Catalytic asymmetric synthesis of 3-(α-hydroxy-ss-carbonyl) oxindoles by a ScIII-catalyzed direct aldol-type reaction

Article abstract of DOI:10.1002/chem.200903471

A direct catalytic asymmetric aldol-type reaction of 3-substituted- 2oxindoles with glyoxal derivatives and ethyl trifluoropyruvate, catalyzed by a chiral N,N-'-dioxide-Sc(OTf)₃ (Tf=trifluoromethanesulfonyl) complex, has been developed that tolerates a wide range of substrates. The reaction proceeds in good yields and excellent enantioselectivities (up to 93% yield, 99:1 diastereomeric ratio (dr), and > 99 % enantiomeric excess (ee)) under mild conditions, to deliver 3-(α-hydroxy-ss-carbonyl) oxindoles with vicinal quaternary-tertiary or quaternaryquaternary stereocenters. Even with 1 mol% catalyst loading or on scaleup (10 mmol of starting material), maintenance of ee was observed, which showed the potential value of the catalyst system. In studies probing the reaction mechanism, a positive nonlinear effect was observed and Sc^III-based enolate intermediates were detected by using ESIMS. On the basis of the experimental results and previous reports, a possible catalytic cycle was assumed.

Full text of DOI:10.1002/chem.200903471

DOI: 10.1002/chem.200903471
Catalytic Asymmetric Synthesis of 3-(a-Hydroxy-b-carbonyl) Oxindoles by a
ScIII-Catalyzed Direct Aldol-Type Reaction
Ke Shen, Xiaohua Liu, Ke Zheng, Wei Li, Xiaolei Hu, Lili Lin, and Xiaoming Feng*[a]
Abstract: A direct catalytic asymmetric
aldol-type reaction of 3-substituted-2-
oxindoles with glyoxal derivatives and
ethyl trifluoropyruvate, catalyzed by a
>99% enantiomeric excess (ee)) under
mild conditions, to deliver 3-(a-hy-
droxy-b-carbonyl) oxindoles with vici-
nal quaternary–tertiary or quaternary–
quaternary stereocenters. Even with
1 mol% catalyst loading or on scaleup
(10 mmol of starting material), mainte-
nance of ee was observed, which
showed the potential value of the cata-
lyst system. In studies probing the reac-
tion mechanism, a positive nonlinear
effect was observed and ScIII-based
enolate intermediates were detected by
using ESIMS. On the basis of the ex-
perimental results and previous reports,
a possible catalytic cycle was assumed.
chiral N,N’-dioxide–ScACHTNUTRGNEUNG(OTf)3 (Tf=tri-
fluoromethanesulfonyl) complex, has
been developed that tolerates a wide
range of substrates. The reaction pro-
ceeds in good yields and excellent
enantioselectivities (up to 93% yield,
99:1 diastereomeric ratio (dr), and
Keywords: aldol reaction · asym-
metric catalysis · enolates · ox-
AHCTUNGTREGiNNNU ndoles · scandium
Introduction
is only one successful transformation between oxindoles and
ethyl trifluoropyruvate, disclosed by Shibata and co-workers,
using a chiral amine base as a catalyst.[12] Although some
progress has been made, demand for other practical process-
es remains in this area. Herein, we report our efforts to ad-
dress these issues by employing a scandium–N,N’-dioxide
complex to catalyze the direct asymmetric aldol-type reac-
tion of 3-substituted-2-oxindoles 1 with dicarbonyl com-
pounds 2 to produce 3-(a-hydroxy-b-carbonyl) oxindoles 3
with vicinal quaternary–tertiary or quaternary–quaternary
stereocenters in up to >99% ee and 99:1 dr.
In recent years, great effort has been devoted to the asym-
metric reactions of nucleophilic 3-substituted oxindoles[1–5]
to construct oxindoles containing an asymmetric quaternary
stereogenic center[6] at the C3 position. These compounds
have potential medicinal interest, owing to the unique bio-
logical activities of such natural products.[7] Typically, con-
struction of 3-(a-hydroxy-b-carbonyl) oxindoles is attractive
and imperative because several pharmaceutical candidates,
taking surugatoxin, for instance,[8] contain this structural
motif. The direct catalytic asymmetric aldol reaction[9] is a
powerful and atom-economical method for the synthesis of
such oxindoles. However, the higher pKa values of the a-
proton of oxindoles relative to activated carbonyl com-
pounds[10] and the inconvenience of creating contiguous,
sterically hindered, chiral carbon stereocenters[11] made this
catalytic asymmetric aldol reaction difficult. To date, there
Results and Discussion
As revealed in Scheme 1, the aldol adduct from the reaction
of an oxindole and a dicarbonyl compound was expected to
be obtained by means of nucleophilic attack through two
possible activation models: 1) the 3-substituted-2-oxindole
could be activated through an in situ generated O–metal
enolate,[13] or 2) the activation of the dicarbonyl substrate
might be facilitated through bidentate chelation with a
metal complex. Further work was directed toward a detailed
examination of these models.
