Catalytic asymmetric formal [3+3] cycloaddition of an azomethine ylide with 3-indolylmethanol: Enantioselective construction of a six-membered piperidine framework

Article abstract of DOI:10.1002/chem.201304187

A catalytic asymmetric formal [3+3] cycloaddition of 3-indolylmethanol and an in situ-generated azomethine ylide has been established to construct a chiral six-membered piperidine framework with two stereogenic centers. This approach not only represents the first enantioselective cycloaddition of isatin-derived 3-indolylmethanol, but also has realized an unusual enantioselective formal [3+3] cycloaddition of azomethine ylide rather than its common [3+2] cycloadditions. Besides, this protocol combines the merits of a multicomponent reaction and organocatalysis, which efficiently assembles a variety of isatin-derived 3-indolylmethanols, aldehydes, and amino esters into structurally diverse spiro[indoline-3,4′-pyridoindoles] with one all-carbon quaternary stereogenic center in high yields and excellent enantioselectivities (up to 93 % yield, >99 % enantiomeric excess (ee)). Although the diastereoselectivity of the reaction is generally moderate, most of the diastereomers can be separated by using column chromatography followed by preparative TLC. An unusual addition: The first catalytic asymmetric formal [3+3] cycloaddition of isatin-derived 3-indolylmethanol with an in situ-generated azomethine ylide has been established to construct a chiral six-membered piperidine framework with two stereogenic centers. This approach represents the first enantioselective cycloaddition of isatin-derived 3-indolylmethanol, and it has realized an unusual enantioselective formal [3+3] cycloaddition of the azomethine ylide (DCEa =a 1,2-dichloroethane). Copyright

Full text of DOI:10.1002/chem.201304187

DOI: 10.1002/chem.201304187
Full Paper
&
Asymmetric Synthesis
Catalytic Asymmetric Formal [3+3] Cycloaddition of an
Azomethine Ylide with 3-Indolylmethanol: Enantioselective
Construction of a Six-Membered Piperidine Framework
Feng Shi,* Ren-Yi Zhu, Wei Dai, Cong-Shuai Wang, and Shu-Jiang Tu*[a]
Abstract: A catalytic asymmetric formal [3+3] cycloaddition
of 3-indolylmethanol and an in situ-generated azomethine
ylide has been established to construct a chiral six-mem-
bered piperidine framework with two stereogenic centers.
This approach not only represents the first enantioselective
cycloaddition of isatin-derived 3-indolylmethanol, but also
has realized an unusual enantioselective formal [3+3] cyclo-
addition of azomethine ylide rather than its common [3+2]
cycloadditions. Besides, this protocol combines the merits of
a multicomponent reaction and organocatalysis, which effi-
ciently assembles a variety of isatin-derived 3-indolylmetha-
nols, aldehydes, and amino esters into structurally diverse
spiro[indoline-3,4’-pyridoindoles] with one all-carbon quater-
nary stereogenic center in high yields and excellent enantio-
selectivities (up to 93% yield, >99% enantiomeric excess
(ee)). Although the diastereoselectivity of the reaction is gen-
erally moderate, most of the diastereomers can be separated
by using column chromatography followed by preparative
TLC.
Introduction
The enantioselective 1,3-dipolar cycloaddition (1,3-DC) of azo-
methine ylides to dipolarophiles has proven to be a powerful
tool to synthesize chiral nitrogenous heterocycles.[1] As a result,
catalytic asymmetric azomethine ylide-involved 1,3-DCs have
achieved elegant developments.[2,3] However, in most of these
cycloadditions, electron-deficient olefins[2] or alkynes[3] were
employed as dipolarophiles to react with azomethine ylides
through a [3+2] reaction mode, which always delivered five-
membered heterocycles such as pyrrolidines [Eq. (1)]. In sharp
contrast, the enantioselective construction of six-membered
heterocycles through azomethine ylide-involved cycloadditions
has met with little success [Eq. (2)].
Recently, the two groups of Waldmann[4] and Wang[5] inde-
pendently discovered the catalytic asymmetric [6+3] cycload-
ditions of azomethine ylides with fulvenes to provide stereose-
lective piperidine derivatives [Eq. (3)]. During the preparation
of this manuscript, Wang, Guo et al., also reported a chiral CuI-
complex catalyzed cross-1,3-DC between pyrazolidinium ylides
and azomethine ylides to give stereoselective 1,2,4-triazinane
frameworks [Eq. (4)].[6] In spite of these creative works, the cat-
alytic asymmetric azomethine ylide-involved cycloadditions
leading to enantioenriched six-membered heterocyclic archi-
tectures are still underdeveloped and thus highly desirable be-
cause of the importance of chiral six-membered heterocycles.
3-Indolylmethanols have recently been recognized as a type
of active reaction components for their characteristic of being
easily transformed into vinyliminium or carbocation intermedi-
ates in the presence of a Lewis or Brønsted acid (LA or BH),[7]
which are resonance structures of a delocalized cation. Howev-
[a] Prof. Dr. F. Shi, R.-Y. Zhu, W. Dai, C.-S. Wang, Prof. S.-J. Tu
School of Chemistry and Chemical Engineering
Jiangsu Normal University
Xuzhou, 221116 (China)
Fax: (+86)516-83500065
E-mail: fshi@jsnu.edu.cn
laotu@jsnu.edu.cn
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201304187.
