An empirical model for stereochemical control in the cyclization of cyclopropanetricarboxylic acid esters

Article abstract of DOI:10.1246/bcsj.20190096

Here, we report an empirical model for diastereoselective cyclopropanation of fumarate/maleate diesters with chloroacetate, sulfonium ylide, or ammonium ylide. With symmetrical fumarate/maleate diesters, cyclopropanation was found to proceed with a high level of diastereoselectivity in favor of the chiral isomer. In contrast, production of the meso isomer was observed in 3848% diastereoselectivity when unsymmetrical fumarate/maleate was employed. An improved synthesis of (N-desmethy)dysibetaine CPa in both racemic and enantiomeri-cally pure forms was furthermore achieved. Configurational analysis by experimental and calculated ¹³C NMR data is also reported.

Full text of DOI:10.1246/bcsj.20190096

Document type: Article
An Empirical Model for Stereochemical Control in the Cyclization
of Cyclopropanetricarboxylic Acid Esters
Kento Tanaka, Hitomi Manabe, Raku Irie, and Masato Oikawa*
Yokohama City University, 22-2 Seto, Kanazawa-ku, Yokohama, Kanagawa 236-0027, Japan
E-mail: moikawa@yokohama-cu.ac.jp
Received: April 4, 2019; Accepted: April 30, 2019; Web Released: June 1, 2019
Masato Oikawa
Masato Oikawa graduated from Hokkaido University in 1988 under supervision of Prof. Shirahama. After
working several years at Nippon Soda Co., he returned to Hokkaido University and obtained his doctor
degree in 1994 under supervision of Prof. Ichihara. He then moved to Osaka University as an assistant
professor (Prof. Kusumoto), and further moved to Tohoku University as an associate professor in 2003 (Prof.
Sasaki). During 2001–2002, he worked at ICCB (Harvard University, Prof. Schreiber) for postdoctoral
research. Since 2009, he is at Yokohama City University, where he got promotion to a full professor in 2010.
He now studies synthetic development of neuroactive agents.
Abstract
Here, we report an empirical model for diastereoselective
cyclopropanation of fumarate/maleate diesters with chloroace-
tate, sulfonium ylide, or ammonium ylide. With symmetrical
fumarate/maleate diesters, cyclopropanation was found to pro-
ceed with a high level of diastereoselectivity in favor of the
chiral isomer. In contrast, production of the meso isomer was
observed in 38­48% diastereoselectivity when unsymmetrical
fumarate/maleate was employed. An improved synthesis of (N-
desmethy)dysibetaine CPa in both racemic and enantiomeri-
cally pure forms was furthermore achieved. Configurational
analysis by experimental and calculated 13C NMR data is also
reported.
NMe3
NMe3
*
*
OOC
COOH
OOC
*
CONH2
(
–)–dysibetaine CPa
dysibetaine CPb
relative stereochemistry
*
Figure 1. Cyclopropane natural products from Micronesian
marine sponge.9
control had not been well investigated in cyclopropanations
between fumarate/maleate diesters, and some Michael donors
such as chloroacetate esters, sulfonium ylides, and ammo-
11
12
1
3
nium ylides. In this work, we found that the stereochemistry
can be controlled by the combination of different ester groups,
and a model is proposed, i.e., cis,trans,trans-chiral cyclo-
propane can be conveniently and selectively constructed using
symmetrical fumarate diester and chloroacetate ester through 1)
Michael reaction followed by 2) spontaneous cyclization
without conformational change of the intermediary Michael
adduct. Here, the methodology was applied to an improved,
Keywords: Cyclopropanation j Diastereoselectivity
j
Dysibetaine CPa
Introduction
Cyclopropanes are an important class of carbocycles. These
1
are 1) often found in natural products, 2) used for conforma-
2
10b
tional restriction in medicinal chemistry, and 3) also used as
fewer-step synthesis of rac-(N-desmethyl)dysibetaine CPa.
an intermediate that shows unique reactivity in multi-step
Furthermore, enantioselective synthesis of (R,R)- and (S,S)-
(N-desmethyl)dysibetaine CPa was attempted using quinidine
derivative as an organocatalyst. Results for configurational
analysis of cyclopropanes, based on both experimental and
calculated C NMR chemical shift values, are also reported
herein.
3
synthesis. Many methodologies have been, therefore, devel-
4
oped for selective construction of cyclopropanes; halomethyl-
5
metal-mediated reactions including Simmons­Smith reaction,
6
13
decomposition of diazoalkane, and Michael reaction followed
by ring-closure. A number of asymmetric syntheses of
7
cyclopropanes have been also reported.8
Cyclopropanation of Diethyl Fumarate (1) with Ethyl
1
0d
During the course of our synthetic study of marine natural
Chloroacetate (2). As has been already reported by us and
9
11
product dysibetaine CPa/CPb (Figure 1), we needed to stereo-
another group, cyclopropanation of diethyl fumarate (1) with
1
0
selectively construct 1,2,3-trisubstituted cyclopropanes. To
our surprise, however, it was soon realized that stereochemical
ethyl chloroacetate (2) cleanly proceeds in 80% yield to give
cis,trans,trans-cyclopropane (CTT) triethyl ester 3 predom-
1314 | Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan

inantly (Scheme 1). Thermodynamically, the CTT isomer 3 is
obviously more stable than the cis,cis,cis-isomer (CCC) 4,
which indeed was not detected in this reaction.14 In all other
experiments in this paper, we did not observe such CCC
isomers (see below). In this study, reactivity and stereoselec-
tivity in the cyclopropanation of a variety of fumarate/maleate
esters with chloroacetate, sulfonium ylide, or ammonium ylide
were examined, as follows.
When diethyl maleate (5) was employed (run 2), again only
CTT-chiral cyclopropane 7 was obtained, but the reaction was
found to be sluggish (10% yield after 24 h).
Cyclopropanation of “Unsymmetrical” Fumarate/Mal-
eate Diesters (9, 10, 11) with Chloroacetates (12, 13).
Table 2 shows the results for cyclopropanation of “unsym-
metrical” fumarate/maleate diesters (9, 10, 11) with ethyl
chloroacetate (12) or methyl chloroacetate (13). It was found
that tert-butyl ethyl fumarate (9) undergoes cyclopropanation
with 12 to give a mixture of nearly comparable amounts of the
CTT-chiral and CTT-meso cyclopropanes 7 and 8 (7/8 =
52:48, run 1) in lower total yield than that in Table 1 (run 1).
Generation of CTT-meso isomer was observed for the first time
here in a series of these reactions. When sterically less demand-
ing tert-butyl methyl fumarate (10) was employed in com-
bination with methyl chloroacetate (13) (run 2), the fraction
of the CTT-meso isomer decreased (14/15 = 62:38). As had
been also observed in Table 1 (run 2), only a trace amount
(4.7%) of cyclopropane products was obtained in run 3,
wherein the maleate 11, isomeric to the fumarate 10 in run 2,
was employed. Quite interestingly, however, the diastereoselec-
tivity (14/15 = 61:39) was found to be comparable to that in
run 2. Similarly, poor yield was observed in the synthesis of
cyclopropanes 7/8, by reaction of tert-butyl ethyl maleate with
ethyl chloroacetate (12) (data not shown in Table 2).
Results
Cyclopropanation of Diethyl Fumarate (1)/Diethyl Mal-
eate (5) with tert-Butyl Chloroacetate (6). According to the
standard procedures employed in Scheme 1,1
0d,11
cyclopropa-
nation of diethyl fumarate (1, 1.0 equiv) with tert-butyl chloro-
acetate (6, 1.35 equiv) was first conducted in the presence of
K CO (2.2 equiv) and BnEt NCl (0.01 equiv) in DMF at
2
3
3
1
4
0 °C (Table 1, run 1). After 24 h, TLC and H NMR analyses
revealed that the fumarate 1 was completely consumed, and
cyclopropane 7 was isolated in 85% yield after chromato-
graphic purification. The planar structure as well as the stereo-
chemistry of 7 was determined by spectroscopic means includ-
1
ing H NMR analysis to be a cyclopropane with CTT-chiral
configuration. No CTT-meso isomer 8 was detected in the
reaction mixture.
O
O
H
O
O
H
O
O
cis
Cyclopropanation with Sulfonium Ylides (16­18). Sul-
trans
O
O
12
cis
cis
H
H
H
H
fonium ylides are promising cyclopropanation reagents. Gen-
2
Cl
O
O
O
O
O
O
O
O
O
O
Table 2. Cyclopropanations of “unsymmetrical” fumarate/
maleate with chloroacetate provide CTT-meso products
with 38­48% diastereoselectivity as minor products.
