Efficient route to (S)-azetidine-2-carboxylic acid

Article abstract of DOI:10.1271/bbb.69.1892

A new and efficient route to (S)-azetidine-2-carboxylic acid (>99.9% ee) in five steps and total yield of 48% via malonic ester intermediates was established. As the key step, efficient four-membered ring formation (99%) was achieved from dimethyl (S)-(1′-methyl)benzylaminomalonate by treating with 1,2-dibromoethane (1.5 eq) and cesium carbonate (2 eq) in DMF. Krapcho dealkoxycarbonylation of dimethyl (1′S)-1-(1′-methyl) benzylazetidine-2,2-dicarboxylate, the product of this cyclization procedure, proceeded with preferential formation (2.7:1, 78% total yield) of the desired (2S,1′S)-monoester, with the help of a chiral auxiliary which was introduced on the nitrogen atom. The undesired (2R,1′S)-isomer could be converted to that with proper stereochemistry, by a deprotonation and subsequent reprotonation step. Finally, lipase-catalyzed preferential hydrolysis of the (2S,1′S)-monoester and subsequent deprotection provided enantiomerically pure (S)-azetidine-2-carboxylic acid in a 91% yield from the mixture of (2S,1′S)- and (2R,1′S)-isomers.

Full text of DOI:10.1271/bbb.69.1892

Biosci. Biotechnol. Biochem., 69 (10), 1892–1897, 2005
Efficient Route to (S)-Azetidine-2-carboxylic Acid
y
Yasuhiko FUTAMURA, Masayuki KUROKAWA, Rika OBATA, Shigeru NISHIYAMA, and Takeshi SUGAI
Department of Chemistry, Keio University, 3-14-1 Hiyoshi, Yokohama 223-8522, Japan
Received May 2, 2005; Accepted July 12, 2005
A new and efficient route to (S)-azetidine-2-carboxylic
acid (>99:9% ee) in five steps and total yield of 48% via
malonic ester intermediates was established. As the key
step, efficient four-membered ring formation (99%) was
achieved from dimethyl (S)-(10-methyl)benzylaminomal-
onate by treating with 1,2-dibromoethane (1.5 eq) and
cesium carbonate (2 eq) in DMF. Krapcho dealkoxy-
carbonylation of dimethyl (10S)-1-(10-methyl)benzylaze-
tidine-2,2-dicarboxylate, the product of this cyclization
procedure, proceeded with preferential formation
(2.7:1, 78% total yield) of the desired (2S,10S)-mono-
ester, with the help of a chiral auxiliary which was
introduced on the nitrogen atom. The undesired
(2R,10S)-isomer could be converted to that with proper
stereochemistry, by a deprotonation and subsequent re-
protonation step. Finally, lipase-catalyzed preferential
hydrolysis of the (2S,10S)-monoester and subsequent
deprotection provided enantiomerically pure (S)-azeti-
dine-2-carboxylic acid in a 91% yield from the mixture
of (2S,10S)- and (2R,10S)-isomers.
precursor (4a), and 2) the asymmetric Krapcho deal-
koxycarbonylation12) assisted by an ꢀ-methylbenzyl
chiral auxiliary.13)
Toward this end, DBU-catalyzed transesterification of
known aminomalonic ester 4b14) provided 4a (86%).15)
It was found from systematic studies that the proper
combination of the amine and haloester, in regard to
steric hindrance, nucleophilicity, electrophilicity, and
stability, was very important. For example, an attempt to
directly prepare 4a from (S)-ꢀ-methylbenzylamine and
dimethyl bromomalonate only resulted in very low
yield. An ꢀ-methyl substituent in the benzylamine
moiety increased the nucleophilicity, in spite of the
steric hindrance. Indeed, benzylamine16) itself showed
much lower reactivity toward alkylation.
The first keystep, azetidine ring formation, was
successful (99%) by applying 1,2-dibromoethane (1.5
eq) and cesium carbonate (2 eq) in DMF. The only
detected by-product was an eight-membered ring com-
pound (5). Steric hindrance of the diester moiety
affected the rate of cyclization. When the substrate
was changed to diethyl ester 4b, C-alkylated intermedi-
ate 6 was isolated in a substantial amount, and the
addition of tetra-n-butylammonium iodide (TBAI) was
necessary to ensure cyclization to azetidine 3b (95%).
Efficient deprotonation of the malonic ester was also
essential to promote the first step of cyclization. When a
weaker base, potassium carbonate, was applied to 4b,
the only isolable product was imine 7, which originated
from the initial attack of the amine lone-pair electron on
the bromine atom of 1,2-dibromoethane, accompanied
with ethylene formation and subsequent dehydrobromi-
nation of the resulting N-bromo intermediate.