Chem. Eur. J. 2014, 20, 2597 – 2604
2597
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
er, previous reports were focused on nucleophilic substitutions
of 3-indolylmethanols [Eq. (5)],[8] and hardly any cycloadditions
of 3-indolylmethanols have been found in the literature, not to
mention their enantioselective transformations [Eq. (6)].[9] Thus,
the 3-indolylmethanol-involved cycloaddition, in particular, its
enantioselective transformation has become a formidable chal-
lenge.
ed by chiral phosphoric acid to undergo enantioselective
formal [3+3] cycloadditions, leading to the formation of six-
membered piperidine framework and the production of the
spiro[indoline-3,4’-pyridoindole] skeleton with multiple stereo-
genic centers.
In this work, we present the first catalytic asymmetric formal
[3+3] cycloadditions of isatin-derived 3-indolylmethanols and
in situ-generated azomethine
ylide, which directly assembles
isatin-derived
3-indolylmetha-
nols, aldehydes, and amino
esters into chiral spiro[indoline-
3,4’-pyridoindoles] with two ste-
reogenic centers including one
all-carbon quaternary center in
high yields and excellent enan-
tioselectivities (up to 93% yield,
>99% enantiomeric excess (ee)).
Considering this challenge and the great demand for azome-
thine ylide-involved cycloadditions leading to six-membered
heterocycles, we wondered whether isatin-derived 3-indolyl-
methanols could be employed as dipolarophiles to undergo
catalytic asymmetric formal [3+3] cycloadditions with azome-
thine ylides in the presence of a chiral catalyst, thereby afford-
ing stereoselective spiro[indoline-3,4’-pyridoindoles], in which
a six-membered piperidine framework would be constructed
with one all-carbon quaternary stereogenic center [Eq. (7)].
Figure 1. Selected natural products and bioactive compounds containing
the core structures of spiro[indoline-3,4’-pyridoindoles]
Results and Discussion
The initial attempt to validate our hypothesis commenced with
a three-component reaction of N-benzyl isatin-derived 3-indo-
lylmethanol 1a, 4-nitrobenzaldehyde 2a, and diethyl 2-amino-
malonate 3a in the presence of 10 mol% of chiral phosphoric
acid (CPA) 4a in toluene at 308C (Table 1, entry 1), which
smoothly proceeded through the designed formal [3+3]
reaction mode although the yield and the stereoselectivity is
moderate.
On the other hand, our targeted formal [3+3] cycloaddition
products, that is, spiro[indoline-3,4’-pyridoindoles], contain two
scaffolds of tetrahydropyridoindole and spiro[indoline-piperi-
dine], which exist in a variety of natural products[10] and phar-
maceuticals[11] with significant bioactivities such as being
MDM2-p53 inhibitors[11a] and sodium channel blockers[11b]
(Figure 1), thereby holding great synthetic importance.
The screening of catalysts revealed that CPA 4 f with the
bulky 9-phenanthrenyl group at the 3,3’-positions of the BINOL
backbone exhibited higher catalytic activity than the others
with regard to enantioselective control (Table 1, entry 6 vs. 1–
5), delivering chiral spiro[indoline-3,4’-pyridoindole] 5aaa with
quaternary stereogenic center in 85% ee. The subsequent eval-
uation on solvents in the presence of CPA 4 f disclosed that
chlorine-containing alkanes were superior to toluene in terms
of enantioselectivity and reactivity (Table 1, entries 7–10 vs.
entry 6). Among them, 1,2-dichloroethane (DCE) delivered the
We have recently realized a series of chiral phosphoric acid-
catalyzed cycloadditions[12] for the synthesis of enantioenriched
heterocycles.[13] Inspired by this success and the fact that there
have been few reports either on enantioselective cycloaddi-
tions of 3-indolylmethanols or on [3+3] cycloadditions of azo-
methine ylides, we envisioned that isatin-derived 3-indolylme-
thanols and azomethine ylides could be simultaneously activat-
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2598
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
(Table 2, entries 7–11). However, increasing or de-
creasing the stoichiometry of 3-indolylmethanol 1a
had no obvious effect on the reaction (Table 2, en-
tries 7–9), and using an excess of 2a and 3a at the
same time did not benefit the reaction with regard
to stereoselectivity (entries 10 and 11). So, the reac-
tion conditions as shown in Table 2, entry 1 were fi-
nally selected as the most suitable one for its perfor-
mance in delivering high yields and excellent
enantioselectivity.