We probed the optimal reaction conditions using 3-
methyl-2-oxindole (1a) and phenylglyoxal (2a) as model
substrates. N,N’-Dioxides have been applied to many asym-
metric procedures because they permit a tunable electronic
[a] K. Shen, Dr. X. Liu, K. Zheng, W. Li, X. Hu, Dr. L. Lin,
Prof. X. Feng
Key Laboratory of Green Chemistry & Technology
Ministry of Education
College of Chemistry, Sichuan University
Chengdu 610064 (P.R. China)
Fax : (+86)2885418249
E-mail: xmfeng@scu.edu.cn
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200903471.
3736
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 3736 – 3742

FULL PAPER
than isopropyl substituents, afforded much less satisfactory
results (36% ee; Table 1, entry 8). When the amide moiety
was replaced by an aliphatic ring, both the enantioselectivity
and the yield decreased dramatically (Table 1, entry 11).
When N,N’-dioxide L8 (derived from l-ramipril) and L9
(derived from l-proline) were used instead of l-pipecolic
acid derived ligand L5, the enantioselectivities decreased to
83 and 57%, ee respectively (Table 1, entries 12 and 13
versus entry 9). Therefore, the L5–ScACTHNUTRGENUG(N OTf)3 system was
chosen to assess other reaction parameters.
Scheme 1. Proposed activation models of the direct aldol-type reaction
between 3-substituted-2-oxindoles and dicarbonyl compounds.
and steric chiral environment.[14] Our investigations began
with screening several chiral Lewis acid catalysts that were
generated in situ from metal salts and N,N’-dioxide L1 to
evaluate their ability to promote the addition of 1a to 2a.
Inspiringly, Sc
the best results (Table 1, entry 1). Other metal sources, such
as Y(OTf)3, La(OTf)3, Ti(OiPr)4, and Cu(OTf)2 resulted in
ACHTUNGTRENNUNG(OTf)3 (Tf=trifluoromethanesulfonyl) gave
A
E
N
ACHTUNGTRENNUNG
poor reactivity and selectivity (Table 1, entries 2–5). Further
optimization of the reaction conditions was aimed at explor-
ing the efficiency of ScACHTNUGTRNEUNG(OTf)3 paired with other N,N’-dioxide
ligands L. We found that the amide moiety in the N,N’-diox-
ide ligands had a significant effect on the enantioselectivity
(Table 1, entries 6–11). Significant enhancement of the reac-
tion was observed upon addition of N,N’-dioxide L5. The
aldol addition was effectively promoted by a 1:1 L5–Sc-
ACHTUNGTRENNUNG(OTf)3 complex in CH2Cl2 at 08C and gave oxindole 3aa in
52% yield, 75:25 dr, and 96% ee (Table 1, entry 9). Ligand
L4, with dimethyl substituents on the phenyl ring rather
Encouraged by the initial results, various solvents were
tested in the presence of L5–ScACHTNUGTRENUNG(OTf)3 (5 mol%). The results
indicated that the reaction solvent played an important role
in governing the rate and enantioselectivity of the reaction.
Chlorinated alkanes were investigated because CH2Cl2 was
initially found to be the optimal solvent, however, no superi-
or result was obtained (Table 2, entry 1 versus entries 2–5).
Other solvents, such as ethyl acetate, THF, diethyl ether,
and toluene, also gave less satisfactory results (Table 2, en-
Table 1. Central metal and ligand effects on the catalytic asymmetric
aldol-type reaction of 3-methyl-2-oxindole (1a) and phenylglyoxal (2a)
under the indicated conditions.[a]
Table 2. Solvent and concentration effects on the catalytic asymmetric
aldol-type reaction of 3-methyl-2-oxindole (1a) and phenylglyoxal (2a).[a]
Entry
Metal
Sc(OTf)3
(OTf)3
La(OTf)3
Ti(OiPr)4
Cu(OTf)2
(OTf)3
(OTf)3
(OTf)3
(OTf)3
(OTf)3
(OTf)3
(OTf)3
(OTf)3
L
Yield [%][b]
dr
ee [%][c]
Entry
Solvent
Yield [%][b]
dr
ee [%][c]
1
2
3
4
5
6
7
8
G
L1
L1
L1
L1
L1
L2
L3
L4
L5
L6
L7
L8
L9
20
12
8
7
9
17
12
36
52
23
20
57
55
74:26
70:30
72:28
73:27
84:16
69:31
60:40
73:27
75:25
65:35
84:16
50:50
72:28
23
À2
5
À5
race.
12
race.
36
96
33
4
83
57
1
2
3
4
5
6
7
8
CH2Cl2
CHCl3
52
50
30
40
36
72
70
77
71
55
79
75:25
83:17
86:14
91:9
86:14
68:32
60:40
55:45
55:45
74:26
83:17
96
80
97
87
95
55
36
70
75
95
98
Y
G
G
Cl2CHCHCl2
ClCH2CHCl2
CH3CCl3
ethyl acetate
THF
Et2O
toluene
CH2Cl2
CH2Cl2
G
G
Sc
Sc
Sc
Sc
Sc
Sc
Sc
Sc
E
N
R
9
N
9
10
11
12
13
T
10[d]
11[e]
N
R
[a] Unless noted otherwise, the reactions were conducted on a 0.1 mmol
scale at 0.1m, using equimolar amounts of 1a and 2a, in CH2Cl2, under
Ar, at 08C. [b] Isolated yield. [c] The ee value of the major diastereomer
is given, determined by chiral HPLC analysis. [d] CH2Cl2 was used with-
out prior distillation. [e] The reaction was carried out in CH2Cl2 (2.0 mL)
for 24 h.