Key words: cyclic amino acid; azetidine-2-carboxylate;
azetidine-ring formation; Krapcho deal-
koxycarbonylation; diastereofacially selec-
tive protonation
(S)-Azetidine-2-carboxylic acid (1a), a non-proteo-
genic cyclic amino acid, is the key component of
deoxymugineic acid and nicotianamine, which are
potent plant-origin promoters for the uptake of iron
from soil.1–3) Acid 1a also works as the starting material
for a nicotinic receptor tracer.4) Preparation of the pure
(S)-enantiomer has been achieved by synthesis from
chiral pools,5–7) optical resolution,8,9) and enzyme-
catalyzed kinetic resolution.10,11) Among these, Sumi-
tomo’s group has reported the separation of diaster-
eomers (2S,10S)- and (2R,10S)-2a,9) which had been
prepared in an equimolar ratio from racemic methyl 2,4-
dibromobutanoate. Our synthesis, inspired by the fore-
going intermediates, is shown in Scheme 1. The key
feature is the use of the (10S)-1-(10-methyl)benzylazeti-
dine-2,2-dicarboxylate ester (3a), a cyclic malonic acid
ester intermediate, which facilitated 1) the formation of
a four-membered ring, taking advantage of efficient
alkylation at the ꢀ-position of the acyclic malonic ester
Conventional Krapcho dealkoxycarbonylation on 3a
by applying sodium chloride in aqueous DMSO12) at as
high a temperature as 160 ꢀC only resulted in a moderate
yield of 2a (65%), probably due to undesired hydrolysis
of either the starting material or the product. The
attempted dealkoxycarbonylation of more stable diethyl
ester 3b under similar conditions was very slow. This
situation was improved (78%) by the combined use of
lithium chloride in DMSO17) under strictly anhydrous
conditions in the presence of molecular sieves 3A.
Instead of water, 2,6-di-tert-butyl-4-methylphenol
(BHT) was added as a proton source. The yields of
y
To whom correspondence should be addressed. Fax: +81-45-566-1697; E-mail: sugai@chem.keio.ac.jp

(S)-Azetidine-2-carboxylate
1893
Scheme 1. Reagents: a) DBU, MeOH, 86%; b) Cs2CO3 (2.0 eq), BrCH2CH2Br (1.5 eq), 99%; c) LiCl (3.2 eq), BHT (3.2 eq), 140 ꢀC, 78%;
d) 1) LDA, THF, ꢁ78 ꢀC; 2) aq. NH4Cl, ꢁ78 ꢀC, 82%.
(2S,10S)-2a and (2R,10S)-2a were 57% and 21%,
respectively, and both could be easily separated by
silica gel column chromatography. On the other hand,
attempts at selective hydrolysis18,19) of one of the two
esters prior to decarboxylation of the ꢀ-alkoxycarbonyl
acid only resulted in a complex mixture of highly polar,
water-soluble materials.
We have some comments on the preferential forma-
tion (2.7:1) of desired (2S,10S)-2a. The intermediate
from dealkoxycarbonylation was an enolate, and two
possible stable conformers, A and B, which are
postulated by means of the Newman projection, are
shown in Fig. 1. Protonation from the less sterically
hindered sides, the re-face on A and the si-face on B,
would provide (2S,10S)-2a and (2R,10S)-2a, respectively.
Gauche repulsion between the phenyl group in the ꢀ-
methylbenzyl chiral auxiliary and the ꢀ-oriented hydro-
gen atom at the azetidine C-4 position would reduce the
stability of conformer B, and protonation would then
preferentially proceed on conformer A. As a high
Fig. 1.

1894
Y. FUTAMURA et al.
Scheme 2. Reagents: a) Chirazyme L-2, H2O, quant.; b) H2, Pd-C, EtOH–H2O, 75%.; c) same as b, 91% from the mixture of (2S,10S)- and
(2R,10S)-2a.
temperature was required for the dealkoxycarbonylation
(160 ꢀC), the difference between the foregoing two
pathways is rather small (2.7:1). However, when the
same enolate was independently generated20) by the
action of LDA in THF on product 2a, protonation with
an aqueous ammonium chloride solution proceeded in a
6:1 ratio at 0 ꢀC. This ratio was further improved to 6.7:1
under a lower (ꢁ78 ꢀC) temperature. In turn, by treating
with sodium methoxide or DBU in refluxing methanol,
no interconversion between the two dealkoxycarbon-
ylation products, (2S,10S)-2a and (2R,10S)-2a, occurred.
