8
1
1
1
1
1
2
2
2
3
3
1
2
1
2
TBDPS
TBDPS
TBDPS
TBDPS
TBDPS
Bn
Bn
Bn
PMB
PMB
TBDPS
Bn
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
EtMgBr
BF3·OEt2
AlCl3
ZnBr2
TiCl4
HMPA
none
ZnBr2
HMPA
none
ZnBr2
ZnBr2
ZnBr2
ZnBr2
ZnBr2
3
5
8
3
8
5
8
8
5
5
5
8
8
8
9a, 10a
9a, 10a
9a, 10a
9a, 10a
9a, 10a
11a, 12a
11a, 12a
11a, 12a
13a, 14a
13a, 14a
9b, 10b
11b, 12b
9c, 10c
85 (15)c
53 (37)c
99
1.4:1
1:1.5
1:3.1
1:2.8
1:1.1
1:4.3
1:7.2
1.3:1
1:5.7
1:9.3
1:3.0
1:10
98
82 (18)c
71 (23)c
92
58 (42)c
75 (19)c
95
9
10
11
12
13
14
80 (19)c
64 (30)c
86
EtMgBr
i-BuMgBr
i-BuMgBr
TBDPS
Bn
1:5.9
1:20
11c, 12c
91
a
Reaction was performed with 2 equiv. of Grignard reagent and 2 equiv. of additive in CH2Cl2 at 278 8C. A trace of other solvent, toluene/THF in entries 1–
10, Et2O in entries 11 and 12, and THF in entries 13 and 14, was included in the reaction system: see Section 3.
Determined by 1H NMR.
b
c
Yield of the substrate recovered is given in parenthesis.
Table 4. Lewis acid-promoted and nitroaldol-type nucleophilic addition reactions to the cyclopropanecarbaldehyde 1a
Entry
Substrate
Pg
Nucleophile
Promoter
Solvent
Temp (8C)
Time (h)
Products
Yield (%)
Syn/antib
1
2
3
1
1
1
TBDPS
TBDPS
TBDPS
CH2vCHCH2TMS
CH2vC(Ph)OTMS
i-PrNO2 (solvent)
BF3·OEt2
BF3·OEt2
Et3N
CH2Cl2
CH2Cl2
—
278
278
rt
48
2
340
9d, 10d
9e, 10e
9f, 10f
93
84 (4)c
80 (10)c
7.3:1
1:1.1
1:1.0
a
Reaction was performed with 8 equiv. of nucleophile and 3 equiv. of promoter (entry 1), 1 equiv. of nucleophile and 1 equiv. of promoter (entry 2), or
1 equiv. of promoter (entry 3).
b
c
Determined by 1H NMR.
Yield of the substrate recovered is given in parenthesis.
by Reiser and co-workers (Table 1, entries 1 and 2, syn/
Grignard reaction of the carbamoyl-type substrate 4 with
MeMgBr in the presence of excess (10 equiv.) of HMPA in
THF, the anti-selectivity was somewhat lower (entry 3, syn/
anti¼1:15) compared with reactions without HMPA
anti¼99:1). On the other hand, in the nitroaldol reaction of 1
with an i-PrNO2/Et3N the syn/anti ratio was changed
depending on the reaction time, and it gave a 1:1 syn/anti-
mixture of the addition product after 14 days. This is
because the stereochemical outcome in the nitroaldol
reaction is not kinetically controlled under the reaction
conditions.23
2.6. Reaction of the trans-phenyl-substituted substrates
We previously speculated that the highly anti-selective
additions of Grignard reagents to the cyclopropane-
carbaldehydes I would occur via the bisected s-trans-
pathway.8 However, the above-mentioned stereochemical
results with the structurally simplified cyclopropane-
carbaldehydes 1–3 in the present study suggest that the
anti-products would not be formed via the bisected s-trans
but rather via the chelation-controlled pathway, as
suggested by Reiser.10b To confirm the reaction mechanism,
we planned to examine the reactions further with the trans-
phenyl-substituted substrate 4 used in the previous studies8
and its related substrate 5, in which the carbamoyl group of
4 was replaced with a TBDPS–O–CH2-group. Substrate 5
was prepared from the known lactone 15, as shown in
Scheme 4.
Scheme 4. Preparation of the O-TBDPS-protected substrate 5.
The reactions were performed with MeMgBr or Me3Al as
the nucleophile and the effect of HMPA and ZnBr2 on the
reaction was investigated (Scheme 5). The results are
summarized in Table 5. The effect of HMPA on the
stereoselectivity of the reaction was first investigated. In the
Scheme 5. Nucleophilic additions to the cis-cyclopropanecarbaldehydes 4
and 5.

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6695
Table 5. Addition reactions of MeMgBr or Me3Al to the cyclopropanecarbaldehydes 4 and 5a
Entry
Substrate
R
Nucleophile
Additive
Solvent
Temp (8C)
Time (h)
Products
Yield (%)
Syn/antib
1c
2
3
4
5
6
7
8
4
4
4
4
4
4
4
5
5
5
CONEt2
CONEt2
CONEt2
CONEt2
CONEt2
CONEt2
CONEt2
CH2OTBDPS
CH2OTBDPS
CH2OTBDPS
MeMgBr
MeMgBr
MeMgBr
MeMgBr
MeMgBr
Me3Al
MeMgBr
MeMgBr
MeMgBr
MeMgBr
None
None
HMPA
HMPA
None
None
ZnBr2
None
None
ZnBr2
THF
CH2Cl2
THF
220
220
220
220
66
2
2
2
2
2
12
2
8
8
8
18, 19
18, 19
18, 19
18, 19
18, 19
18, 19
19
20, 21
20, 21
21
95
1:23
1:24
1:15
1:3.9
1:24
1:2.8
Anti only
2.8:1
1.3:1
67 (29)d
81 (10)b
77
THF
THF
CH2Cl2
THF
CH2Cl2
CH2Cl2
78 (19)b
62 (6)b
48 (49)b
85 (12)b
77 (14)b
93
66
220
278
278
278
9
10
Anti only
a
Reaction was performed with 2 equiv. of a nucleophile and 10 equiv. (entry 3) or 2 equiv (entries 7 and 10) of additive. A trace of other solvent, toluene/THF
in entries 2, 7, 9 and 10, and hexane in entry 6, was included in the reaction system: see Section 3.
Determined by 1H NMR.
The data were taken from Ref. 8b.
Yield of the substrate recovered is given in parenthesis.
b
c
d
Table 6. 13C NMR chemical shifts (d) of the substrate 4 in the absence or the presence of ZnBr2 (1 equiv.)a
C1
C2
C3
C4
CON
NCH2
CH3
Ph
4
198.3
b
37.4
37.4
0
40.7
41.4
20.7
20.0
24.1
24.1
167.7
171.9c
24.2
40.0, 42.0
43.9, 44.9
23.9, 22.8
12.5, 13.1
12.4, 12.7
þ0.1,þ0.4
126.3, 127.8, 129.3, 138.6
127.7, 129.3, 129.9, 135.0
4þZnBr2
—
—
Ddd
a
Measured at 100 MHz in CD2Cl2 at room temperature.
Not detected.
Observed as a broad peak.
d (4)2d (4þZnBr2).
b
c
d
(entries 1 and 2, syn/anti¼1:23 and 1:24). When HMPA was
used as the reaction solvent, the anti-selectivity clearly
decreased to give a syn/anti-mixture of 18 and 19 in a 1:3.9
ratio (entry 4). Reactions with Me3Al, a relatively
ineffective chelate-forming agent compared with Grignard
reagents, was next carried out. Although the reaction did not
proceed at room temperature, the addition products were
obtained with low anti-selectivity (entry 6, syn/anti¼1:2.8)
by heating 4 with Me3Al in THF under reflux. On the other
hand, similar heating of 4 with MeMgBr as the nucleophile
again gave the anti-product highly selectively (entry 5, syn/
anti¼1:24). When ZnBr2 was used as the additive in the
Grignard reaction, as in the above cases, the stereo-
selectivity effectively increased to give the anti-alcohol 19
as the sole product (entry 7).
shifts of the CO and the NCH2 signals of the carbamoyl
moiety of 4 were observed, which suggests that coordination
of Zn2þ to the carbamoyl CO occurs during the course of the
reaction.
The reactions of the silyloxymethyl-type substrate 5 were
next investigated. In contrast to the high-anti-selectivity in
the case of 4 (entries 1 and 2), the Grignard reaction of 5
without an additive was almost non-stereoselective (entries
8 and 9, syn/anti¼2.8:1 and 1.3:1). However, when ZnBr2
was present with MeMgBr in the reaction system, the
stereoselectivity was dramatically improved to furnish the
anti-product 21 exclusively in high yield (entry 10).21
We measured the 13C NMR spectrum of the carbamoyl-type
substrate 4 in the presence of 1 equiv. of ZnBr2 and
compared the data with those of 4 without the additive. As
shown in Table 6, in the presence of ZnBr2, clear down-field
Figure 9. Reaction mechanisms proposed for nucleophilic addition to
cyclopropanecarbaldehydes via the chelation-controlled pathway (A) and
the bisected conformation-controlled pathway (B).