Table 1. The screening of catalysts and solvents.[a]
To provide a variety of chiral spiro[indoline-3,4’-pyr-
idoindoles] with structural diversity, we then carried
out the study on the substrate scope of the formal
[3+3] cycloaddition. First, the applicability of the re-
action for various N-substituted isatin-derived 3-indo-
lylmethanols 1 was examined. As shown in Table 3,
this protocol is amenable to a series of N-benzyl
isatin-derived 3-indolylmethanols 1a–h with electron-
donating or -withdrawing substituents linked to the
benzyl group (Table 3, entries 1–8), offering the de-
sired spiro-products in high yields (67–93%) and with
excellent enantioselectivities (97–>99% ee). General-
ly speaking, 3-indolylmethanols 1 with electronically
rich substituents on the benzyl moiety exhibited
better reactivity and enantioselectivity than those
with electronically poor substituents (Table 3, en-
tries 2–5 vs. 6–8), and 1c with a para-methyl (pMe)
group delivered the highest enantioselectivity of
>99% ee (entry 3). It seemed that the position of the
methyl group exerted some effect on the enantiose-
Entry
4
Solvent
Yield
[%][b]
d.r.[c]
ee
[%][d]
1
2
3
4
5
6
7
8
4a
4b
4c
4d
4e
4 f
4 f
4 f
4 f
4 f
4 f
4 f
toluene
toluene
toluene
toluene
toluene
toluene
CH2Cl2
CHCl3
CCl4
DCE
THF
CH3CN
57 (29)[e]
74 (37)[e]
58 (39)[e]
85 (58)[e]
79 (62)[e]
69 (53)[e]
90 (57)[e]
83 (46)[e]
70 (45)[e]
85 (54)[e]
–
50:50
50:50
67:33
68:32
78:22
77:33
63:37
56:44
64:36
63:37
–
56(41)[f]
68(29)[f]
45
80
69
85
97
83
87
98
–
9
10
11[g]
12[g]
–
–
–
[a] Unless indicated otherwise, the reaction was carried out in 0.1 mmol scale in sol-
vent (1.5 mL) with 3 ꢁ MS (100 mg) for 24 h, and the ratio of 1a/2a/3a was 1.5:1.2:1.
[b] Isolated total yields of two diastereomers. [c] The diastereomeric ratio (d.r.) was de-
termined by HPLC. [d] The ee value refers to the major diastereomer 5aaa and was
determined by HPLC. [e] The yield in parentheses refers to that of the major diastereo-
mer 5aaa. [f] The ee value in parentheses refers to another diastereomer 5aaa’. [g] No
desired reaction occurred.
Table 2. Further optimization of the reaction conditions.[a]
product 5aaa in high yield of 85% and excellent
enantioselectivity of 98% ee, albeit with a moderate
diastereomeric ratio (d.r.) of 63:37 (Table 1, entry 10).
However, using tetrahydrofuran (THF) and acetonitrile
as the reaction media failed to facilitate the formal
[3+3] cycloaddition, which just gave the azomethine
ylide generated from the condensation of 2a and 3a
(Table 1, entries 11 and 12). Thus, DCE was chosen as
the solvent of choice for further investigation.
Entry
1a/2a/3a
MS
T
Yield
[%][b]
d.r.[c]
ee
[%][d]
[8C]
The further optimization of reaction conditions
was concentrated on changing other reaction para-
meters such as additives, temperature, and the ratio
of the reagents to improve the diastereoselectivity of
the reaction without the sacrifice of the enantioselec-
tivity and the yield (Table 2). The screening of the
molecular sieves (MS) found that no other MS was
better than 3 ꢁ MS (Table 2, entries 2 and 3 vs.
entry 1). Raising or lowering the reaction temperature
could hardly improve the diastereoselectivity (Table 2,
entries 4–6 vs. entry 1), and the yield was decreased
dramatically at 08C (entry 6). At last, the ratio of
the reagents was carefully tuned, which included
a change in the stoichiometry of both 3-indolylme-
thanol and the in situ-generated azomethine ylide
1
2
3
4
5
6
7
8
1.5:1.2:1
1.5:1.2:1
1.5:1.2:1
1.5:1.2:1
1.5:1.2:1
1.5:1.2:1
2:1.2:1
3:1.2:1
1:1.2:1
1:2.4:2
1:3.6:3
3 ꢁ
4 ꢁ
5 ꢁ
3 ꢁ
3 ꢁ
3 ꢁ
3 ꢁ
3 ꢁ
3 ꢁ
3 ꢁ
3 ꢁ
30
30
30
45
15
0
30
30
30
30
30
85 (54)[e]
68 (41)[e]
29 (18)[e]
79 (47)[e]
70 (44)[e]
27 (17)[e]
88 (51)[e]
89 (52)[e]
76 (48)[e]
94 (56)[e]
83 (48)[e]
63:37
60:40
62:38
60:40
63:37
64:36
58:42
58:42
63:37
60:40
58:42
98
97
92
97
97
97
98
96
97
97
97
9
10
11
[a] Unless indicated otherwise, the reaction was carried out on a 0.1 mmol scale cata-
lyzed by 10 mol% 4 f in DCE (1.5 mL) with MS (100 mg) for 24 h. [b] Isolated total
yields of two diastereomers. [c] Determined by HPLC. [d] The ee value refers to the
major diastereomer 5aaa and was determined by HPLC. [e] The yield in parentheses
refers to that of the major diastereomer 5aaa.
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2599
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
product 5maa in perfect enantioselectivity of >99%
(Table 4, entry 3). However, in the case of the indole
core, the position of the substituents exerted an op-
posite effect on the stereoselectivity. Namely, C5’-sub-
stituted 3-indolylmethanols were superior to C6’- and
C7’-substituted ones with regard to stereoselectivity
(Table 4, entries 10 and 11 vs. 8 and 9), whereas C6’-
substituted 3-indolylmethanol delivered higher enan-
tioselectivity than the C7’-substituted one (Table 4,
entry 9 vs. 8).