This means that re-abstraction of the ꢀ-proton from the
resulting products in the presence of a proton source was
not feasible, even under basic conditions at high
temperature. The combined results suggest that the ratio
of the two products was determined under kinetic
control, and not under thermodynamic control, as
Krapcho dealkoxycarbonylation proceeded under nearly
neutral conditions.
Based on the foregoing comments, the property of the
proton source would be important under non-aqueous
conditions. Among such bulky and hydrophobic candi-
dates as BHT, 2,4,6-trichlorophenol and Amberlyst-15,
only BHT worked well as was shown before. The
similar phenol, (S)-binaphthol, also promoted the reac-
tion (23%), but the existing chirality of the proton
source21) did not affect the product ratio (2.8:1).
by (2R,10S)-2b (ca. 3%). This was eventually converted
to (S)-1 (91% from the mixture of 2a) with >99:9% ee,
after recrystallization at the final stage.10) As already
stated, undesired (2R,10S)-2a was epimerized into a
6.7:1 mixture of (2S,10S)- and (2R,10S)-2a in an 82%
yield.
In conclusion, a new and efficient route to (S)-
azetidine-2-carboxylic acid (1a) via the malonic ester
intermediates, 3a and 4a, was established.
Experimental
All boiling point (bp) and melting point (mp) data are
uncorrected. IR spectra were measured as films for oil
and as KBr disks for solids with a Jasco FT/IR-410
1
spectrometer. H- and 13C-NMR spectra were measured
with a Jeol JNM GX-270 or GX-400 spectrometer.
High-resolution mass spectra were recorded by a Jeol
JMS-700 spectrometer. HPLC data were recorded by an
SSC-5410 liquid chromatograph (Senshu Scientific Co.,
Ltd.). Optical rotation data were recorded on a Jasco
DIP 360 polarimeter. Silica gel 60 (spherical, 100–
210 mm) from Kanto Chemical Co. was used for column
chromatography. Preparative TLC was performed with
E. Merck silica gel 60 F256 plates (0.5 mm thickness,
No. 5744).
Product (2S,10S)-2a isolated by chromatography was
hydrolyzed with Candida antarctica lipase B (Chira-
zyme L-2)8) under neutral conditions to provide acid
(2S,10S)-2b in a 75% yield. Finally, catalytic hydro-
genolysis of the chiral auxiliary enabled (S)-azetidine-2-
carboxylic acid (1a) to be obtained in quantitative yield
(Scheme 2). Its high ee (>99:9%) was confirmed by an
HPLC analysis of corresponding ester (S)-1b, to which a
benzoyl group had been introduced on the free amine.
Lipase-catalyzed hydrolysis also worked in a diaster-
eoselective manner on the 2.7:1 mixture of (2S,10S)- and
(2R,10S)-2a to give (2S,10S)-2b with some contamination
Dimethyl (10S)-(10-methyl)benzylaminomalonate (4a).
Known crude diester 4b14) (1.30 g, 4.68 mmol) was
dissolved in MeOH (65 ml), and 1,8-diazabicyclo-
[5.4.0]undec-7-ene (DBU, 70 ml, 0.47 mmol) was added
dropwise at room temperature, before the mixture was
stirred under reflux for 3 h. The reaction mixture was
concentrated in vacuo. The residue was poured into a
phosphate buffer (0.1 M, pH 8.0, 100 ml), and the
mixture was extracted three times with ethyl acetate.
The extracts were combined, washed with brine, dried
with anhydrous sodium sulfate, and concentrated in
vacuo to give 4a (1.01 g, 86%, from diethyl bromomal-

(S)-Azetidine-2-carboxylate
1895
onate) as a yellow oil. This was employed for the next
step without further purification.
A small portion was purified by preparative TLC
added, and the mixture was heated to 40 ꢀC for 7 h. A
similar workup and purification to that described for 3a
gave 3b as a pale yellow oil (417 mg, 95%). IR ꢁmax
cmꢁ1: 1761, 1732, 1252, 1101; NMR ꢂH (270 MHz,
CDCl3): 1.26 (6H, t, J ¼ 7:1 Hz), 1.28 (3H, d, J ¼
6:4 Hz), 2.39–2.55 (2H, m), 3.00 (1H, ddd, J ¼ 3:0, 6.5,
7.9 Hz), 3.14 (1H, ddd, J ¼ 7:4, 7.4, 7.9 Hz), 3.73 (1H,
q, J ¼ 6:4 Hz), 4.06–4.39 (4H, m), 7.14–7.34 (5H, m).