K CO
BnEt3NCl
DMF, 40 °C
2
3
diethyl fumarate (1)
3
4
trans
cis
24 h
cis,trans,trans–isomer
cis,cis,cis–isomer
O
O
O
O
H
O
(CTT)
(CCC)
R
dr = 100:0
00% conversion yield ( H NMR)
0% isolated yield
O
1
H
H
H
H
1
Cl 12, 13 a
unsymmetrical
fumarate/maleate
diester (9, 10, 11)
8
O
R
O
R
O
R
O
O
O
O
O
R
K
2
CO
3
Scheme 1. Formation of cyclopropane triethyl ester gener-
ally proceeds with exclusive CTT diastereoselectivity.1
BnEt
DMF, 40 °C
3
NCl
trans
CTT–chiral isomer
(R = ethyl)
4 (R = methyl)
cis
0d,11
CTT–meso isomer
7
8 (R = ethyl)
15 (R = methyl)
1
Table 1. Cyclopropanations of both fumarate and maleate
with chloroacetate proceed with high CTT-chiral
diastereoselectivity.
Diastereo-
selectivity
chiral/meso)
R for
Run Unsymmetrical diester
Results
b
chloroacetate
(
O
O
O
O
H
O
O
O
O
6
H
H
H
H
diethyl fumarate (1)
or
Cl
(1.35 equiv)
O
O
O
O
O
50% conversion
44% isolation
1
ethyl (12)
7/8 = 52:48
diethyl maleate (5)
O
O
O
O
O
K
BnEt
2
CO
3
(2.2 equiv)
(1.0 equiv)
3
NCl (0.01 equiv)
DMF, 40 °C
4 h
cis
tert–butyl ethyl
fumarate (9)
7
trans
8
2
CTT–chiral isomer CTT–meso isomer
O
O
Diastereoselectivity
Results a
Run
Symmetrical diester
(
chiral/meso)
O
2
74% conversion
methyl (13)
14/15 = 62:38
50% isolation
O
O
O
tert–butyl methyl
fumarate (10)
1
00% conversion
85% isolation
1
7/8 = >95:5
O
O
O
O
diethyl fumarate (1)
O
O
4% conversion
3
methyl (13)
14/15 = 61:39
4.7% isolation
O
O
tert–butyl methyl
maleate (11)
1
5% conversion
0% isolation
2
O
O
7/8 = >95:5
1
diethyl maleate (5)
a
b
For equivalency, see Table 1.
Conversion yield was determined from 1H NMR spectrum.
a
Conversion yield was determined from 1H NMR spectrum.
Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan | 1315

Table 3. Cyclopropanations of fumarate with various sul-
fonium ylides proceed with acceptable CTT-chiral
diastereoselectivity.
Table 4. Cyclopropanations of maleate with sulfonium ylide
proceed with acceptable CTT-chiral diastereoselectivity in
high yield.
O
O
R
O
O
O
O
O
H
O
O
O
O
H
O
O
R
R
S
S
O
1
(
6–18
2.9 equiv)
H
H
H
H
18 a
H
H
H
H
O
R
O
R
O
O
O
O
O
O
O
O
O
O
O
O
R
O
O
R
R
O
O
R
O
O
O O
(
CH2Cl)2
5 °C
(CH
2
Cl)
2
dialkyl maleate
6
trans
CTT–chiral isomer
(R = ethyl)
14 (R = methyl)
cis
diethyl fumarate (1)
1.0 equiv)
(5, 21)
65 °C
trans
CTT–chiral isomer
(R = ethyl)
9 (R = benzyl)
(R = tert–butyl)
cis
CTT–meso isomer
(
CTT–meso isomer
7
8 (R = ethyl)
15 (R = methyl)
3
—
1
7
20 (R = benzyl)
8 (R =tert–butyl)
Diastereoselectivity
chiral/meso)
Run
R for maleate
Results
b
(
Diastereoselectivity
chiral/meso)
Run
R for ylide
ethyl (16)
Results a
ethyl (5)
diethyl maleate)
100% conversion
94% isolation
(
1
7/8 = 92:8
(
1
00% conversion
9% isolation
b
110d,13
—
methyl (21)
dimethyl maleate)
100% conversion
99% isolation
9
2
10e
14/15 = 89:11
(
5
0% conversion
2
benzyl (17)
tert–butyl (18)
19/20 = 92:8
7/8 = 92:8
41% isolation
a
For equivalency, see Table 3.
b
Conversion yield was determined from 1H NMR spectrum.
100% conversion
3
91% isolation
O
O
a
Conversion yield was determined from 1H NMR spectrum.
b
The diastereoselectivity for CTT/CCC isomers (3/4) were 100:0.
S
No chiral center is present in these molecules bearing triethyl ester.
22
3.0 equiv)
O
(
O
14
15
O
CTT–chiral
isomer
CTT–meso
isomer
(
CH
2 2
Cl)
erally, the ylide is prepared prior to use, and cyclopropanation
O
65 °C
14/15 = 55:45
is conducted at elevated temperature in (CH Cl)2 without
2
tert–butyl methyl
fumarate (10)
6
5% conversion
50% isolation
additives. Structurally complex cyclopropane with sensitive
functional groups can be synthesized with sulfonium ylide-
mediated cyclopropanation, owing to the high chemoselec-
(
1.0 equiv)
Scheme 2. Cyclopropanation of unsymmetrical fumarate
with sulfonium ylide provides CTT-meso product as a
minor product (45% diastereoselectivity).
1
0d
12
tivity. We
and another group have previously reported
that cyclopropanation of diethyl fumarate (1) with ethyl
sulfanylideneacetate (16) gives CTT cyclopropane 3 quantita-
tively (Table 3, run 1). The complete CTT selectivity, shown
above in the chloroacetate-mediated cyclopropanations
O
O
OH
2
CO H
TFA
3
BH •THF
H
H
O
O
CH
2
Cl
2
EtO
2
C
CO
2
Et
THF
(
Scheme 1, and Tables 1, 2), was again observed here in the
EtO
2
C
2
CO Et
O
O
10e
combinations using two other sulfonium ylides 17 (run 2)
23
24 (known)10b
trans
1
0b
and 18 (run 3). Interestingly, in both cases, the CTT-chiral
isomers, 19 (run 2) and 7 (run 3), were preferably generated
over the meso isomers in reasonable (41% yield, 92% purity)
and high yields (91%, 92% purity), respectively.
50% (2 steps)
chiral isomer 7
prepared in one step
85%, see Table 1)
(
3
NH Cl
known steps10b
Even from maleate diesters, surprisingly, the CTT-chiral
isomer was found to be selectively obtained, when sulfonium
ylide was employed (Table 4). Thus, tert-butyl sulfanylidene-
HO
2
C
2
CO H
1) CBr , PPh , CH Cl , 64%
4
3
2
2
rac-(N-desmethyl)
dysibetaine CPa
2
) NaN
2
3
, TBAI, DMF, 80 °C, 85%
, Boc O, MeOH, THF, 67%
3) H , Pd(OH)
2
2
((±)–25)
1
0b
acetate (18)
reacts with diethyl maleate (5, run 1) and
4) 6 M HCl, 90 °C, 100%
GABA analog
dimethyl maleate (21, run 2)1 to selectively give rise to CTT-
chiral cyclopropanes 7 (94% yield, 92% purity) and 14 (99%
yield, 89% purity), respectively. The higher isolation yield in
run 1 (94%) than that in chloroacetate-mediated reaction
0e
Scheme 3. Seven-step synthesis of rac-(N-desmethyl)-
dysibetaine CPa ((«)-25).
(
Table 1, run 2, 10%) would be due to the higher reactivity
Synthesis of rac-(N-Desmethyl)dysibetaine CPa ((«)-25)
as an Application. With the ester group-controlled cyclo-
propanation shown above, we decided to demonstrate natural
product synthesis. Thus, synthesis of N-desmethyl analog 25 of
dysibetaine CPa, a Micronesian sponge-derived natural prod-
15
of the sulfonium ylide.