A
[a] Unless noted otherwise, the reactions were conducted on a 0.1 mmol
scale at 0.1m, using equimolar amounts of 1a and 2a, in CH2Cl2, under
Ar, at 08C. [b] Isolated yield. [c] The ee value of the major diastereomer
is given, determined by chiral HPLC analysis. Race.=racemic mixture.
Chem. Eur. J. 2010, 16, 3736 – 3742
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3737

X. Feng et al.
Table 4. Direct catalytic asymmetric aldol-type reaction of oxindole 1a
with glyoxal derivatives 2.[a]
tries 6–9, respectively). It should be noted that this process
tolerated the presence of moisture. When CH2Cl2 was used
without prior distillation, no clear decrease in ee value was
observed in this case (Table 2, entry 10). Dilution was also a
key factor and it was found that the reactivity was increased
by reducing the reaction concentration to 0.05m (Table 1,
entry 11).
Subsequently, the effect of additives and catalyst loading
was examined and the results are presented in Table 3. In-
troduction of 3 ꢁ molecular sieves could further shorten the
Entry R1
Product t [h] Yield [%][b] dr
ee [%][c]
1
2
3
4
5
6
7
8
C6H5
3aa
3ab
3ac
3ad
3ae
3af
12
12
12
12
16
16
16
24
24
24
24
24
24
24
24
36
36
36
36
24
79
82
89
90
89
88
82
75
88
85
88
92
85
77
83
83
80
75
78
75
86:14
90:10
85:15 >99
91:9
96:4
98
99
2-MeC6H4
3-MeC6H4
4-MeC6H4
3-MeOC6H4
4-MeOC6H4
3,4-(MeO)2C6H3 3ag
2-FC6H4
4-FC6H4
3-ClC6H4
4-ClC6H4
3,4-Cl2C6H3
4-BrC6H4
3-NO2C6H4
4-NO2C6H4
1-naphthyl
2-naphthyl
2-furyl
97
98
97
95
95
99
95
99
94
98
94
95
99
99
97
99
92
Table 3. Effect of additives and catalyst loading on the catalytic asym-
metric aldol-type reaction of 3-methyl-2-oxindole (1a) and phenylglyoxal
(2a).[a]
96:4
95:5
3ah
3ai
3aj
3ak
3al
3am
3an
3ao
3ap
3aq
3ar
3as
3at
73:27
83:17
79:21
97:3
62:38
96:4
63:37
85:15
94:6
9[d]
10
11[d]
12
13[d]
14
15
16
17
18
19
20[e]
Entry
Additive
Yield [%][b]
dr
ee [%][c]
1
2
3
4
5
6
7
3 ꢁ MS (2 mg)
4 ꢁ MS (2 mg)
5 ꢁ MS (2 mg)
Mg2SO4 (2 mg)
H2O (20 mL)
iPrOH
phenol
HCOOH
3 ꢁ MS (2 mg)
79
82
80
83
77
67
38
86:14
78:24
79:21
80:20
82:18
77:23
75:25
–
98
85
98
98
96
51
77
–
99:1
56:44
65:35
79:21
2-thienyl
Me
[a] For details, see the Supporting Information. [b] Isolated yield. [c] The
ee value of the major diastereomer is given, determined by chiral HPLC
analysis. [d] The reaction was carried out in CH2Cl2 and Cl2CHCHCl2
(2.0 mL; 1:1 v/v). [e] The reaction was carried out in ClCH2CHCl2 and
toluene (2.0 mL; 1:1 v/v).
8
n.r.[d]
70
9[e]
81:19
96
[a] Unless noted otherwise, the reactions were conducted on a 0.1 mmol
scale at 0.05m, using equimolar amounts of 1a and 2a, in CH2Cl2, under
Ar, at 08C. [b] Isolated yield. [c] The ee value of the major diastereomer
is given, determined by chiral HPLC analysis. [d] No reaction.
[e] 1 mol% catalyst loading was used (0.5 mmol scale).
and 99% ee, respectively (Table 4, entries 18 and 19). Ali-
phatic glyoxal 2t resulted in poor yield and dr when the re-
action was performed in CH2Cl2 (40% yield, 49:50 dr).
Modification of the solvent was effective in improving the
results to 75% yield, 79:21 dr, and 92% ee (Table 4,
entry 20).
reaction time to 12 h (Table 3, entry 1). To our surprise, the
reaction was not sensitive to additional water (Table 3,
entry 5). Acidic additive, formic acid, shut down the catalyt-
ic ability of the L5–Sc
(OTf)3 complex (Table 3, entry 8). The
Moreover, the scope of the aldol-type reaction was ex-
tended successfully to a variety of 3-substituted-2-oxindoles
1a–j. Oxindoles with saturated aliphatic substituents at the
C3 position generally provided the corresponding products
in excellent ee (98–>99%, Table 5, entries 1–4).[15] With a 3-
allyl substituent on the oxindole, the enantiomeric excess of
the desired product was still excellent (Table 5, entry 5),
which is meaningful because the ally group in compound
3ea is a useful handle for further functional group manipu-
lation.[2c,3b,c] Variation of the C3-substituent to benzyl or 2-
naphthylmethyl was also possible to give the expected prod-
ucts 3 fa and 3ga in >98% ee (Table 5, entries 6–7). Signifi-
cantly, oxindole 1h, which has a 2-thienylmethyl group at
C3, also afforded an excellent ee (>99%, Table 5, entry 8).