When the reaction was quenched prior to cyclization
under a low dose of TBAI, the C-alkylated form (6) was
isolated as an intermediate: NMR ꢂH (270 MHz, CDCl3):
1.10 (3H, t, J ¼ 7:1 Hz), 1.22 (3H, t, J ¼ 7:1 Hz), 1.34
(3H, d, J ¼ 6:8 Hz), 2.14–2.26 (1H, m), 2.43–2.54 (1H,
m), 3.05–3.17 (2H, m), 3.73 (1H, q, J ¼ 6:8 Hz), 7.17–
7.26 (5H, m).
[developed with hexane–EtOAc (4:1) with a trace of
22
triethylamine]. ½ꢀꢂ
ꢁ64:5ꢀ (c 1.1, EtOH); IR ꢁmax
D
cmꢁ1: 3338, 1757, 1739, 1225, 1153; NMR ꢂH (270
MHz, CDCl3): 1.33 (3H, d, J ¼ 6:6 Hz), 2.18 (1H, br),
3.62 (3H, s), 3.69 (3H, s), 3.71 (1H, q, J ¼ 6:6 Hz), 3.87
(1H, s), 7.15–7.29 (5H, m); NMR ꢂC (67.5 MHz, CDCl3):
24.46, 52.67 (ꢃ2), 56.41, 62.54, 126.70 (ꢃ2), 127.24,
128.36 (ꢃ2), 143.47, 168.42, 169.33. The IR and NMR
spectra were in good accordance with those of the
racemate reported previously.14)
Dimethyl (10S)-1-(10-methyl)benzylazetidine-2,2-di-
carboxylate (3a). Cesium carbonate (134 mg, 0.41
mmol) was added to a solution of 4a (45.7 mg, 0.18
mmol) in DMF (0.7 ml), and the mixture was stirred for
10 min at room temperature. 1,2-Dibromoethane (23 ml,
0.267 mmol) was added, and the reaction mixture was
stirred for 14 h at room temperature and then for 22 h at
40 ꢀC. The reaction mixture was next diluted with a
phosphate buffer (0.1 M, pH 7.5) and extracted with
ethyl acetate. The extract was washed with brine, dried
over anhydrous sodium sulfate, and concentrated in
vacuo to give a crude product. Purification by prepara-
tive TLC [developed with hexane–EtOAc (4:1) with a
trace of triethylamine] afforded 3a (49.9 mg, 99%) as a
yellow oil. Bulb-to-bulb distillation gave an analytical
Another by-product with an imine structure (7): NMR
ꢂH (270 MHz, CDCl3): 1.47 (3H, d, J ¼ 6:4 Hz), 1.23
(3H, t, J ¼ 7:3 Hz), 1.24 (3H, t, J ¼ 7:3 Hz), 4.24 (2H,
q, J ¼ 7:3 Hz), 4.27 (2H, q, J ¼ 7:3 Hz), 4.58 (1H, q,
J ¼ 6:6 Hz), 7.10–7.28 (5H, m).
Methyl (2S,10S)-1-(10-methyl)benzylazetidine-2-car-
boxylate (2a) and methyl (2R,10S)-1-(10-methyl)benzyl-
azetidine-2-carboxylate (2a). Lithium chloride (23 mg,
0.54 mmol), BHT (120 mg, 0.55 mmol) and molecular
sieves 3A (150 mg) were added to a solution of 3a
(47 mg, 0.17 mmol) in DMSO (0.25 ml), and the mixture
was stirred for 1 h at room temperature. The reaction
mixture was then heated at 140 ꢀC for a further 2 h and,
after cooling, was diluted with a phosphate buffer
(pH 6.0, 0.1 M) and extracted three times with ethyl
acetate. The combined organic layer was washed with
brine, dried over anhydrous sodium sulfate, and con-
centrated in vacuo. The residue was purified by
preparative thin-layer column chromatography [devel-
oped with hexane–EtOAc (2:1) with a trace of trieth-
ylamine] to afford a mixture of (2S,10S)-2a and (2R,10S)-
2a (total 27.2 mg, 78%) as a pale yellow oil. The
sample. Bp 160–180 ꢀC at 0.15 mmHg (bath temp.);
21
½ꢀꢂ
ꢁ112:5ꢀ (c 1.0, EtOH); IR ꢁmax cmꢁ1: 1763,
D
1734, 1255, 1103; NMR ꢂH (270 MHz, CDCl3): 1.26
(3H, d, J ¼ 6:3 Hz), 2.41–2.56 (2H, m), 2.98 (1H, ddd,
J ¼ 3:0, 6.5, 7.9 Hz), 3.12 (1H, ddd, J ¼ 7:4, 7.4,
7.9 Hz), 3.70 (1H, q, J ¼ 6:3 Hz), 3.73 (3H, s), 3.76 (3H,
s), 7.20–7.33 (5H, m); NMR ꢂC (67.5 MHz, CDCl3):
21.83, 25.87, 48.40, 52.10, 52.17, 61.86, 72.99, 126.98,
127.47 (ꢃ2), 127.92 (ꢃ2), 142.67, 169.40, 170.80. Anal.