Table 3. The applicability of the reaction for various N-substituted isatin-derived 3-in-
dolylmethanols 1.[a]
Entry
5
R
Yield
[%][b]
d.r.[c]
ee
[%][d]
Finally, the substrate scope with respect to aromat-
ic aldehydes 2 was explored by the reactions with 3-
indolylmethanol 1a and amino ester 3a under the
optimized reaction conditions. As indicated in
Table 5, various aromatic aldehydes 2 substituted
with electron-poor, -neutral, or -rich groups were em-
ployed to the reaction, which formed a variety of
azomethine ylides in situ to participate in the desired
formal [3+3] cycloadditions, giving chiral spiro[indo-
line-3,4’-pyridoindoles] 5 with structural diversity in
high yields (73–89%). However, it is evident that the
electronic nature of aldehydes imposed some impact
on the stereoselectivity of the reaction. Indeed, aro-
matic aldehydes substituted with electron-withdraw-
ing groups delivered higher enantioselectivities (80
to >99% ee) than those substituted with electron-
donating or neutral groups, whereas the latter
offered better diastereoselectivities (90:10 to
1
2
3
4
5
6
7
8
9
10
5aaa
5baa
5caa
5daa
5eaa
5 faa
5gaa
5haa
5iaa
C6H5CH2 (1a)
85 (54)[e]
93 (60)[e]
79 (49)[e]
76 (46)[e]
80 (50)[e]
70 (43)[e]
67 (40)[e]
81 (50)[e]
58 (34)[e]
83 (42)[e]
63:37
64:36
62:38
60:40
62:38
62:38
60:40
62:38
58:42
50:50
98
98
>99
98
98
p-tBuC6H4CH2 (1b)
p-MeC6H4CH2 (1c)
m-MeC6H4CH2 (1d)
o-MeC6H4CH2 (1e)
p-FC6H4CH2 (1 f)
p-BrC6H4CH2 (1g)
m-ClC6H4CH2 (1h)
Ph (1i)
97
97
98
89
5jaa
Me (1j)
95 (54[f])
[a] Unless indicated otherwise, the reaction was carried out in 0.1 mmol scale cata-
lyzed by 10 mol% 4 f in DCE (1.5 mL) with 3 ꢁ MS (100 mg) for 24 h, and the ratio of
1/2a/3a was 1.5:1.2:1. [b] Isolated total yields of two diastereomers. [c] Determined
by 1H NMR spectroscopy except for 5aaa. [d] The ee value refers to the major diaste-
reomer 5 and was determined by HPLC. [e] The yield in parentheses refers to that of
the major diastereomer 5. [f] The ee value in parentheses refers to another diastereo-
mer 5’.
lectivity, because pMe-substituted 3-indolylmethanols
1c offered a higher ee value than its ortho- and
Table 4. The generality of the reaction for various 3-indolylmethanols 1 bearing differ-
ent substituents at the phenyl rings.[a]
meta-methyl-substituted
counterparts
(Table 3,
entry 3 vs. 4 and 5). In addition, N-aryl- and N-alkyl
isatin-derived 3-indolylmethanols as exemplified by
1i and 1j could also be utilized to the cycloaddition
to generate the corresponding spiro-products, but
with decreased reactivity and stereoselectivity
(Table 3, entries 9 and 10).
Next, we investigated the generality of the reaction
for N-benzyl isatin-derived 3-indolylmethanols 1 bear-
ing different substituents at the phenyl rings of both
isatin and indole skeletons. As illustrated in Table 4,
this approach is applicable to a wide scope of 3-indo-
lylmethanols with various substituents at different
position of the two phenyl rings, offering structurally
diverse spiro[indoline-3,4’-pyridoindoles] in moderate
to high yields of 42–90% and with good to excellent
enantioselectivities of 79 to >99% ee. Basically, the
position of the substituents had an obvious influence
on the stereoselectivity of the reaction. As for the
isatin moiety, it was found that C5-substituted 3-in-
dolylmethanols were inferior to C6- and C7-substitut-
ed ones in stereoselective control (Table 4, entries 6
and 7 vs. 1–5), and C7-substituted 3-indolylmethanols
were superior to C6-substituted ones in terms of dia-
stereoselectivity (entries 1–3 vs. 4–5). Among them,
7-methyl 3-indolylmethanol 1m offered the spiro-
Entry
5
R1/R2
Yield
[%][b]
d.r.[c]
ee
[%][d]
1
2
3
4
5
6
7
8
5kaa
5laa
7-F/H (1k)
7-Br/H (1l)
7-CH3/H (1m)
6-Cl/H (1n)
6-Br/H (1o)
79 (56)[e]
42 (31)[e]
71 (48)[e]
54 (33)[e]
90 (59)[e]
42 (22)[e]
56 (30)[e]
66 (22)[e]
77 (44)[e]
66 (40)[e]
77 (28)[e]
71:29
74:26
67:33
62:38
66:34
53:47
53:47
33:67
57:43
60:40
37:63
97
96
>99
98
98
94
93
79 (35)[f]
87
5maa
5naa
5oaa
5paa
5qaa
5raa
5saa
5taa
5uaa
5-Cl/H (1p)
5-CH3/H (1q)
6-Br/7’-CH3 (1r)
6-Br/6’-CH3 (1s)
6-Br/5’-CH3 (1t)
6-Br/5’-F (1u)
9
10
11
94
96 (68)[f]
[a] Unless indicated otherwise, the reaction was carried out in 0.1 mmol scale cata-
lyzed by 10 mol% 4 f in DCE (1.5 mL) with 3 ꢁ MS (100 mg) for 24 h, and the ratio of
1/2a/3a was 1.5:1.2:1. [b] Isolated total yields of two diastereomers. [c] Determined
1
by H NMR spectroscopy. [d] The ee value refers to the major diastereomer 5 and was
determined by HPLC. [e] The yield in parentheses refers to that of the major diastereo-
mer 5. [f] The ee value in parentheses refers to another diastereomer 5’.