6696
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
The above described results clearly show that in the case of
the carbamoyl-type substrates I (in Scheme 1A) the anti-
products are not formed via the bisected pathway but rather
via the unusual 7-membered 1,4-chelation-controlled
pathway, as shown in Figure 9A.
forming bond and the cyclopropane C–C bond are in an
almost planar arrangement. This kind of transition state
stabilization in nucleophilic attack on carbonyl carbons by
the antibonding orbital interaction is well known as the
Cieplak theory.26 Accordingly, we initially speculated that
it was possible for the nucleophilic addition to cyclo-
propanecarbaldehydes to occur via the attack on the less
hindered side of the bisected s-cis and/or s-trans-confor-
mation.
2.7. Discussion
In this study, the stereoselectivity in nucleophilic addition
reactions to the cis-substituted cyclopropanecarbaldehydes
was investigated experimentally in a systematic manner,
where a series of structurally simplified substrates was used
to make the results easy to be interpreted. The confor-
mations of the substrates, which are very important for
considering the mode of the reactions, were also analyzed.
Therefore, the results obtained from this study would
contribute to the understanding of the reaction pathways.
The stereoselectivity in the Grignard reaction of the
O-TBDPS-protected substrate 1 without an additive was
low, as summarized in Table 2. Ab initio calculations
showed that the bisected s-cis- (or s-cis like) and s-trans
conformers are quite stable compared with the other
conformations. Although the X-ray and NOE analyses of
the substrate 1 suggested that it might prefer the s-trans
conformation, ab initio calculations with the model
compounds in this study and the previous experimental
and theoretical calculation studies5 indicated that the energy
difference between the two conformers was rather small.
This may account for the low stereoselectivity in the
nucleophilic addition to cyclopropanecarbaldehydes under
non-chelation conditions. Nucleophilic attacks on carbonyl
carbons are known to occur at an angle of about 1098,27 and
the trajectory of the nucleophile to the s-trans conformer
could be interfering with the cyclopropane ring, as shown in
Figure 9B.10b Therefore, the reaction via the s-cis
intermediate might be preferred to give selectively the
syn-product 9a in the reactions using THF as the reaction
solvent (Table 2, entries 5–8), although the selectivity was
not high. In the reactions with non-Lewis basic CH2Cl2 as
the solvent, the anti-product via the chelation-controlled
pathway might have increased at least to some extent,
compared with the reactions in Lewis basic THF, to result in
the somewhat anti-selective outcome (Table 2, entries 1–4).
The Lewis acid-promoted allylation with CH2vCHCH2-
TMS of 1 occurred with considerable syn-selectivity (syn/
anti¼7.3:1, Table 4, entry 1), compared to the allylation
using the Grignard reagent (syn/anti¼1:1.7 and 2.5:1,
Table 2, entries 4 and 8). Although both CH2vCHCH2TMS
and CH2vCHCH2MgBr are known to react with aldehydes
at the g-position, the reaction transition state with the
former reagent would be more electron deficient.28 There-
fore, the bisected s-cis and s-trans transition states in the
Lewis acid-promoted allylation could be effectively
stabilized by the p-sp– hyperconjugation due to the strong
electron-donating feature of the cyclopropane ring, where
the nucleophilic attack to the s-cis would be preferred
because of the above described steric reason, as shown in
Figure 9B, to result in the syn-selective outcome.
The conformation of the transition state and the inter-
mediate can be strongly influenced by conformational
effects, which stabilize the ground state conformation.24 In
nucleophilic additions to cyclopropanecarbaldehydes, the
conformations of the transition state and the ground state
may be similar for the following characteristic stereo-
electronic reason. Cyclopropanecarbaldehydes prefer the
bisected s-cis and/or s-trans conformations. With the Walsh
model,25 which is effective for understanding the conjugat-
ing ability of the cyclopropane ring, this conformational
stability is explained by effective interaction between the
p-orbitals on the cyclopropyl carbon and the adjacent pp
(CvO) orbital because of their planar arrangement.1a,b,5h
We calculated the rotational barrier around the H2–C2–C1–
O1 dihedral angle on the model cyclopropanecarbaldehydes.
The calculations showed that the bisected s-cis (or s-cis like)
and s-trans conformers are the only two minimum energy
conformers, in which the orbital interaction effectively
lowers the energy of the compounds. Such orbital
interaction should also stabilize the transition state of the
nucleophilic attack, as shown in Figure 10. In the course of
the attack, the p-orbital on the cyclopropyl carbon, which
can be characterized as a strong p-donor,1a,b,5h interacts
with the antibonding orbital (sp–) of the electron-deficient
incipient bond between the nucleophile and the carbonyl
carbon. The p-sp– orbital interaction is maximized in the
bisected conformation, in which the orbitals of the newly
In contrast to these results, nucleophilic hydride reduction
of cyclopropyl ketones with a cis-substituent occurs highly
stereoselectively, which can be explained by the bisected
s-cis transition state reaction model.7 The different stereo-
selectivities would be dependent on the relative stability of
the bisected s-cis and s-trans conformations: the s-cis and
the s-trans have similar stability in the aldehydes, whereas
the s-cis is significantly more stable than the s-trans in the
ketones.7
Figure 10. The transition state model for the nucleophilic attack on the
cyclopropanecarbaldehydes: s-trans (A) and s-cis (B).
The reactions using a variety of additives demonstrated that

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6697
use of effective chelation forming Lewis acids, such as
TiCl4 or ZnBr2, increased the amount of the anti-products,
as summarized in Table 3. The oxygen of the O-Bn group is
known to form a chelate more efficiently than the oxygen of
O-silyl groups.13 Thus, the effect of ZnBr2 was significantly
increased when the O-Bn or O-PMB protected substrates 2
and 3 were used. These results suggest that the anti-products
can be selectively obtained in the nucleophilic additions to
cyclopropanecarbaldehydes via the 7-membered 1,4-chela-
tion-controlled pathway.
Dauben (Scheme 1C).9 However, these are in contrast to the
highly syn-selective results obtained by Reiser using the
substrate IV (Scheme 1B and Table 1).10 We initially
speculated that the different stereoselectivity might be due
to the different reaction conditions used in the two studies,
for example, the Grignard reactions and the Lewis acid-
promoted reactions with silicon nucleophiles. Therefore, the
reactions of our simplified substrate 1 under conditions
similar to those used by Reiser10c were next examined.
However, the Lewis acid-promoted nucleophilic additions
to 1 were moderate or almost non-stereoselective, as
summarized in Table 4. The X-ray crystallographic analysis
of IV showed that it assumes the bisected s-cis conformation
in the solid state, where the adjacent C-formyl carbonyl and
the N-formyl carbonyl groups are in the antiparallel
arrangement due probably to stereoelectronic reasons.10c
The substrate IV also has unusually high reactivity as an
electrophile; it smoothly reacted with TMSCN without any
promoter to give the corresponding addition products in
high yields, and its nitroaldol product was obtained as the
N-formyl-migrated O-formate.10c These results suggest that
the substrate IV may be a unique cyclopropanecarbox-
aldehyde, which is very stable and which is also particularly
reactive to nucleophiles in the s-cis conformation. This may
explain why the reaction of IV proceeds with exceptionally
high stereoselectivity, while stereoselectivity in non-
chelation-controlled additions to carbonyl compounds are
usually not so high.13
The chelating additive ZnBr2 was again effective in the
reactions with the phenyl-substituted substrates 4 and 5
(Table 5). Under non-chelation conditions, typically seen
when HMPA was used as the solvent (entry 4), the
stereoselectivity of the nucleophilic addition to the carba-
moyl-type substrate 4 was low. However, the substrate 4
seems to chelate quite effectively with the Grignard reagent
to give the anti-product highly selectively. The Grignard
reaction of 4 gave the chelation-controlled anti-product
selectively even in the presence of 10 equiv. of HMPA
(entry 3). When the carbamoyl group was replaced with a
CH2O-TBDPS group, the chelating ability of the substrate
was diminished dramatically (entries 8 and 9, syn/
anti¼2.8:1 and 1.3:1, respectively), whereas addition of
ZnBr2 to the Grignard reaction system of the O-TBDPS
substrate 5 resulted in complete anti-selectivity via the
chelation-controlled pathway. The stereoselectivity-
improving effect of the trans-phenyl substituent on the
cyclopropane ring in 4 or 5 is significant because the
corresponding des-phenyl substrate 1, in contrast, showed
only low anti-selectivity (Table 3, entry 3, syn/anti¼1:3.1)
under the same MeMgBr/ZnBr2/CH2Cl2 conditions. The
phenyl group, probably due to steric demand, may restrict
the three-dimensional location of the CH2O-TBDPS or
CONEt2 group attached to the same carbon in the
cyclopropane ring to facilitate chelate formation. These
results suggest that not only the substituent at the position
cis but also the substituent at the position trans to the formyl
group can significantly affect the stereoselectivity in
nucleophilic additions to cyclopropanecarboxaldehydes. A
number of studies on 1,2- and 1,3-chelations have
demonstrated the non-chelating nature of the silyl ether
oxygen.13 However, surprisingly, the highly stereoselective
chelation-controlled Grignard addition occurred in the
We consider that reactions under non-chelation-controlled
conditions likely proceed mainly via the bisected confor-
mation-like transition state stabilized by the characteristic
orbital interaction described above. However, the syn-
selective stereochemical results of the nucleophilic
additions are also in accord with those predicted by
the Felkin–Anh model, as described by Reiser and
co-workers.10 Therefore, it is very important, from a
synthetic organic chemistry standpoint, that both the
bisected and the Felkin–Anh models predict the same
stereochemical outcome.
