.
Angewandte
Communications
+
[a]
of the Fc/Fc (Fc = ferrocene) groups in A was slightly
Table 2: Screening of solvents.
negative compared to that of the corresponding ferrocene
(
see Figure S2). Additionally, redox peaks of tertiary amines
were not observed in the CV curve. Therefore, it would be
possible to selectively oxidize the ferrocene moiety. After
[
b]
[c]
[d]
Entry
Solvents
Yield [%]
anti/syn
ee [%]
oxidizing A with ferrocenium tetrafluoroborate, the EPR
+
spectrum of the corresponding A [BF ] revealed a feature at
1
2
3
4
5
6
7
8
9
toluene
THF
90
91
94
93
90
92
91
88
86
82
65
91:9
92:8
95:5
95:5
92:8
94:6
91:9
89:11
89:11
94:6
95:5
93
92
94
93
92
90
90
94
90
95
95
4
g = 4.38 (see Figure S3), a result consistent with the presence
k
[
12]
CH Cl
of ferrocenium species. The broadening of the g feature
2
2
?
CHCl3
DCE
may result from the interference of the bulky group of
+
A [BF ].
4
CH CN
3
To examine the catalytic activity of A, the direct aldol
reaction of cyclohexanone and p-nitrobenzaldehyde was
conducted in the presence of 10 mol% A and 15 mol%
oxidant at room temperature (Table 1). Initially, different
1,4-dioxane
DMF
DMSO
MeOH
[
e]
10
[
f]
11
H O
2
[
a] The reaction was performed with cyclohexanone (0.6 mmol),
Table 1: Screening of the oxidants.
+
p-nitrobenzaldehyde (0.2 mmol), A (10 mol%), and [FeCp ] BF
2
4
(
15 mol%) in 0.2 mL solvent at room temperature. [b] Yield of isolated
1
product. [c] Determined by H NMR analysis of the crude reaction
mixture. [d] Determined by HPLC analysis. [e] 8% by-product. [f] 9% by-
product. DCE=1,2-dichloroethane, DMF=N,N-dimethylformamide,
DMSO=dimethylsulfoxide, THF=tetrahydrofuran.
[
a]
[b]
[c]
[d]
Entry
Oxidants
Yield [%]
anti/syn
ee [%]
[
e]
1
2
3
4
5
–
I2
CAN
DDQ
38
92
94
93
94
42:58
63:37
80:20
80:20
90:10
24
90
92
90
93
Table 3: Control experiments.
+
[FeCp ] BF
2
4
[a] Reaction conditions: cyclohexanone (0.6 mmol), p-nitrobenzaldehyde
(
0.2 mmol), A (10 mol%), and oxidant (15 mol%) at room temperature
1
for 48 h. [b] Yield of isolated product. [c] Determined by H NMR
spectroscopy. [d] Determined by HPLC analysis. [e] Reaction time: 72 h.
Cp=cyclopentadienyl.
Entry
Catalyst
Yield [%]
anti/syn
ee [%]
[
[
a]
b]
+
1
2
3
4
5
6
7
8
8
8
A/[FeCp ] BF
B
A/TFA
A/TfOH
C
94
92
78
95:5
96:4
60:40
54:46
–
94
92
91
24
–
2
4
oxidants were tested. Rather poor results were obtained in the
48
[
c]
absence of an oxidant (Table 1, entry 1). When I , CAN
trace
n.r.
88
5
38
2
(
CAN = ceric ammonium nitrate), and DDQ (DDQ = 2,3-
C/DDQ
–
–
+
C/[FeCp ] BF
86:14
42:58
92:8
67:33
92
24
94
92
dichloro-5,6-dicyano-1,4-benzoquinone) were selected as oxi-
dants, the desired products were obtained with excellent
2
4
[
d]
a
b
c
A
[
d]
+
A/[FeCp ] BF
2
4
yields and ee values, but with poor diastereoselectivities
[d]
+
A/[FeCp ] BF /Na S O
52
+
2
4
2
2
4
(
entries 2–4). [FeCp ] BF (ferrocenium tetrafluoroborate)
2
4
[
(
a] Reaction conditions: p-nitrobenzaldehyde (0.2 mmol), cyclohexanone
provided satisfactory results in terms of diastereoselectivity
90:10 d.r.), enantioselectivity (93% ee), and yield (94%;
entry 5).
Various solvents were then screened (Table 2). The results
showed that the reaction tolerated a number of solvents
including highly polar organic solvents, water, and other
nonpolar organic solvents (entries 1–11). Finally, dichloro-
methane was chosen as the optimal reaction medium.
+
0.6 mmol), A (0.02 mmol), [FeCp ] BF (0.03 mmol), and 0.2 mL
2 4
(
1
1
CH Cl at room temperature for 48 h. [b] 10%mol B. [c] C=(S,S)-N ,N -
dipropylcyclohexane diamine. [d] p-Nitrobenzaldehyde (0.2 mmol),
cyclohexanone (0.6 mmol), A (0.02 mmol), and 0.2 mL CH Cl at room
temperature for 5 h (entry 8a). Then [FeCp ] BF (0.03 mmol) was added
2 4
and the reaction run for an additional 12 h (entry 8b). Then Na S O
2 2 4
2
2
2
2
+
(
(
0.035 mmol) was added and the reaction run for an additional 36 h
entry 8c). TFA=trifluoroacetic acid, Tf=trifluoromethanesulfonyl.
Control experiments were performed to confirm the
catalytic mechanism (Table 3). Excellent results were
obtained for the ferrocenophane-B-catalyzed reaction
when DDQ was added, no product was obtained (entries 5
and 6), thus suggesting redox control did not work in this case
(entry 6 versus Table 1, entry 4). In contrast, the combined
(
entry 2). In contrast, poor stereoselectivities were obtained
with A/TFA and A/TfOH (entries 3 and 4), thus indicating
that a protonated amino moiety did not work well as
a stereocontrolling hydrogen-bonding unit in this case, and
that the ferrocenium species, likely serving as Lewis acid,
played an important role in the stereocontrol. Using (S,S)-
+
use of C/[FeCp ] BF led to slightly inferior results in terms of
2
4
both activity and stereoselectivity (Table 3, entry 7 versus
entry 1). It is known that Lewis acids may also effect
[7k]
stereocontrol in enamine catalysis. Considering the inert-
1
1
+
N ,N -dipropylcyclohexanediamine (C) as the catalyst
ness of DDQ, it was believed that [FeCp ] BF mainly served
2
4
resulted in only trace amounts of the desired product, and
as a Lewis acid to facilitate the reaction, and further
2
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 1 – 5
These are not the final page numbers!