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2600
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
dehyde 2 and amino ester 3a. Then, as illustrated in
the transition states (TS) I–II, the formal [3+3] cyclo-
addition of intermediate 6 or 7 with azomethine
ylide 8 might proceed through a sequential Michael
addition (TS-I) and Pictet-Spengler reaction (TS-II) to
form the desired six-membered heterocyclic skeleton.
In the transition states of this reaction, CPA 4 f acted
as a Brønsted acid/Lewis base bifunctional catalyst to
simultaneously activate both intermediate 7 and azo-
methine ylide 8 by a hydrogen-bonding interaction.
This activation led to an enantioselective formal
[3+3] cycloaddition due to the chiral environment
created by (R)-BINOL backbone and the bulky 3,3’-(9-
phenanthrenyl)-substituents of CPA 4 f, thereby offer-
ing the experimentally observed (1’S, 3S)-configured
product 5.
Table 5. The substrate scope of aromatic aldehydes.[a]
Entry
5
R
Yield
[%][b]
d.r.[c]
ee
[%][d]
1
2
3
4
5aaa
5aba
5aca
5ada
5aea
5afa
5aga
4-NO2 (3a)
3-NO2 (3b)
4-CN (3c)
3,4-Cl2 (3d)
4-Cl (3e)
85 (54)[e]
82 (43)[e]
73 (40)[e]
80 (38)[e]
89 (51)[e]
86 (77)[e]
88 (88)[e]
63:37
52:48
55:45
48:52
57:43
90:10
>95:5
98
80 (31)[f]
96
>99 (75)[f]
94 (83)[f]
81
5
6[g]
7[g]
H (3 f)
4-CH3O (3g)
68
As shown in Figure 2, the two diastereomers of
(1’S, 3S)-5ada and (1’S, 3R)-5ada’ have different ab-
solute configuration at the 3-position but have the
same configuration at the 1’-position, which indicat-
ed that there existed little chiral induction and dis-
crimination in the first step of Michael addition. As il-
lustrated in TS-I and TS-I’, The two stereo-modes of
nucleophilic substitution of azomethine ylide 8 to vi-
nyliminium 7 had a small difference in steric hin-
[a] Unless indicated otherwise, the reaction was carried out in 0.1 mmol scale cata-
lyzed by 10 mol% 4 f in DCE (1.5 mL) with 3 ꢁ MS (100 mg) for 24 h, and the ratio of
1a/2/3a was 1.5:1.2:1. [b] Isolated total yields of two diastereomers. [c] Determined
1
by H NMR spectroscopy. [d] The ee value refers to the major diastereomers 5 and was
determined by HPLC. [e] The yield in parentheses refers to that of the major diastereo-
mer 5. [f] The ee value in parentheses refers to another diastereomer 5’. [g] The reac-
tion was catalyzed by 10 mol% 4e in CHCl3 (1.5 mL).
>99:1 d.r.) than the former (Table 5, entries 1–5 vs. 6 and 7).
Besides, the position of the substituents also had an effect on
the stereoselectivity, since p-nitrobenzaldehyde 3a was far su-
perior to its meta-substituted counterpart 3b in stereoselective
control (Table 5, entry 1 vs. 2). It should be mentioned that
most of the diastereomers generated from the formal [3+3] cy-
cloadditions can be separated by using column chromatogra-
phy followed by preparative TLC (see the Supporting Informa-
tion). However, compounds 5aba and 5aea were obtained as
inseparable mixtures of two diastereomers, which can hardly
be separated even by preparative TLC.
drance and could isomerize to each other, thereby resulting
in unsatisfactory diastereoselectivities in some cases. On
the contrary, the second step of the Pictet–Spengler
reaction of intermediate 9 (TS-II) was proven to be the key
step to establish a highly enantioselective formal [3+3] cyclo-
addition. So, chiral induction and discrimination between
CPA 4 f and substrates mainly occurred in the second step,
leading to the production of (1’S, 3S)-5 with excellent
enantioselectivities.
