Table 1. Optimization of the asymmetric ring opening.[a]
dines.[6] Recently, the catalytic asymmetric meso-aziridine
ring opening with sulfur-based nucleophiles has also been
developed,[7] the chiral b-aminosulfur adducts of such reac-
tions have considerable usefulness both in drug develop-
ment and asymmetric catalysis (chiral ligand synthesis).[8]
However, the sulfur-based nucleophiles used in the enantio-
selective ring-opening reaction are mainly confined to thio-
phenols and their derivatives, and this substrate scope limi-
tation is a large obstacle of such methodology for the syn-
thesis of diverse chiral b-aminosulfur adducts. We envi-
sioned that an activated meso-aziridine would prefer to un-
dergo the sulfur-based nucleophilic attack of the thiooxazole
intermediate I2 rather than the enolate I1 generated from
the a-isothiocyanato imide (see Scheme 1) and such a reac-
tion not only can extend the scope of thiol nucleophiles in
the aziridine ring-opening process, but also provides b-ami-
nothiooxazole compounds with significant biological activi-
ties (Figure 1).[9] To explore the feasibility of the proposed
transformation, aziridine 1a and a-isothiocyanato imide 2a
were chosen for the model reaction. Gratifyingly, in the
presence of K2CO3 and catalytic tetra-n-butylammonium
Entry
1
Cat.
Base
T
Yield
[%][b]
ee
[8C]
[%][c]
1
2
3
4
5
6
7
8
1a
1a
1a
1a
1a
1a
1a
1b
1c
1d
1e
1a
1a
1a
4a
4a
4b
4c
4d
4e
4 f
4 f
4 f
4 f
4 f
4 f
4 f
4g
K2CO3 (aq) 30% wt.
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
K2CO3 (s)
Cs2CO3 (s)
Cs2CO3 (s)
Cs2CO3 (s)
RT
RT
RT
RT
RT
RT
RT
RT
RT
RT
RT
RT
À40
À25
93
98
99
85
93
95
96
95
94
98
99
99
93
96
40
50
48
43
19
55
66
3
36
47
35
70
92
92
9
10
11
12
13[d]
14[e]
Figure 1. Bioactive 2-thiooxazole or b-aminothiooxazole containing com-
pounds.
[a] Unless otherwise specified, the reaction was performed on a 0.1 mmol
scale with 1 (1.1 equiv), 2 (1.0 equiv), base (3.0 equiv), and catalyst
(20 mol%) in solvent (4 mL). The reaction time was 1 h. [b] Yield of iso-
lated product. [c] Determined by HPLC analysis on a chiral stationary
phase. [d] The reaction time is 72 h. [e] The reaction time is 18 h.
bromide (TBAB), the reaction proceeded smoothly in tol-
uene at room temperature, providing the adduct 3aa in
97% yield (Scheme 2).[10]
Encouraged by the successful synthesis b-aminothiooxa-
zole adduct 3aa, we decided to develop an asymmetric ver-
sion of such a transformation, since the acquirement of opti-
cally active products would be vital for its application on the
synthesis of bioactive compounds as well as chiral ligands.
As it was demonstrated above that the reaction proceeded
efficiently by using phase-transfer catalysis (PTC), our initial
investigation began with the reaction between meso-aziri-
dines 1 and isothiocyanato imide 2a in PTC conditions at
room temperature in the presence of a chiral quaternary
ammonium salt at a 20 mol% loading (Table 1).[11,12] To our
delight, readily accessible cinchona alkaloid-type catalyst 4a
in combination with solid K2CO3 gave the desired product
in excellent yield and a promising 50% enantiomeric excess
(ee) (Table 1, entry 2). However, the further steric tuning of
the catalyst 4a did not achieve a more satisfactory result
(entries 3, 4). To address the need for enhanced enatioselec-
tivity, we then screened various cinchona alkaloid-based di-
meric or trimeric quaternary ammonium salts, which may
provide a stronger chiral environment for the reaction tran-
sition state.[13] It was recognized that the distance between
the quaternary ammonium catalytic centers was crucial for
the stereocontrol, since although 4d afforded a very poor
result (entry 5), the catalyst 4e with a relatively shorter dis-
tance between ammonium ion gave the product in an in-
creased 55% ee (entry 6). The trimeric catalyst 4 f with a
bulkier environment gave the product with the best enatio-
selectivity (66% ee, entry 7). Variation of the protecting
group on meso-aziridines 1 had significant effect on the
result in terms of stereochemical control (entries 7–11), and
the 3,5-di-tert-butylbenzoyl-protected substrate 1b furnished
the product in an almost racemic form (entry 8). It was re-
vealed that the strong electron-withdrawing 3,5-dinitroben-
zoyl group was the most suitable one in this process. A fur-
ther increased ee was achieved when Cs2CO3 was used
(entry 12). Lowering the temperature had a dramatically in-
cremental effect on the enatioselectivity. When the reaction
&
2
&
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 0000, 00, 0 – 0
ÝÝ
These are not the final page numbers!