Table 1. Evaluation of catalysts and optimization of reaction conditions.[a]
lin-1,2-diol 3, which would be accessed by reduction of the
corresponding aldehyde 4. The key building block 4 could
be obtained directly from the enantioselective alkylation re-
action of 3-hydroxyoxindoles 5 with aldehydes 6 catalyzed
by a chiral amine. Thus, the asymmetric catalytic alkylation
reACHTUNGTRENNUNGaction of 2-alkyloxyacetaldehyde with 3-hydroxyoxindole
turned out to be a key step in the total synthesis of (+)-glio-
cladin C and was studied initially.
Organocatalytic asymmetric alkylation reactions have
been accomplished in high levels of stereoselectivity by
using chiral secondary amine catalysts.[12] However, the ex-
tension of these protocols to the reaction between 2-alkylox-
yacetaldehyde 6 and 3-hydroxyoxindole 5a failed to give
synthetically useful results (see the Supporting Information).
The cinchona-based primary amines have been proven to be
successful in the control of stereoselectivity in either en-
Entry
7
Acid
TFA
6
4
Yield [%][b]
d.r.[c]
ee [%][d]
ACHTUNGTRENNUNG HCATUNGTREN(NUGN 9-
amine or iminium catalysis.[13] In particular, 9-amino
1
2
3
4
5
6
7a
7b
7c
7d
7d
7d
7d
7d
7d
7d
7d
7d
7d
6a
6a
6a
6a
6b
6c
6c
6c
6c
6c
6c
6c
6c
4aa
4aa
4aa
4aa
4ba
4ca
4ca
4ca
4ca
4ca
4ca
4ca
4ca
31
30
37
48
36
43
47
35
–
32
42
62
80
1:1
1:1
2:1
2:1
2:1
2:1
4:1
6:1
–
3:1
6:1
7:1
8:1
À15
À23
68
73
58
75
77
63
–
65
87
90
94
TFA
TFA
TFA
TFA
TFA
TFA
8a
deoxy) epicinchona alkaloids, easily derived from natural re-
sources, have enabled the stereoselective functionalization
of a variety of sterically demanding carbonyl compounds.[14]
Inspired by these achievements, we envisaged that the cin-
chona alkaloid-based amines would be able to catalyze the
alkylation of 3-hydroxyoxindoles 5 with aldehydes 6 in a
stereoselective manner.
Thus, we first examined the a-alkylation reaction between
3-hydroxyoxindole 5a, which can be transformed into a vi-
nylogous iminACHTUNGTRENNUNGium intermediate under acidic condi-
7[e]
8[e]
9[e]
10[e]
11[e]
12[e]
13[f]
(R)-8a
8b
8c
8d
8d
tions,[11,12c,h,15] with 2-(benzyloxy)acetaldehyde 6a by using
cinchona alkaloid-based amines as chiral catalysts (Table 1).
However, the reaction gave the desired product in only
31% yield, together with poor diastereo- and enantioselec-
tivity in the presence of 10 mol% of the primary amine 7a
and 30 mol% of trifluoroacetic acid (TFA; Table 1, entry 1).
Neither the yield nor the enantioselectivity was significantly
improved when the bifunctional primary amine 7b was em-
ployed as catalyst (Table 1, entry 2). To our delight, the use
of amine 7c, which is derived from quinidine, was able to
dramatically enhance the enantioselectivity (68% ee;
Table 1, entry 3). The stereoselectivity was further improved
by using 7d as a chiral amine catalyst (Table 1, entry 4). The
OPMB-acetaldehyde 6c (PMB=para-methoxybenzyl) pro-
vided comparably higher enantioselectivity than its counter-
parts (Table 1, entries 4–6). Conducting the reaction at 158C
gave higher levels of diastereo- and stereoselectivities
(Table 1, entry 7). Xie et al.,[16] Melchiorre et al.,[14k,17] and
List et al.[18] found that the use of chiral phosphoric acid as a
cocatalyst dramatically enhances the stereochemical control
of primary amine catalysts of type 7. Therefore, a variety of
1,1’-bi-2,2’-naphthol (BINOL)-derived phosphoric acids[19]
were evaluated as the cocatalyst in this reaction (Table 1,
entries 8–12). Indeed, the addition of chiral phosphoric acid
has considerable influence on the stereoselectivity. The (S)-
phosphoric acid was proven to be a matched cocatalyst,
whereas the R enantiomer completely inhibited the catalytic
activity of amine (Table 1, entry 9). The matched/mismatch-
ed catalyst pair combinations (Table 1, entries 8 and 9)
strongly suggest the existence of cooperative dual activa-
[a] Reaction conditions: The reaction was conducted on a 0.1 mmol scale
with 6 (2 equiv) in commercial chloroform (undistilled CHCl3). Aldehyde
6 was added to a solution of 5a and the catalyst through a syringe pump
over a period of 10 h at 308C. [b] Yield of isolated products. [c] Diaster-
eomeric ratio (d.r.) determined by 1H NMR spectroscopy. [d] The ee
value was determined by HPLC analysis. [e] At 158C. [f] At 08C.
tion.[14k] Among the phosphoric acids screened, 8d turned
out to be the best cocatalyst to give 62% yield, 7:1 d.r., and
90% ee (Table 1, entry 12). The best results (80% yield,
8:1 d.r., and 94% ee) were achieved when the reaction was
carried out at 08C (Table 1, entry 13).
More interestingly, when the model reaction of 5a with
6c was conducted in distilled chloroform, both the yield and
enantioselectivity dramatically dropped (Table 2, entries 1
and 2). Considering that the commercial chloroform (undis-
tilled) contains a trace amount of ethanol (0.3–1%) as sta-
bilizer, we suspected that the alcohol might play a certain
role in the catalysis. Thus, different alcohols were examined
as additives of the distilled chloroform solvent for the re-
AHCTUNGTREGaNNUN ction. As anticipated, the addition of ethanol gave relative-
ly higher enatioselectivity, which matches the results ob-
tained from the reactions conducted in commercial chloro-
form (Table 2, entries 1, 3–5). These results demonstrated
that the alcohol indeed plays an important role in the for-
À
mation of a new C C bond and thereby influences the ster-
eoselectivity. To gain insight into the specific role of ethanol,
an alkylation reaction of 3-ethoxyoxindole 5a’, which was
derived from 3-hydroxyoxindole 5a and ethanol under
acidic conditions [Eq. (1)], with the aldehyde 6c was carried
3320
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 3319 – 3323