To investigate the role of the NÀH group in the indole
moiety of 3-indolylmethanol and to testify our proposed acti-
vation mode, we performed a control experiment by using
As illustrated in Figure 2, the absolute configuration of spi-
ro[indoline-3,4’-pyridoindole] 5ada was unambiguously deter-
mined to be (1’S, 3S) by X-ray analysis on its single crystal in
>99% ee.[14] The configurations of other spiro[indoline-3,4’-pyr-
idoindoles] 5 were assigned by analogy. Besides, to gain some
insight into the reaction pathway related to stereoselective
control, we also tested the absolute configuration of another
diastereomer 5ada’ to be (1’S, 3R) by the same method
(Figure 2).[14]
N-benzyl-protected 3-indolylmethanol 1v as
a substrate
(Scheme 2). However, no desired cycloaddition product 5vaa
was observed under the optimal reaction conditions. Instead,
a Michael addition product 9vaa was generated in 38% yield,
which indicated that the NÀH group of the indole moiety
played a crucial role in the reaction to generate formal [3+3]
cycloaddition products. On the other hand, the formation of
product 9vaa supported our hypothesis that this formal [3+3]
cycloaddition proceeded through a sequential Michael addi-
tion/Pictet–Spengler reaction pathway rather than a concerted
process.
On the basis of our experimental results and previous re-
ports on catalytic asymmetric 3-indolylmethanol-involved reac-
tions[8a,g–h] as well as 1,3-DCs of azomethine ylides catalyzed by
CPA,[3a,13a] we suggested a possible reaction mechanism and re-
lated transition states to explain the chemistry and stereo-
chemistry of this catalytic asymmetric three-component formal
[3+3] cycloaddition (Scheme 1). Initially, the treatment of
isatin-derived 3-indolylmethanol 1 with CPA 4 f afforded the
corresponding carboncation 6 or vinyliminium 7 intermediate,
which transformed into each other in fast equilibrium and
served as an efficient electrophilic reagent to react with azo-
methine ylide 8 generated in situ from the condensation of al-
Conclusion
We have established the first catalytic asymmetric formal [3+3]
cycloaddition of isatin-derived 3-indolylmethanol and in situ-
generated azomethine ylide. This approach not only represents
the first enantioselective cycloaddition of isatin-derived 3-indo-
lylmethanol, but has also realized an unusual enantioselective
formal [3+3] cycloaddition of azomethine ylide rather than its
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2601
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Scheme 1. Proposed reaction mechanism and transition states.
common [3+2] cycloadditions, thus constructing a chiral six-
membered piperidine framework with two stereogenic centers.
Besides, this protocol efficiently assembles a variety of isatin-
derived 3-indolylmethanols, aldehydes, and amino ester into
structurally diverse spiro[indoline-3,4’-pyridoindoles] with one
all-carbon quaternary stereogenic center in high yields and ex-
cellent enantioselectivities (up to
93% yield, >99% ee). Although
the diastereoselectivity of the re-
action is generally moderate,
most of the diastereomers can
be separated by using column
chromatography followed by
preparative TLC. So, this strategy
provides the first enantioselec-
tive construction of this type of
spiro-architecture, which exists
in many natural products and
pharmaceuticals.
Scheme 2. Control experiment involving N-benzyl-protected 3-indolylmethanol 1v.
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2602
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
7.86 (m, 2H), 7.75–7.68 (m, 3H), 7.67–7.62 (m, 2H), 7.25
(d, J=3.9 Hz, 2H), 7.21 (d, J=8.1 Hz, 1H), 6.95–6.88 (m,
2H), 6.74 (d, J=7.9 Hz, 1H), 6.71–6.63 (m, 1H), 6.17 (d,
J=3.4 Hz, 1H), 5.27 (d, J=15.0 Hz, 1H), 5.11 (d, J=
15.0 Hz, 1H), 4.28–4.20 (m, 1H), 4.12–4.04 (m, 1H), 3.84
(d, J=3.5 Hz, 1H), 3.76–3.68 (m, 1H), 3.64–3.56 (m, 1H),
1.19 (t, J=7.1 Hz, 3H), 0.67 ppm (t, J=7.1 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=175.2, 167.9, 166.6, 148.2,
148.1, 142.4, 136.3, 134.8, 133.4, 131.8, 130.7, 129.9,
128.4, 124.9, 124.2, 123.1, 122.3, 122.1, 120.0, 118.3,
111.1, 108.9, 108.1, 70.9, 61.8, 61.7, 54.6, 52.3, 44.0, 13.8,
13.2 ppm; IR (KBr): n˜ =3350, 2928, 1726, 1602, 1520,
1444, 1201, 841, 741 cmÀ1; ESI FTMS m/z calcd for
C37H31BrN4O7: 721.1298 [C37H31BrN4O7ÀH]À; found:
721.1337.
Acknowledgements
We are grateful for financial support from National
Natural Science Foundation of China (21372002 and
21232007), Open Foundation of Jiangsu Key Labora-
tory of Green Synthetic Chemistry for Functional Ma-
terials (K201314), PAPD of Jiangsu Higher Education
Institutions, and Graduate Students Foundation of
Jiangsu Normal University (2013YYB123).
Keywords: asymmetric synthesis · cycloaddition ·
enantioselectivity
·
multicomponent reactions
·
organocatalysis · ylides
[1] a) F. J. Sardina, H. Rapoport, Chem. Rev. 1996, 96, 1825–
1872; b) I. Coldham, R. Hufton, Chem. Rev. 2005, 105, 2765–
2809; c) G. Pandey, P. Banerjee, S. R. Gadre, Chem. Rev. 2006,
106, 4484–4517; d) V. Nair, T. D. Suja, Tetrahedron 2007, 63,
12247–12275; e) W. H. Pearson in Studies in Natural Product
Chemistry, Vol. 1 (Ed.: Atta-Ur-Rahman), Elsevier, Amsterdam,
1998, p. 323.