2.8. Conclusion
This study showed that the anti-selective Grignard additions
to cis-substituted cyclopropanecarboxaldehydes occurred
via the unusual 7-membered 1,4-chelation-controlled path-
way. The results have great importance because they
suggest that 1,4-chelation-controlled stereoselective
addition reactions can indeed be realized. Under non-
chelation conditions, the syn-products, which are likely to
be formed via the bisected s-cis conformation-like transition
state effectively stabilized by the characteristic orbital
interaction, were produced with moderate stereoselectivity.
The predictability of stereochemical outcome based on the
chelation or the bisected s-cis reaction model make these
reactions highly useful.
O-TBDPS-protected substrate
7-membered 1,4-chelation pathway.
5
via the unusual
The present study showed that, in the nucleophilic additions
to the cis-substituted cyclopropanecarbaldehydes, the high
anti-selectivity can be realized via the 7-membered
chelation-controlled pathway. Although a number of highly
stereoselective additions to carbonyl compounds via 5- and
6-membered chelation-controlled pathways are known,
there is no precedent for such a highly stereoselective
addition reaction via a 7-membered 1,4-chelation-
controlled pathway.10b,13–15 The cis-substituted cyclo-
propane structure seems to facilitate formation of the
unusual 7-membered chelate.
3. Experimental
On the other hand, under non-chelation conditions, the syn-
products were produced in moderate selectivity, which is in
accord with the previous moderate syn-selective results by
3.1. General experimental methods
Melting points are uncorrected. NMR spectra were recorded

6698
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
at 400 or 500 MHz (1H) and at 100 or 125 MHz (13C), and
are reported in ppm downfield from Me4Si. The H NMR
room temperature for 30 min, the mixture was cooled to the
temperature indicated in the Table, and then a solution of
Grignard reagent [2 equiv, (MeMgBr, 3 M in THF or 1.4 M
in toluene/THF (3:1); EtMgBr 3.0 M in Et2O; CH2-
vCHCH2MgBr, 1.0 M in Et2O; iPrMgCl, 2.0 M in THF;
iBuMgBr, 2.0 M in THF; Me3Al, 1.0 M in hexane] was
added slowly. The mixture was stirred at the same
temperature for 0.5–8 h, and then MeOH was added.
After additional stirring for 10 min at ambient temperature,
the resulting mixture was evaporated. The residue was
partitioned between AcOEt and H2O, and then the organic
layer was washed with brine, dried (Na2SO4), and
evaporated. The residue was purified by column chroma-
tography (silica gel; AcOEt/hexane 1:9) to give a syn/anti-
mixture of the corresponding Grignard addition products,
the syn/anti ratio of which was determined by the 1H NMR
spectrum. The 1H NMR data of the products 9a–c, 10a–c,
11a, 12a, 18 and 19 were in accord with those of the
authentic samples previously synthesized by another
method.7 The stereochemistries of other Grignard addition
products were determined as described below.19,20
1
assignments indicated were in agreement with COSY
spectra. Mass spectra were obtained by electron ionization
(EI) or the fast atom bombardment (FAB) method. Thin-
layer chromatography was performed on Merck coated plate
60F254. Silica gel chromatography was performed with
Merck silica gel 5715 or 9385 (neutral). Reactions were
carried out under an argon atmosphere.
3.1.1. (1S,2R)-1-Hydroxymethyl-2-(4-methoxybenzyl-
oxymethyl)cyclopropane (8). After stirring a mixture of 7
(102 mg, 0.30 mmol) and NaH (60% in paraffin liquid,
18 mg, 0.45 mmol) in THF (5 mL) at 0 8C for 1 h, PMBnCl
(163 mL, 1.20 mmol) was added, and the resulting mixture
was stirred at the same temperature for 10 min and at room
temperature for further 1 h. After addition of MeOH, the
mixture was partitioned between AcOEt and H2O, and the
organic layer was washed with brine, dried (Na2SO4), and
evaporated. A mixture of the residue and TBAF (1.0 M in
THF, 0.60 mL, 0.60 mmol) in THF (10 mL) was stirred at
room temperature for 12 h, and then the resulting mixture
was partitioned between AcOEt and H2O. The organic layer
was washed with brine, dried (Na2SO4), evaporated, and
purified by column chromatography (silica gel; AcOEt/
hexane 1:4 then 1:2) to give 8 (51 mg, 77%) as a colorless
liquid. 1H NMR (500 MHz, CDCl3): d¼0.20 (1H, m), 0.80
(1H, m), 1.25–1.38 (2H, m), 3.09–3.18 (3H, m), 3.81 (3H,
s), 3.89 (1H, dd, J¼5.5, 10.7 Hz), 3.93 (1H, m), 4.44 (1H, d,
J¼10.7 Hz), 4.52 (1H, d, J¼11.4 Hz), 6.90 (2H, d,
J¼8.6 Hz), 7.30 (2H, d, J¼8.6 Hz); LR-MS (EI): m/z (%):
222 (6) [Mþ]; elemental analysis calcd (%) for C13H18O3: C
70.24, H 8.16; found: C 70.16, H 8.10.
3.2.1. (1S,2R)-2-t-Butyldiphenylsilyloxymethyl-1-[(S)-1-
hydroxy-3-butenyl]cyclopropane (9d) and (1S,2R)-2-t-
butyldiphenylsilyloxymethyl-1-[(R)-1-hydroxy-3-
butenyl]cyclopropane (10d). The mixture of 9d and 10d
(163 mg) was separated by column chromatography (silica
gel; AcOEt/hexane 1:19) to give 9d (79 mg) and 10d
(70 mg) in a pure form.
1
Compound 9d. H NMR (500 MHz, CDCl3): d¼0.27 (1H,
m), 0.72 (1H, m), 1.05 (9H, s), 1.07 (1H, m), 1.23 (1H, m),
1.76 (1H, br s), 2.38 (1H, m), 2.66 (1H, m), 3.41 (1H, m),
3.52 (1H, dd, J¼8.8, 11.2 Hz), 3.87 (1H, dd, J¼5.8,
11.2 Hz), 5.14 (2H, m), 5.90 (1H, m), 7.37–7.45 (6H, m),
7.65–7.69 (4H, m); 13C NMR (125 MHz, CDCl3): d¼7.45,
18.4, 19.1, 22.5, 26.82, 42.3, 64.2, 71.0, 117.7, 127.6, 127.7,
129.6, 133.7, 135.0, 135.5, 135.6; LR-MS (FAB): m/z (%):
381 (13) [(MþH)þ]; elemental analysis calcd (%) for
C24H32O2Si: C 75.74, H 8.47; found: C 75.42, H 8.32.
3.1.2. (1S,2R)-1-Formyl-2-(4-methoxybenzyloxymethyl)-
cyclopropane (3). To a solution of oxalyl chloride (35 mL,
0.40 mmol) in CH2Cl2 (8 mL) was added a solution of
DMSO (57 mL, 0.80 mmol) in CH2Cl2 (1 mL) at 278 8C
slowly over 30 min, and then a solution of 8 (45 mg,
0.20 mmol) in CH2Cl2 (1 mL) was added dropwise. The
resulting mixture was stirred at the same temperature for
1 h, and then Et3N (225 mL, 16.0 mmol) was added. After
stirring the resulting mixture at the same temperature for
further 30 min, aqueous saturated NH4Cl and then CH2Cl2
were added. The organic layer separated was washed with
brine, dried (Na2SO4), evaporated, and purified by column
chromatography (silica gel; AcOEt/hexane 1:15) to give 3
as a colorless oil (28 mg, 64%). 1H NMR (500 MHz,
CDCl3): d¼1.24 (1H, m), 1.32 (1H, m), 1.83 (1H, m), 2.03
(1H, m), 3.39 (1H, dd, J¼8.6, 10.3 Hz), 3.46 (1H, dd,
J¼5.8, 10.3 Hz), 3.80 (3H, s), 4.38 (1H, d, J¼11.5 Hz), 4.41
(1H, d, J¼11.5 Hz), 6.87 (2H, d, J¼8.6 Hz), 7.23 (2H, d,
J¼8.6 Hz), 9.44 (1H, d, J¼4.7 Hz); 13C NMR (125 MHz,
CDCl3): d¼12.36, 23.62, 26.82, 55.23, 67.55, 72.56,
113.81, 129.36, 130.02, 159.27, 200.42; LR-MS (EI): m/z
(%): 220 (7) [Mþ]; elemental analysis calcd (%) for
C13H16O3: C 70.89, H 7.32; found: C 70.67, H 7.53.