Found: C, 64.81; H, 6.94; N, 4.97%. Calcd. for
C15H19NO4: C, 64.97; H, 6.91; N, 5.05%.
1
diastereomeric ratio (2.7:1) was estimated from the H-
NMR spectra as shown next. These two components
were further purified by the same preparative TLC and
subsequent bulb-to-bulb distillation.
Eight-membered by-product resulting from the double
C- and N-alkylation (5): NMR ꢂH (400 MHz, CDCl3):
1.37 (6H, d, J ¼ 6:8 Hz), 2.27 (2H, ddd, J ¼ 5:8, 9.8,
15.1 Hz), 2.56 (2H, ddd, J ¼ 5:9, 9.6, 15.1 Hz), 3.12
(2H, ddd, J ¼ 5:8, 9.6, 9.9 Hz), 3.23 (2H, ddd, J ¼ 5:9,
9.8, 9.9 Hz), 3.46 (6H, s), 3.63 (1H, q, J ¼ 6:8 Hz), 3.74
(6H, s), 7.20–7.35 (10H, m); FAB-HRMS (Mþ þ Naþ,
m=z): 577.2502; calcd. for C30H38N2O8Na, 577.2526.
(2S,10S)-2a (Rf 0.40): Bp 150–170 ꢀC at 0.18 mmHg
22
(bath temp.); ½ꢀꢂ
ꢁ118:7ꢀ (c 1.1, EtOH); IR ꢁmax
D
cmꢁ1: 1753, 1730, 1196, 1174; NMR ꢂH (270 MHz,
CDCl3): 1.21 (3H, d, J ¼ 6:6 Hz), 2.16 (1H, dddd,
J ¼ 2:8, 8.0, 8.4, 8.6 Hz), 2.25 (1H, dddd, J ¼ 8:0, 8.0,
8.4, 8.6 Hz), 2.78 (1H, ddd, J ¼ 8:0, 8.0, 8.0 Hz), 3.09
(1H, ddd, J ¼ 2:8, 8.0, 8.0 Hz), 3.43 (1H, q, J ¼ 6:6 Hz),
3.74 (3H, s), 3.74 (1H, dd, J ¼ 8:4, 8.4 Hz), 7.18–7.33
(5H, m); NMR ꢂC (67.5 MHz, CDCl3): 20.85, 20.99,
49.62, 51.87, 63.89, 67.21, 127.04, 127.30 (ꢃ2), 128.14
(ꢃ2), 142.27, 173.44. Anal. Found: C, 71.26; H, 7.74; N,
6.35%. Calcd. for C13H17NO2: C, 71.21; H, 7.81; N,
Diethyl (10S)-1-(10-methyl)benzylazetidine-2,2-dicar-
boxylate (3b). Cesium carbonate (2.11 g, 6.47 mmol)
was added to a solution of 4b (408 mg, 1.42 mmol) in
DMF (5.0 ml), and the mixture was stirred for 10 min at
room temperature. 1,2-Dibromoethane (0.25 ml, 2.90
mmol) and tetra-n-butylammonium iodide (TBAI,
55 mg, 0.15 mmol) were then added, and the reaction
mixture was stirred for 15.5 h at room temperature.