To our delight, 3-phenyl-2-oxindole (1i) was also a suitable
substrate for this reaction (98% ee, Table 5, entry 9). The
catalytic system was also effective for electron-deficient 5-
bromo-2-oxindole (Table 5, entry 10). The effect of N-substi-
tution of the 2-oxindole on the reaction was further exam-
ined. For N-Boc-2-oxindole 1k and N-methyl-2-oxindole 1l,
catalyst loading was then evaluated. Even with 1 mol% cat-
alyst loading, maintenance of the ee value, although a slight
decrease in dr, was observed (Table 3, entry 9). Extensive
screening showed that the optimal conditions were as fol-
lows: L5–Sc
ACHTUNGTRENNUNG(OTf)3 complex (5 mol%; 1:1 molar ratio of L5–
Sc(OTf)3), 1a (0.1 mmol), and 2a (0.1 mmol) in CH2Cl2
ACHTUNGTRENNUNG
(2.0 mL) at 08C. These conditions were subsequently used
to determine the generality of this process.
Evaluation of the reaction scope revealed that high levels
of enantioselectivity could be achieved for a wide range of
glyoxal derivatives 2 (92–>99% ee; Table 4). Aromatic
rings substituted with both electron-donating and -withdraw-
ing groups (Table 4, entries 2–15) afforded products in 75–
92% yield, 62:38–97:3 dr, and 94–>99% ee. The naphthyl
glyoxals 2p and 2q reacted smoothly with 1a and gave the
desired products with excellent diastereo- and enantioselec-
tivity (up to 99:1 dr, 99% ee; Table 4, entries 16 and 17).
The catalyst system was also applicable to heteroaryl glyox-
als, which delivered the corresponding adducts in up to 97
3738
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 3736 – 3742

Asymmetric Synthesis of 3-(aHydroxy-b-carbonyl) Oxindoles
FULL PAPER
Table 5. Direct catalytic asymmetric aldol-type reaction of oxindoles 1 with phenylglyoxal (2a).[a]
To gain insight into the reac-
tion mechanism, control experi-
ments were carried out and the
results are summarized in
Table 6. Initially, product 3aa
was obtained in less than 10%
yield using ScACHTNUTRGNENG(U OTf)3 as a Lewis
acid catalyst (Table 6, entry 2).
In contrast, significant improve-
ment was observed upon addi-
Entry
R2
R4
Prod.
t [h]
Yield [%][b]
dr
ee [%][c]
1
2
3
4
5
6
Me
Et
nPr
nBu
Allyl
Bn
H
H
H
H
H
H
3aa
3ba
3ca
3da
3ea
3 fa
12
12
12
16
16
24
79
81
82
82
80
72
86:14
87:13
89:11
83:17
86:14
64:36
98
99
>99
99
99
99
tion
of
N,N’-dioxide
L5
(Table 6, entry 4 versus 2). In
view of this phenomenon, the
enantioselective addition with
L5–Sc
ligand-accelerated
(OTf)3 should be
a
process.[17]
7
8
H
H
3ga
24
75
77:23
98
This result demonstrated that
L5 works well as a Brønsted
base, adjusting the Lewis acidi-
ty of the scandium. The experi-
mental procedure also had a
great effect on the reaction.
When 2a was added prior to 1a
(procedure B), enantioselectivi-
ty decreased from 98 to 69% ee
3ha
24
88
70:30
>99
9
Ph
H
H
3ia
3ja
3ka
3la
24
16
12
12
62
90
trace
<10
80:20
75:25
–
98
82
–
10
11
12
Me
Me
Me
Boc[d]
Me
–
–
[a] For details, see the Supporting Information. [b] Isolated yield. [c] The ee value of the major diastereomer is
given, determined by chiral HPLC analysis. [d] Boc=tert-butyloxycarbonyl.
Table 6. Control experiments.[a]
low product levels were detected in the reaction with 2a
(Table 5, entries 11 and 12), which indicated that the NH
subunit of the oxindole was necessary for this reaction.
Inspiringly, the L5–ScACTHNUTRGNE(UGN OTf)3 complex was also effective
for reaction with ethyl trifluoropyruvate (Scheme 2a). Oxin-
dole 4a was isolated, which bore a tertiary a-trifluoromethyl
alcohol moiety,[1a] and could provide access to novel drug
candidates.[16] To exploit the potential of the current catalyst
system, the reaction was scaled up to 10 mmol of starting
material (1.47 g of 1a and 1.34 g of 2a) in the presence of
5 mol% of L5–ScACHTUNGTRENNUNG(OTf)3 as the catalyst. Encouragingly, the
reaction proceeded smoothly and good yield, dr, and excel-
lent ee were maintained (Scheme 2b).