Compound 10d. 1H NMR (500 MHz, CDCl3): d¼0.13 (1H,
m), 0.72 (1H, m), 1.04 (1H, m), 1.06 (9H, s), 1.20 (1H, m),
2.49 (2H, m), 3.25–3.42 (2H, m), 3.89 (1H, br s), 4.09 (1H,
dd, J¼5.8, 11.6 Hz), 5.06–5.17 (2H, m), 5.98 (1H, m),
7.34–7.46 (6H, m), 7.64–7.73 (4H, m); 13C NMR
(125 MHz, CDCl3): d¼8.6, 17.0, 19.0, 23.0, 26.8, 41.1,
65.4, 72.5, 116.5, 127.8, 127.8, 129.8, 129.9, 132.9, 133.0,
135.4, 135.5, 135.6; LR-MS (FAB): m/z (%): 381 (13)
[(MþH)þ]; elemental analysis calcd (%) for C24H32O2Si: C
75.74, H, 8.47; found: C 75.52, H 8.40.
3.2.2. The mixture of (1S,2R)-2-benzyloxymethyl-1-[(S)-
1-hydroxypropyl]cyclopropane (11b) and (1S,2R)-2-
benzyloxymethyl-1-[(R)-1-hydroxypropyl]cyclopropane
(12b). 1H NMR (500 MHz, CDCl3): d¼0.22 (0.9H, m), 0.34
(0.1H, m), 0.81 (1H, m), 0.97 (0.3H, t, J¼7.4 Hz), 0.99
(2.7H, t, J¼7.4 Hz), 1.10 (1H, m), 1.25 (1H, m), 1.55–1.69
(2H, m), 3.10 (0.9H, m), 3.17 (0.9H, dd, J¼10.3, 10.8 Hz),
3.20 (0.1H, m), 3.41 (0.1H, dd), 3.79 (0.1H, dd), 3.95 (0.9H,
dd, J¼5.5, 10.3 Hz), 4.47 (1H, m), 4.54 (1H, m), 7.26–7.36
(5H, m); HR-MS (EI) calcd C14H20O2 220.1463, found
220.1453 (Mþ).
3.2. General procedure for the Grignard addition to
aldehydes 1–5 (Tables 2, 3, and 5)
After stirring a mixture of an aldehyde 1–5 (0.10 mmol) and
an additive (2 equiv., if required) in a solvent (2 mL) at

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6699
3.2.3. The mixture of (1S,2R)-2-benzyloxymethyl-1-[(S)-
1-hydroxy-3-methylbutyl]cyclopropane (11c) and
(1S,2R)-2-benzyloxymethyl-1-[(R)-1-hydroxy-3-methyl-
into the mixture of 9c/10c (32 mg, 89%) as described above
for the mixture of 13a/14a.
1
butyl]cyclopropane (12c). H NMR (500 MHz, CDCl3):
3.5.1. (1S,2R)-2-t-Butyldiphenylsilyloxymethyl-1-[(S)-1-
hydroxy-3-oxo-3-phenylpropyl]cyclopropane (9e) and
(1S,2R)-2-t-butyldiphenylsilyloxymethyl-1-[(R)-1-
hydroxy-3-oxo-3-phenylpropyl]cyclopropane (10e). To a
solution of 1 (34 mg, 0.10 mmol) and CH2vCH(Ph)OTMS
(21 mL, 0.10 mmol) in CH2Cl2 (2 mL) was added BF3·Et2O
(13 mL, 0.10 mmol) at 278 8C, and the mixture was stirred
at the same temperature for 2 h. After addition of aqueous
saturated NaHCO3, the mixture was partitioned between
AcOEt and H2O, and the organic layer was washed with
brine, dried (Na2SO4), and evaporated. The residue was
purified by column chromatography (silica gel; AcOEt/
hexane 1:9) to give 9e (18 mg) and 10e (20 mg) as a
colorless oil, respectively.
d¼0.19 (0.95H, m), 0.35 (0.05H, m), 0.80 (1H, m), 0.88–
0.94 (6H, m), 1.06 (1H, m), 1.33 (1H, m), 1.38 (1H, m), 1.60
(1H, m), 1.87 (1H, m), 3.17 (1H, dd, J¼10.5, 10.8 Hz), 3.23
(1H, m), 3.61 (0.05H, dd), 3.64 (0.95H, br s), 3.82 (0.05H,
dd), 3.94 (0.95H, dd, J¼5.5, 10.5 Hz), 4.51–4.60 (2H, m,),
7.27–7.36 (5H, m); HR-MS (EI) calcd C16H24O2 248.1776,
found 248.1769 (Mþ).
3.2.4. (1S,2R)-1-[(S)-1-Hydroxyethyl]-2-(4-methoxy-
benzyloxymethyl)cyclopropane (13a) and (1S,2R)-1-
[(R)-1-hydroxyethyl]-2-(4-methoxybenzyloxymethyl)-
cyclopropane (14a). The mixture of 13a and 14a (65 mg)
was separated by column chromatography (silica gel;
AcOEt/hexane 1:19) to give 13a (10 mg) and 14a (50 mg)
in a pure form.
1
Compound 9e. H NMR (500 MHz, CDCl3): d¼0.35 (1H,
m), 0.81 (1H, m), 0.96 (9H, s), 1.20 (1H, m), 1.28 (1H, m),
3.34 (1H, dd, J¼9.2, 17.9 Hz), 3.39 (1H, br s), 3.40 (1H, dd,
J¼5.3, 11.5 Hz), 3.76 (1H, dd, J¼2.1, 17.9 Hz), 3.92 (1H,
ddd, J¼2.1, 2.3, 9.2 Hz), 4.02 (1H, dd, J¼5.0, 11.5 Hz),
7.31–7.34 (4H, m), 7.37–7.45 (4H, m), 7.53–7.59 (3H, m),
7.62–7.64 (2H, m), 7.92–7.96 (2H, m); 13C NMR
(125 MHz, CDCl3): d¼7.9, 18.1, 19.1, 21.9, 26.8, 46.0,
64.5, 68.5, 127.7, 127.7, 128.2, 128.5, 129.6, 129.7, 133.4,
133.5, 133.5, 135.4, 135.5, 137.7, 201.0; LR-MS (FAB): m/z
(%): 459 (25 [(MþH)þ]; elemental analysis calcd (%) for
C29H34O3Si·H2O: C 73.07, H 7.11; found: C 72.80, H 7.11.
Compound 13a. 1H NMR (500 MHz, CDCl3): d¼0.31 (1H,
m), 0.80 (1H, m), 3.32 (1H, dd, J¼8.5, 10.1 Hz), 3.44 (1H,
m), 3.59 (1H, dd, J¼6.5, 10.1 Hz), 3.81 (3H, s), 4.40 (1H, d,
J¼11.6 Hz), 4.48 (1H, d, J¼11.6 Hz), 6.86–6.89 (2H, m),
7.24–7.27 (2H, m); HR-MS (EI) calcd C14H20O3 236.1412.
found 236.1422 (Mþ).
Compound 14a. 1H NMR (500 MHz, CDCl3): d¼0.17 (1H,
m), 0.78 (1H, m), 1.08 (1H, m), 1.25 (1H, m), 1.30 (3H, d,
J¼6.1 Hz), 3.13 (1H, dd, J¼9.6, 10.5 Hz), 3.33 (1H, m),
3.80 (3H, s), 3.93 (1H, dd, J¼5.5, 10.5 Hz), 4.44 (1H, d,
J¼11.2 Hz), 4.51 (1H, d, J¼11.2 Hz), 6.88 (2H, d,
J¼8.5 Hz), 7.27 (2H, d, J¼8.5 Hz); HR-MS (EI) calcd
C14H20O3 236.1412, found 236.1432 (Mþ).
