conformation, and two types of diastereomers (cis and
trans) are derived from the orientations of the azido and
carboxyl groups. The synthesis of β-azidocyclopropane
esters must address diastereoselectivity as well as enantios-
electivity. We are more interested in the cis-β-azidocyclo-
propane esters as they could be regarded as precursors of
cis-β-aminocyclopropane carboxylic acids (cis-β-ACCs),
which were widely used in peptide syntheses.3 Further-
more, the β-azidocyclopropane esters have two potential
chelating groups oriented in the same direction, offering
the possibility of chiral ligand design.
enantioselectivities were observed when R-substituted styr-
enes were used for the cyclopropanation.8
To test our proposal, the first attempt to attain the cis-
β-azidocyclopropane ester employed the cyclopropana-
tion of the benzyl diazophenylacetate (BnDPA) 1a and
R-azido-styrene 2a (Table 1). Several commercially avail-
able chiral dirhodium catalysts were screened (Table 1,
entries 1À4). Among them, Rh2(S-DOSP)4 and Rh2
(S-PTAD)4 gave positive results in hexane. The Rh2
(S-DOSP)4-catalyzed (0.2 mol %) cyclopropanation
of diazo ester 1a in the presence of azido alkene 2a
(5 equiv) at room temperature afforded the cis-β-azidocy-
clopropane ester 3aa in 79% yield (entry 3), and the dia-
stereoselectivity (96:4 dr) as well as the enantioselectivity
(85% ee) were controlled. The product from the Rh2(S-
PTAD)4-catalyzed reaction was obtained in an even higher
yield (89%) and enantioselectivity (89% ee); however,
the diastereoselectivity (83:17 dr) was imperfect (entry 4).
It should be noted that opposite enantioselectivity
was observed in this case, and a similar situation was
reported by Davies’s group with the cyclopropanation of
4-chlorostyrene withR-aryl-R-diazoketone.9 Further experi-
ments using Rh2(S-DOSP)4 and Rh2(S-PTAD)4 in hexane
at À5 °C revealed that Rh2(S-DOSP)4 was better than Rh2
(S-PTAD)4 (entry 5 vs 6). The yield of 3aa was improved to
95%, and the diastereoselectivity (98:2 dr) and enantio-
selectivity (89% ee) were slightly increased. If the tem-
perature was lowered (À10 °C), insolubility of diazo
ester 1a took place. After the catalyst was identified as
Rh2(S-DOSP)4, various solvents including diethyl ether,
toluene, DCM, and THF were investigated at À5 °C
(entries 7À10); however, no better results could be obtained
than that in hexane.
The asymmetric cyclopropanation of N-protected
enamines gave trans-β-aminocyclopropane esters in poor
diastereo- and enantioselectivities using chiral dirhodium
catalysts.4 Moreover, the attempted cyclopropanation of a
Boc-protected enamine by Doyle’s catalyst led to no
conversion.4a As prior solvents for many cyclopropana-
tions with rhodium catalysts were hydrocarbons, we con-
sidered that the low yields and diastereoselectivities as well
as enantioselectivities for the cyclopropanation of the
N-protected enamines with the diazo esters by chiral
dirhodium catalysts might be due to the following situa-
tions: the steric hindrance from the protection group, the
nucleophilicity of the protected-enamine nitrogen, and
poor solubility of the protected enamine in hydrocarbons.
The azido group was smaller than the protected-amino
group, and it generally showed weak nucleophilicity when
it existed in organic azide under neutral conditions.6 More-
over the alkyl azide was revealed to be better incorporated
with hydrocarbon solvents than protected enamine. With
this in mind, we envisioned that the azido alkenes would be
more suitable substrates than the protected-enamine ana-
logues for stereocontrolled cyclopropanation, which
would generate high enantioenriched β-N-functionalized
cyclopropane esters. In order to attain the more confor-
mationally restricted β-azidocyclopropane esters, we decided
to install two quaternary carbons on the cyclopropane ring.
One of the quaternary carbons was designed to attach to the
carbonyl group, and the other one was bound to the azido
nitrogen. The cis-β-azidocyclopropane esters would be
obtainedwhenR-azido-styrene analogueswereused, asthe
ester was the directing group in the cyclopropanation. For
the preparation of the enantioenriched β-azidocyclopropane
esters to be successful, there were two central issues to be
addressed: (1) could the azido alkenes be suitable for
cyclopropanation as no such alkenes were employed
previously;7 (2) would the cyclopropanation proceed in
a good stereocontrolled manner in the presence of a
chiral dirhodium catalyst, as poor diastereo- and
Table 1. Optimization of the Cyclopropanationa of 1a and 2a
t
yieldb
(%)
drc
eed
(%)
entry
Rh
solvent
(°C)
(c:t)
1
A
B
C
D
C
D
C
C
C
C
hexane
hexane
hexane
hexane
hexane
hexane
Et2O
rt
61
96:4
67:33
96:4
83:17
98:2
83:17
99:1
95:5
87:13
À
83
82
85
89
89
85e
78
81
52
À
2
rt
83
3
rt
79
4
rt
89
5
6
À5
À5
À5
À5
À5
À5
95
32
7
63
8
toluene
DCM
THFf
87
9
80
10
trace
(5) For the preparation of racemic β-azidocyclopropane esters in
moderate yield by the Michael initiated ring closure reaction, see: (a)
Mangelinckx, S.; Kimpe, N. D. Synlett 2006, 369–374. (b) Su, J.; Qiu, G.;
Liang, S.; Hu, X. Synth. Commun. 2005, 35, 1427–1433.
(6) The azido group shows nucleophilicity under acidic conditions;
ꢀ
for a selected example, see: Aube, J.; Milligan, G. L. J. Am. Chem. Soc.
a Cyclopropanation of diazo ester 1a (0.2 mmol) and azido alkene
2a (1.0 mmol) in the presence of chiral dirhodium catalyst (0.0004
mmol) for 3 h under the conditions mentioned in the table. b Isolated
yield after purification. c Determined from the crude reaction mixture
by 1H NMR. d Determined by chiral HPLC. e Enantioselectivity of
the product was observed to be opposite to that of other catalysts.
f Trace product was obtained after 24 h.
1991, 113, 8965–8966.
(7) We had examined the cyclopropanation of 2-nitroethenylbenzene
with various diazo esters using many dirhodium catalysts but failed to
obtain any desired β-nitrocyclopropane ester.
Org. Lett., Vol. 14, No. 9, 2012
2247