Figure 2. X-ray structures of 5ada and 5ada’.
[2] For early reports on metal-catalyzed 1,3-DCs, see: a) J. M.
Longmire, B. Wang, X. Zhang, J. Am. Chem. Soc. 2002, 124,
13400–13401; b) A. S. Gothelf, K. V. Gothelf, R. G. Hazell,
K. A. Jørgensen, Angew. Chem. 2002, 114, 4410–4412; Angew. Chem. Int.
Ed. 2002, 41, 4236–4238; for reviews on metal-catalyzed 1,3-DCs, see:
c) C. Nꢂjera, J. M. Sansano, Angew. Chem. 2005, 117, 6428–6432; Angew.
Chem. Int. Ed. 2005, 44, 6272–4276; d) L. M. Stanley, M. P. Sibi, Chem.
Rev. 2008, 108, 2887–2902; e) J. Adrio, J. C. Carretero, Chem. Commun.
2011, 47, 6784–6794; for early reports on organocatalyzed 1,3-DCs,
see: f) J. L. Vicario, S. Reboredo, D. Badꢃa, L. Carrillo, Angew. Chem. 2007,
119, 5260–5262; Angew. Chem. Int. Ed. 2007, 46, 5168–5170; g) I. Ibra-
hem, R. Rios, J. Vesely, A. Cordova, Tetrahedron Lett. 2007, 48, 6252–
6257; h) M.-X. Xue, X.-M. Zhang, L.-Z. Gong, Synlett 2008, 691–694; i) Y.-
K. Liu, H. Liu, W. Du, L. Yue, Y.-C. Chen, Chem. Eur. J. 2008, 14, 9873–
9877; for reviews on organo-catalyzed 1,3-DCs, see: j) A. Moyano, R.
Rios, Chem. Rev. 2011, 111, 4703–4832; k) H. Pellissier, Tetrahedron 2012,
68, 2197–2232.
Experimental Section
Representative experimental procedures
Synthesis of (1’S, 3S)-diethyl 1-(4-bromobenzyl)-1’-(4-nitrophen-
yl)-2-oxo-1’,2’-dihydrospiro[indoline-3,4’-pyrido[3,4-b]indole]-
3’,3’(9’H)-dicarboxylate (5gaa):
A solution of aldehyde 2a
(0.12 mmol), amino ester 3a (0.1 mmol), the catalyst 4 f
(0.01 mmol), and 3 ꢁ molecular sieves (100 mg) in DCE (0.5 mL)
was stirred at 308C for 30 mins. Next, a solution of 3-indolylmetha-
nols 1g (0.15 mmol) in DCE (1 mL) was added. After being stirred
at 308C for 24 h, the reaction mixture was filtered to remove mo-
lecular sieves and the solid powder was washed with ethyl acetate.
The resultant solution was concentrated under the reduced pres-
sure to give the residue, which was purified through flash column
chromatography on silica gel (eluent: petroleum ether/ethyl ace-
tate=6:1), then preparative thin layer chromatography (toluene/
ethyl acetate=5:1); total yield: 67%; 60:40 d.r.; yield of 5 gaa:
40% as a sticky oil. [a]D20 = +268 (c=0.6, CHCl3); enantiomeric
excess: 97%, as determined by HPLC (Daicel Chirapak AD-H,
[3] a) F. Shi, S.-W. Luo, Z.-L. Tao, L. He, J. Yu, S.-J. Tu, L.-Z. Gong, Org. Lett.
2011, 13, 4680–4683; b) F. Shi, Z.-L. Tao, J. Yu, S.-J. Tu, Tetrahedron:
Asymmetry 2011, 22, 2056–2064; c) F. Shi, G.-J. Xing, W. Tan, R.-Y. Zhu, S.
Tu, Org. Biomol. Chem. 2013, 11, 1482–1489; d) F. Shi, R.-Y. Zhu, X.
Liang, S.-J. Tu, Adv. Synth. Catal. 2013, 355, 2447–2458.
[4] M. Potowski, J. O. Bauer, C. Strohmann, A. P. Antonchick, H. Waldmann,
Angew. Chem. 2012, 124, 9650–9654; Angew. Chem. Int. Ed. 2012, 51,
9512–9516.
hexane/isopropanol=70:30, flow rate 1.0 mLminÀ1
254 nm): tR =19.107 min (major), tR =45.533 min (minor); 1H NMR
(400 MHz, [D6]acetone): d=9.98 (s, 1H), 8.37–8.27 (m, 2H), 7.92–
, T=308C,
[5] Z.-L. He, H.-L. Teng, C.-J. Wang, Angew. Chem. 2013, 125, 3006–3010;
Angew. Chem. Int. Ed. 2013, 52, 2934–2938.
