tivity (syn/anti = 91:9) and enantioselectivity (84% ee) were
obtained at À108C (Table 1, entry 4). In entries 5–7, other
group 2 metals were utilized.[10,11] Although Ca(O-iPr)2 and
Ba(O-iPr)2 gave poor results, Sr(O-iPr)2 gave 4a in an 86%
yield (Table 1, entry 6). In addition, the use of Sr(O-iPr)2/
Schiff base 1b led to an unexpected reversal of the
diastereoselectivity, and the anti adduct was obtained in
high diastereoselectivity (syn/anti = 6:94) and enantioselec-
tivity (92% ee, anti-4a) at ambient temperature.
The substrate scope of the strontium- and magnesium-
catalyzed reactions is summarized in Table 2. The reaction
temperature and solvent were optimized for each ketimine,
and the best results are reported. Absolute and relative
configurations of the products were unequivocally deter-
mined by X-ray crystallographic analysis (see the Supporting
Information).[12] The enantiofacial selectivity of the reaction
of the ketimines (3) catalyzed by Sr/1b was opposite to that of
the reaction catalyzed by Mg/1b. The results of the strontium-
catalyzed anti-selective reaction are shown in Table 2,
entries 1–12. Aryl ketimines 3a–3i gave products with high
anti selectivity and high enantioselectivity at either room
temperature or À58C (Table 2, entries 1–9). Good yields were
achieved even with ketimines bearing an electron-donating
group at the para position, such as 4-methyl imine 3e (Table 2,
entry 5) and 4-methoxy imine 3g (entry 7). However, a
strongly electron-donating 4-dimethylamine group had
adverse effects on the yield, but high anti selectivity and
enantioselectivity were maintained (Table 2, entry 8).
Heteroaryl ketimines 3j–3l were also applicable, and the
products were obtained in good to high enantioselectivity
(Table 2, entries 10–12), although slightly lower anti selectiv-
ity was observed (entries 10–12). The results of the magne-
sium-catalyzed syn-selective reaction are shown in entries 13–
20. Although the enantioselectivity was slightly lower than
with the strontium catalyst, except for ketimine 3m (Table 2,
entry 18), high syn selectivity was achieved in all cases
(Table 2, entries 13–20). Unfortunately, the present Sr/1b
and Mg/1b systems were not applicable to other ketimines,
such as aryl ethyl ketimines and aliphatic ketimines, because
of the lower reactivity of these ketimines. The a-ethyl-a-
isothiocyanato ester also showed
much lower reactivity and stereose-
Table 2: Stereodivergent direct catalytic asymmetric Mannich-type reaction of a-methyl-a-isothiocya-
nato ester 2 to ketimines 3.[a]
lectivity than the a-methyl-a-iso-
thiocyanato ester 2; this result is
possibly due to the severe steric
hindrance in the construction of
vicinal tetrasubstituted carbon ste-
reocenters.[13] Further optimization
studies, such as ligand modifications
to overcome the severe steric hin-
drance, and expansion of the scope
of the ketimines as well as a-iso-
thiocyanato esters, are ongoing.
Entry
Cat.
R
3
T
[8C]
t
[h]
Yield[e]
[%]
d.r.[f]
syn/anti
4
ee[g]
[%]
1[b]
2[b]
3[b]
4[b]
5[c]
6[c]
7[c]
8[c]
Sr
Sr
Sr
Sr
Sr
Sr
Sr
Sr
4-BrC6H4
4-ClC6H4
4-FC6H4
4-CF3C6H4
4-MeC6H4
3-MeC6H4
4-MeOC6H4
4-Me2NC6H4
3a
3b
3c
3d
3e
3 f
RT
RT
RT
RT
RT
RT
RT
RT
48
48
48
48
20
24
24
69
86
82
71
85
97
99
91
45
6:94
10:90
6:94
11:89
6:94
8:92
4:96
4:96
anti-4a
anti-4b
anti-4c
anti-4d
anti-4e
anti-4 f
anti-4g
anti-4h
92
87
90
92
95
93
97
97
Investigations to obtain a pre-
liminary insight into the structural
differences of the two catalysts by
1
using H NMR spectroscopy failed.
The 1H NMR spectra of the Bu2Mg/
1b (1:1) complex and the Sr(O-
iPr)2/1b (1:1) complex were com-
plicated, which is possibly due to
the oligomeric structures of the
catalysts.[14] Circular dichroism
(CD) spectra of the Bu2Mg/1b
(1:1) complex and the Sr(O-iPr)2/
1b (1:1) complex provided an
insight into the differences between
the aggregates of the two catalysts
(Figure 2).[15] The CD spectrum of
Schiff base 1b was clearly different
from those of Mg/1b and Sr/1b,
thus suggesting that a chiroptically
different aggregate was formed in
each metal/1b solution. In addition,
clear differences between Mg/1b
and Sr/1b in the 210–250 nm
region can be ascribed to the differ-
ence in the dihedral angle of the
binaphthyl unit in Mg/1b and Sr/
3g
3h
9[c]
Sr
3i
À5
47
76
6:94
anti-4i
95
10[c]
11[c]
12[c]
13[b]
14[b]
15[b]
16[d]
Sr
Sr
Sr
Mg
Mg
Mg
Mg
2-thienyl
3-thienyl
2-furyl
4-BrC6H4
4-ClC6H4
4-FC6H4
3-MeC6H4
3j
3k
3l
3a
3b
3c
3 f
0
À5
48
48
48
48
48
44
48
70
74
84
87
90
96
99
13:87
12:88
17:83
91:9
92:8
93:7
anti-4j
anti-4k
anti-4l
syn-4a
syn-4b
syn-4c
syn-4 f
90
92
84
84
85
84
82
À10
À10
À10
0
À25
90:10
17[d]
Mg
3i
À5
17
96
92:8
syn-4i
81
18[b]
19[d]
20[b]
Mg
Mg
Mg
2-naphthyl
3-thienyl
2-furyl
3m
3k
3l
0
À25
À5
48
48
48
99
80
70
93:7
93:7
93:7
syn-4m
syn-4k
syn-4l
95
81
80
[a] Reaction conditions: ketimine 3 (1.0 mmol), 2 (2.0 equiv), Sr(O-iPr)2 (10 mol%; entries 1–12) or
Bu2Mg (10 mol%; entries 13–20), (R)-Schiff base 1b (10 mol%), molecular sieves (5 ꢀ; 200 mg). [b] The
reaction was run in CHCl3 (0.2m). [c] The reaction was run in CHCl3/THF (2:1; 0.17m). [d] The reaction
was run in THF (0.2m). [e] The yield of 4 after isolation and purification by column chromatography on
silica gel. The yield of (syn-4 + anti-4) is shown. [f] Determined by 1H NMR analysis of the crude
mixture. [g] Determined by HPLC analysis on a chiral stationary phase. The enantiomeric excess of anti-4
is shown in entries 1–12 and that of syn-4 is shown in entries 13–20.
Angew. Chem. Int. Ed. 2011, 50, 4382 –4385
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4383