Compound 10e. 1H NMR (500 MHz, CDCl3): d¼0.24 (1H,
m), 0.66 (1H, m), 1.05 (9H, s), 1.20–1.29 (2H, m), 3.25 (1H,
dd, J¼5.5, 15.8 Hz), 3.47 (1H, dd, J¼7.1, 15.8 Hz), 3.48
(1H, dd, J¼8.6, 10.6 Hz), 3.97 (1H, br s), 4.06 (1H, m), 4.10
(1H, dd, J¼5.3, 10.6 Hz), 7.38–7.48 (8H, m), 7.56 (1H, m),
7.67–7.73 (4H, m), 8.01 (2H, d, J¼7.1 Hz); 13C NMR
(125 MHz, CDCl3): d¼8.7, 17.6, 19.1, 23.3, 26.8, 45.5,
65.2, 69.4, 127.8, 127.8, 128.3, 128.5, 129.8, 129.9, 132.9,
133.0, 135.5, 135.6, 137.4, 198.7; LR-MS (FAB): m/z (%):
459 (28) [(MþH)þ]; elemental analysis calcd (%) for
C29H34O3Si·0.8H2O: C 73.63, H 7.58; found: C 73.68, H
7.54.
3.3. Conversion of the mixture of 13a/14a into the
mixture of 9a/10a
A mixture of 13a and 14a (22 mg, 95 mmol) and 10% Pd-
charcoal (5 mg) in MeOH (1 mL) was stirred under
atmospheric pressure of hydrogen gas at room temperature
for 1 h, and then the catalyst was filtered off. The filtrate was
evaporated, and a mixture of the residue, TBDPSCl (26 mL,
95 mmol) and imidazole (6.4 mg, 95 mmol) in DMF (1 mL)
was stirred at room temperature for 5 h. After addition of
MeOH, the resulting mixture was partitioned between
AcOEt and H2O, and the organic layer was washed with
brine, dried (Na2SO4), and evaporated. The residue was
purified by column chromatography (silica gel; AcOEt/
hexane 1:9) to give a mixture of 9a and 10a (30 mg, 90%) as
an oil.
Reaction of 1 and CH2vCHCH2TMS/BF3·Et2O. To a
solution of 1 (135 mg, 0.40 mmol) and CH2vCHCH2TMS
(509 mL, 3.2 mmol) in CH2Cl2 (5 mL) was added BF3·Et2O
(152 mL, 1.2 mmol) at 278 8C, and the mixture was stirred
at the same temperature for 48 h. After addition of saturated
NaHCO3, the mixture was partitioned between AcOEt and
H2O, and the organic layer was washed with brine, dried
(Na2SO4), and evaporated. The residue was purified by
column chromatography (silica gel; AcOEt/hexane 1:19) to
give 9d (124 mg, 81%) and 10d (17 mg, 11%) as a colorless
oil, respectively.
3.4. Conversion of the mixture of 11b/12b into the
mixture of 9b/10b
The mixture of 11b/12b (14 mg, 64 mmol) was converted
into the mixture of 9b/10b (22 mg, 95%) as described above
for the mixture of 13a/14a.
3.5.2. (1S,2R)-2-t-Butyldiphenylsilyloxymethyl-1-[(S)-1-
hydroxy-3-methyl-3-nitropropyl]cyclopropane (9f) and
(1S,2R)-2-t-butyldiphenylsilyloxymethyl-1-[(R)-1-
hydroxy-3-methyl-3-nitropropyl]cyclopropane (10f). A
mixture of 1 (169 mg, 0.50 mmol) and Et3N (70.3 mL,
0.50 mmol) in 2-nitropropane (2 mL) was stirred at room
temperature for 1 week. After addition of aqueous saturated
3.5. Conversion of the mixture of 11c/12c into the
mixture of 9c/10c
The mixture of 11c/12c (23 mg, 91 mmol) was converted

6700
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
NH4Cl, the mixture was partitioned between AcOEt and
H2O, and the organic layer was washed with brine, dried
(Na2SO4), and evaporated. The residue was purified by
column chromatography (silica gel; AcOEt/hexane 1:19
then 1:9) to give stereoisomers 9f (48 mg, 22%) and 10f
(48 mg, 22%) as a colorless oil, respectively, the 10-con-
configurations of which were not determined. Isomer eluted
was added slowly. The resulting mixture was stirred at the
same temperature for 2 h, and then Et3N (21.4 mL,
152 mmol) was added. After stirring the resulting mixture
at the same temperature for a further 30 min, aqueous
saturated NH4Cl and then CH2Cl2 was added. The organic
layer separated was washed with brine, dried (Na2SO4),
evaporated, and purified by column chromatography (silica
gel; AcOEt/hexane 1:99) to give 5 as white crystals (1.95 g,
25%). M.p. (hexane/iPr2O) 64–65 8C; 1H NMR (500 MHz,
CDCl3): d¼0.96 (9H, s), 1.48 (1H, dd, J¼4.9, 8.1 Hz), 1.78
(1H, dd, J¼4.9, 5.4 Hz), 2.29 (1H, ddd, J¼5.2, 5.4, 8.1 Hz),
3.77 (1H, d, J¼11.1 Hz), 4.05 (1H, d, J¼11.1 Hz), 7.19–
7.22 (2H, m), 7.25–7.28 (2H, m), 7.30–7.35 (6H, m), 7.37–
7.41 (3H, m), 7.50–7.51 (2H, m), 9.63 (1H, d, J¼5.2 Hz);
13C NMR (125 MHz, CDCl3): d¼18.6, 19.1, 26.67, 34.3,
41.1, 66.5, 127.4, 127.5, 127.7, 128.3, 129.5, 129.7,
129.7, 132.8, 133.0, 135.4, 135.5, 142.0, 200.0; LR-MS
(EI): m/z (%): 414 (2) [Mþ]; elemental analysis calcd
(%) for C27H30O2Si: C 78.22, H 7.29; found: C 78.12, H
7.45.
1
early; H NMR (500 MHz, CDCl3): d¼0.35 (1H, m), 0.81
(1H, m), 0.96 (9H, s), 1.20 (1H, m), 1.28 (1H, m), 3.34 (1H,
dd, J¼9.2, 17.9 Hz), 3.39 (1H, br s), 3.40 (1H, dd, J¼5.3,
11.5 Hz), 3.76 (1H, dd, J¼2.1, 17.9 Hz), 3.92 (1H, ddd,
J¼2.1, 2.3, 9.2 Hz), 4.02 (1H, dd, J¼5.0, 11.5 Hz), 7.31–
7.34 (4H, m), 7.37–7.45 (4H, m), 7.53–7.59 (3H, m), 7.62–
7.64 (2H, m), 7.92–7.96 (2H, m); 13C NMR (125 MHz,
CDCl3): d¼5.8, 18.3, 19.0, 19.2, 20.2, 24.3, 26.8, 63.3, 73.5,
92.4, 127.7, 127.8, 129.8, 133.4, 133.4, 135.6, 135.6; HR-
MS (FAB) calcd C24H34NO4Si 428.2257, found 428.2283
((MþH)þ). Isomer eluted late; 1H NMR (500 MHz,
CDCl3): d¼0.24 (1H, m), 0.66 (1H, m), 1.05 (9H, s),
1.20–1.29 (2H, m), 3.25 (1H, dd, J¼5.5, 15.8 Hz), 3.47
(1H, dd, J¼7.1, 15.8 Hz), 3.48 (1H, dd, J¼8.6, 10.6 Hz),
3.97 (1H, br s), 4.06 (1H, m), 4.10 (1H, dd, J¼5.3, 10.6 Hz),
7.38–7.48 (8H, m), 7.56 (1H, m), 7.67–7.73 (4H, m), 8.01
(2H, d, J¼7.1 Hz); 13C NMR (125 MHz, CDCl3): d¼9.6,
16.1, 18.5, 19.0, 20.6, 23.1, 26.8, 65.0, 76.56, 91.4,
127.8, 127.9, 130.0, 132.5, 132.6, 135.5, 135.6; HR-MS
(FAB) calcd C24H34NO4Si 428.2257, found 428.2251
((MþH)þ).
3.5.5. (1R,2S)-2-t-Butyldiphenylsilyloxymethyl-2-[(S)-
hydroxyethyl]-1-phenylcyclopropane (20) and (1R,2S)-
2-t-butyldiphenylsilyloxymethyl-2-[(R)-hydroxyethyl]-1-
phenylcyclopropane (21). The mixture of 20 and 21
(160 mg) was separated by column chromatography (silica
gel; AcOEt/hexane 1:19) to give 20 (96 mg) and 21 (58 mg)
in a pure form, respectively.
