H. Zheng et al. / Tetrahedron Letters 48 (2007) 7934–7937
7935
a remarkable influence on the results of the reaction.
ferred that the methyl group on nitrogen might benefit
to the enantioselectivity of the reaction. In search of
more effective catalysts, we synthesized N-methyl deriv-
atives 2c, 2d, and 2e. As we expected, introduction of
methyl group on nitrogen of catalysts 1a and 1b led to
significant increase in enantioselectivity (Table 1, entries
1–4, vs 17–21). Meanwhile, it resulted in inversion of the
configuration of the product. However, N-methyl deriv-
ative 2e directed the reaction with almost racemic prod-
uct (Table 1, entries 22 and 23). It appears that the effect
of the structure of catalyst on the reaction is rather com-
plicated. Moreover, methylation of the hydroxyl group
of 2a resulted in an obvious drop in enantioselectivity
(Table 1, entry 24). It seems that the H-bonding between
catalyst and substrate also makes an important contri-
(
1
1R,2S)-1,2-Diphenyl-2-aminoethanol derived catalyst
a gave the product with good yields but poor enantio-
selectivities (Table 1, entries 1 and 2). Its diastereomer
b provided the product with slightly higher ee values
Table 1, entries 3 and 4). 1c which has only one chiral
center bonded to amino group gave much better result
Table 1, entry 5) in dichloromethane. However, 1d
1
(
(
which has an isopropyl group on carbon 1 gave only
racemic product (Table 1, entry 8). Catalyst 1e which
has only one chiral center bonded to hydroxy group
gave almost racemic product (Table 1, entry 9). There-
fore, it can be concluded that the chirality of the carbon
bonded to amino group is determinant for the enantio-
selectivity of the reaction.
6
b
bution to enantioselectivity of the reaction.
Then catalysts 2a and 2b derived from ephedrine and
pseudoephedrine were employed in the reaction. Fortu-
nately, we found that 2a resulted in much higher enantio-
selectivities (Table 1, entries 11 and 14) in chloroform.
Its diastereomer 2b was obviously inferior to 2a but
superior to 1a–e (Table 1, entries 15 and 16). We in-
The solvents also show dramatic effect on the reaction.
Generally, chlorinated solvents are advantageous to
the reaction. For most of the catalysts, chloroform is
superior to dichloromethane except for 1c, which
resulted in much higher enantioselectivity in dichloro-
methane than in chloroform (Table 1, entries 5 and 6).
When the reaction was carried out in toluene, both the
reactivity and the enantioselectivity dropped greatly
Table 1. Survey of chiral catalysts for the hydrosilylation of ketimine
4a
(
Table 1, entries 7 and 12). We conjectured that the
Ph
Ph
p–p interaction of catalyst with substrate was blocked
in toluene, which led to low enantioselectivity. The use
of non-protonic polar solvent THF also led to decrease
in enantioselectivity (Table 1, entry 13). Perhaps it is due
to the competitive coordination of THF with
trichlorosilane.
N
HN
*
HSiCl3, Solvent
Cat*
4a
5a
a
b
c
Entry
Catalyst
Solvent
Yield (%)
ee (%)
Afterwards, under the optimal reaction conditions, a
1
2
3
4
5
6
7
8
9
1a
1a
1b
1b
1c
1c
1c
1d
1e
2a
2a
2a
2a
2a
2b
2b
2c
2c
2d
2d
2e
2e
3
CH
CHCl
CH Cl
CHCl
CH Cl
CHCl
Toluene
2
Cl
2
2
2
90
88
85
86
84
91
80
90
76
81
90
40
88
88
82
90
87
89
88
87
87
85
89
5 (S)
20 (S)
31 (S)
30 (S)
65 (S)
18 (S)
0
variety of ketimines 4a–y were reduced with HSiCl in
3
3
CHCl in the presence of 20 mol % 2a at À10 °C (Table
3
2
2). Most of the aromatic ketimines were reduced
smoothly to provide corresponding products in good
yields and enantioselectivities. However, ketimine 4k,
which has a chloro substituent in the ortho position of
the phenyl group of ketone moiety, was reduced in high
yield but poor enantioselectivity (Table 2, entry 11).
Tetralone derived ketimine 4o was reduced in high ee
value (Table 2, entry 15), while reduction of indanone-
derived ketimine 4n gave only moderate enantioselectiv-
ity (Table 2, entry 14). It seems that the conformation
difference between six member ring and five member
ring makes the transition structures of reactions of the
two substrates much different. Meanwhile, a-ketimino
ester 4t was reduced to give protected a-phenylglycine
in good yield and moderate enantioselectivity (Table 2,
entry 20). Furthermore, moderate yields and enantio-
selectivities were afforded in reduction of aliphatic
ketimines (Table 2, entries 24 and 25). In general, higher
enantioselectivities were resulted with N-PMP ketimines
than with N-Ph ketimines. It is noteworthy that
N-benzyl ketimines (Table 2, entries 21–23) could also
be reduced in good yields and enantioselectivities.
3
2
3
CHCl
CHCl
3
0
3
6 (S)
68 (R)
86 (R)
37 (R)
70 (R)
10
11
12
13
14
15
16
17
19
20
21
22
23
24
2 2
CH Cl
CHCl
3
Toluene
THF
d
CHCl
CH Cl
CHCl
CH Cl
CHCl
CH Cl
CHCl
CH Cl
3
92 (R)
2
2
2
2
2
46 (R)
73 (R)
51 (R)
55 (R)
34 (R)
55 (R)
0
3
2
3
2
3
2
CHCl
CHCl
3
5 (R)
70 (R)
3
a
Unless specified otherwise, reactions were carried out with 10 mol %
of catalyst and 2.0 equiv of HSiCl
organic solvent at 0 °C for 16 h.
Isolated yield based on the imine.
3
on 0.2 mmol scale in 2.0 mL of
b
c
The ee values were determined by chiral HPLC. The configuration of
the product was determined by comparison of the HPLC data with
the literature data.
According to the literature,10 N-PMP amines could un-
1
0
6f
8
dergo oxidative deprotection by CAN or TCCA to
give free amines. Furthermore, hydrogenolysis of the
benzyl group of the addition product 5v catalyzed by
d
20 mol % catalyst was used and the reaction was carried out at
À10 °C for 24 h.