Entry Sc
(OTf)3
L5
[mol%]
Procedure Yield
[%][b]
dr
ee
[mol%]
[%][c]
1
2
3
4
5
6
none
5
none
5
5
5
none
none
5
5
5
5
A
A
A
A
B
n.r.[d]
<10
<5
79
–
–
–
–
–
race.[e]
86:14 98
77:23 69
84:16 97
45
C
80
[a] Unless noted otherwise, the reactions were conducted on a 0.1 mmol
scale at 0.05m, using equimolar amounts of 1a and 2a, in CH2Cl2. Proce-
dure A: 1a and 2a were added sequentially at 08C. Procedure B: 2a was
stirred with the catalyst for 1 h at 08C before 1a was added. Procedur-
e C: 1a was stirred with the catalyst for 1 h at 08C before 2a was added.
[b] Isolated yield. [c] The ee value of the major diastereomer is given, de-
termined by chiral HPLC analysis. [d] No reaction. [e] Racemic mixture.
(Table 1, entry 5 versus entry 4). The inferior result caused
by procedure B might be ascribed to the strong bidentate
coordination of 2a to L5–ScACHTNUTRGNENG(U OTf)3, which could develop a
different chiral environment prior to the formation of the
chiral ScIII–enolate. Hence, the key step to accomplish this
reaction should be the preferential formation of the enolate,
Scheme 2. a) Aldol-type reaction of 1a with ethyl trifluoropyruvate
induced by L5–Sc
ACHTUNGRTEN(NUNG OTf)3, and the activation pathway a in
under the catalysis of L5–ScACHTUNGTRENUN(NG OTf)3 complex; b) Scaleup of the asymmet-
ric catalytic direct aldol-type reaction of 1a with 2a.
Scheme 1 is favorable.
Chem. Eur. J. 2010, 16, 3736 – 3742
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3739

X. Feng et al.
For direct proof of the rationality of the proposed ScIII-
based enolate activation, we investigated the aldol-type re-
action using ESIMS, which enabled us to identify the critical
intermediates in the reaction mixture. The sample was pre-
pared from ScACHTUNGTRENNUNG(OTf)3 and L5 in a 1:1 ratio, in the presence
of 1a (2.0 equiv relative to the catalyst) in CH2Cl2. The
spectrum of the sample obtained from the reaction mixture
after stirring for 1 h revealed ions at m/z 1138.3318 and
2277.5900, which correspond to enolate intermediates
[5+H]+ and [6+H]+ (Figure 1). Moreover, characteristic sig-
nals at m/z 1122.3566 and 1272.3114 were observed in the
ESIMS spectrum of ions [7ÀOTf]+ and [7+H]+ (Figure 2).
These signals were consistent with the fragmentation of the
aldol-type adduct before scission with the L5–ScIII complex,
which indicated that these signals should correspond to the
intermediacy of the aldol adduct. The above MS results con-
firmed our proposed activation models, which involved a
metal–enolate intermediate in the aldol-type reaction mech-
anism, as shown in Scheme 1.
Figure 2. ESIMS spectrum of the aldol-type reaction 3 h after stirring 2a
with L5–ScACTHNUGTRNEUNG(OTf)3 and 1a in CH2Cl2. HRMS: m/z calcd: 1122.4668
[7ÀOTf]+, 1272.4261 [7+H]+; found: 1122.3566, 1272.3114.
As a result of the detected dimer species 6 (Figure 1), a
nonlinear effects (NLE) study[18] was carried out. Asymmet-
ric amplification leading to both diastereomers was observed
(Figure 3). The (+)-NLE should be a consequence of the
Figure 3. Chiral amplification in the direct aldol-type reaction of 1a with
~
&
2a, catalyzed by 5 mol% L5–ScACTHNUTRGNE(NUG OTf)3. : major diastereomer, : minor
diastereomer.
formation of an oligomeric species of the L5–ScACTHNUGTRENUNG(OTf)3 cata-
lyst, rather than a unique monomeric structure. This obser-
vation was also consistent with the results of the ESIMS
study above. Thus, it was considered that the active catalytic
species was a less-stable homochiral complex. When the
concentration was decreased from 0.1 to 0.05m, higher yield
could be obtained (Table 2, entry 11 versus 1), which might
be ascribed to the shift in equilibrium between oligomeric
and monomeric species.
Figure 1. ESIMS spectrum of the aldol-type reaction 1 h after stirring 1a
with L5–Sc
ACHTUNGTRENNUNG
(OTf)3 in CH2Cl2. HRMS: m/z calcd: 1138.3893 [5+H]+,
2277.7714 [6+H]+; found: 1138.3318, 2277.5900.
3740
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 3736 – 3742

Asymmetric Synthesis of 3-(aHydroxy-b-carbonyl) Oxindoles
FULL PAPER
128.1, 124.5, 122.7, 109.6, 53.1, 21.1 ppm; HRMS (EI): m/z calcd for
C17H15NO3+Na+: 304.0944 [M+Na]+; found: 304.0947. The ee was deter-
mined by HPLC analysis by using a chiralcel OD-H column, hexane/2-
propanol (90:10), flow rate=1.0 mLminÀ1, 254 nm. Retention time (tR)
(major diastereomer, 98% ee)=10.20 min (major enantiomer), tR =
12.03 min (minor enantiomer); tR (minor diastereomer)=15.64 min
(major enantiomer), tR =17.08 min (minor enantiomer).
In light of the above experimental results, we assumed
that
a catalytic cycle was operating for this reaction
(Scheme 3). The in situ formed catalyst 8 (real active spe-
Typical experimental procedure for the reaction of oxindole 1 with ethyl
trifluoropyruvate: A mixture of L5 (5 mol%), ScACTHNURGTNEUNG(OTf)3 (5 mol%), 3 ꢁ
molecular sieves (2 mg), and oxindole 1a (0.1 mmol) was stirred in
ClCH2CH2Cl (2.0 mL) at 358C under Ar for 1 h. Ethyl trifluoropyruvate
(0.11 mmol) was then added at 358C. The reaction mixture was stirred at
358C and monitored by TLC. The crude product was purified by flash
chromatography directly (petroleum ether/ethyl acetate, 3:1) to afford 4a
as a white solid,[1a] as an inseparable diastereomeric mixture. 1H NMR
(400 MHz, CDCl3): d=8.74 (s, 1H), 7.44–7.46 (d, J=7.6 Hz, 1H), 7.25–
7.29 (t, J=7.7 Hz, 1H), 7.05–7.09 (t, J=7.6 Hz, 1H), 6.91–6.93 (d, J=
7.6 Hz, 1H), 5.38 (s, 1H), 4.39–4.44 (q, J=7.2 Hz, 1H), 4.14–4.29 (m,
1H), 1.58 (s, 3H), 1.10–1.12 ppm (t, J=7.2 Hz, 3H); 13C NMR (100 MHz,
CDCl3): d=179.3, 167.9, 139.9, 129.7, 129.0, 126.2, 125.8, 124.5, 123.0,
110.1, 63.5, 51.6, 19.7, 13.7 ppm. The ee was determined by HPLC analy-
sis using a chiralcel OJ-H column, hexane/2-propanol (93:7), flow rate=
1.0 mLminÀ1
, 254 nm. tR (major diastereomer, 91% ee)=11.62 min
Scheme 3. Proposed catalytic cycle.
(major enantiomer), tR =25.31 min (minor enantiomer); tR (minor diaste-
reomer)=11.66 min (major enantiomer), tR =15.99 min (minor enantio-
mer).
cies), which was generated from L5 and ScACTHNURTGNENG(U OTf)3, deproton-
ated the a-position of 1a to give chiral scandium enolate 9
in situ. Thus formed, the chiral scandium enolate reacted
with phenylglyoxal (2a) through a bidentate chelation to
form intermediate 10. Subsequent nucleophilic attack and
protonation by HOTf afforded the aldol adduct 3aa and the
regenerated catalyst.
Acknowledgements
We appreciate the National Natural Science Foundation of China (no.
20732003), PCSIRT (no. IRT0846), and the Major State Basic Research
and Development Program (no. 2010CB833300) for financial support.
We also thank Sichuan University Analytical & Testing Center for NMR
spectroscopy analysis and the State Key Laboratory of Biotherapy for
HRMS analysis.
Conclusion
The direct catalytic asymmetric aldol-type reaction of 3-sub-
stituted-2-oxindoles with glyoxal derivatives and ethyl tri-
fluoropyruvate was successfully established through ScIII-
[1] For examples of Aldol and Mannich reactions of 2-oxindoles, see:
a) S. Ogawa, N. Shibata, J. Inagaki, S. Nakamura, T. Toru, M. Shiro,
Angew. Chem. 2007, 119, 8820–8823; Angew. Chem. Int. Ed. 2007,
46, 8666–8669; b) X. Tian, K. Jiang, J. Peng, W. Du, Y. C. Chen,
Org. Lett. 2008, 10, 3583–3586; c) L. Cheng, L. Liu, H. Jia, D.
Wang, Y. J. Chen, J. Org. Chem. 2009, 74, 4650–4653.
[2] For examples of fluorination and hydroxylation of 2-oxindoles, see:
a) T. Ishimaru, N. Shibata, T. Horikawa, N. Yasuda, S. Nakamura, T.
Toru, M. Shiro, Angew. Chem. 2008, 120, 4225–4229; Angew. Chem.
Int. Ed. 2008, 47, 4157–4161; b) Y. Hamashima, T. Suzuki, H.
Takano, Y. Shimura, M. Sodeoka, J. Am. Chem. Soc. 2005, 127,
10164–10165; c) D. Sano, K. Nagata, T. Itoh, Org. Lett. 2008, 10,
1593–1595.
based enolate activation. The L5–ScACTHNUTRGENUN(G OTf)3 complex effi-
ciently promoted the aldol addition, and resulted in 3-(a-hy-
droxy-b-carbonyl) oxindoles with vicinal quaternary–tertiary
or quaternary–quaternary stereocenters, in up to 93% yield,
99:1 dr, and >99% ee under mild conditions. Application of
N,N’-dioxide–metal complexes to other reactions involving
oxindoles as nucleophiles is ongoing in our laboratory.
Experimental Section
[3] For examples of 1,3-dipolar cycloaddition of 2-oxindoles, see: X. H.
Chen, Q. Wei, S. W. Luo, H. Xiao, L. Z. Gong, J. Am. Chem. Soc.
2009, 131, 13819–13825.
Typical experimental procedure for the reaction of oxindoles 1 with
glyoxal derivatives 2: A mixture of L5 (5 mol%), Sc
(OTf)3 (5 mol%),
[4] For examples of asymmetric allylic alkylation and amination reac-
tions of 2-oxindoles, see: a) B. M. Trost, M. U. Frederiksen, Angew.
Chem. 2005, 117, 312–314; Angew. Chem. Int. Ed. 2005, 44, 308–
310; b) B. M. Trost, Y. Zhang, J. Am. Chem. Soc. 2006, 128, 4590–
4591; c) B. M. Trost, M. K. Brennan, Org. Lett. 2006, 8, 2027–2030;
d) B. M. Trost, Y. Zhang, J. Am. Chem. Soc. 2007, 129, 14548–
14549; e) B. M. Trost, Y. Zhang, Chem. Eur. J. 2010, 16, 296–303;
f) K. Jiang, J. Peng, H. L. Cui, Y. C. Chen, Chem. Commun. 2009,
3955–3957; g) T. Bui, M. Borregan, C. F. Barbas III, J. Org. Chem.
2009, 74, 8935–8938; h) L. Cheng, L. Liu, D. Wang, Y. J. Chen, Org.
Lett. 2009, 11, 3874–3877; i) Z. Q. Qian, F. Zhou, T. P. Du, B. L.
Wang, M. Ding, X. L. Zhao, J. Zhou, Chem. Commun. 2009, 6753–
and 3 ꢁ molecular sieves (2 mg) was stirred in CH2Cl2 (1.0 mL) at 358C
under Ar for 1 h. Oxindole 1a (0.1 mmol) and phenylglyoxal 2a
(0.1 mmol) were then added sequentially at 08C, followed by the addition
of CH2Cl2 (1.0 mL). The reaction was stirred at 08C and monitored by
TLC. The crude product was purified by flash chromatography directly
(petroleum ether/ethyl acetate, 2:1) to afford 3aa as a colorless gel, as an
1
inseparable diastereomeric mixture. H NMR (400 MHz, CDCl3): d=7.97
(s, 1H), 7.42–7.50 (m, 4H), 7.27–7.31 (m, 2H), 6.95–7.02 (m, 1H), 6.91–
6.94 (m, 1H), 6.63–6.65 (d, J=7.6 Hz, 1H), 5.37–5.39 (d, J=6.4 Hz, 1H),
4.19–4.21 (d, J=6.4 Hz, 1H), 1.62 ppm (s, 3H); 13C NMR (100 MHz,
CDCl3): d=200.9, 179.9, 139.7, 135.0, 134.0, 128.8, 128.6, 128.5, 128.4,
Chem. Eur. J. 2010, 16, 3736 – 3742
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3741

X. Feng et al.
6755; j) S. Mouri, Z. H. Chen, H. Mitsunuma, M. Furutachi, S. Mat-
sunaga, M. Shibasaki, J. Am. Chem. Soc. 2010, 132, 1255–1257.
[5] For examples of the Michael addition of 2-oxindoles, see: a) P. Gal-
zerano, G. Bencivenni, F. Pesciaioli, A. Mazzanti, B. Giannichi, L.
Sambri, G. Bartoli, P. Melchiorre, Chem. Eur. J. 2009, 15, 7846–
7849; b) T. Bui, S. Syed, C. F. Barbas III, J. Am. Chem. Soc. 2009,
131, 8758–8759; c) Y. Kato, M. Furutachi, Z. Chen, H. Mitsunuma,
S. Matsunaga, M. Shibasaki, J. Am. Chem. Soc. 2009, 131, 9168–
9169; d) R. J. He, C. Ding, K. Maruoka, Angew. Chem. 2009, 121,
4629–4631; Angew. Chem. Int. Ed. 2009, 48, 4559–4561; e) G. Ben-
civenni, L. Y. Wu, A. Mazzanti, B. Giannichi, F. Pesciaioli, M. P.
Song, G. Bartoli, P. Melchiorre, Angew. Chem. 2009, 121, 7336–
7339; Angew. Chem. Int. Ed. 2009, 48, 7200–7203; f) R. J. He, S.
Shirakawa, K. Maruoka, J. Am. Chem. Soc. 2009, 131, 16620–16621;
g) X. Li, Z. G. Xi, S. Z. Luo, J. P. Cheng, Org. Biomol. Chem. 2010,
8, 77–82.
[6] For reviews of catalytic asymmetric construction of all-carbon qua-
ternary stereocenters, see: a) B. M. Trost, C. Jiang, Synthesis 2006,
369–396; b) Quaternary Stereocenters: Challenges and Solutions for
Organic Synthesis (Eds.: J. Christoffers, A. Baro), Wiley, New York,
2005; c) J. Christoffers, A. Baro, Adv. Synth. Catal. 2005, 347, 1473–
1482; d) C. J. Douglas, L. E. Overman, Proc. Natl. Acad. Sci. USA
2004, 101, 5363–5367; e) D. J. Ramon, M. Yus, Curr. Org. Chem.
2004, 8, 149–183; f) P. G. Cozzi, R. Hilgraf, N. Zimmermann, Eur. J.
Org. Chem. 2007, 5969–5994.
[7] For selected examples, see: a) R. M. Williams, R. J. Cox, Acc. Chem.
Res. 2003, 36, 127–139; b) A. K. Franz, P. D. Dreyfuss, S. L. Schreib-
er, J. Am. Chem. Soc. 2007, 129, 1020–1021; c) J. T. Mohr, M. R.
Krout, B. M. Stoltz, Nature 2008, 455, 323–332.
[8] a) M. Asakawa, K. Miyazawa, Jpn. J. Toxicol. 1998, 11, 361–366;
b) S. Inoue, K. Okada, H. Tanino, K. Hashizume, H. Kakoi, Tetrahe-
dron 1994, 50, 2729–2752; c) K. Okada, Y. Mizuno, H. Tanino, H.
Kakoi, S. Inoue, Chem. Lett. 1989, 703–706; d) S. Inoue, K. Okada,
H. Tanino, H. Kakoi, Tetrahedron Lett. 1986, 27, 5225–5228;
e) D. A. Brown, J. Garthwaite, E. Hayashi, S. Yamada, Br. J. Phar-
macol. 1976, 58, 157–159.
[12] See reference [1a]. The direct aldol-type reaction of 3-substituted-2-
oxindoles with ethyl trifluoropyruvate was developed by chiral
amine catalysis:
[13] For examples of a metal enolate formed in situ in direct aldol reac-
tion, see: a) S. Saito, S. Kobayashi, J. Am. Chem. Soc. 2006, 128,
8704–8705; b) A. Yamaguchi, S. Matsunaga, M. Shibasaki, J. Am.
Chem. Soc. 2009, 131, 10842–10843; c) T. Yoshino, H. Morimoto, G.
Lu, S. Matsunaga, M. Shibasaki, J. Am. Chem. Soc. 2009, 131,
17082–17083.
[14] For selected examples, see: a) K. Zheng, J. Shi, X. H. Liu, X. M.
Feng, J. Am. Chem. Soc. 2008, 130, 15770; b) Z. P. Yu, X. H. Liu,
Z. H. Dong, M. S. Xie, X. M. Feng, Angew. Chem. 2008, 120, 1328;
Angew. Chem. Int. Ed. 2008, 47, 1308; c) D. H. Chen, Z. L. Chen, X.
Xiao, Z. G. Yang, L. L. Lin, X. H. Liu, X. M. Feng, Chem. Eur. J.
2009, 15, 6807–6810.
[15] The aldol reaction of oxindole 1 with a 3-isopropyl group was quite
sluggish and retro aldol reaction was observed during column chro-
matography purification.
[16] a) J. C. Adkins, S. Noble, Drugs 1998, 56, 1055–1064; b) M. Barker,
M. Clackers, D. A. Demaine, D. Humphreys, M. J. Johnston, H. T.
Jones, F. Pacquet, J. M. Pritchard, M. Salter, S. E. Shanahan, P. A.
Skone, V. M. Vinader, I. Uings, I. M. McLay, S. J. F. Macdonald, J.
Med. Chem. 2005, 48, 4507–4510.
[17] For reviews on ligand-accelerated catalysis, see: a) D. J. Berrisford,
C. Bolm, K. B. Sharpless, Angew. Chem. 1995, 107, 1159–1171;
Angew. Chem. Int. Ed. Engl. 1995, 34, 1059–1070; b) J. W. Faller,
A. R. Lavioe, J. Parr, Chem. Rev. 2003, 103, 3345–3367.
[18] For reviews on NLE studies, see: a) D. Guillaneux, S. H. Zhao, O.
Samuel, D. Rainford, H. B. Kagan, J. Am. Chem. Soc. 1994, 116,
9430–9439; b) M. Avalos, R. Babiano, P. Cintas, J. L. Jimꢂnez, J. C.
Palacios, Tetrahedron: Asymmetry 1997, 8, 2997–3017; c) C. Girard,
H. B. Kagan, Angew. Chem. 1998, 110, 3088–3127; Angew. Chem.
Int. Ed. 1998, 37, 2922–2959; d) D. G. Blackmond, Acc. Chem. Res.
2000, 33, 402–411.
[9] For a review of direct aldol reactions, see: Modern Aldol Reactions
(Ed.: R. Mahrwald), Wiley-VCH, Weinheim, 2004.
[10] The pKa value of 3-substituted oxindoles can be expected to be sub-
stantially higher, considering the pKa value of oxindole (pKa ꢀ18.2),
compared with a-arylated cyanoacetates and b-dicarbonyl com-
pounds (pKa ꢀ13).
[11] For reviews of the construction of contiguous sterically hindered
À
chiral carbon stereocenters in C C bond-forming reactions, see:
E. A. Peterson, L. E. Overman, Proc. Natl. Acad. Sci. USA 2004,
101, 11943–11948.
Received: December 18, 2009
Published online: February 19, 2010
3742
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 3736 – 3742