1
3.5.3. (1R,2S)-1,2-Dihydroxymethyl-1-phenylcyclo-
propane (16). A mixture of 158 (3.48 g, 20.0 mmol) and
NaBH4 (1.51 g, 40.0 mmol) in MeOH/THF (1:2, 150 mL)
was stirred at room temperature for 3 h, and then AcOH was
added. The resulting mixture was evaporated, and the
residue was partitioned between AcOEt and H2O. The
organic layer was washed with brine, dried (Na2SO4),
evaporated, and purified by column chromatography (silica
gel; AcOEt/hexane 1:4, then 1:2) to give 16 (3.56 g, 100%)
as a colorless oil. 1H NMR (500 MHz, CDCl3): d¼0.75 (1H,
dd, J¼5.0, 5.4 Hz), 1.06 (1H, dd, J¼5.0, 8.6 Hz), 1.66 (1H,
m), 2.96 (1H, br s), 2.38 (2H, m), 3.52 (1H, d, J¼11.9 Hz),
4.09–4.14 (2H, m), 7.23 (1H, n), 7.31 (2H, m), 7.39 (2H,
m); 13C NMR (125 MHz, CDCl3): d¼16.8, 25.8, 32.3, 63.5,
67.5, 126.6, 128.4, 129.2, 143.9; LR-MS (EI): m/z (%): 160
(19) [(M2H2O)þ]; elemental analysis calcd (%) for
C11H14O2: C 74.13, H 7.92; found: C 74.03, H 7.93.
Compound 20. H NMR (500 MHz, CDCl3): d¼0.78 (1H,
dd, J¼4.9, 5.8 Hz), 0.95 (9H, s), 0.99 (1H, dd, J¼4.8,
8.5 Hz), 1.40 (1H, ddd, J¼5.4, 8.5, 8.8 Hz), 1.61 (3H, d,
J¼6.2 Hz), 3.60 (1H, d, J¼9.0 Hz), 3.75 (1H, dq, J¼6.1,
9.8 Hz), 3.94 (1H, d, J¼9.0 Hz), 7.17–7.21 (4H, m), 7.25–
7.41 (9H, m), 7.50–7.52 (2H, m); 13C NMR (125 MHz,
CDCl3): d¼14.9, 19.1, 24.1, 26.7, 32.4, 33.5, 68.4, 68.6,
126.6 127.5, 127.6, 128.0, 129.4, 129.6 130.1, 132.8, 133.4,
135.5, 135.5, 144.3; LR-MS (EI): m/z (%): 412 (13)
[(M2H2O)þ]; Anal. Calcd for C28H34O2Si: C, 78.09; H,
7.96; found: C, 77.89; H, 8.03.
1
Compound 21. H NMR (500 MHz, CDCl3): d¼0.65 (1H,
dd, J¼4.9, 5.8 Hz), 0.93 (1H, dd, J¼4.9, 9.1 Hz), 0.96 (9H,
s), 1.42 (3H, d, J¼6.1 Hz), 1.56 (1H, ddd, J¼5.8, 9.1,
9.8 Hz), 3.58 (1H, d, J¼11.2 Hz), 3.67 (1H, dq, J¼6.1,
9.8 Hz), 4.08 (1H, d, J¼11.2 Hz), 4.14 (1H, br s), 7.04 (2H,
d, J¼7.0 Hz), 7.11 (2H, t, J¼7.6 Hz), 7.23–7.46 (9H, m),
7.58 (2H, d, J¼7.6 Hz); 13C NMR (125 MHz, CDCl3):
d¼16.1, 21.7, 26.7, 32.2, 32.7, 69.4, 69.7, 126.8, 127.5,
127.8, 128.2, 129.5, 129.8. 130.5, 131.2, 131.7, 135.4,
135.5, 143.98; LR-MS (EI): m/z (%): 412 (3) [(M2H2O)þ];
elemental analysis calcd (%) for C28H34O2Si: C 78.09, H
7.96; found: C 77.87, H 8.07.
3.5.4. (1R,2S)-1-t-Butyldiphenylsilyloxymethyl-2-for-
myl-1-phenylcyclopropane (5). After stirring a mixture
of 16 (3.56 g, 20.0 mmol) and NaH (60% in paraffin liquid,
880 mg, 22.0 mmol) in THF (50 mL) at 220 8C for 1 h,
TBDPSCl (5.20 mL, 20.0 mmol) was added, and the
resulting mixture was stirred at the same temperature for
1 h. After addition of MeOH, the mixture was partitioned
between AcOEt and H2O. The organic layer was washed
with brine, dried (Na2SO4), evaporated and purified by
column chromatography (silica gel; AcOEt/hexane 1:19) to
give a mixture of the mono-silylated products (8.12 g, 97%)
as a colorless oil. To a solution of oxalyl chloride (3.32 mL,
36.0 mmol) in CH2Cl2 (5 mL) was added a solution of
DMSO (5.39 mL, 76.0 mmol) in CH2Cl2 (30 mL) at 278 8C
slowly over 30 min, and then a solution of the mono-
silylated products (7.92 g, 19.0 mmol) in CH2Cl2 (10 mL)
3.6. General procedure for preparing the MTPA esters
A mixture of an alcohol (0.10 mmol), (S)- or (R)-MTPA
(28 mg, 0.12 mmol), EDC·HCl (23 mg, 0.12 mmol) and
DMAP (3.7 mg, 30 mmol) in CH2Cl2 (3 mL) was stirred at
room temperature for 12 h, and then MeOH was added. The
mixture was evaporated, and the residue was partitioned
between AcOEt and H2O. The organic layer was washed
with brine, dried (Na2SO4), evaporated, and purified by

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6701
column chromatography (silica gel; AcOEt/hexane 1:49) to
give the corresponding MTPA ester as a colorless oil.
(1H, m, H-10), 7.34640–7.68560 (18H, m, aromatic),
7.97540–8.01100 (2H, m, aromatic); HR-MS (FAB) calcd
C39H41F3NaO5Si 697.2573, found 697.2559 ((MþNa)þ).
3.6.1. (S)-MTPA ester of 20. Yield 99% (64 mg); 1H NMR
(500 MHz, CDCl3)
d
0.924628 (1H, dd, H-3a,
3.6.6. (S)-MTPA ester of 9e. Yield 100% (68 mg); 1H
NMR (500 MHz, CDCl3) d 0.55832 (1H, m, H-3a), 0.85980
(1H, m, H-3b), 0.87000 (9H, s, –C(CH3)3), 1.40686 (1H, m,
H-2), 1.48887 (1H, m, H-1), 3.50240 (3H, s, –OMe),
J3a,3b¼4.9 Hz, J3a,2¼6.0 Hz), 0.96584 (9H, s, –C(CH3)3),
1.02113 (1H, dd, H-3b, J3b,3a¼4.9 Hz, J3b,2¼8.4 Hz),
00
1.56472 (1H, ddd, H-2, J2,3a¼6.0 Hz, J2,3b¼8.4 Hz, J2,1
¼
10.0 Hz), 1.71023 (3H, d, H-200, J2 , 1 ¼6.3 Hz), 3.58820
3.57450 (1H, dd, H-100a, J1 a,2¼10.2 Hz, J1 a,1 b¼11.4 Hz),
00
00
00
00
00
(3H, s, –OMe), 3.69080 (1H, d, H-10a, J1 a, 1 b¼11.1 Hz),
3.64490 (1H, dd, H-20a, J2 a,1 ¼1.2 Hz, J2 a,2 b¼18.3 Hz),
0
0
0
0
0
0
4.02150 (1H, d, H-10b, J1 b,1 a¼11.1 Hz), 5.21330 (1H, dq,
3.93480 (1H, dd, H-20b, J2 b, 1 ¼10.3 Hz, J2 b,2 a¼18.3 Hz),
0
0
0
0
0
0
H-100, J1 , 2 ¼6.3 Hz, J1 , 2¼10.0 Hz), 7.17580–7.21600
(4H, m, aromatic), 7.24580–7.42600 (12H, m, aromatic),
7.52860–7.57840 (4H, m, aromatic); HR-MS (FAB) calcd
C38H41F3NaO4Si 669.2624, found 669.2621 ((MþNa)þ).
4.01630 (1H, dd, H-100b, J1 b,2¼4.9 Hz, J1 b,1 a¼11.4 Hz),
5.67300 (1H, m, H-10), 7.35080–7.67320 (18H, m,
aromatic), 7.91920–7.93620 (2H, m, aromatic); HR-MS
(FAB) calcd C39H41F3NaO5Si 697.2573, found 697.2565
((MþNa)þ).