The cyclopropanation outcome of unsymmetrical diester
with sulfonium ylide was comparable to that observed in the
chloroacetate-mediated reaction shown above. Namely, as
shown in Scheme 2, cyclopropanation of 10 using methyl
sulfanylideneacetate (22)1 gave CTT-chiral isomer 14 and
CTT-meso isomer 15 in 50% isolated yield in a ratio of 55:45,
which are comparable to those shown in Table 2 (run 2).
9
uct, in racemic form was demonstrated in 5 steps fewer than
0e
10b
our previous synthesis, from cyclopropane 7 (see Table 1,
run 1) as follows (Scheme 3). Acidic hydrolysis of the tert-
butyl ester of 7 by TFA, followed by chemoselective reduction
1316 | Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan

O
Cl
2
Cl
6
H
H
2
CO tBu
O
tBuO C
tBuO
2
C
CO
2
tBu
1) TFA, CH2Cl2
Br
26
2
EtO C
diethyl
EtO
2
C
H
EtO
2
C
H
(R)
2
CO Et
fumarate
(R)
Et 2) BH
•THF
THF
H
CO
2
Et
(1)
quinidine
derivative 27
Cs CO , CH CN
0 °C, 36%
EtO
2
C
CO
2
3
H
CO
2
Et
7
1
Michael adduct A1
CTT–chiral isomer
2
8
3
3
(R,R)–7
27.8% ee
59% (2 steps)
lk (Re,Re)
X
NPhth
Scheme 5. Our proposed mechanism for cyclopropanation of
symmetrical” fumarate with chloroacetate (for Table 1,
1) phthalimide
“
(S)
2
EtO C
(S)
2
CO Et
DEAD, PPh
PhH, 94%
3
run 1).
2
EtO C
2
CO Et
(
(
S,S)–23 (X = CO
S,S)–24 (X = CH
2
H)
OH)
2) chiral HPLC
separation
(S,S)–28 (>99.9% ee)
2
structure of the major cyclopropane (R,R)-7 was confirmed
after leading to the final product ((S,S)-25) with known
NH
3
Cl
10d
spectroscopic properties,
by the reactions illustrated in
H
NH
2
NH
2
HN
O
6 M HCl
Scheme 4; the carboxylic acid (S,S)-23, derived from (R,R)-7,
was reduced by BH3¢THF by following the procedure in the
racemate synthesis shown in Scheme 3. To introduce nitrogen
functionality, the alcohol (S,S)-24 then underwent Mitsunobu
(
S)
CO
2
Et
90 °C
87%
HO
2
C
(
S)
CO
2
H
EtOH
7%
H
8
(S,S)–2510d
(
S,S)–29
reaction with phthalimide (PPh , DEAD, PhH, rt) in 94% yield.
3
Attempts to recrystallize the phthalimide (S,S)-28 as well as all
other intermediates in order to improve the enantiomeric purity
were unsuccessful, however, optical resolution of 27.8% ee of
OMe
N
MeO
(
(
S,S)-28 by chiral HPLC was found to be practical to furnish
S,S)-28 and enantiomeric (R,R)-28 in enantiomerically pure
N
form for each isomers (see the Supporting Information). The
major and minor phthalimides (S,S)-28 and (R,R)-28, were
independently deprotected (hydrazine, EtOH; 6 M HCl) to
provide (S,S)-(N-desmethyl)dysibetaine CPa ((S,S)-25) with
unnatural configuration, and the (R,R)-congener ((R,R)-25, not
shown in Scheme 4 but included in the Experimental section),
respectively, in a stepwise manner via lactam (S,S)-29/(R,R)-
29. Overall yields were 9.7% and 5.5% for (S,S)-25 and (R,R)-
25, respectively, for total 6 steps each.
quinidine derivative 27
Scheme 4. Asymmetric synthesis of (S,S)-(N-desmethyl)-
dysibetaine CPa hydrochloride ((S,S)-25). For the synthesis
of natural (R,R)-congener ((R,R)-25), see the Experimental
section.
1
0d
of the generated carboxyl group (BH3¢THF) gave alcohol 24 in
0% yield over 2 steps. Owing to the selective cyclopropana-
5
tion shown in Table 1, the three-step synthesis of alcohol 24
from diethyl fumarate (Scheme 3, 42.5% total yield) apparently
improved the efficiency of our previous synthesis of (N-
desmethyl)dysibetaine CPa ((«)-25),1 wherein a longer 8 step
procedure (8.5% yield from maleic anhydride) had been
required for the synthesis of alcohol 24.
Discussion
Cyclopropanation Mechanisms. Cyclopropanations using
chloroacetate or sulfanylideneacetate are generally thought to
proceed through Michael addition to the acceptor followed by
0b
11,12,18
intramolecular nucleophilic substitution (S 2).
Based on
N
Synthesis of Optically Pure, Both Enantiomers of (N-
Desmethyl)dysibetaine CPa (25) by Chiral Ammonium
Ylide. We further attempted asymmetric synthesis of (N-
desmethyl)dysibetaine CPa (25) by enantioselective cyclo-
propanation using quinidine derivative as an organocatalyst
the stereochemical outcomes observed in the present study, we
deduced more detailed mechanism of these cyclopropanations,
as follows. First, rapid rotation of the C­C bond in the Michael
product is obviously involved prior to cyclopropane formation
(e.g. run 1 vs run 2 in Table 1, see below for discussion).
Second, the diastereoselectivity in the cyclopropanation seems
to be, therefore, controlled by intramolecular steric repulsion
between substituents in the Michael adduct. Third, the leaving
groups (Cl, sulfonium, ammonium) do not play a crucial role in
the diastereoselectivity.
Our proposed mechanisms are shown in Schemes 5­8. The
diastereoselective cyclopropanation of fumarate acceptor 1 (see
Table 1, run 1) to give CTT-chiral isomer 7 can be explained
by a rather simple mechanism drawn in Scheme 5. Here, the
Michael adduct intermediate A1 is generated through Re/Re
face reaction (lk topicity) and directly cyclizes to form 7,
without large conformational change.
(
Scheme 4). The organocatalytic reaction is mediated by chiral
1
6
16,17
ammonium ylide, which is empirically
expected to give
cyclopropane with (R,R)-CTT-chiral configuration. Thus,
diethyl fumarate (1) was first reacted with tert-butyl bromo-
1
6
acetate (26) in the presence of the quinidine derivative 27 and
Cs CO in CH CN at 80 °C to give cyclopropane (R,R)-7 in
2
3
3
3
6% yield. Gratifyingly, no diastereomers other than CTT-
chiral cyclopropane 7 were detected. Disappointingly, how-
ever, chiral HPLC analysis of 23, obtained after removal of the
tert-butyl ester, showed that the enantiomeric purity was only
2
7.8% ee in favor of the (S,S)-enantiomer ((S,S)-23) illustrated
in Scheme 4. Note that (R,R)-7 corresponds to (S,S)-23, that
is generated after TFA treatment. All attempts to improve the
efficiency were unsuccessful. For example, diethyl maleate (5)
for the organocatalytic reaction provided (R,R)-7 with higher
enantiomeric purity (39.4% ee) but in lower yield (5%). The
In contrast, as shown in Scheme 6, cyclopropanation of
maleate 5 (see Table 1, run 2) seems to involve large confor-
mational change by rotation of a C­C bond in the Michael
adduct conformer A2 generated through Re/Re face reaction
Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan | 1317

Michael adduct A2
cyclization of A3 is a disfavored process and is not observed,
due to the steric repulsion between CO tBu and CO Et groups.
Minor
pathway
Cl
Cl
H
6
H
2
2
8
tBuO
2
C
2
tBuO C
CTT–meso
isomer
For cyclopropanation of “unsymmetrical” fumarate acceptor
, nonselective formation of CTT-chiral isomer 7 and CTT-
2
EtO C
CO
2
Et
EtO
2
C
H
CO
H
2
Et
9
H
H
meso isomer 8 was observed (7/8 = 52:48, see Table 2, run 1).
As shown in Scheme 8 (chiral/meso processes), this can be
explained by nonstereoselective Michael addition (both lk and
ul topicities) to form diastereomeric Michael adducts B/C
followed by cyclization. In these processes, two Michael
adducts B/C would contain the same level of unfavorable
intramolecular steric interactions, between sterically demand-
ing tert-butyl ester and other functional groups such as
ethoxycarbonyl or chloro group.