Another portion of TBAI (479 mg, 1.30 mmol) was
1
6.39%. The H-NMR spectrum was identical with that
reported previously.9)
(2R,10S)-2a (Rf 0.15): Bp 150–170 ꢀC at 0.18 mmHg
19
(bath temp.); ½ꢀꢂ
þ48:9ꢀ (c 1.1, EtOH); IR ꢁmax
D

1896
Y. FUTAMURA et al.
21
200 ꢀC (dec.) [lit.7) 210 ꢀC]; ½ꢀꢂ D ꢁ121:7ꢀ (c 0.5, H2O)
cmꢁ1: 1745, 1196, 1173; NMR ꢂH (270 MHz, CDCl3):
1.27 (3H, d, J ¼ 6:6 Hz), 2.12 (1H, dddd, J ¼ 3:0, 8.1,
8.8, 9.2 Hz), 2.28 (1H, dddd, J ¼ 9:2, 9.2, 9.2, 9.2 Hz),
2.99 (1H, ddd, J ¼ 8:1, 8.1, 9.2 Hz), 3.31 (3H, s), 3.34
(1H, q, J ¼ 6:6 Hz), 3.56 (1H, ddd, J ¼ 3:0, 8.1, 9.2
Hz), 3.56 (1H, dd, J ¼ 8:8, 9.2 Hz), 7.17–7.27 (5H, m);
NMR ꢂC (67.5 MHz, CDCl3): 19.74, 20.85, 50.74, 51.45,
64.40, 68.09, 127.31, 127.79 (ꢃ2), 127.86 (ꢃ2), 141.42,
172.38. Anal. Found: C, 71.13; H, 7.78; N, 6.30%.
Calcd. for C13H17NO2: C, 71.21; H, 7.81; N, 6.39%.
The NMR spectrum was identical with that reported
previously.9)
21
[lit.7) ½ꢀꢂ
ꢁ120ꢀ (c 1, H2O); IR ꢁmax cmꢁ1: 2976,
D
2688, 2517, 1637, 1593, 1410, 1306, 1288, 1254; NMR
ꢂH (270 MHz, D2O): 2.28–2.44 (1H, m), 2.54–2.69 (1H,
m), 3.75 (1H, ddd, J ¼ 6:0, 10.0, 10.0 Hz), 3.91 (1H,
ddd, J ¼ 8:8, 10.0, 10.0 Hz), 4.62 (1H, dd, J ¼ 8:3,
10.2 Hz). All the physical and spectral properties were
coincident with those of a commercially available
authentic sample (Sigma, A0760).
A small portion of (S)-1a was converted to corre-
sponding N-benzoyl methyl ester 1b by successively
treating with benzoyl chloride in an aqueous sodium
hydroxide solution and then with diazomethane in
diethyl ether in the conventional manner to determine
its enantiomeric excess (ee). HPLC: column, Daicel
Chiralcel OD, 0:46 cm ꢃ 25 cm; eluent, hexane-isoprop-
yl alcohol = 7/1; flow rate, 0.5 ml/min; Rt 63.4 min
[for (S)-1b as a single peak]. Rt 56.7 min for (R)-1b.
Isomerization of (2R,10S)-2a. An LDA solution was
prepared by adding n-butyllithium (2.7 M in hexane,
150 ml, 0.41 mmol) to a solution of diisopropylamine
(117 ml, 0.83 mmol) in THF (0.5 ml) at 0 ꢀC. After
cooling to ꢁ78 ꢀC, a solution of (2R,10S)-2a (45.4 mg,
0.21 mmol) in THF (0.5 ml) was added, and the mixture
was stirred for 2 h. The reaction was quenched by adding
MeOH at ꢁ78 ꢀC, and the conventional workup and
purification already described to give a mixture of
(2S,10S)- and (2R,10S)-2a (37.1 mg, 82%, 6.7:1).
(S)-Azetidine-2-carboxylic acid (1a) from a mixture of
(2S,10S)- and (2R,10S)-2a. Chirazyme L-2 (c–f, 117 mg)
was added to a mixture of 2a [115 mg, total 0.52 mmol,
(2S,10S)/(2R,10S) = 2.7:1 by NMR] in water (2.5 ml),
which had been ultrasonically emulsified in advance.
The mixture was stirred for 10 h at 17 ꢀC. The reaction
mixture was filtered, and the resulting filtrate was
extracted several times with toluene to remove the
unreacted materials. Concentration of the combined
organic layer and purification by preparative thin-layer
column chromatography yielded 2a (23.4 mg, 20%
recovery). Its NMR spectrum showed that the unreacted
material was the pure (2R,10S)-isomer.
The aqueous layer was concentrated in vacuo to give
(2S,10S)-2b (108 mg, quant.) as a pale yellow amorphous
solid. This contained (2R,10S)-2b (ca. 3%), judging from
its NMR spectrum with the signal at ꢂH 3.54 (1H, ddd,
J ¼ 8:9, 8.9, 8.9 Hz). This was hydrogenated in a similar
manner to that just described to give crude 1a (51.1 mg,
quant.). Preferential crystallization from methanol
(0.5 ml) afforded (S)-1a (35.2 mg, 91% from the mixture
of 2a), whose ee was >99:9% after conversion to 1b as
already described. The acid (S)-1a recovered from
the mother liquor of the final crystallization showed
87.5% ee.