00
00
00
00
00
00
3.6.2. (R)-MTPA ester of 20. Yield 97% (63 mg); 1H NMR
(500 MHz, CDCl3) d 0.90778 (1H, dd, H-3a, J3a,3b¼5.2 Hz,
J3a,2¼6.2 Hz), 0.93488 (1H, dd, H-3b, J3b,3a¼5.2 Hz,
J3b,2¼8.4 Hz), 0.96786 (9H, s, –C(CH3)3), 1.46095 (1H,
3.7. X-ray crystallographic data of 117
˚
C21H26O2Si, M¼338.52, Monoclinic, P21, a¼10.652 (3) A,
˚
˚
00
ddd, H-2, J2,3a¼6.2 Hz, J2,3b¼8.4 Hz, J2,1 ¼10.0 Hz),
b¼9.022 (3) A, c¼10.632 (3) A, b¼104.66 (2)8, V¼1033.4
1.78783 (3H, d, H-200, J2 ,1 ¼6.3 Hz), 3.58200 (3H, s,
(4) A , Z¼2, Dcalc¼1.137 Mg cm23. Cell parameters were
3
˚
00 00
–OMe), 3.66570 (1H, d, H-10a, J1 a, 1 b¼11.1 Hz), 4.01160
determined and refined from 26 reflections in the range
26.58,u,30.08. A colorless crystal (0.60£0.40£0.30 mm3)
was mounted on a Mac Science MXC18 diffractometer with
0
0
(1H, d, H-10b, J1 b, 1 a¼11.1 Hz), 5.21920 (1H, dq, H-100,
0
0
00 00
00
J1 ,2 ¼6.3 Hz, J1 ,2¼10.0 Hz), 7.17260–7.20000 (4H, m,
aromatic), 7.23840–7.42600 (12H, m, aromatic), 7.53340–
7.55540 (4H, m, aromatic); HR-MS (FAB) calcd C38H41-
F3NaO4Si 669.2624, found 669.2628 ((MþNa)þ).
˚
graphite-monochromated Cu Ka radiation (l¼1.54178 A).
Data collection using the v/2u scan technique gave 1662
reflections at room temperature, 1522 unique, of which 1522
with I.3.00 s(I) reflections were used in calculations. The
intensities were corrected for the Lorentz, polarization, and
the extinction effect, but not for the absorption. The
structure was solved by the direct method and refined by
full-matrix least squares technique using maXus (ver. 4.3)
as the computer program. The non-hydrogen atoms were
refined anisotropically. The hydrogen atoms were included
by calculation, but these positions were not refined. The R
value was 0.059.
3.6.3. (S)-MTPA ester of 9d. Yield 95% (64 mg); 1H NMR
(500 MHz, CDCl3) d 0.42900 (1H, m, H-3a), 0.72092 (1H,
m, H-3b), 1.06144 (9H, s, –C(CH3)3), 1.25006 (1H, 0m,
H-2), 1.29378 (1H, m, 0 H-1), 2.58350 (1H, m, H-2 0a0 ),
2.77450 (1H, m, H-2 b), 3.57770 (1H, dd, H-1 a,
00
00
00
J1 a, 2¼8.9 Hz, J1 a, 1 b¼11.6 Hz), 3.59040 (3H, s, –OMe),
3.96970 (1H, dd, H-100b, J1 b, 2¼5.0 Hz, J1 b, 1 a¼11.6 Hz),
4.98260 (2H, m, H-40), 5.03270 (1H, m, H-10), 5.67000 (1H,
m, H-30), 7.35520-7.45580 (9H, m, aromatic), 7.57560–
7.59120 (2H, m, aromatic), 7.66320–7.70360 (4H, m,
aromatic); HR-MS (FAB) calcd C34H39F3O4Si 597.2648,
found 597.2663 ((MþH)þ).
00
00
00
3.8. Calculations
All ab initio and density functional theory (DFT) calcu-
lations were performed using the GAUSSIAN98 program17
on an SGI O2 workstation. The O1–C1–C2–H2 dihedral
angle of i or ii was rotated from 0 to 3608 at the intervals of
208, and the conformations were optimized at RHF/
3-21G(d). Finally, single point energies were calculated at
RB3LYP/6-31G(d). The bisected s-cis and trans conformers
of i and ii were fully optimized at RB3LYP/6-31G(d) and
their single point energies also were calculated by RB3LYP/
6-31G(d).
1
3.6.4. (R)-MTPA ester of 9d. Yield 100% (68 mg); H
NMR (500 MHz, CDCl3) d 0.40896 (1H, m, H-3a), 0.62952
(1H, m, H-3b), 1.06042 (9H, s, –C(CH3)3), 1.13647 (1H,0m,
H-2), 1.24478 (1H, m, 0 H-1), 2.64880 (1H, m, H-2 0a0 ),
2.85370 (1H, m, H-2 b), 3.53680 (1H, dd, H-1 a,
00
00
00
J1 a,2¼8.6 Hz, J1 a,1 b¼11.4 Hz), 3.55900 (3H, s, –OMe),
3.96080 (1H, dd, H-100b, J1 b,2¼5.2 Hz, J1 b,1 a¼11.4 Hz),
5.05600 (1H, m, H-10), 5.10380 (2H, m, H-40), 5.82830 (1H,
m, H-30), 7.35260–7.45420 (9H, m, aromatic), 7.53420–
7.57000 (2H, m, aromatic), 7.66420–7.69700 (4H, m,
aromatic); HR-MS (FAB) calcd C34H39F3O4Si 597.2648,
found 597.2621 ((MþH)þ).
00
00
00
Acknowledgements
This investigation was supported by a Grant-in-Aid for
Creative Scientific Research (13NP0401) (A. M.), a Grant-
in-Aid for Scientific Research (S. S.), and Research
Fellowships for Young Scientists (H. A.) from the Japan
Society for the Promotion of Science. We are grateful to Ms.
H. Matsumoto, A. Maeda, and S. Oka (Center for
Instrumental Analysis, Hokkaido University) for technical
assistance with NMR, MS, and elemental analyses, and to
Daiso Co., Ltd for the gift of chiral epichlorohydrins.
3.6.5. (R)-MTPA ester of 9e. Yield 95% (64 mg); 1H NMR
(500 MHz, CD2Cl2) d 0.61320 (1H, m, H-3a), 0.78480 (1H,
m, H-3b), 0.85940 (9H, s, –C(CH3)3), 1.27920 (1H, m, H-
2), 1.34960 (1H, m, H-1), 3.48350 (3H, s, –OMe), 3.56180
(1H, dd, H-100a, J1 a, 2¼9.7 Hz, J1 a,1 b¼11.1 Hz), 3.69960
00
00
00
(1H, dd, H-20a, J2 a,1 ¼1.3 Hz, J2 a, 2 b¼18.4 Hz), 3.98900
0
0
0
0
(1H, dd, H-100b, J1 b,2¼4.8 Hz, J1 b, 1 a¼11.1 Hz), 4.01190
00
00
00
(1H, dd, H-20b, J2 b, 1 ¼10.6 Hz, J2 b,2 a¼18.4 Hz), 5.63820
0
0
0
0

6702
Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
8. (a) Ono, S.; Shuto, S.; Matsuda, A. Tetrahedron Lett. 1996, 37,
References and notes
221–224. (b) Shuto, S.; Ono, S.; Hase, Y.; Kamiyama, N.;
Takada, H.; Yamashita, K.; Matsuda, A. J. Org. Chem. 1996,
61, 915–923.
1. (a) de Meijere, A. Angew. Chem., Int. Ed. 1979, 18, 809–826.
(b) Wong, H. N. C.; Hon, M.-Y.; Tse, C.-W.; Yip, Y.-C.;
Tanko, J.; Hudlicky, T. Chem. Rev. 1989, 89, 165–198.
(c) Stammer, C. H. Tetrahedron 1990, 46, 2231–2254. (d) In
Small Ring Compounds in Organic Synthesis VI. Topic in
Current Chemistry 207; de Meijere, A., Ed.; Springer: Berlin,
1999.
9. Dauben, W. G.; Dinges, J.; Smith, T. C. J. Org. Chem. 1993,
58, 7635–7637.
10. (a) Bubert, C.; Reiser, O. Tetrahedron Lett. 1997, 38,
4985–4988. (b) Mengel, A.; Reiser, O. Chem. Rev. 1999,
¨
99, 1191–1223. (c) Bohm, C.; Schimmerl, M.; Bubert, C.;
Zabel, M.; Labahn, T.; Parisini, E.; Reiser, O. Eur. J. Org.
Chem. 2000, 2955–2965.
2. Reviews on stereoselective synthesis of cyclopropanes:
(a) Singh, V. K.; DattaGupta, A.; Sekar, G. Synthesis 1997,
137–149. (b) Doyle, M. P.; Protopopova, M. N. Tetrahedron
1998, 54, 7919–7946. (c) Lebel, H.; Marcoux, J.-F.; Molinaro,
C.; Charette, A. B. Chem. Rev. 2003, 103, 977–1050.
3. (a) Kazuta, Y.; Matsuda, A.; Shuto, S. J. Org. Chem. 2002, 67,
1669–1677. (b) Kazuta, Y.; Hirano, K.; Natsume, K.;
Yamada, S.; Kimura, R.; Matsumoto, S.; Furuichi, K.;
Matsuda, A.; Shuto, S. J. Med. Chem. 2003, 46, 1980–1988.
(c) Yamaguchi, K.; Kazuta, Y.; Abe, H.; Matsuda, A.; Shuto,
S. J. Org. Chem. 2003, 68, 9255–9262, and references therein.
4. (a) Shuto, S.; Ono, S.; Hase, Y.; Kamiyama, N.; Matsuda, A.