The mechanistic discussion shown in Scheme 8 can be
similarly applied to the case of methyl ester (see Table 2,
run 2), from the fact that reactivity of sterically less demand-
ing methyl ester-containing unsymmetrical fumarate 10 (50%
yield) is higher than that in run 1 (44% yield). The rather
higher ratio of the product 14 (14/15 = 62:38) in run 2 would
further indicate that the chiral process is essentially preferred
over the meso process, which is in accord with the model in
Scheme 5 for symmetrical fumarate 1, that gives chiral cyclo-
propane 7 predominantly.
5
lk (Re,Re)
Bond rotation to
reduce non-bonding
interactions
Major pathway
7
Michael
adduct A1
CTT–chiral
isomer
Scheme 6. Our proposed mechanism for cyclopropanation of
symmetrical” maleate with chloroacetate (for Table 1,
run 2).
“
Cl
H
CO
2
tBu
2
CO tBu
EtO
2
C
CO
H
2
Et
2
EtO C
2
CO Et
H
CCC isomer
Michael adduct A3
Scheme 7. The possible third Michael adduct A3 leading to
an imaginary CCC isomer.
chiral process
It is of interest to note that unsymmetrical maleate diester 11
(see Table 2, run 3) is remarkably unreactive, as compared to
the fumarate diester 10 (run 2). Intermolecular steric interac-
tions between sterically demanding maleate diester 11 and
chloroacetate 13 would prevent the first Michael reaction, as
judged from the fact that both components were recovered
quantitatively after the reaction.
Cl
12
Cl
H
H
2
CO Et
EtO
2
C
EtO
2
C
2
EtO C
EtO
2
C
H
EtO
2
C
H
2
CO tBu
H
CO
2
tBu
H
2
CO tBu
7
9
Michael adduct B
CTT–chiral isomer
lk (Re,Re)
Stereochemical Analysis by Experimental and Calculat-
ed 13C Chemical Shift Values.
Configuration of cyclo-
propanation products in this paper was determined on the basis
meso process
1
13
Cl
H
12
Cl
H
of the structural symmetry as judged by the H and C NMR
CO
2
Et
2
EtO C
EtO
2
C
10e
spectra. The assignments were finally confirmed by leading
2
CO tBu
H
C
CO
H
2
tBu
H
C
CO
H
2
tBu
to dysibetaine analog (see Schemes 3 and 4).
Here, we furthermore attempted to determine the config-
uration of cyclopropanes in this paper by comparison of the
2
EtO C
EtO
2
EtO
2
8
9
13
Michael adduct C
CTT–meso isomer
C
ul (Re,Si)
chemical shift value with those expected by density functional
theory (DFT) calculation, to investigate the possibility for
development of a rapid and reliable method for structural anal-
ysis of even more complex cyclopropanes. Our initial attempts,
Scheme 8. Our proposed mechanism for cyclopropanation of
unsymmetrical” fumarate with chloroacetate (for Table 2,
run 1).
“
1
3
however, were not very encouraging, since C chemical shift
values for no cyclopropanetricarboxylic acid esters (3, 4, 7,
8, 14, 15, 19, and 20 in Schemes 1­2 and Tables 1­4) in this
paper were well reproduced with B3LYP/6-31G* model
(
lk topicity), in order to release non-bonding interactions to
give rise to cyclopropane 7 predominantly via the conformer
A1. The low yield (10%) would be due to severe interactions
between sterically demanding coupling partners (the nucleo-
phile and the (Z)­alkene) in the first intermolecular Michael
2
0
(Spartan ’18 software; Wavefunction, Irvine, CA, U.S.A.).
It is now suspected that structurally flexible substituents on the
cyclopropane core may interfere with precise estimation of the
1
9
reaction.
1
3
It should be also noted here that, although we did not
observe production of CCC cyclopropane in this reaction
C NMR values.
We, therefore, next tried analysis of the derivatives, and
(
Table 1, run 2), we cannot exclude presence of the diastereo-
found that γ-butyrolactam-fused cyclopropane 29 (see
Scheme 4) showed good correlation between experimental
meric Michael adduct intermediate A3 (Scheme 7) as a precur-
sor that bears three alkoxycarbonyl groups arranged on the
same face (e.g. α face as shown in Scheme 7) in these ther-
modynamically equilibrated processes, since we previously
observed formation of CCC cyclopropane in the cyclopropa-
1
3
and calculated C NMR values, as follows. Here, two
diastereomeric candidates, (1S,3S)-29 (correct structure) and
(1S,3R)-29a (incorrect structure) (Figure 2), were subjected to
the calculation; 1) conformational search, 2) structural opti-
mizations with B3LYP-D3/6-31G* model, 3) calculations of
nation of cyclic imide acceptor, which obviously includes the
intermediate corresponding to A3.1
0b,14
13
In general, however,
the C NMR chemical shifts with B3LYP/6-31G* model, 4)
1318 | Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan

6
6
2
H
2
H
all combinations, CTT-chiral cyclopropane was found to be
predominantly provided, which allowed us to propose the
reaction mechanisms for the diastereoselectivity. Based on the
diastereoselective cyclopropanation, an improved racemate
synthesis of (N-desmethyl)dysibetaine CPa ((«)-25), a γ-
aminobutyric acid (GABA) analog of marine-derived natural
cyclopropane, was achieved in fewer steps than our previous
synthesis. Furthermore, asymmetric synthesis of both enan-
tiomers of 25 has been accomplished employing cinchona
alkaloid-mediated cyclopropanation followed by optical reso-
lution by chiral HPLC. Beside the satisfactorily high diastereo-
selectivity, the observed poor enantioselectivity indicated the
limitation of the current organocatalytic methodology.
HN
HN
4
(3S)
8
4
(3R)
8
O
O
O
5
O
(1S)
5
H
7
H
7
(1S)
O
O
(
1S,3S)–29
correct
(1S,3R)–29a
incorrect
Figure 2. A set of diastereomers subjected to the NMR
calculations for configurational analysis.
1
0b
Table 5. Comparison of experimental and calculated
C NMR chemical shift values of (1S,3S)-29 (correct
structure) and (1S,3R)-29a (incorrect structure).
13
Calculatedb
Position Experimentala
(1S,3S)-29
(1S,3R)-29a
Structurally symmetric meso compounds are increasingly
(
correct structure) (incorrect structure)
22
recognized as useful intermediates for asymmetric synthesis.
1
2
3
4
5
6
7
8
28.0
23.5
25.8
170.8
175.3
43.9
61.3
14.2
27.7
24.9
26.1
170.7
172.8
43.2
61.7
14.2
1.05
99.5
27.8
24.8
25.7
169.6
171.5
41.2
61.8
14.2
1.78
0.5
For example, we have recently reported asymmetric synthesis
of N-desmethyl analog of marine natural product dysibetaine
CPb, employing enantioselective methanolysis of meso-
1
0e
succinic anhydride fused to cyclopropane.
Regardless of
the usefulness, however, meso-selective cyclopropanation has
not yet been realized. From Scheme 2 and Table 2 in this work,
some unreactive Michael acceptors under harsh reaction condi-
tions are now expected to possibly produce meso-cyclopropane
selectively. Experiments are currently in progress to discover
conditions for meso-selective cyclopropanation, using acetate
Michael donor with fumarate/maleate Michael acceptor.
Configuration of substituted cyclopropanes is often chal-
RMS/ppm
DP4+/%
a
13
Experimental C NMR data were collected at 100 MHz in
1
3
lenging to determine. We showed here that C NMR chemical
shift values of γ-butyrolactam-fused cyclopropane, in combi-
nation with the DFT calculation values, are of use to determine
the stereochemistry. The strategy would be applied to structur-
ally more complex cyclopropanes whose configuration cannot
b
13
CDCl3. Calculated C NMR data were obtained employing
B3LYP-D3/6-31G*// B3LYP/6-31G* model. For details, see
text and the Supporting Information.
correction of the 13C NMR values on the Boltzmann weighting
calculated with a single point energy calculation employing
B3LYP-D3/6-31G*// B3LYP/6-31G* model to give theoret-
ical shifts. It should be noted that parameters regarding solvents
were not added in these calculations. The results are summa-
rized in Table 5. The theoretical values obtained for C4 and C6
in (1S,3S)-29 (correct structure) were found to agree with the
experimental data, whereas (1S,3R)-29a (incorrect structure)
shows obviously larger differences for these carbons. The
correct (1S,3S)-29 afforded smaller root mean square devia-
tion (RMS), 1.05 ppm, than the incorrect isomer (1S,3R)-29a
1.78 ppm). The calculations also suggest that solvent effects
are not critically operative in these models.