(2S,10S)-1-(10-Methyl)benzylazetidine-2-carboxylic
acid (2b) from pure (2S,10S)-2a. Chirazyme L-2 (c–f,
40 mg) was added to (2S,10S)-2a (36.4 mg, 0.12 mmol)
in water (1.5 ml), which had been ultrasonically emulsi-
fied in advance. The mixture was stirred for 2 h at 40 ꢀC.
The reaction mixture was then filtered, and the filtrate
was extracted with diethyl ether. The aqueous layer was
concentrated in vacuo to give (2S,10S)-2b (38.7 mg,
20
quant) as a pale yellow amorphous solid. ½ꢀꢂ
ꢁ48:9ꢀ
D
(c 1.0, EtOH); IR ꢁmax cmꢁ1: 3411, 1626, 1404; NMR
ꢂH (270 MHz, CDCl3): 1.69 (3H, d, J ¼ 6:9 Hz), 2.40
(1H, dddd, J ¼ 3:6, 9.3, 9.6, 9.9 Hz), 2.66 (1H, dddd,
J ¼ 9:3, 9.3, 9.6, 9.9 Hz), 3.34 (1H, ddd, J ¼ 9:3, 9.3,
9.3 Hz), 3.89 (1H, ddd, J ¼ 3:6, 9.3, 9.3 Hz), 4.25 (1H,
q, J ¼ 6:9 Hz), 4.42 (1H, dd, J ¼ 9:6, 9.6 Hz), 7.34–7.52
(5H, m); NMR ꢂC (67.5 MHz, CDCl3): 18.57, 21.04,
47.43, 65.12, 66.86, 121.94, 129.08 (ꢃ2), 129.13 (ꢃ2),
135.38, 170.84; FAB-HRMS (Mþ þ Naþ, m=z):
1
228.0974; calcd. for C12H15NO2Na, 228.1001. The H-
NMR spectrum was identical with that reported pre-
viously.9)
Acknowledgment
(S)-Azetidine-2-carboxylic acid (1a). N-Protected
amino acid 2b (112 mg, 0.55 mmol) was dissolved in a
mixture of water and ethanol (1:1, 2.5 ml). Palladium on
carbon (10%, 120 mg) was added, and the mixture was
vigorously stirred under hydrogen for 46 h. The resulting
suspension was filtered, and the solid residue was
washed with water. The solvent was evaporated in
vacuo, and the solid residue (41.8 mg, 75%) was
ultrasonically suspended in methanol (0.5 ml), before
the mixture was left to stand overnight at ꢁ20 ꢀC. The
solid was recovered and washed with methanol to afford
1a (30.0 mg, 54%) as colorless fine needles. Mp above
This work formed part of the 21st century COE
program (Keio LCC) from the Ministry of Education,
Culture, Sports, Science, and Technology of Japan. This
work was also supported by the collaborative program
entitled ‘‘Solution-Oriented Research for Science and
Technology (SORST)’’ of Japan Science and Technol-
ogy Corporation, and we express our sincere thanks to
Professors Keisuke Suzuki and Takashi Matsumoto, and
to Dr. Ken Ohmori of the Tokyo Institute of Technol-
ogy. Part of this work was also supported by grant-aid
for scientific research (No. 14560084).

(S)-Azetidine-2-carboxylate
1897
esters and related compounds in dipolar aprotic media—
part 1. Synthesis, 805–822 (1982).
References
´
1) Ripperger, H., and Schreiber, K., The ‘‘normalizing
factor’’ for the tomato mutant chloronerva. XXV (1).
Synthesis of 2-epinicotianamine and nicotianamine. J.
Prakt. Chem., 329, 231–234 (1987).
2) Shioiri, T., Irako, N., Sakakibara, S., and Hamada, Y., A
new efficient synthesis of nicotianamine and 20-deoxy-
mugineic acid. Heterocycles, 44, 519–530 (1997).
3) Miyakoshi, K., Oshita, J., and Kitahara, T., Expeditious
synthesis of nicotianamine and 20-deoxymugineic acid.
Tetrahedron, 57, 3355–3360 (2001).
4) Brenner, E., Baldwin, R. M., and Tamagnan, G.,
Synthesis of a new precursor to the nicotinic receptor
tracer 5-IA-85380 precursor using trimethylsilyl iodide
as deblocking agent. Tetrahedron Lett., 45, 3607–3610
(2004).
5) Kondo, T., and Ueyama, N., process for producing
optically active azetidine-2-carboxylic acid. PCT Int.
Appl. WO 2003014081 (2003); CA 138: 170068.
6) Takano, N., Ushio, H., and Seko, S., Preparation of
optically active N-substituted azetidine-2-carboxylic
acids. Jpn. Kokai Tokkyo Koho JP 10120648 (1998);
CA 129: 41069.
7) Miyoshi, M., Sugano, H., Fujii, T., Ishihara, T., and
Yoneda, N., A novel synthesis of optically active
azetidine-2-carboxylic acids. Chem. Lett., 5–6 (1973).
8) Rodebaugh, R. M., and Cromwell, N. H., Resolution of
DL-azetidine-2-carboxylic acid. J. Heterocycl. Chem., 6,
993–994 (1969).
9) Ushio, H., Takano, N., Honda, Y., Seko, S., and Hazama,
M., Preparation of optically active N-alkylbenzylazeti-
dine-2-carboxylic acids. Eur. Pat. Appl. EP 827954
(1998); CA 128: 230235.
10) Kudo, J., Hazama, M., and Hirata, N., Process for
producing optically active azetidine-2-carboxylic acid
derivative. Eur. Pat. Appl. EP 855446 (1998); CA 129:
160706.
11) Starmans, W. A. J., Doppen, R. G., Thijs, L., and
Zwanenburg, B., Enzymatic resolution of methyl N-
alkyl-azetidine-2-carboxylates by Candida antarctica
lipase-mediated ammoniolysis. Tetrahedron: Asymme-
try, 9, 429–435 (1998).
13) Juaristi, E., Leon-Romo, J. L., Reyes, A., and Escalante,
J., Recent applications of ꢀ-phenylethylamine (ꢀ-PEA)
in the preparation of enantiopure compounds. Part 3: ꢀ-
PEA as chiral auxiliary. Part 4: ꢀ-PEA as chiral reagent
in the stereodifferentiation of prochiral substrates.
Tetrahedron: Asymmetry, 10, 2441–2495 (1999).
14) Sugiyama, S., Watanabe, S., Inoue, T., Kurihra, R., Itou,
T., and Ishii, K., A novel dynamic kinetic resolution
accompanied by intramolecular transesterification:
asymmetric synthesis of 4-hydroxymethyl-2-oxazolidi-
none from serinol derivatives. Tetrahedron, 59, 3417–
3425 (2003).
15) Husinec, S., Juranic, I., Liobera, A., and Porter, A. E. A.,
Bis(methoxycarbonyl)carbene insertion into N–H bonds:
a facile route to N-substituted aminomalonic esters.
Synthesis, 721–723 (1988).
´
16) Simig, G., Doleschall, G., Hornyak, G., Fetter, J.,
Lempert, K., Nyitrai, J., Huszthy, P., and Gizur, T.,
Simple and condensed ꢃ-lactams-I. The application of
diketene in ꢃ-lactam synthesis. The synthesis and
functional group manipulations of diethyl 3-acetyl-4-
oxoacetidine-2,2-dicarboxylate. Tetrahedron, 40, 479–
484 (1984).
17) Qian, X., Zheng, B., Burke, B., Saindane, M. T., and
Kronenthal, D. R., A Stereoselective synthesis of BMS-
262084, an azetidinone-based tryptase inhibitor. J. Org.
Chem., 67, 3595–3600 (2002).
18) Toshima, H., and Ichihara, A., Practical stereoselective
syntheses of all four stereoisomers of coronamic acid (2-
ethyl-1-aminocyclopropane-1-carboxylic acid). Biosci.
Biotechnol. Biochem., 59, 497–500 (1995).
19) Niwayama, S., Highly efficient selective monohydrolysis
of symmetric diesters. J. Org. Chem., 65, 5834–5836
(2000).
20) Wasserman, H. H., Lipshutz, B. H., Tremper, A. W., and
Wu, J. S., Preparation of ꢃ-lactams from azetidine-2-
carboxylic acids and esters. J. Org. Chem., 46, 2991–
2999 (1981).
21) Duhamel, L., Duhamel, P., and Plaquevent, J.-C.,
Enantioselective protonations: fundamental insights and
new concepts. Tetrahedron: Asymmetry, 15, 3653–3691
(2004).
12) Krapcho, P. A., Synthetic applications of dealkoxycar-
bonylations of malonate esters, ꢃ-keto esters, ꢀ-cyano