Tetrahedron Lett. 1996, 37, 641–644. (b) Shuto, S.; Ono, S.;
Hase, Y.; Ueno, Y.; Noguchi, T.; Yoshii, K.; Matsuda, A.
J. Med. Chem. 1996, 39, 4844–4852. (c) Shuto, S.; Ono, S.;
Imoto, H.; Yoshii, K.; Matsuda, A. J. Med. Chem. 1998, 41,
3507–3514. (d) Noguchi, T.; Ishii, K.; Imoto, H.; Otubo, Y.;
Shuto, S.; Ono, S.; Matsuda, A.; Yoshii, K. Synapse 1999, 31,
87–96. (e) Uchino, S.; Watanabe, W.; Nakamura, T.; Shuto,
S.; Kazuta, Y.; Matsuda, A.; Nakazima-Iijima, S.; Kohsaka,
S.; Kudo, Y.; Mishina, M. FEBS Lett. 2001, 69, 2007–2015.
(f) Kazuta, Y.; Tsujita, R.; Ogawa, K.; Hokonohara, T.;
Yamashita, K.; Morino, K.; Matsuda, A.; Shuto, S. Bioorg.
Med. Chem. 2002, 10, 1777–1791. (g) Kazuta, Y.; Tsujita, R.;
Uchino, S.; Kamiyama, N.; Mochizuki, D.; Yamashita, K.;
Ohmori, Y.; Yamashita, A.; Yamamoto, T.; Uchino, S.;
Kohsaka, S.; Matsuda, A.; Shuto, S. J. Chem. Soc., Perkin
Trans. 1 2002, 1199–1212. (h) Ono, S.; Ogawa, K.;
Yamashita, K.; Yamamoto, T.; Kazuta, Y.; Matsuda, A.;
Shuto, S. Chem. Pharm. Bull. 2002, 50, 966–968. (i) Kazuta,
Y.; Tsujita, R.; Yamashita, K.; Uchino, S.; Kohsaka, S.;
Matsuda, A.; Shuto, S. Bioorg. Med. Chem. 2002, 10,
3829–3848.
11. Stereoselective Grignard addition to a C-cyclopropylaldo-
nitrone has also been reported: (a) Kazuta, Y.; Shuto, S.;
Matsuda, A. Tetrahedron Lett. 2000, 41, 5373–5377.
(b) Kazuta, Y.; Shuto, S.; Abe, H.; Matsuda, A. J. Chem.
Soc., Perkin Trans. 1 2001, 599–604.
12. Critcher, D. J.; Connolly, S.; Wills, M. J. Org. Chem. 1997, 62,
6638–6657.
13. (a) Reetz, M. T. Angew. Chem., Int. Ed. 1984, 23, 556–569.
(b) Huryn, D. M. In Carbanions of alkali and alkaline earth
cations; (ii) selectivity of carbonyl addition reactions.
Comprehensive Organic Synthesis; Schreiber, S. L., Ed.;
Pergamon: Oxford, 1991; Vol. 1, pp 49–75. (c) Reetz, M. T.
Acc. Chem. Res. 1993, 26, 462–468.
14. 1,4-Chelation-controlled addition to an aldehyde in moderate
diastereoselectivity (1:5–1:6) has been reported: Reetz, M. T.;
Kessler, K.; Schidtberger, S.; Westermann, J.; Steinback, R.
Angew. Chem. 1983, 22, 989–991.
15. 1,3-Asymmetric inductions in enolate alkylation reactions
have been thought to occur via a 1,5-chelation pathway:
(a) Narasaka, K.; Ukaji, Y. Chem. Lett. 1986, 81–84. (b) Ukaji,
Y.; Narasaka, K. Bull. Chem. Soc. Jpn 1988, 61, 571–572.
16. Ohtani, I.; Kusumi, T.; Kashman, Y.; Kakisawa, H. J. Am.
Chem. Soc. 1991, 113, 4092–4096.
17. The X-ray crystallographic data have been deposited at the
Cambridge Crystallographic Data Centre (CCDC 238520).
18. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A., Jr.; Stratmann, R. E.; Burant, J. C.;
Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin, K. N.;
Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.;
Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford,
S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari,
K.; Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov,
B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.;
Challacombe, M.; Gill, P.M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Andres, J. L.; Gonzalez, C.; Head-Gordon, M.;
Replogle, E. S.; Pople, J.A. GAUSSIAN98, Revision A.6:
Gaussian, Inc., Pittsburgh PA, 1998.
5. (a) Bartell, L. S.; Carroll, B. L.; Guillory, J. P. Tetrahedron
Lett. 1964, 5, 705–708. (b) Bartell, L. S.; Guillory, J. P.
J. Chem. Phys. 1965, 43, 647–654. (c) Volltrauer, H. N.;
Schwendemann, R. H. J. Chem. Phys. 1971, 260–267.
(d) Pelissier, M.; Serafini, A.; Devanneaux, J.; Tocanne, J. F.
Tetrahedron 1971, 27, 3271–3284. (e) Tocanne, J. F.
Tetrahedron 1972, 28, 389–416. (f) Aroney, M. J.;
Calderbank, K. E.; Stootman, H. J. J. Chem. Soc., Perkin II
1973, 1365–1368. (g) Crasnier, F.; Labarre, J. F.; Cousse, H.;
D’Hinterland, L. D.; Mouzin, G. Tetrahedron 1975, 31,
825–829. (h) Pawar, D. M.; Noe, E. A. J. Org. Chem. 1998,
63, 2850–2853, and references therein.
19. Optical purity of the substrates was more than 95% ee: see
Refs. 3a,8b.
6. (a) Descotes, G.; Menet, A.; Collonges, F. Tetrahedron 1973,
29, 2931–2935. (b) Meyers, A. I.; Romine, J. L.; Fleming,
S. A. J. Am. Chem. Soc. 1988, 110, 7245–7247. (c) Lautens,
M.; Delanghe, P. H. M. J. Org. Chem. 1995, 60, 2474–2487.
(d) Yokomatsu, T.; Yamagishi, T.; Suemune, K.; Abe, H.;
Kihara, T.; Soeda, S.; Shimeno, H.; Shibuya, S. Tetrahedron
2000, 56, 7099–7108.
20. The 1H NMR data of 9a–c, 10a–c, 11a and 12a were in
accord with those of the authentic samples previously
synthesized by another method: see Ref. 7.
21. The stereochemistries of 9d,e, 10d,e, 20, and 21 were
determined by the modified Mosher’s method.
22. The stereochemistries of 11b and 12b, 11c and 12c, and 13a
and 14a were determined by their conversion into the
corresponding TBDPS-protecting congeners 9b and 10b, 9c
and 10c, and 9a and 10a, respectively.
7. Kazuta, Y.; Abe, H.; Yamamoto, T.; Matsuda, A.; Shuto, S.
J. Org. Chem. 2003, 68, 3511–3521.

Y. Kazuta et al. / Tetrahedron 60 (2004) 6689–6703
6703
23. Nitroaldol reactions are know to be reversible, see: Wade,
P. A.; Murray, J. K., Jr.; Shah-Patel, S.; Palfery, B. A.; Carroll,
P. J. J. Org. Chem. 2000, 65, 7723–7730.
25. Walsh, A. D. Nature (London) 1947, 159, 712–713.
26. (a) Cieplak, A. S. J. Am. Chem. Soc. 1981, 103, 4540–4552.
(b) Cieplak, A. S.; Tait, B. D.; Johnson, C. R. J. Am. Chem.
Soc. 1989, 111, 8447–8462. (c) Cieplak, A. S. Chem. Rev.
1999, 99, 1265–1336.
24. For examples see: (a) Pothier, N.; Goldstein, S.;
Deslongchamps, P. Helv. Chim. Acta 1992, 75, 604–620.
(b) Romero, J. A. C.; Tabacco, S. A.; Woerpel, K. A. J. Am.
Chem. Soc. 2000, 122, 168–169. (c) Abe, H.; Shuto, S.;
Matsuda, A. J. Am. Chem. Soc. 2001, 11, 870–882.
(d) Tamura, S.; Abe, H.; Matsuda, A.; Shuto, S. Angew.
Chem., Int. Ed. 2003, 42, 1021–1023.
27. (a) Burgi, H. B.; Dunitz, J. D.; Shefter, E. J. Am. Chem. Soc.
1973, 95, 5065–5067. (b) Burgi, H. B.; Dunitz, J. D.; Lehn,
J. M.; Wipff, G. Tetrahedron 1974, 30, 1563–1572. (c) Burgi,
H. B. Angew. Chem., Int. Ed. 1975, 14, 460–473.
28. Yamamoto, Y.; Asao, N. Chem. Rev. 1993, 93, 2207–2293.