COMMUNICATION
Table 1. Optimization of enantioselective conjugate addition of cyclohex-
99% yield and with 99% ee (Table 1, entry 16). To our de-
light, the catalyst loading could be decreased from 10 to
2 mol% without decreasing the yield or enantioselectivity
(Table 1, entry 17). Notably, this process could also tolerate
air and moisture. The experimental procedure was very
simple, since neither ligand nor metal could promote the re-
action, therefore preparing the catalyst beforehand was un-
necessary (Table 1, entries 17–19). The experimental proce-
dure was as follows: the substrate, ligand, and metal were
weighed in a dry reaction tube followed by addition of
CH2Cl2 and the reaction mixture was stirred at room tem-
perature for 12 h.
Under the optimized reaction conditions (Table 1,
entry 17), the substrate scope for asymmetric conjugate ad-
dition of cyclic diketones to various b,g-unsaturated a-ke-
toesters was examined and the results are summarized in
Table 2. Substrate 1b with an ethyl group on the ester
moiety (R2) gave a 97% yield with 98% ee (Table 2,
entry 3). Additionally, dimedone (2b) is also a good nucleo-
phile for the reaction, and gave the product in 99% yield
with 97% ee (Table 2, entry 2). The substrates with electron-
withdrawing or donating groups at the meta-, para-, or
ortho-position of the aromatic ring were well tolerated in
terms of yield and enantioselectivity (up to 99% yield and
ane-1,3-dione (2a) to 1a.[a]
Entry
Ligand
Metal
La(OTf)3
Yb(OTf)3
(OTf)3
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
(OTf)2
Solvent
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
8
L1
L1
L1
L1
L1
L1
L1
L1
L2
L3
L4
L5
L6
L7
L8
L9
L9
L9
–
G
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
PhMe
CHCl3
THF
Et2O
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
94
78
95
85
41
79
40
77
99
77
80
71
57
92
90
99
11(S)[e]
E
41(S)[e]
Y
N
31(S)[e]
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
Cu
–
N
75(R)[e]
R
7
E
7
41
13
67
49
27
0
T
R
9
R
10
11
12
13
14
15
16
17[d]
18[d]
19[d]
R
N
G
R
0
E
65
69
99
99(R)[e]
–
G
G
C
99
no reaction
no reaction
Cu
G
–
[a] Unless otherwise noted, the reactions were performed with 2a
(0.10 mmol), 1a (0.12 mmol), N,N’-dioxide (0.01 mmol), metal
(0.01 mmol) in solvent (1.0 mL) at room temperature for 12 h. [b] Yield
of the isolated product. [c] Determined by HPLC analysis. [d] 2 mol%
Table 2. Substrate scope for catalytic asymmetric conjugate addition of
catalyst loading (0.004 mmol L9 and 0.004 mmol CuACTHUNRGTNEUNG(OTf)2), 1a
cyclic diketones to b,g-unsaturated a-ketoesters.[a]
(0.2 mmol), 2a (0.2 mmol) in CH2Cl2 (1.0 mL) at room temperature for
12 h. [e] Entries 1–3 S, the others R, the absolute configuration was deter-
mined by comparison to literature data.[6c]
choice of solvent had a pronounced effect on reactivity and
enantioselectivity. Other organic solvents including toluene,
CHCl3, THF, and Et2O all showed highly deleterious effects
on reactivity and enantioselectivity (less than 41% ee,
Table 1, entries 5–8).
Entry[a]
R1, R2
Product
Yield [%][b]
ee [%][c]
1
2
3
4
5
6
7
Ph, Me (1a)
Ph, Me (1a)
Ph, Et (1b)
3-MeC6H4, Me (1c)
4-MeC6H4, Me (1d)
3-MeOC6H4, Me (1e)
4-MeOC6H4, Me (1 f)
3a
99
99
97
99
93
99
91
99 (R)[e]
97 (R)[e]
98 (R)[e]
99
99
98
3a’[d]
3b
3c
Next we focused on the optimization of the reaction con-
ditions to improve the efficiency of CuACTHNUTRGENUG(N OTf)2 with other li-
3d
3e
3 f
gands. Various N,N’-dioxide ligands with different chiral
backbone moieties and amines were investigated. The steric
effect of the amide moiety played a crucial role in the enan-
tioselectivity of the reaction. Decreasing the steric hindrance
of the amide moiety led to a dramatic decrease in the enan-
tioselectivity (Table 1, entries 9–11).When L5 and L6 were
used (derived from aniline and benzylamine, respectively), a
racemic mixture was obtained (Table 1, entries 12 and 13,
respectively). The chiral backbone of the N,N’-dioxides also
had significant impact on the enantioselectivity. Neither L7
(derived from l-pipecolic acid) nor L8 (derived from l-ram-
ipril acid) gave better results than the l-proline-derived L1
(Table 1, entries 14 and 15). Surprisingly, remarkable im-
provement on activity and enantioselectivity was achieved
by shortening the linkage between the two chiral backbones.
Previously, we have found that the N,N’-dioxide ligand is op-
timized by incorporating a three-carbon linkage. However,
N,N’-dioxide L9 containing a two-carbon linkage gave 3a in
99
8
3g
96
99
9
4-PhC5H4, Me (1h)
2-ClC6H4, Me (1i)
3-ClC6H4, Me (1j)
4-ClC6H4, Me (1k)
2,4-Cl2C6H3, Me (1l)
3-BrC6H4, Me (1m)
4-BrC6H4, Me (1n)
4-FC6H4, Me (1o)
2-naphthyl, Me (1p)
2-thienyl, Me (1q)
PhCH=CH, Me (1r)
3h
3i
3j
3k
3l
3m
3n
3o
3p
3q
3r
95
96
99
90
73
93
87
99
99
92
67
99
91
99
99
93
99
98
98
99
98
95
10
11
12
13
14
15
16
17
18
19
[a] Unless otherwise noted, the reactions were performed with 2a
(0.20 mmol), (0.20 mmol), N,N’-dioxide L9 (2 mol%), Cu(OTf)2
(2 mol%) in CH2Cl2 (1.0 mL) and the reaction was stirred at room tem-
perature for 12 h. [b] Yield of the isolated product. [c] Determined by
HPLC analysis. [d] Dimedone 2b instead of 2a. [e] The absolute configu-
ration was determined by comparison to literature data.[6b,c]
1
ACHTUNGTRENNUNG
Chem. Eur. J. 2011, 17, 1118 – 1121
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1119