To give solid proof for the analysis based on the RMS
values, we furthermore analyzed the C NMR values using
DP4+ probability statistics.21 The results, 99.5% and 0.5%
probabilities for (1S,3S)-29 and (1S,3R)-29a, respectively,
were consistent with the fact that (1S,3S)-29 bears correct
be readily determined by other means.
23
Experimental
General. Details of the experimental techniques and the
1
0b
apparatus are summarized in our previous paper. ESI mass
spectra were measured with an Exactive Focus Hybrid
Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Sci-
entific, San Jose, CA).
1-(tert-Butyl) 2,3-Diethyl (2S*,3S*)-Cyclopropane-1,2,3-
tricarboxylate (7) (Table 1, run 1): To a stirred suspension
of potassium carbonate (35.0 mg, 0.255 mmol) and benzyl
triethylammonium chloride (0.264 mg, 0.00116 mmol) in DMF
(0.066 mL) at 40 °C were added a solution of diethyl fumarate
(1, 20.0 mg, 0.116 mmol) and tert-butyl chloroacetate (6,
23.6 mg, 0.157 mmol) in DMF (0.050 mL). After stirring at
(
13
40 °C for 24 h, H O (1 mL) was added and the mixture extract-
2
ed with EtOAc/hexane (1:1, 3 © 1 mL). Combined extracts
were dried over Na SO and concentrated under reduced pres-
2
4
relative configuration. These results were consistent with the
sure. The residue was purified by column chromatography on
silica gel (1 g, EtOAc/hexane = 1:9) to give cyclopropane
tert-butyl diethyl ester 7 (28.2 mg, 0.0985 mmol, 85%) as a
colorless oil: IR (neat) 2981, 1727, 1639, 1370, 1308, 1150
3
observed vicinal coupling constants for (S,S)-29 ( J
=
H1,H2
3
3
6
.1 Hz, J
= 2.7 Hz, J
= 2.7 Hz; see the Experimen-
H2,H3
H1,H3
tal section).
¹
1 1
cm ; H NMR (400 MHz, CDCl3) δ 4.16 (dd, J = 7.0, 7.0 Hz,
H), 4.13 (dd, J = 7.0, 7.0 Hz, 2 H), 2.69 (t, J = 8.0 Hz, 1 H),
Conclusion
2
In this paper, we reported reactivity and selectivity in the
cyclopropanation of fumarate/maleate diesters in combination
with chloroacetate, sulfonium ylide, or ammonium ylide. In
2.48 (dd, J = 10.0, 10.0 Hz, 1 H), 2.43 (dd, J = 10.0, 10.0 Hz,
1 H), 1.41 (s, 9 H), 1.24 (t, J = 7.0 Hz, 3 H), 1.24 (t, J = 7.0 Hz,
1
3
3 H); C NMR (100 MHz, CDCl ) δ 170.3, 167.5, 166.6, 82.0,
3
Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan | 1319

6
1.5, 61.4, 29.5, 28.7, 28.0 (©3), 25.4, 14.1, 14.1; HRMS (ESI,
(500 g, EtOAc/hexane = 1:9) to give fumarate 10 (20.0 g,
0.107 mol, 35%), maleate 11 (19.2 g, 0.103 mol, 34%), and the
mixture (10/11 = 46:54, 5.29 g, 0.0284 mol, 9.3%) as colorless
oils. The spectroscopic data for fumarate 10 and maleate 11
were identical to those reported.
Maleate 11 can be quantitatively isomerized to fumarate 10
as follows. Thus, a mixture of maleate 11 (5.00 g, 0.0269 mol),
NBS (5.27 g, 0.0296 mol), and AIBN (442 mg, 2.69 mmol) in
+
positive) calcd for C H O Na [(M + Na) ] 309.1314, found
3
14
22
6
09.1311.
2
5
26
1
-(tert-Butyl) 2,3-Diethyl (2S*,3S*)-Cyclopropane-1,2,3-
tricarboxylate (7) (Table 1, run 2): With the same procedure
as for the synthesis of 7 (Table 1, run 1), 7 (3.32 mg, 0.0116
mmol, 10%) was obtained as a colorless oil starting from
diethyl maleate (5, 20.0 mg, 0.116 mmol), 6 (23.6 mg, 0.157
mmol), potassium carbonate (35.3 mg, 0.255 mmol), and
benzyl triethylammonium chloride (0.264 mg, 0.00116 mmol).
The chromatographic and spectroscopic data were identical to
those of authentic sample shown above.
CCl (135 mL) was heated to reflux with stirring for 2 h. Insol-
4
uble materials were removed by filtration, and H O (200 mL)
2
was added to the filtrate. The organic layer was separated, and
aqueous layer was extracted with CCl (2 © 100 mL). The com-
4
tert-Butyl Ethyl Fumarate (9): To a stirred solution of
fumaric acid monoethyl ester (3.00 g, 20.8 mmol) in tert-
butanol (42 mL) were added Boc2O (6.80 g, 31.2 mmol) and
DMAP (25.0 mg, 0.208 mmol). After stirring at 50 °C for 11 h,
the mixture was concentrated under reduced pressure to remove
bined organic layer was dried over Na SO and concentrated
2
4
under reduced pressure. The residue was purified by column
chromatography on silica gel (50 g, EtOAc/hexane = 1:9) to
give fumarate 10 (5.00 g, 0.0269 mol, 100%) as a colorless oil.
1-(tert-Butyl) 2,3-Dimethyl (2S*,3S*)-Cyclopropane-
1,2,3-tricarboxylate (14) and 1-(tert-Butyl) 2,3-Dimethyl
tert-butanol. Then saturated aqueous NH Cl (50 mL) was added
4
and the mixture was extracted with CHCl (50 mL). The extract
(1r*,2R*,3S*)-Cyclopropane-1,2,3-tricarboxylate
(15):
3
was washed with saturated aqueous NaHCO (50 mL), dried
(Table 2, run 2) With the same procedure used for the synthesis
of 7 (Table 1, run 1), an inseparable mixture of 14 and 15
(14/15 = 62:38, 13.9 g, 0.0538 mol, 50%) was obtained as a
colorless oil starting from tert-butyl methyl fumarate (10,
20.0 g, 0.107 mol), 13 (23.2 g, 0.214 mol), potassium carbonate
(32.5 g, 0.235 mol), and benzyl triethylammonium chloride
(244 mg, 1.07 mmol).
3
over Na SO , filtered, and concentrated under reduced pressure.
2
4
The residue was purified by column chromatography on silica
gel (50 g, EtOAc/hexane = 1:9) to give tert-butyl ester 9
(
3.88 g, 19.4 mmol, 93%) as a yellow oil. The spectroscopic
24
data were identical to those reported.
1
-(tert-Butyl)
2,3-Diethyl
(2S*,3S*)-Cyclopropane-
1
,2,3-tricarboxylate (7) and 1-(tert­Butyl) 2,3-Diethyl
Data for an inseparable mixture of 14 and 15 (14/15 =
62:38): IR (neat) 2980, 1725, 1630, 1380, 1300, 1150 cm .
Selected data for the major CTT-chiral isomer 14: H NMR
¹
1
(
(
1r*,2R*,3S*)-Cyclopropane-1,2,3-tricarboxylate
(8)
1
Table 2, run 1): With the same procedure as for the synthe-
sis of 7 (Table 1, run 1), an inseparable mixture of 7 and 8
7/8 = 52:48, 2.44 g, 0.00852 mol, 44%) was obtained as a
colorless oil starting from tert-butyl ethyl fumarate (9, 3.88 g,
.0194 mol), 12 (4.75 g, 0.0388 mol), potassium carbonate
(400 MHz, CDCl ) δ 3.69 (s, 3 H), 3.69 (s, 3 H), 2.71 (t, J =
3
(
5.6 Hz, 1 H), 2.50 (dd, J = 10.1, 5.6 Hz, 1 H), 2.46 (dd, J = 9.8,
1
3
5.3 Hz, 1 H), 1.41 (s, 9 H); C NMR (100 MHz, CDCl ) δ
3
0
170.7, 167.9, 166.5, 82.1, 52.5, 52.5, 29.5, 28.5, 27.9 (©3),
(
(
5.36 g, 0.0388 mol), and benzyl triethylammonium chloride
44.0 mg, 0.194 mmol). The chromatographic and spectro-
25.3.
1
Selected data for the minor CTT-meso isomer 15: H NMR
scopic data of CTT-chiral isomer 7 were identical to those of
authentic sample shown above.
Selected data for the minor CTT-meso isomer 8: H NMR
(400 MHz, CDCl ) δ 3.70 (s, 6 H), 2.67 (t, J = 5.6 Hz, 1 H),
2.47 (d, J = 5.6 Hz, 2 H), 1.43 (s, 9 H); C NMR (100 MHz,
3
1
3
1
CDCl ) δ 168.9 (©2), 168.3, 82.2, 52.4 (©2), 28.0 (©2), 27.9
3
(400 MHz, CDCl3) δ 4.10­4.03 (m, 4 H), 2.58 (t, J = 5.6 Hz,
(©3), 26.8.
1
H), 2.38 (d, J = 5.6 Hz, 2 H), 1.37 (s, 9 H), 1.18 (t, J = 6.8
(Table 2, run 3) With the same procedure used for the
synthesis of 7 (Table 1, run 1), an inseparable mixture of 14
and 15 (14/15 = 61:39, 1.13 g, 4.38 mmol, 4.7%) was obtained
as a colorless oil starting from tert-butyl methyl maleate (11,
17.5 g, 0.0940 mol), 13 (20.0 g, 0.188 mol), potassium carbon-
ate (26.0 g, 0.188 mol), and benzyl triethylammonium chloride
(214 mg, 0.939 mmol). The chromatographic and spectroscopic
data were in good agreement with those shown above.
1-Benzyl 2,3-Diethyl (2S*,3S*)-Cyclopropane-1,2,3-tri-
carboxylate (19) (Table 3, run 2): To a stirred solution of
diethyl fumarate (1, 10.5 mg, 0.0610 mmol) in 1,2-dichloro-
ethane (2.6 mL) at 65 °C was added dropwise a solution of
1
3
Hz, 6 H); C NMR (100 MHz, CDCl ) δ 168.9, 167.6 (©2),
3
8
1.6, 61.2 (©2), 28.1 (©2), 27.7 (©3), 26.5, 14.0 (©2).
tert-Butyl Methyl Fumarate (10) and tert-Butyl Methyl
Maleate (11):
A solution of maleic anhydride (30.0 g,
0
.306 mol) in methanol (612 mL) was stirred at rt for 21 h. The
mixture was then concentrated under reduced pressure to give
maleic acid monomethyl ester (39.8 g, 0.306 mol) as a colorless
oil, which was sufficiently pure and was used for the next
reaction without purification.
To a stirred solution of maleic acid monomethyl ester (39.8 g,
0
.306 mol) thus obtained above in tert-butanol (306 mL) were
1
0e
added Boc O (86.4 g, 0.396 mol) and DMAP (374 mg, 3.06
ylide 17 (44.1 mg, 0.176 mmol). The mixture was stirred for
23 h, and was then concentrated under reduced pressure. The
residue was purified by column chromatography on silica gel
(1.5 g, EtOAc/hexane = 1:9) to give an inseparable mixture of
cyclopropane benzyl diethyl esters 19 and 20 (19/20 = 92:8,
8.0 mg, 0.025 mmol, 41%) as a colorless oil: IR (neat) 2980,
2
mmol). After stirring at 50 °C for 74 h, the mixture was con-
centrated under reduced pressure to remove tert-butanol. Then
saturated aqueous NH4Cl (100 mL) was added and the mixture
was extracted with Et O (2 © 150 mL). Combined extracts were
washed with saturated aqueous NaHCO (150 mL), dried over
2
3
¹1 1
Na SO , filtered, and concentrated under reduced pressure. The
1726, 1189 cm ; H NMR (400 MHz, CDCl ) δ 7.38­7.28 (m,
2
4
3
residue was purified by column chromatography on silica gel
5 H), 5.13 (s, 2 H), 4.15 (q, J = 7.1 Hz, 2 H), 4.06 (q, J = 7.1
1320 | Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan

Hz, 2 H), 2.78 (t, J = 5.6 Hz, 1 H), 2.57 (dd, J = 10.0, 5.7 Hz,
1-(tert-Butyl) 2,3-Diethyl (2R,3R)-Cyclopropane-1,2,3-
tricarboxylate ((R,R)-7) (Scheme 4): To a stirred suspension
1
1
1
2
H), 2.52 (dd, J = 10.0, 5.6 Hz, 1 H), 1.26 (t, J = 7.1 Hz, 3 H),
.18 (t, J = 7.1 Hz, 3 H); 13C NMR (100 MHz, CDCl ) δ 169.8,
of quinidine derivative 27 (47.2 mg, 0.139 mmol) and Cs CO
16
3
2
3
67.3, 167.3, 135.2, 128.4, 128.3, 128.3, 67.2, 61.5, 61.4, 28.5,
8.3, 25.6, 14.0, 13.9.
(272 mg, 0.836 mmol) in CH CN (1.4 mL) at 80 °C was slowly
3
added a solution of diethyl fumarate (1, 120 mg, 0.697 mmol)
and tert-butyl bromoacetate (26, 163 mg, 0.836 mmol) in
CH3CN (1.0 mL) over 6 h. After stirring for 22 h, to the mix-
1
-(tert-Butyl) 2,3-Diethyl (2S*,3S*)-Cyclopropane-1,2,3-
tricarboxylate (7) (Table 3, run 3): With the same procedure
used for the synthesis of 19 (Table 3, run 2), an inseparable
mixture of cyclopropane tert-butyl diethyl esters 7 and 8 (7/8 =
ture were added EtOAc (3 mL) and saturated aqueous NH Cl
4
(2 mL), and organic layer was separated. The aqueous layer
was extracted with EtOAc (3 © 2 mL), and combined organic
layer was dried over Na SO and concentrated under reduced
9
2:8, 75.8 mg, 0.265 mmol, 91%) was obtained as a colorless
oil starting from diethyl fumarate (1, 50.0 mg, 0.290 mmol) and
a solution of ylide 181 (188 mg, 0.871 mmol). The chromato-
graphic and spectroscopic data were in good agreement with
those shown above.
2
4
0b
pressure. The residue was purified by column chromatography
on silica gel (8 g, EtOAc/hexane = 8:92) to give cyclopropane
(R,R)-7 (72.0 mg, 0.251 mmol, 36%) as a yellowish oil. The
spectroscopic data for (R,R)-7 were identical to those shown
above. The absolute stereochemistry was determined later by
leading to known compound ((S,S)-25), and the enantiomeric
purity was also later determined as 27.8% ee (see below).
(2S,3S)-2,3-Bis(ethoxycarbonyl)cyclopropane-1-carboxylic
Acid ((S,S)-23) and Diethyl (1S,2S)-3-(Hydroxymethyl)-
cyclopropane-1,2-dicarboxylate ((S,S)-24) (Scheme 4):
According to the same procedure for the synthesis of racemic
23 and 24 shown above, deprotection of optically active tert-
butyl ester (R,R)-7 (3.23 g, 11.3 mmol) gave crude carboxylic
acid (S,S)-23 (2.94 g), which was then reduced to give alcohol
(S,S)-24 (1.43 g, 6.61 mmol, 59% for 2 steps) as a yellowish oil.
The spectroscopic data for (S,S)-23 and (S,S)-24 were identi-
cal to those shown above. The enantiomeric purity was deter-
mined as 27.8% ee by chiral HPLC analysis (0.46 © 25 cm
CHIRALPAK IF column, EtOH/hexane/TFA = 10:90:0.1, 1.0
mL/min, 40 °C, tR 9, 11 min) of carboxylic acid (S,S)-23.
1
-(tert-Butyl) 2,3-Diethyl (2S*,3S*)-Cyclopropane-1,2,3-
tricarboxylate (7) (Table 4, run 1): With the same procedure
used for the synthesis of 19 (Table 3, run 2), an inseparable
mixture of cyclopropane tert-butyl diethyl esters 7 and 8 (7/8 =
9
2:8, 77.9 mg, 0.272 mmol, 94%) was obtained as a colorless
oil starting from diethyl maleate (15, 50.0 mg, 0.290 mmol) and
a solution of ylide 18 (188 mg, 0.871 mmol). The chromato-
graphic and spectroscopic data were in good agreement with
those shown above.
1
-(tert-Butyl) 2,3-Dimethyl (2S*,3S*)-Cyclopropane-
1
,2,3-tricarboxylate (14) (Table 4, run 2): With the same
procedure used for the synthesis of 19 (Table 3, run 2), an
inseparable mixture of cyclopropane tert-butyl dimethyl esters
1
4 and 15 (14/15 = 89:11, 88.8 mg, 0.344 mmol, 99%) was
obtained as a colorless oil starting from dimethyl maleate (21,
0.0 mg, 0.347 mmol) and a solution of ylide 18 (225 mg, 1.04
5
mmol). The chromatographic and spectroscopic data were in
good agreement with those shown above.
Diethyl
(1S,2S)-3-((1,3-Dioxoisoindolin-2-yl)methyl)-
1
-(tert-Butyl) 2,3-Dimethyl (2S*,3S*)-Cyclopropane-
cyclopropane-1,2-dicarboxylate ((S,S)-28) and Diethyl
(1R,2R)-3-((1,3-Dioxoisoindolin-2-yl)methyl)cyclopropane-
1,2-dicarboxylate ((R,R)-28): To a stirred solution of alcohol
(S,S)-24 (517 mg, 2.39 mmol) in benzene (24.0 mL) at rt were
added triphenylphosphine (941 mg, 3.59 mmol), phthalimide
(704 mg, 4.78 mmol), and diethyl azodicarboxylate (2.2 M in
toluene, 1.63 mL, 3.59 mmol). After 24 h, the mixture was con-
centrated under reduced pressure to give a residue, which was
purified by column chromatography on silica gel (20 g, EtOAc/
hexane = 29:71) to give N-alkyl phthalimide (S,S)-28 (773 mg,
2.24 mmol, 94%) as a yellowish oil. Purification by chiral
HPLC (0.46 © 25 cm CHIRALPAK IC column, EtOH/hex-
1
(
(
,2,3-tricarboxylate (14) and 1-(tert-Butyl) 2,3-Dimethyl
1r*,2R*,3S*)-Cyclopropane-1,2,3-tricarboxylate
(15)
Scheme 2): With the same procedure used for the synthesis
of 19 (Table 3, run 2), an inseparable mixture of cyclopropane
dimethyl esters 14 and 15 (14/15 = 55:45, 41.4 mg, 0.160
mmol, 52%) was obtained as a colorless oil starting from tert-
butyl methyl fumarate (10, 57.8 mg, 0.310 mmol) and a solu-
tion of ylide 22 (162 mg, 0.930 mmol). The chromatographic
and spectroscopic data were in good agreement with those
shown above.
1
0e
Diethyl (1R*,2R*)-3-(Hydroxymethyl)cyclopropane-1,2-
dicarboxylate (24) (Scheme 3):
To a stirred solution of
ane = 10:90, 1.0 mL/min, 40 °C, t 17 min) gave enantiomeri-
R
tert-butyl ester 7 (1.00 g, 3.49 mmol) in CH Cl (3.49 mL) at
cally pure (S,S)-28 (>99.9% ee). The enantiomer ((R,R)-28)
eluted at tR 20 min was also collected and determined as
>99.9% ee.
2
2
0
°C was added TFA (3.49 mL). After stirring at rt for 1 h, the
reaction mixture was concentrated under reduced pressure to
give crude carboxylic acid 23 (765.9 mg) as a yellow oil.
To a stirred solution of the crude carboxylic acid 23 (765.9
mg, 3.33 mmol) thus obtained above in THF (11.9 mL) at 0 °C
2
2:5
Data for the major N-alkyl (S,S)-phthalimide (S,S)-28: ½ꢀꢀ
D
+26.6 (c 0.257, CHCl ); IR (neat) 2982, 1774, 1718, 1189
3
¹
1 1
cm ; H NMR (400 MHz, CDCl ) δ 7.85­7.81 (m, 2 H), 7.72­
3
was added BH ¢THF (1.0 M in THF, 13.3 mL, 13.3 mmol).
7.68 (m, 2 H), 4.25 (dd, J = 14.2, 7.1 Hz, 2 H), 4.13­4.08 (m,
2 H), 4.06 (dd, J = 14.4, 7.4 Hz, 1 H), 3.89 (dd, J = 14.4, 7.4
Hz, 1 H), 2.41 (t, J = 5.3 Hz, 1 H), 2.32 (dd, J = 9.4, 4.8 Hz,
1 H), 2.09 (m, 1 H), 1.30 (t, J = 7.1 Hz, 3 H), 1.23 (t, J = 7.1
3
After 2 h, the reaction mixture was diluted with CHCl (10 mL),
3
washed with saturated aqueous NH Cl (12 mL), dried over
4
Na2SO4, and concentrated under reduced pressure. The resi-
due was purified by column chromatography on silica gel
1
3
Hz, 3 H); C NMR (100 MHz, CDCl ) δ 171.0, 169.4, 167.9
3
(
16 g, EtOAc/hexane = 5:5) to give alcohol 24 (375.9 mg, 1.74
(©2), 134.0 (©2), 132.0 (©2), 123.3 (©2), 61.4, 61.2, 34.5,
27.0, 26.9, 26.4, 14.1, 14.1; HRMS (ESI, positive) calcd for
mmol, 50%) as a yellowish oil. The spectroscopic data for 24
were identical to those we have reported previously.1
0b
+
C H NO [(M + H) ] 346.1285, found 346.1285.
1
8
20
6
Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan | 1321

Data for the minor N-alkyl (R,R)-phthalimide (R,R)-28:
22:3
References
½
ꢀꢀ
¹26.3 (c 0.237, CHCl ). Other data were identical to
3
D
those for the congener (S,S)­28.
1
D. Y.-K. Chen, R. H. Pouwer, J.-A. Richard, Chem. Soc.
Ethyl
(1S,5R,6S)-2-Oxo-3-azabicyclo[3.1.0]hexane-6-
Rev. 2012, 41, 4631.
carboxylate ((S,S)-29): To a stirred solution of phthalimide
2
a) K. Shimamoto, M. Ishida, H. Shinozaki, Y. Ohfune,
J. Org. Chem. 1991, 56, 4167. b) A. Reichelt, S. F. Martin, Acc.
Chem. Res. 2006, 39, 433.
(
S,S)-28 (8.47 mg, 0.0245 mmol) in EtOH (0.818 mL) at rt was
added hydrazine hydrate (0.00955 mL, 0.196 mmol). After
9 h, the mixture was concentrated under reduced pressure to
3
a) B.-L. Lu, L. Dai, M. Shi, Chem. Soc. Rev. 2012, 41,
4
3
2
318. b) M. Shi, J.-M. Lu, Y. Wei, L.-X. Shao, Acc. Chem. Res.
012, 45, 641. c) L. Jiao, Z.-X. Yu, J. Org. Chem. 2013, 78, 6842.
give a residue, which was purified by column chromatog-
raphy on silica gel (0.6 g, EtOAc/hexane = 40:60) to give
butyrolactam (S,S)-29 (3.59 mg, 0.0212 mmol, 87%) as a
yellowish oil: ½ꢀꢀ
d) T. F. Schneider, J. Kaschel, D. B. Werz, Angew. Chem., Int. Ed.
2014, 53, 5504. e) H. K. Grover, M. R. Emmett, M. A. Kerr, Org.
Biomol. Chem. 2015, 13, 655. f) D. Rosa, A. Nikolaev, N. Nithiy,
A. Orellana, Synlett 2015, 26, 441. g) A. P. Thankachan, K. S.
Sindhu, K. K. Krishnan, G. Anilkumar, Org. Biomol. Chem. 2015,
2
D
3:6
+79.9 (c 0.180, CHCl ); IR (KBr)
3
¹
1 1
3
5
1
1
1
7
4
210, 1715, 1680, 1186 cm ; H NMR (400 MHz, CDCl ) δ
3
.53 (br, 1 H), 4.14 (q, J = 7.1 Hz, 1 H), 4.14 (q, J = 7.1 Hz,
H), 3.60 (dd, J = 10.9, 5.5 Hz, 1 H), 3.43 (d, J = 10.9 Hz,
H), 2.45 (ddd, J = 6.1, 5.5, 2.7 Hz, 1 H), 2.34 (d, J = 6.1 Hz,
H), 1.79 (dd, J = 2.7, 2.7 Hz, 1 H), 1.25 (br dd, J = 7.1,
1
3, 8780. h) M. A. A. Walczak, T. Krainz, P. Wipf, Acc. Chem.
Res. 2015, 48, 1149. i) L. Yu, M. Liu, F. Chen, Q. Xu, Org.
Biomol. Chem. 2015, 13, 8379.
1
3
.1 Hz, 3 H); C NMR (100 MHz, CDCl ) δ 175.3, 170.8, 61.3,
3
4
H. Lebel, J.-F. Marcoux, C. Molinaro, A. B. Charette,
3.9, 28.0, 25.8, 23.5, 14.2; HRMS (ESI, positive) calcd for
Chem. Rev. 2003, 103, 977.
+
C H NO [(M + H) ] 170.0812, found 170.0812.
5
5323.
6
H. E. Simmons, R. D. Smith, J. Am. Chem. Soc. 1958, 80,
M. T. Rispens, E. Keller, B. de Lange, R. W. J. Zijlstra,
18
12
3
Ethyl
(1R,5S,6R)-2-Oxo-3-azabicyclo[3.1.0]hexane-6-
carboxylate ((R,R)-29): According to the same procedure for
the synthesis of (S,S)-29 shown above, deprotection of phthal-
imide (R,R)-28 (7.32 mg, 0.0212 mmol) gave butyrolactam
B. L. Feringa, Tetrahedron: Asymmetry 1994, 5, 607.
7
1353.
8
E. J. Corey, M. Chaykovsky, J. Am. Chem. Soc. 1965, 87,
(
½
R,R)-29 (3.35 mg, 0.0198 mmol, 93%) as a yellowish oil:
2
D
3:0
a) M. P. Doyle, M. N. Protopopova, Tetrahedron 1998, 54,
ꢀꢀ
¹79.3 (c 0.168, CHCl ). Other data were identical to
3
7
919. b) G. Bartoli, G. Bencivenni, R. Dalpozzo, Synthesis 2014,
6, 979.
R. Sakai, K. Suzuki, K. Shimamoto, H. Kamiya, J. Org.
Chem. 2004, 69, 1180.
a) M. Oikawa, S. Sasaki, M. Sakai, R. Sakai, Tetrahedron
those for the congener (S,S)-29.
4
(
S,S)-(N-Desmethyl)dysibetaine CPa Hydrochloride
(S,S)-25): A suspension of butyrolactam ethyl ester (S,S)-
9 (4.32 mg, 0.0255 mmol) in hydrochloric acid (1 M, 0.600
9
(
2
1
0
mL) was stirred at 70 °C for 16 h. The mixture was then concen-
trated by blowing air to give a residue, which was purified by
column chromatography on silica gel (ODS, 0.6 g, H O) to give
(
4
(
2
(
2
Lett. 2011, 52, 4402. b) M. Oikawa, S. Sasaki, M. Sakai, Y.
Ishikawa, R. Sakai, Eur. J. Org. Chem. 2012, 5789. c) M. Sakai, Y.
Ishikawa, S. Takamizawa, M. Oikawa, Tetrahedron Lett. 2013, 54,
2
S,S)-(N-desmethyl)dysibetaine CPa hydrochloride ((S,S)-25,
5
911. d) M. Sakai, K. Tanaka, S. Takamizawa, M. Oikawa,
2
D
2:3
.33 mg, 0.0221 mmol, 87%) as a yellow solid: ½ꢀꢀ
+58.4
Tetrahedron 2014, 70, 4587. e) K. Tanaka, M. Sakai, S.
Takamizawa, M. Oikawa, Chem. Lett. 2015, 44, 253.
1
c 0.317, H O); H NMR (400 MHz, D O) δ 3.25­3.13 (m, 2 H),
2 2
.25 (dd, J = 9.3, 4.8 Hz, 1 H), 2.12 (t, J = 5.6 Hz, 1 H), 1.95
11
S. I. Kozhushkov, A. Leonov, A. de Meijere, Synthesis
2003, 956.
V. K. Aggarwal, H. W. Smith, G. Hynd, R. V. H. Jones, R.
1
3
m, 1 H); C NMR (100 MHz, D O) δ 175.1, 173.7, 36.8, 27.3,
2
1
0b
6.9, 24.3. Other data were identical to those reported by us.
R,R)-(N-Desmethyl)dysibetaine CPa Hydrochloride
(R,R)-25): According to the same procedure for the synthesis
1
2
(
Fieldhouse, S. E. Spey, J. Chem. Soc., Perkin Trans. 1 2000, 3267.
13 C. D. Papageorgiou, S. V. Ley, M. J. Gaunt, Angew. Chem.,
Int. Ed. 2003, 42, 828.
(
of (S,S)-25 shown above, acidic hydrolysis of butyrolactam
ethyl ester (R,R)-29 (4.86 mg, 0.0287 mmol) gave (S,S)-(N-
desmethyl)dysibetaine CPa hydrochloride ((R,R)-25, 5.27 mg,
14 However, we have previously found that cyclopropanation
10b
of cyclic imide gives CCC­isomer. Works are in progress to
control the CTT/CCC selectivity.
2
D
3:4
0
.0269 mmol, 94%) as a yellow solid: ½ꢀꢀ
¹58.4 (c 0.264,
1
5
We speculate that, because of the absence of the counter
H O). Other data were identical to those for the congener (S,S)-
2
2
5 shown above, as well as those reported by us.1
0b
cation species, sulfonium ylide is more reactive than chloroacetate
which is accompanied by sterically demanding benzyltriethylam-
monium cation.
This work was financially supported by Grant-in-Aid for
1
6
C. D. Papageorgiou, M. A. Cubillo de Dios, S. V. Ley, M. J.
Scientific Research (26282216) from the Ministry of Educa-
tion, Science, Sports, Culture and Technology, Japan. We also
thank Ms. Mihoko Yamaguchi (Thermo Fisher Scientific K.K.)
for collecting HRMS data of some compounds.
Gaunt, Angew. Chem., Int. Ed. 2004, 43, 4641.
1
1
7
8
Y. Sugeno, Y. Ishikawa, M. Oikawa, Synlett 2014, 25, 987.
a) D. Janardanan, R. B. Sunoj, J. Org. Chem. 2007, 72, 331.
b) S. S. Bhella, M. Elango, M. P. S. Ishar, Tetrahedron 2009, 65,
40. c) H. K. Hall, Jr., K. Childers, T. Centeno-Hall, C. Contreras,
2
Supporting Information
B. Mazel, H. Menon, V. Nguyen, J. Robertson, R. Bates,
Tetrahedron Lett. 2012, 53, 4725. d) R. Herchl, M. Waser,
Tetrahedron Lett. 2013, 54, 2472.
Characterization data for new intermediates, HPLC profiles
for chiral resolution, and 13C NMR calculations are included.
This material is available on https://doi.org/10.1246/bcsj.
1
9
Another possible reason might be ascribed to the lower
2
0190096.
1322 | Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan

reactivity of maleates in the first Michael reaction; i.e., 1,3-allylic
strains in the maleic system, which would diminish conjugation at
the C­C and C­O double bonds, reduce electrophilicity of the
olefinic part.
Ram-de-Víu, Chem.®Eur. J. 2012, 18, 13920. d) A. Enríquez-
García, E. P. Kündig, Chem. Soc. Rev. 2012, 41, 7803.
23 The DP4+ methodology has been actively used for
structure elucidation of complex natural products with only one
set of experimental data, see; N. Grimblat, A. M. Sarotti, Chem.®
Eur. J. 2016, 22, 12246.
24 B. Neises, W. Steglich, Org. Synth. 1985, 63, 183.
25 A. Matsumoto, Y. Sano, M. Yoshioka, T. Otsu, J. Polym.
Sci., Part A: Polym. Chem. 1996, 34, 291.
2
0
Tutorial and User’s Guide, Spartan ’18 for Windows,
Macintosh and Linux, Wavefunction Inc., 2018.
N. Grimblat, M. M. Zanardi, A. M. Sarotti, J. Org. Chem.
015, 80, 12526.
a) I. Atodiresei, I. Schiffers, C. Bolm, Chem. Rev. 2007,
07, 5683. b) M. D. Díaz de Villegas, J. A. Gálvez, P. Etayo, R.
2
1
2
2
2
1
26 O. Meth-Cohn, C. Moore, H. C. Taljaard, J. Chem. Soc.,
Perkin Trans. 1 1988, 2663.
Badorrey, P. Lopez-Ram-de-Viu, Chem. Soc. Rev. 2011, 40, 5564.
c) M. D. Díaz-de-Villegas, J. A. Gálvez, R. Badorrey, M. P. López-
Bull. Chem. Soc. Jpn. 2019, 92, 1314–1323 | doi:10.1246/bcsj.20190096
© 2019 The Chemical Society of Japan | 1323