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2603
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[6] a) M.-C. Tong, X. Chen, H.-Y. Tao, C.-J. Wang, Angew. Chem. 2013, 125,
12603–12606; Angew. Chem. Int. Ed. 2013, 52, 12377–12380; b) H. Guo,
H. Liu, F.-L. Zhu, R. Na, H. Jiang, Y. Wu, L. Zhang, Z. Li, H. Yu, B. Wang, Y.
Xiao, X.-P. Hu, M. Wang, Angew. Chem. 2013, 125, 12873–12877; Angew.
Chem. Int. Ed. 2013, 52, 12641–12645.
Phytochemistry 1989, 28, 308–309; c) A. Itoh, T. Tanahashi, N. Nagakura,
T. Nishi, Phytochemistry 2003, 62, 359–369.
[11] a) J. Zhang, Z. Zhang, U. S. Pat. Appl. Publ. 2010, US 20100210674 A1
20100819; b) M. Chafeev, S. Chowdhury, J. Fu, R. Kamboj, D. Hou, S. Liu,
PCT Int. Appl., 2008, WO 2008046084 A2 20080417; c) O.-G. Berge, A.
Claesson, B.-M. Swahn, PCT Int. Appl., 2001, WO 2001005790 A1
20010125.
[7] For a review, see: A. Palmieri, M. Petrini, R. R. Shaikh, Org. Biomol. Chem.
2010, 8, 1259–1270.
[8] For enantioselective substitutions of 3-indolylmethanols, see: a) Q.-X.
Guo, Y.-G. Peng, J.-W. Zhang, L. Song, Z. Feng, L.-Z. Gong, Org. Lett.
2009, 11, 4620–4623; b) P. G. Cozzi, F. Benfatti, L. Zoli, Angew. Chem.
2009, 121, 1339–1342; Angew. Chem. Int. Ed. 2009, 48, 1313–1316;
c) B. Han, Y.-C. Xiao, Y. Yao, Y.-C. Chen, Angew. Chem. 2010, 122, 10387–
10389; Angew. Chem. Int. Ed. 2010, 49, 10189–10191; d) D.-S. Wang, J.
Tang, Y.-G. Zhou, M.-W. Chen, C.-B. Yu, Y. Duan, G.-F. Jiang, Chem. Sci.
2011, 2, 803–806; e) J. Xiao, K. Zhao, T.-P. Loh, Chem. Asian J. 2011, 6,
2890–2894; f) J. Xiao, Org. Lett. 2012, 14, 1716–1719; g) L. Song, Q.-X.
Guo, X.-C. Li, J. Tian, Y.-G. Peng, Angew. Chem. 2012, 124, 1935–1938;
Angew. Chem. Int. Ed. 2012, 51, 1899–1902; h) C. Guo, J. Song, J.-Z.
Huang, P.-H. Chen, S.-W. Luo, L.-Z. Gong, Angew. Chem. 2012, 124,
1070–1074; Angew. Chem. Int. Ed. 2012, 51, 1046–1050.
[9] For a sole racemic [4+3] cycloaddition, see: a) X. Han, H. Li, R. P.
Hughes, J. Wu, Angew. Chem. 2012, 124, 10536–10539; Angew. Chem.
Int. Ed. 2012, 51, 10390–10393; for enantioselective reactions, see: b) B.
Xu, Z.-L. Guo, W.-Y. Jin, Z.-P. Wang, Y.-G. Peng, Q.-X. Guo, Angew. Chem.
2012, 124, 1083–1086; Angew. Chem. Int. Ed. 2012, 51, 1059–1062; c) J.
Huang, S. Luo, L. Gong, Acta Chim. Sin. 2013, 71, 879–883.
[12] For early examples, see: a) T. Akiyama, J. Itoh, K. Yokota, K. Fuchibe,
Angew. Chem. 2004, 116, 1592–1594; Angew. Chem. Int. Ed. 2004, 43,
1566–1568; b) D. Uraguchi, M. Terada, J. Am. Chem. Soc. 2004, 126,
5356–5357; for reviews, see: c) T. Akiyama, Chem. Rev. 2007, 107, 5744–
5758; d) M. Terada, Chem. Commun. 2008, 4097–4112; e) M. Terada,
Synthesis 2010, 1929; f) J. Yu, F. Shi, L.-Z. Gong, Acc. Chem. Res. 2011, 44,
1156–1171.
[13] a) F. Shi, Z.-L. Tao, S.-W. Luo, S.-J. Tu, L.-Z. Gong, Chem. Eur. J. 2012, 18,
6885–6894; b) F. Shi, G.-J. Xing, R.-Y. Zhu, W. Tan, S. Tu, Org. Lett. 2013,
15, 128–131; c) F. Shi, W. Tan, R.-Y. Zhu, G.-J. Xing, S.-J. Tu, Adv. Synth.
Catal. 2013, 355, 1605–1622.
[14] CCDC-967946 (5ada) and CCDC-965690 (5ada’) contain the supple-
mentary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
Received: October 27, 2013
[10] a) S. P. Hollinshead, M. L. Trudell, P. Skolnick, J. M. Cook, J. Med. Chem.
1990, 33, 1062–1069; b) J. Garnier, J. Mahuteau, M. Plat, C. Merienne,
Published online on February 2, 2014
Chem. Eur. J. 2014, 20, 2597 – 2604
www.chemeurj.org
2604
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim