Asymmetric synthesis of indolines through intramolecular shifting of aromatic sulfinyl groups. Role of the π,π-stacking interactions in these unusual SNAr processes

Article abstract of DOI:10.1021/ja8096527

Cyclization of N-aryl substituted 1-aryl-2[2-p-tolylsulfinyl]phenyl propylamines under LDA, LHMDS, or KHMDS provides a new approach for synthesizing optically pure 2,3-disubstituted indolines. Both the scope and the limitations of this method have been investigated. The π,π-stacking interactions are crucial for these unprecedented intramolecular S_NAr processes, in which a sulfinyl group located on a slightly deactivated ring is substituted by amide anions under mild conditions. X-ray and NMR proofs supporting these π,π-stacking interactions are presented.

Full text of DOI:10.1021/ja8096527

Published on Web 06/15/2009
Asymmetric Synthesis of Indolines through Intramolecular
Shifting of Aromatic Sulfinyl Groups. Role of the π,π-Stacking
Interactions in these Unusual SNAr Processes
Jose´ L. Garc´ıa Ruano,*,† Alejandro Parra,† Vanesa Marcos,† Carlos del Pozo,‡
Silvia Catala´n,‡ Silvia Monteagudo,‡ Santos Fustero,*,‡,§ and Ana Poveda|
Departamento de Qu´ımica Orga´nica (C-I), UniVersidad Auto´noma de Madrid, Cantoblanco,
28049 Madrid, Spain, Departamento de Qu´ımica Orga´nica, UniVersidad de Valencia, Burjassot,
46100 Valencia, Spain, Centro de InVestigacio´n, Pr´ıncipe Felipe, E-46013-Valencia, Spain, and
SerVicio Interdepartamental de InVestigacio´n (RMN), UniVersidad Auto´noma de Madrid,
Cantoblanco, 28049 Madrid, Spain
Received December 15, 2008; E-mail: joseluis.garcia.ruano@uam.es; santos.fustero@uv.es
Abstract: Cyclization of N-aryl substituted 1-aryl-2[2-p-tolylsulfinyl]phenyl propylamines under LDA, LHMDS,
or KHMDS provides a new approach for synthesizing optically pure 2,3-disubstituted indolines. Both the
scope and the limitations of this method have been investigated. The π,π-stacking interactions are crucial
for these unprecedented intramolecular SNAr processes, in which a sulfinyl group located on a slightly
deactivated ring is substituted by amide anions under mild conditions. X-ray and NMR proofs supporting
these π,π-stacking interactions are presented.
Scheme 1
Introduction
The importance of the indoline skeletons, present in the
structure of alkaloids1 and other natural products with diverse
biological activities2 and considered as privileged structures3
due to their widespread use as building blocks and as chiral
auxiliaries in asymmetric synthesis,4 has determined that numer-
ous synthetic multistep strategies5 have been developed. How-
ever, the lack of short and efficient methods for obtaining
optically pure substituted indolines has left researchers with the
challenge of finding new, simpler methods for their preparation.
We have recently reported that reactions of 2-p-tolylsulfinyl
benzylcarbanions with N-p-methoxyphenyl (PMP) fluorinated
imines produce indolines with almost complete control of the
stereoselectivity6 (Scheme 1). These reactions involve a tandem
AN/SNAr process. The first step, the nucleophilic addition of
the sulfinylated benzylcarbanion to the fluorinated imine, takes
place in a completely stereoselective manner as previously
observed with other imines.7 Subsequent intramolecular cy-
clization appears to involve a nucleophilic aromatic substitution
of the amide anion intermediate in which the sulfinyl group acts
as the leaving group.
† Departamento de Qu´ımica Orgánica, Univerisidad Auto´noma de
Madrid.
‡ Departamento de Qu´ımica Orgánica, Universidad de Valencia.
§ Centro de Investigacio´n.
| Servicio Interdepartmental de Investigacio´n, Universidad Auto´noma de
Madrid.
(1) (a) Gan, Z.; Reddy, P. T.; Quevillon, S.; Couve-Bonaire, S.; Arya, P.
Angew. Chem., Int. Ed. 2005, 44, 1366. (b) Nicolaou, K. C.; Rao,
P. B.; Hao, J.; Reddy, M. V.; Rassias, G.; Huang, X.; Chen, D. Y.;
Snyder, S. A. Angew. Chem., Int. Ed. 2003, 42, 1753. (c) Ori, M.;
Toda, N.; Takami, K.; Tago, K.; Kogen, H. Angew. Chem., Int. Ed.
2003, 42, 2540.
(5) Recent references representative of the different strategies used for
preparing indolines: (a) Thansandote, P.; Raemy, M.; Rudolph, A.;
Lautens, M. Org. Lett. 2007, 9, 5255. (b) Yamamoto, H.; Pandey, G.;
Asai, Y.; Nakano, M.; Kinoshita, A.; Namba, K.; Imagawa, H.;
Nishizawa, M. Org. Lett. 2007, 9, 4029. (c) Benniaza, R.; Dunet, J.;
Robert, F.; Schenk, K.; Landais, Y. Org. Lett. 2007, 9, 3913. (d) Wang,
Z.; Wan, W.; Jiang, H.; Hao, J. J. Org. Chem. 2007, 72, 9364. (e) Jin,
W.; Trzupek, J. D.; Rayl, T. J.; Broward, M. A.; Veilhauer, G. A.;
Weir, S. J.; Hwang, I.; Boger, D. L. J. Am. Chem. Soc. 2007, 129,
15391. (f) Fuwa, H.; Sasaki, M. Org. Lett. 2007, 9, 3347. (g) Zhang,
G.; Catalano, V. J.; Zhang, L. J. Am. Chem. Soc. 2007, 129, 11358.
(h) Snyder, B. B.; Wu, X. Org. Lett. 2007, 9, 4913. (i) Zhang, H.;
Boonsombat, J.; Padwa, A. Org. Lett. 2007, 9, 279. (j) Matsuda, Y.;
Katahima, M.; Takayama, H. Org. Lett. 2008, 10, 125. (k) Silva-Lo´pez,
C.; Pe´rez-Valado, C.; Rodr´ıguez-Gran˜a, C.; de Lera, A. R. Org. Lett.
2008, 10, 77.
(2) (a) Crich, D.; Banerjee, A. Acc. Chem. Res. 2007, 40, 151. (b) Correa,
A.; Tellitu, I.; Dominguez, E.; SanMartin, R. J. Org. Culghem. 2006,
71, 8316, and references cited therein. (c) Ontoria, J. M.; Marco, S. D.;
Conte, M.; Di Francesco, M. E.; Gardelli, C.; Koch, U.; Matassa, V. G.;
Poma, M.; Steinku¨hler, C.; Volpari, C.; Harper, S. J. Med. Chem. 2004,
47, 6443. (d) Horton, D. A.; Bourne, G. T.; Smythe, M. L. Chem.
ReV. 2003, 103, 893. (e) Hesse, M. Alkaloide; Verlag Helvetica
Chimica Acta Zu¨rich, Wiley-VCH: Weinheim, 2000.
(3) The concept of privileged structures was first introduced by Evans
and recently updated by Patchett. (a) Evans, B. E.; Rittle, K. E.; Bock,
M. G.; Dipardo, R. M.; Freidinger, R. M.; Whitter, W. L.; Lundell,
G. F.; Veber, D. F.; Anderson, P. S.; Chang, R. S. L.; Lotti, V. J.;
Cerino, D. J.; Chen, T. B.; Kling, P. J.; Kunkel, K. A.; Springer, J. P.;
Hirschfield, J. J. J. Med. Chem. 1988, 31, 2235. (b) Patchett, A. A.;
Nargund, R. P. Annu. Rep. Med. Chem. 2000, 35, 289.
(4) Andersson, F.; Hedenstro¨m, E. Tetrahedron: Asymmetry 2006, 1952,
and references cited therein.
(6) Garc´ıa Ruano, J. L.; Alema´n, J.; Catala´n, S.; Marcos, V.; Monteagudo,
S.; Parra, A.; del Pozo, C.; Fustero, S. Angew. Chem., Int. Ed. 2008,
47, 7941.
9
9432 J. AM. CHEM. SOC. 2009, 131, 9432–9441
10.1021/ja8096527 CCC: $40.75  2009 American Chemical Society

Asymmetric Synthesis of Indolines
A R T I C L E S
Scheme 2
mixture of four easily isolated products was obtained (Scheme
2). These included the two starting products, (S)-1 and 2a, in
similar amounts (66% combined yield), along with two new
compounds in the same ratio (34% combined yield). The two
latter compounds, identified with the aid of spectroscopic
methods as p-tolyl disulfide and trans-5a indoline (Scheme 2),
both seem to result from the intramolecular substitution at rt of
the p-Tol-SO- group by the amide anion derived from anti-3a
(the sulfinyl moiety is eliminated as p-tolenesulfenic acid, which
is then readily converted into p-Tol-S-S-p-Tol10). The
amount of the recovered starting products (S)-1 and 2a suggests
that they are formed from the amide anion derived from syn-
4a, which indicates that the nucleophilic addition is reversible
at rt and that, in the case of syn-4a, this reversal is preferred
over its indoline cyclization.
To confirm this hypothesis, we first studied the behavior of
anti-3a (prepared in accordance with the method outlined in
ref 7g) with LDA and LHMDS (easier to handle) at rt. In both
cases, a complete conversion to trans-5a and p-Tol-S-S-p-
Tol was observed after 90 min. This result indicates that
intramolecular substitution of the sulfinyl group is possible from
anti-3a despite slight deactivation of the aromatic ring. In order
to evaluate the scope of this reaction, we reacted various N-aryl
derivatives of anti-1,2-diarylpropylamines (3b-3n) with LH-
MDS (Table 1). Starting from the N-PMP derivatives 3a-3e,
the reactions took place under mild conditions (only 3e required
reflux in THF to produce a 34% conversion). Reaction time
increases were necessary when the electronic density of the ring
joined to C-1 decreased. Thus, the N-phenyl derivatives (3f-3h)
showed the same tendency (entries 6-8), but their reactivity
was lower than that of the PMP derivatives (all of these reactions
were performed at 70 °C). N-p-Chlorophenyl derivatives 3j-3l
required even longer reaction times in refluxing THF to evolve
into the corresponding indolines (entries 10-12), with 3l giving
a conversion of only 40%. Finally, we found that N-p-
cyanophenyl derivatives were not able to react under similar
conditions, even after 48 h in refluxing THF. In all cases in
which the reaction did not take place or the conversion was
incomplete, starting amine 3 was not recovered, but precursor
(S)-1 and imine 2 were, a finding that points to the reversible
character of the nucleophilic addition under these conditions
and the higher stability of the starting products.
When the reactions were performed with KHMDS as base,
the reactivity increased substantially, especially when 18-
crown-6 ether was added (Table 1). Thus, compound 3a reacted
almost instantaneously under these conditions (entry 3) while
3e required 24 h for its conversion to trans-5e (entry 5). Anilines
3g and 3h reacted in less than 5 min at rt (entries 7 and 8) and
3l required just 3 h for complete conversion at 70 °C. Compound
3m, which did not react with LHMDS, evolved with KHMDS
in 24 h (entry 13), but 3n was recovered unaltered after 48 h at
70 °C with KHMDS (entry 14). This substantial increase in
reactivity in the presence of a different base had previously been
observed in the cyclization of fluorinated amines.6 These results
indicate that reactions of amines anti-3 with the indicated bases
constitute a good general method for obtaining trans-5 indolines.
We have also studied the behavior of the amines anti-3o and
anti-3p, derived from 3,4-dimethoxyaniline (see Supporting
Information), with KHMDS, in order to check that their
reactivity was respectively higher than that of 3a and 3e
Considering the interest in new methods for generating
enantiomerically pure substituted indolines, it was necessary to
determine whether our method was applicable to the synthesis
of nonfluorinated indolines, as well as to delimit its possible
structural restrictions. We were also interested in examining two
mechanistic aspects of the SNAr step in more depth. First,
nucleophilic substitution of a sulfinyl group is seldom exploited
in this type of process;8 what enables this transformation in our
work? Second, the sulfinyl group is a moderately deactivated
R-alkylaryl ring; how can the mild conditions9 employed in these
transformations be explained? In this paper, we report on the
results of the reactions of different (2-p-tolylsulfinyl)phenyl
ethylamines with bases to provide a new entry for the asym-
metric synthesis of enantiomerically pure 2,3-disubstituted
indolinesonthebasisofthedirectformationoftheN-C(aromatic)
bond according to a SNAr process. After establishing the
structural requirements needed for the success of this cyclization,
we proposed a significant role for π,π-stacking interactions to
better understand the evolution of these unusual SNAr reactions.
Results and Discussion
When sulfoxide (S)-1 reacted with LDA and the N-PMP
benzylideneimine 2a for 30 min at -78 °C and the resulting
reaction mixture was then protonated at the same temperature,
a 36:64 mixture of anti-3a and syn-4a amines was obtained7g
(Scheme 2). In contrast, when the reaction was maintained at
-78 °C for 30 min and then left for 90 min at room temperature
(rt) before protonation (conditions that are commonly used in
the direct synthesis of indolines from fluorinated imines), a
(7) (a) Garc´ıa Ruano, J. L.; Alema´n, J.; Soriano, J. F. Org. Lett. 2003, 5,
677. (b) Garc´ıa Ruano, J. L.; Alema´n, J. Org. Lett. 2003, 5, 4513. (c)
Garc´ıa Ruano, J. L.; Alema´n, J.; Parra, A. J. Am. Chem. Soc. 2005,
127, 13048. (d) Arroyo, Y.; Meana, A.; Rodr´ıguez, J. F.; Santos, M.;
Sanz-Tejedor, M. A.; Garc´ıa Ruano, J. L. J. Org. Chem. 2005, 70,
3914. (e) Garc´ıa Ruano, J. L.; Aleman, J.; Cid, M. B. Synthesis 2006,
687. (f) Arroyo, Y.; Meana, A.; Rodriguez, J. F.; Santos, M.; Sanz-
Tejedor, M. A.; Garcia Ruano, J. L. J. Org. Chem. 2006, 62, 8525.
(g) Garc´ıa Ruano, J. L.; Aleman, J.; Alonso, I.; Parra, A.; Marcos,
V.; Aguirre, J. Chem.sEur. J. 2007, 13, 6179.
(8) Rossi, R. A.; Pierini, A. B.; Pen˜en˜ory, A. B. Chem. ReV. 2003, 103,
71.
(9) The low number of references on intramolecular SNAr (see ref 8)
always deal with strongly deactivated rings bearing very good leaving
groups (-F or -NO2). See: (a) Jawdosiuk, M.; Kmiotek-Skarzinska,
I.; Czarnecka, E. Pol. J. Chem. 1978, 52, 1837. (b) Fekner, T.; Gallucci,
J.; Chan, M. K. Org. Lett. 2003, 5, 4795. (c) Buttero, P.; del; Molteri,
G.; Papagni, A.; Pilati, T. Tetrahedron 2003, 59, 5259. (d) Domingos,
J. B.; Longhinotti, E.; Bunton, C. A.; Faruk, N. J. Org. Chem. 2003,
68, 7051. (e) Bowman, W. R.; Heaney, H.; Smith, P. H. G. ArkiVoc
2003, 434. (f) Crowley, B. M.; Mori, Y.; McComas, C. C.; Tang, D.;
Boger, D. L. J. Am. Chem. Soc. 2004, 126, 4310.
(10) Davis, F. A.; Arthur, J. F.; Upender, K. N. J. Am. Chem. Soc. 1978,
78, 2844.
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9433

A R T I C L E S
Garc´ıaRuano et al.
Table 1. Results Obtained in the Cyclization of the Amines anti-3a-3q
Table 2. Results Obtained in the Tandem-Cyclization of the
Amines anti-3a-3m
-CF3 group is similar to that of the -CN group (entry 9). The
synthesis of the anti amines derived from 4-trifluoromethyl
benzaldehyde and 4-methoxy or 3,4-dimethoxy anilines was not
possible (reactions only yield compounds with the syn stereo-
chemistry, see Supporting Information).
In order to study the role of substrate relative configuration
on the course of the reaction, we next examined the behavior
of 4a (the syn isomer of 3a) in the presence of LHMDS (Scheme
3). After 1.5 h at rt (which was the time required for complete
evolution of anti-3a; see entry 3, Table 1), syn-4a had
completely disappeared and a mixture of the starting products
(S)-1 and 2a, along with a small amount of the indoline trans-
5a, had been formed. When the reaction time was increased to
5 h, compound trans-5a was the only product isolated in the
reaction (70% yield). Identical results were obtained with syn-
4f at 40 °C after 10 h, which afforded trans-5f as the only
product (67% isolated yield). These results indicate that the syn
isomers 4a and 4f cannot cyclize into their corresponding cis
indolines. In contrast, they must reach an equilibrium with their
corresponding anti-3 isomers (through the starting imines 2a
or 2f and the carbanion derived from (S)-1), which in turn would
be converted into the trans indolines 5a or 5f. This provokes a
shift of the equilibrium between anti-3 and syn-4 toward the
former, which finally leads to complete conversion of syn-4 to
indolines trans-5 (Scheme 3).
(containing a 4-MeOC6H4 group). Reaction of anti-3o was
almost instantaneous at rt (entry 15) whereas anti-3p required
24 h at reflux (entry 17) to be converted in their corresponding
indolines. These results were similar to those obtained from 3a
(entry 3) and 3e (entry 5). However, when the reaction of a 1:1
mixture of anti-3e and anti-3p with more than 2.0 equiv of
KHMDS was followed by 1H NMR it could be confirmed that
reactivity of anti-3p is around 1.8 times higher than that of the
anti-3e. Finally, reaction of anti-3q with KHMDS was unsuc-
cessful (entry 17), indicating that the negative influence of the
This behavior is highly interesting because reactions of (S)-1
with N-arylimines are not usually completely stereoselective,
yielding instead a mixture of syn and anti diastereoisomers. As
both evolve into the same trans indoline through treatment with
9
9434 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

Asymmetric Synthesis of Indolines
A R T I C L E S
Scheme 3
Scheme 4
Scheme 5
trans-5e, which could also be a consequence of the relative
reactivity of their corresponding precursors (see above).
When we studied the behavior of the N-(2,4,6-trimethoxy-
phenyl) amino sulfoxides syn-4o and syn-4p in their reactions
with LHMDS, we observed that only imines 2o and 2p were
obtained along with (S)-1 (Scheme 3). The trans indolines 5o
and 5p, which presumably arise in the cyclization of the anti
isomers 3o and 3p, were not detected. These results were not
wholly unexpected, since the reactions of 2o and 2p with (S)-1
are completely stereoselective, yielding only the syn-4o and syn-
4p isomers.11 The anti isomers required for cyclization are never
formed.7g
Reaction of the syn isomers 4a or 4f with KHMDS, which
had proven to be more efficient for cyclization than LHMDS,
were also unsuccessful. However, in the presence of this base,
the starting aminosulfoxides were recovered unaltered. Cis
indolines and the products resulting in the retro-addition were
never detected, even when the reaction was halted before it had
finished. The same absence of reactivity was observed when
the syn amines obtained in reactions of 4-trifluoromethyl
benzaldehyde and 4-methoxy or 3,4-dimethoxy anilines were
treated with KHMDS. These results not only confirm that cis
indolines cannot be formed by means of this procedure but also
indicate that equilibrium between isomers 3 and 4 does not take
place in the presence of KHMDS.12
Once we had clarified that substrate need to have an anti
configuration for successful intramolecular SNAr to indolines,
we examined the role of other structural features. We first
studied the behavior of anti N-sulfinyl amines 6a and 6b, which
were easily prepared through the reaction of Li-(S)-1 with
N-sulfinylimines,7a-f in the presence of LHMDS (Scheme 4.
Both amines were recovered unaltered after 10 h, which
indicates that no reactions take place, probably due to the stable
and consequently less nucleophilic character of the N-sulfinyl
amide anions with respect to the N-aryl amide anions. We then
studied reactions of primary amines 7a and 7b, which were
obtained by means of N-S cleavage of 6a and 6b with TFA/
MeOH,7a in analogous conditions, but the results were similarly
LHMDS, the reaction mixture, without previous separation,
could be stereoselectively transformed into only one indoline.
This was confirmed starting from the mixture syn-4a + anti-
3a, which was obtained through reaction of 2a with (S)-1. After
5 h with LHMDS, only compound trans-5a was obtained.
Moreover, these results suggest that the tandem AN/SNAr
processes used for synthesizing fluorinated indolines (the AN
in these reactions is usually completely stereoselective, yielding
only anti imines) can also be applied to the synthesis of
nonfluorinated indolines, despite the fact that the first step, which
entails nucleophilic addition, affords mixtures of syn and anti
amines. To corroborate this hypothesis, we reacted (S)-1 with
the N-phenyl benzylideneimine 2g, as well as with other imines
bearing electron-donating (2a and 2f) and electron-withdrawing
(2i) groups at their aromatic rings. All of these reactions gave
satisfactory results (Table 2), which indicates that optically pure
N-aryl trans indolines can be prepared through direct reaction
of the imines with (S)-1, involving AN/SNAr tandem processes.
It is remarkable that the yield of trans-5p is higher than that of
(11) We have also tried to obtain unsuccessfully the indolines containing
p-CF3-C6H4 groups at C-2 by reaction with KHMDS of the syn amines
obtained from 4-trifluoromethyl benzaldehyde and 4-methoxy or 3,4-
dimethoxy anilines.
(12) The reversibility of the nucleophilic addition in the presence of lithium
bases (LDA or LHMDS) could be due to the higher stability of the
lithium benzylcarbanions, which form chelated species that are scarcely
stabilized for potassium benzylcarbanions.
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9435

A R T I C L E S
Garc´ıaRuano et al.
Figure 1. Structural preorganization in the TS leading to indolines.
Scheme 6
unsuccessful. This time, mixtures of compound (S)-1 with
aldehydes 8a or 8b, which arise from the hydrolysis of imines
2a or 2b, were isolated (Scheme 4), but indolines were not
detected even after adding TMDA or 12-crown-4 to increase
the reactivity of the lithium anion amide. Reactions with
KHMDS (or KHMDS/18-crowm-4) were also unsuccessful,
even at refluxing THF, with recovery of the starting products
7a and 7b unaltered.
Taking into account that the nucleophilic power of the amide
anion derived from free amines 7a and 7b must be higher than
that of the N-aryl derivatives 3a and 3b, the different reactivity
of both types of substrates seems to be related to the presence
of the aryl group at the nitrogen. In order to confirm this, we
studied the reaction of fluorinated amines 9 and 11 with
KHMDS (Scheme 5). Whereas compound 9 was transformed
into the indoline under mild conditions,6 compound 11, obtained
from 9 through CAN oxidation, was recovered unaltered after
several hours at 70 °C. This finding indicates that indoline 12
cannot be obtained from free amine 11, which further suggests
that the presence of the N-aryl group is required for the success
of this cyclization.
interaction in the transition states involving the anti-3 isomers,
which are not there when the relative stereochemistry is altered.
The magnitude of this interaction is larger when the electronic
density of the ring joined to C-1 increases, thus explaining the
significant reactivity differences observed when the substituents
at this ring are changed (Table 2).14
To account for all these experimental results, we propose that
the double π,π-stacking interaction shown in Figure 1 is
responsible for the efficiency of this SNAr reaction for com-
pounds anti-3. The ortho-disubstituted ring (A in Figure 1),
which is deactivated by the sulfinyl group and therefore acts as
an acceptor, interacts with the N-aryl group (B in Figure 1),
which is usually activated and therefore acts as the donor.
Moreover, the deactivated p-tolyl ring joined to the sulfur (C
in Figure 1) can interact favorably with the ring joined to C-1
(D in Figure 1) only in the case of the anti isomers 3. In contrast,
syn-4 causes the ring D to be arranged in a gauche position
with respect to the methyl and ring A (Figure 1), which is not
overly favored.
The ready cyclization observed in fluorinated anti amines6
suggests that -CF3 is also able to interact favorably with the
p-methylphenyl group joined to the sulfur (C ring in Figure 1).
Taking into account the electronic deficiency of this ring and
the fact that the -CF3 is an electron-rich group; a donor-acceptor
interaction between both (similar to π,π-stacking) can be
postulated to explain the experimental results (Figure 1).
According to the theoretical calculations performed by
Uneyama,15 the electrostatic interaction of a CF3 functional-
ity with aryl groups is repulsive when the rings are negatively
charged, but it becomes attractive when their electronic density
Taken together, these results suggest an SNAr mechanism.13
The influence of the substituents at the N-aryl ring on the
reactivity, which is generally higher for compounds with
electron-donating groups, can be explained by taking into
account their effect on the nucleophilic strength of the amide
ion, which is lowered by electron withdrawing groups. In this
sense, the lack of reactivity observed for the N-sulfinyl amides
derived from 6 is also understandable. However, several facts
point to a more complex situation. For example, the lack of
reactivity of the amide anions 7 cannot be attributed to their
nucleophilic features, which are more propitious than those of
the N-aryl derivatives anti-3 from both an electronic and steric
perspective, but rather to the absence of the N-aryl group, which
seems to be required for cyclization. This suggests that N-aryl
groups are involved in some sort of interaction able to stabilize
the transition state of the reaction.
(13) Two additional tests were made to corroborate this mechanism. The
first test involved reacting anti-3b with LHMDS in the presence of
TEMPO. The results obtained were identical to those shown in entry
1 of Table 1, indicating that the radical inhibitor has no influence,
which thus excludes a radical process. In contrast, the reaction of (S)-
15 (Scheme 6) with LHMDS did not work, as could be expected from
the negative influence of the methoxy groups on the course of the
SNAr reaction.
(14) As this aryl group is not directly joined to the nitrogen, the influence
of its electronic density on the nucleophilic strength of the amide anion
should be very low.
(15) Katagiri, T.; Yamaji, S.; Handa, M.; Irie, M.; Uneyama, K. Chem.
Commun 2001, 2054.
In addition, the syn or anti stereochemistry of the starting
amides is crucial for the success of the reaction, as can be seen
from the fact that the syn isomers 4 do not react, although this
feature is not related to the nucleophilic power of the amide
anion. This, too, indicates the existence of a stabilizing
9
9436 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

Asymmetric Synthesis of Indolines
A R T I C L E S
Figure 2. Evolution of (SSS)-13 and (RSS)-13 with KHMDS.
3
is decreased by electron-withdrawing substituents, an observa-
tion that supports our proposal. Thus, syn-fluorinated amines
should not evolve into the corresponding cis indolines. Unfor-
tunately, we were unable to prove this because reaction of (S)-1
with fluorinated aldimines and ketimines yielded only anti
amines. We were, however, able to establish the importance of
the stabilizing CF3/aryl interactions in the course of the
cyclization by studying the behavior of amines syn (RSS)-13
and anti (SSS)-13 (Figure 2), obtained as a mixture in the reaction
of 2-p-tolylsulfinyl toluene and the N-p-methoxyphenyl imine
derived from CF3-CHO.6 These compounds differ in their
configuration at the aminic carbon. Compound (SSS)-13 reacts
with KHMDS in less than 1 h, affording the corresponding
indoline (S)-14, whereas the isomer syn (RSS)-13 remained
unaltered after 24 h under similar conditions. As can be seen
in Figure 2, this behavior can be attributed to the CF3/p-tolyl
interaction, which stabilizes the TS from anti (SSS)-13, but which
is absent in the TS from syn (RSS)-13.
At this point it was necessary to find some proofs supporting
the π,π-stacking interactions we have proposed to explain the
experimental results. A combined experimental and theoretical
approach was adopted, using X-ray diffraction, NMR experi-
ments, and molecular mechanics calculations. We first studied
in detail the X-ray data obtained for compound anti-1.6,16
Inspection of Figure 3 reveals that the p-OMe aniline ring (B)
and the ortho-disubstituted ring (A) exhibit an almost parallel
(face-to-face) geometry with an average separation of ca. 3.9
Å, which strongly suggests the presence of an attractive π,π-
interaction. An anti-type arrangement for the torsion angle value
between the CHs protons at the aliphatic chain was observed
in the solid state. Also, an intramolecular SdO · · · H-N
hydrogen bond was evidenced, although its role to favor the
stacked orientation of the aromatic rings was negligible (see
below).
(see Supporting Information). The J vicinal coupling value
between the CHs protons at the aliphatic chain is 10 Hz, also
suggesting a major anti-type conformation at the corresponding
torsion angle, as that deduced in the crystal structure. The strong
NOEs between the CH(NH) and the aromatic protons adjacent
to the aliphatic linker, in both PMP (H1 of ring B) and
orthosulfinylated (H4 of ring A) aromatic moieties, also
indicated their proximity in space. Nevertheless, the simulta-
neous existence of medium intensity NOEs between the aliphatic
CH(CH3) and the same H1 of ring B and H4 of ring A
hydrogens indicated that, in THF solution, there is certain
mobility around the torsional degress of freedom of the
molecule. In any case, these NMR evidence mostly agree with
the X-ray structure and suggest that, although not unique, the
molecule in solution displays a major conformation, with the
PMP-containing ring and the ortho-disubstituted ring, containing
the sulfinyl moiety, stacked in space.
We have also tried to find evidence of these interactions in
the non fluorinated compounds which are described in this paper.
Thus, the X-ray diffraction study of compound anti-3a was
performed.16 However, in contrast to the observations for the
CF2Cl analogue, the aromatic rings in the solid state structure
of anti-3a do not adopt the π,π-stacked geometry (Figure 4),
although the intramolecular SdO· · ·H-N hydrogen bond was
also present, as well as the anti-type geometry at the CH-CH
aliphatic segment. An intermolecular π,π-stacking interaction
of the A and D rings of different molecules can be observed in
the crystal structure (see Supporting Information). Thus, the
different chemical nature at one of the segments of the molecule
gives rise to the presence of a significantly different geometry,
at least in the solid state.
NMR studies were then performed using experimental
conditions similar to those described above, using diluted
samples of anti-3a and THF-d8 as solvent. The 1H NMR
spectrum at 195 K was fairly similar to that of anti-17, with a
JCH-CH vicinal coupling constant value about 10 Hz. Only a
very minor shielding of the protons at ring B was observed.
In a parallel manner, these observations found support by
NMR studies performed on diluted samples of anti-15 at 223
and 273 K in THF-d8. These experimental conditions were
chosen to minimize any possible intermolecular stacking.
Moreover, the choice of this polar solvent could favor hydro-
phobic interactions, as well as disfavor the intramolecular
SdO· · ·H-N hydrogen bond present in the solid state. Weak
NOE contacts among the hydrogens of the PMP ring (B) and
those of the orthosulfinylated ring (A) were clearly detected
(16) CCDC 730583 and CCDC 689575 contain the supplementary crystal-
lographic data of compound anti-3a and anti-17, respectively. These
data can be obtained free of charge at www.ccdc.cam.ac.uk/conts/
retrieving.html (or from the Cambridge Crystallographic Data Centre,
12 Union Rd., Cambridge CB2 1EZ, UK; fax: (int.) +44(1223)336-
033, e-mail: deposit@ccdc.cam.ac.uk).
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9437

A R T I C L E S
Garc´ıaRuano et al.
Figure 3. View of the X-ray structure of compound anti-17 showing the almost parallel geometry of the rings A and B and the interatomic distances of
some significant nuclei.
in Figure 4) with those at the B ring, as well as a medium-
intensity peak between the hydrogens at A and those at the
aromatic D moiety (see figure 4). Therefore, the discussion will
only be focused on the anti-3b analogue.
As a further step, and since all these experimental data could
not be accounted for any single structure, a conformational
search was performed with the aim of scanning the possible
energy minima available for anti-3b. Twenty three conformers
were found by the MM3* force field by random rotation around
all the torsional degrees of freedom of the molecule. The 10
geometries within a 5 kcal/mol threshold were analyzed on the
basis of the experimental NMR data (Table 1 in the Supporting
Information). In particular, the 6.5 Hz value for the vicinal JCH,CH
coupling is in agreement with a ca. 50:50 equilibrium between
gauche-type and anti-type rotamers at the CH-CH aliphatic
Figure 4. ORTEP of compound anti-3a.
linkage.
The analysis of the NOESY experiment performed under these
conditions permitted to detect several NOE crosspeaks between
different hydrogens at the molecule, including aromatic and
aliphatic protons, with somehow different intensities to those
detected for the CF2Cl analogue (see Table 1 in the Supporting
Information). These experimental evidence pointed out that the
aromatic ring B in anti-3a could be involved in different or
additional interactions than those described above for anti-17.
No crystals were obtained for anti-3b and anti-3f, but the
NMR data permitted deduction of the existence of a strikingly
different behavior for these molecules under analogous experi-
mental conditions (diluted samples in THF-d8 and 195 K).
Strong shielding of the protons at rings A and C with respect
to those in anti-3a (and anti-17) was observed, when the
p-methoxy group was attached to ring D, anti-3b and anti-3f.
This behavior was independent of the presence of a p-methoxy
group at ring B (as in anti-3b), or not (as in anti-3f). Also, for
both anti-3b and anti-3f, the 3J vicinal coupling value between
the CHs protons at the aliphatic chain was much smaller than
for anti-3a and anti-17 (6.5 Hz) strongly suggesting the presence
of a conformational equilibrium at the corresponding torsion
angle. Again, several NOE contacts among the hydrogens of
the different aromatic rings were observed, but the presence of
additional nontrivial NOEs all around the molecule could
obviously not been explained by one unique geometry.
The two lowest energy geometries (conf1 and conf2, Figure
6) are anti-type geometries at the CH-CH section of the
molecule, and display a satisfactory agreement with the
experimental X-ray geometry described above for anti-3a. They
are extended structures with no stacking between the aromatic
rings. They slightly differ at the position of the sulfur atom,
although no SdO· · ·H-N hydrogen bond is formed. Both
geometries can explain the strong observed NOE peak between
the CH(NH) and the ortho A-ring proton.
Additionally, there are also two low energy geometries (conf6
and conf10) with a gauche-type conformation at the CH-CH
fragments which, in turn, can explain the medium intensity NOE
crosspeak between the ortho protons at A and D rings. These
conformers display a π,π-stacking interaction between rings D
and C (Figure 6). The combination of conformers conf1 and/or
conf2 with conformers conf6 and/or conf10, in a ca. 50:50 ratio,
can satisfactorily explain the experimental NOE and J NMR
data for anti-3f and anti-3b.
On the observation of the geometries of the alternative
conformers predicted by the MM3* calculations, the minor
conformers contributing to the conformational equilibria
of compounds anti-3a and anti-17 could be deduced. In
particular, conf3 for anti-3a could explain the observed A4-B1
NOE for this molecule, while a minor presence of conf4 for
anti-17 could support the observed medium A4-CH(CH3) NOE.
Moreover, conf7 is probably also contributing in both molecules
and may account for the experimental B1-CH(CH3) NOE
observed for both anti-3a and anti-17. These three conformers
also display the anti geometry at the CH-CH linkage, required
to enlighten the experimental value of 10 Hz for the corre-
Interestingly, despite the presence of different substitutions
at ring B of anti-3f and anti-3b, their NOE crosspeaks, and
their relative intensities were fairly similar, indicating an
analogous conformational behavior for both compounds. As
example, it is possible to identify a weak (but reliable) NOE
contact among the hydrogens at the orthosulfinylated ring (A
9
9438 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

Asymmetric Synthesis of Indolines
A R T I C L E S
1
Figure 5. Section of the H NMR spectra of the (a) anti-3f, (b) anti-3b, (c) anti-3a, and (d) anti-15 compounds showing the differences in chemical shift
in the aromatic rings for each compound.
Figure 6. Low-energy conformers used to explain the experimental results. (a) Anti-type conformers, conf1 and conf2. (b) Gauche-type conformers, conf6
and conf10.
sponding coupling constant. A more complete description of
the different geometries is given in the Supporting Information.
Then, although subtle structural and chemical differences may
modify the existing conformational equilibrium for this type of
molecules, the X-ray and NMR data support the presence of
conformers exhibiting π,π-stacking between rings A and B (for
anti-15) or C and D (for anti-3f and anti-3b). Therefore, the
proposed TS to explain the evolution of the anti isomers into
indolines (Figure 1) could be stabilized by two π,π-stacking
interactions (A/B and C/D rings) and therefore would be favored
with respect to that TS in which only the A/B interaction were
present, as is the case for the syn isomers. This different
stabilization can account for the observed different behavior for
the anti and syn isomers in the unusual SNAr reaction reported
in this paper.
from reactions between 2-p-tolylsulfinylphenyl ethane and
N-aryl arylideneimines with LHMDS or KHMDS. After a
detailed study of this unusual intramolecular SNAr on mildly
deactivated rings under very mild conditions, we have estab-
lished the scope and limitations of this method, which seems
to require a double π,π-stacking donor-acceptor interaction to
stabilize the transition states that lead to the desired indolines.
These interactions have been evidenced from X-ray and NMR
studies.
Experimental Section
General Method. 1H NMR spectra were acquired at either 200
or 300 MHz and 13C NMR were acquired at either 50 or 75 MHz
unless otherwise indicated. Chemical shifts (δ) are reported in ppm
relative to CDCl3 (7.26 and 77.0 ppm). Low-temperature and 2D
NMR studies were performed at 500 MHz. Mass spectra (MS) were
determined by FAB. Melting points were determined in open
capillary tubes. All reactions were carried out in anhydrous solvents
and under argon atmosphere. THF was distilled from sodium-benzo-
phenone under argon. Toluene was dried under metallic sodium.
Molecular sieves (4 Å) were activated at 110 °C for 4 days. i-Pr2NH
Conclusion
In conclusion, we have described a new entry for the
preparation of enantiomerically pure trans indolines containing
two chiral centers which entails treating the amines resulting
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9439

A R T I C L E S
Garc´ıaRuano et al.
was distilled from KOH. Flash silica gel column chromatography
was performed with silica gel Merk-60 (230-400 mesh). n-BuLi
(2.5 M solution in hexane), LHMDS (1 M in THF), KHMDS (0.5
M in Toluene), and compound 2g (imine, PhCHdN-Ph) were
purchased from Aldrich. Compounds (S)-1, 2b, 2c, 2e, 2f, 2i, 2j,
2k, 2n, 6a-b, 7a-b, anti-9, 10, (S,SS)-13, (S,SR)-13, (S)-14, and
(S)-15 were previously synthesized.6,7
123.1, 124.3, 127.3, 127.5, 129.2, 133.5, 136.9, 136.9, 137.1, 139.1,
150.1, 46.2, 55.4, 77.9, 107.9, 114.6, 118.9, 123.2, 124.4, 127.7,
20
128.6, 128.7, 129.8, 133.1, 136.8, 140.7, 149.9, 156.1. [R]D
+100.2 (c ) 1.0, CHCl3). HRMS (M + H)+ calcd for C22H20ClNO:
349.1233; Found: 349.1232.
(2S,3S)-1-Phenyl-3-methyl-2-(4-methoxyphenyl)indoline (trans-
5f). By means of the process described above at rt, trans-5f was
obtained as a yellow oil in 71% yield from amine anti-3f in a step-
by-step process and in 59% yield with a tandem procedure. 1H NMR
(CDCl3, 300 MHz): 1.40 (d, J ) 6.8 Hz, 3H), 3.14-3.24 (m, 1H),
3.76 (s, 3H), 4.68 (d, J ) 9.2 Hz, 1H), 6.78-6.83 (m, 2H),
6.93-7.01 (m, 2H), 7.08-7.11 (m, 4H), 7.13-7.29 (m, 4H). 13C
NMR (CDCl3, 75 MHz): 18.3, 29.7, 46.4, 55.2, 108.1, 113.9, 119.3,
120.9, 122.4, 123.6, 127.4, 127.9, 128.9, 134.0, 134.3, 143.6, 148.2,
158.9. [R]D20 -76.6 (c ) 1.0, CHCl3). HRMS (M + H)+ calcd for
C22H21NO: 315.2034; Found: 315.2190.
General Procedure for Cyclization of Amines anti-3a-n.
Synthesis of the Indolines trans-5a-n (Table 1). Either LHMDS
or KHMDS (0.55 mmol) was added to a solution of the corre-
sponding amine 3a-n (0.5 mmol) in THF (2 mL) at 0 °C. When
the reaction was complete (see Table 1), the resulting reaction
mixture was hydrolyzed (1 mL of H2O), extracted (3 × 10 mL
Et2O), washed (2 × 10 mL sat. NaCl), and dried (MgSO4), and the
solvent was removed under reduced pressure. Compounds were
purified by means of flash silica gel column chromatography with
hexane as the eluent in each case. Yields are indicated in all cases.
Compounds Listed in Table 1. (2S,3S)-1-(4-Methoxyphenyl)-3-
methyl-2-phenylindoline (trans-5a).6 By means of the process
described above at rt, trans-5a was obtained as a white solid; mp:
152-154 °C, in 79% yield from amine anti-3a in a step-by-step
process and in 62% yield with a tandem procedure. 1H NMR
(CDCl3, 300 MHz): 1.41 (d, J ) 6.8 Hz, 3H), 3.29-3.20 (m, 1H),
3.74 (s, 3H), 4.60 (d, J ) 9.1 Hz, 1H), 6.81-6.71 (m, 4H),
7.11-7.05 (m, 4H), 7.31-7.22 (m, 3H), 7.39 (dd, J ) 6.8, 1.5 Hz,
2H). 13C NMR (CDCl3, 75 MHz): 17.5, 46.2, 55.3, 78.2, 107.8,
114.3, 118.8, 123.2, 124.2, 127.3, 127.4, 127.6, 128.4, 133.4, 137.1,
(2S,3S)-1-Phenyl-3-methyl-2-phenylindoline (trans-5g). By
means of the process described above at rt, trans-5g was obtained
as a yellow oil in 81% yield from amine anti-3g in a step-by-step
process and in 54% yield with a tandem procedure. 1H NMR
(CDCl3, 300 MHz): 1.43 (d, J ) 6.8 Hz, 3H), 3.18-3.27 (m, 1H),
4.74 (d, J ) 7.6 Hz, 1H), 6.80 (t, J ) 1.1 Hz, 1H), 6.85 (t, J ) 1.1
Hz, 1H), 6.91-7.30 (m, 12H). 13C NMR (CDCl3, 75 MHz): 18.7,
46.4, 77.1, 108.3, 119.4, 120.6, 122.3, 123.7, 126.7, 127.3, 127.5,
20
128.6, 128.9, 134.0, 142.4, 143.6, 148.0 [R]D -66.6 (c ) 0.3,
CHCl3). HRMS (M + H)+ calcd for C21H19NO: 285.1508; Found:
285.1518.
20
142.1, 149.9, 155.8. [R]D +151.1 (c ) 1, CHCl3). HRMS (M +
(2S,3S)-1-Phenyl-3-methyl-2-(chlorophenyl)indoline (trans-
5h). By means of the process described above at rt, trans-5h was
obtained as a yellow oil in 76% yield from amine anti-3h in a step-
by-step process. 1H NMR (CDCl3, 300 MHz): 1.41 (d, J ) 6.8 Hz,
3H), 3.15-3.22 (m, 1H), 4.71 (d, J ) 9.2 Hz, 1H), 6.83 (t, J ) 6.6
Hz, 1H), 6.96-7.02 (m, 2H), 7.09-7.14 (m, 4H), 7.14-7.32 (m,
6H). 13C NMR (CDCl3, 75 MHz): 18.3, 46.4, 76.4, 108.3, 119.3,
120.8, 122.6, 123.6, 127.6, 128.1, 128.8, 129.1, 133.0, 133.6, 140.9,
H)+ calcd for C22H21NO: 315.1616; Found: 315.1623.
(2S,3S)-1-(4-Methoxyphenyl)-3-methyl-2-(4-methoxyphenyl)in-
doline (trans-5b). By means of the process described above at rt,
trans-5b was obtained as a yellow oil in 80% yield from amine
1
anti-3b in a step-by-step process. H NMR (CDCl3, 300 MHz):
1.40 (d, J ) 6.8 Hz, 3H), 3.20-3.29 (m, 1H), 3.74 (s, 3H), 3.77
(s, 3H), 4.55 (d, J ) 9.1 Hz, 1H), 6.69 (d, J ) 7.8 Hz, 1H),
6.77-6.84 (m, 5H), 7.04-7.12 (m, 4H), 7.32 (d, J ) 8.6 Hz, 2H).
13C NMR (CDCl3, 75 MHz): 17.2, 46.1, 55.2, 55.3, 77.7, 107.7,
113.8, 114.3, 118.3, 123.1, 124.4, 127.5, 128.5, 133.4, 133.9, 139.9,
143.3, 147.9. [R]D -90.2 (c ) 1.0, CHCl3). HRMS (M + H)+
20
calcd for C22H21NO: 315.2034; Found: 315.2190.
(2S,3S)-1-(4-Chlorophenyl)-3-methyl-2-(4-methoxyphenyl)in-
doline (trans-5j). By means of the process described above at rt,
trans-5j was obtained as a yellow oil in 75% yield from amine
anti-3j in a step-by-step process. 1H NMR (CDCl3, 300 MHz): 1.40
(d, J ) 6.8 Hz, 3H), 3.16-3.226 (m, 1H), 3.77 (s, 3H), 4.62 (d, J
) 9.2 Hz, 1H), 6.81-6.86 (m, 2H), 6.93-6.96 (m, 2H), 7.03-7.18
(m, 6H), 7.24-7.27 (m, 2H). 13C NMR (CDCl3, 75 MHz): 18.1,
46.3, 55.2, 76.2, 108.0, 114.0, 119.7, 122.3, 123.7, 127.3, 127.5,
128.0, 129.0, 133.6, 134.0, 142.1, 147.8, 159.0. [R]D20 -21.0 (c )
0.7, CHCl3). HRMS (M + H)+ calcd for C22H20ClNO: 349.1235;
Found: 349.1224.
20
150.1, 155.8, 158.9. [R]D -57.1 (c ) 0.89, CHCl3). HRMS (M
+ H)+ calcd for C23H23NO2: 345.1728; Found: 345.1728.
(2S,3S)-1-(4-Methoxyphenyl)-3-methyl-2-(4-methylphenyl)in-
doline (trans-5c). By means of the process described above at rt,
trans-5c was obtained as a yellow oil in 72% yield from amine
anti-3c in a step-by-step process. 1H NMR (CDCl3, 300 MHz): 1.35
(d, J ) 6.8 Hz, 3H), 2.26 (s, 3H), 3.15-3.21 (m, 1H), 3.68 (s,
3H), 4.53 (d, J ) 9.2 Hz, 1H), 6.67 (d, J ) 7.8 Hz, 1H), 6.72-6.99
(m, 2H), 7.01-7.06 (m, 6H), 7.06-7.25 (m, 3H). 13C NMR (CDCl3,
75 MHz): 17.4, 21.1, 46.2, 55.3, 77.9, 107.7, 114.3, 118.7, 123.1,
124.2, 127.2, 129.1, 133.4, 136.9, 137.1, 143.5, 150.1, 155.8, 159.4.
[R]D -55.4 (c )1.1, CHCl3). HRMS (M + H)+ calcd for
20
(2S,3S)-1-(4-Chlorophenyl)-3-methyl-2-phenylindoline (trans-
5k). By means of the process described above at rt, trans-5k was
obtained as a yellow oil in 71% yield from amine anti-3k in a step-
by-step process. 1H NMR (CDCl3, 300 MHz): 1.34 (d, J ) 6.8 Hz,
3H), 3.10-3.20 (m, 1H), 4.61 (d, J ) 9.2 Hz, 1H), 6.74 (t, J ) 6.6
Hz, 1H), 6.79-7.13 (m, 5H), 7.10-7.27 (m, 7). 13C NMR (CDCl3,
75 MHz): 18.3, 46.4, 55.2, 108.1, 119.3, 121.9, 123.7, 126.7, 127.6,
128.7, 129.0, 129.2, 130.5, 133.9, 141.8, 142.1, 147.7. [R]D20 -85.3
(c ) 1.0, CHCl3). HRMS (M + H)+ calcd for C21H18Cl: 319.1137;
Found: 319.1128.
C23H23NO: 329.1779; Found: 329.1781.
(2S,3S)-1-(4-Methoxyphenyl)-3-methyl-2-(4-chlorophenyl)in-
doline (trans-5d). By means of the process described above at rt,
trans-5d was obtained as a yellow oil in 68% yield from amine
anti-3a in a step-by-step process and in 58% yield with a tandem
procedure. 1H NMR (CDCl3, 300 MHz): 1.32 (d, J ) 6.8 Hz, 3H),
3.07-3.17 (m, 1H), 3.66 (s, 3H), 4.53 (d, J ) 9.2 Hz, 1H), 6.60
(d, J ) 7.8 Hz, 1H), 6.69-6.71 (m, 3H), 6.95-7.02 (m, 5H),
7.16-7.32 (m, 3H). 13C NMR (CDCl3, 75 MHz): 17.3, 155.8. [R]D
20
-87.2 (c ) 0.78, CHCl3). HRMS (M + H)+ calcd for C23H20N2O:
340.1278; Found: 340.1225.
(2S,3S)-1-(4-Chlorophenyl)-3-methyl-2-(4-chlrophenyl)indo-
line (trans-5l). By means of the process described above at rt, trans-
5l was obtained as a yellow oil in 84% yield from amine anti-3l in
(2S,3S)-1-(4-Methoxyphenyl)-3-methyl-2-(4-cyanophenyl)in-
doline (trans-5e). By means of the process described above at rt,
trans-5e was obtained as a yellow oil in 82% yield from amine
anti-3e in a step-by-step process and in 25% yield with a tandem
procedure. 1H NMR (CDCl3, 300 MHz): 1.39 (d, J ) 6.8 Hz, 3H),
3.17-3.27 (m, 1H), 3.72 (s, 3H), 4.54 (d, J ) 9.2 Hz, 1H),
6.69-6.79 (m, 4H), 7.03-7.09 (m, 4H), 7.24-7.28 (m, 3H). 13C
NMR (CDCl3, 75 MHz): 17.4, 21.1, 46.2, 55.4, 107.7, 114.3, 118.7,
1
a step-by-step process. H NMR (CDCl3, 300 MHz): 1.40 (d, J )
6.8 Hz, 3H), 3.13-3.23 (m, 1H), 4.65 (d, J ) 9.2 Hz, 1H), 6.81 (t,
J ) 6.6 Hz, 1H), 6.86 (d, J ) 1.1 Hz, 1H), 6.93-7.25 (m, 10H).
13C NMR (CDCl3, 75 MHz): 14.1, 22.7, 31.6, 75.8, 108.2, 120.0,
122.1, 123.7, 127.7, 128.1, 129.0, 129.1, 133.3, 133.6, 141.9, 147.7
[R]D -88.9 (c ) 0.8, CHCl3). HRMS (M + H)+ calcd for
20
C21H17Cl2N: 353.0734; Found: 353.0738.
9
9440 J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009

Asymmetric Synthesis of Indolines
A R T I C L E S
(2S,3S)-1-(4-Cyanophenyl)-3-methyl-2-(4-methoxyphenyl)in-
doline (trans-5m). By means of the process described above at rt,
trans-5m was obtained as a yellow oil in 57% yield from amine
anti-3m in a step-by-step process. Compound trans-5m was
4-((2S,3S)-1-(3,4-Dimethoxyphenyl)-3-methylindolin-2-yl)ben-
zonitrile (trans-5p). By means of the process described above at
rt, trans-5o was obtained as a yellow oil in 81% yield from amine
1
anti-3p in a step-by-step process. H NMR (CDCl3, 300 MHz):
1
1.33 (d, J ) 6.8 Hz, 3H), 3.10-3.20 (m, 1H), 3.66 (s, 3H), 4.51
(d, J ) 9.2 Hz, 1H), 6.49-6.53 (m, 2H), 6.61-6.67 (m, 4H),
6.71-6.78 (m, 2H), 6.78-7.54 (m, 3H). 13C NMR (CDCl3, 75
MHz): 17.6, 29.7, 46.3, 55.9, 56.0, 107.2, 108.3, 111.5, 111.7,
114.9, 119.5, 123.7, 126.1, 126.7, 126.9 (2C), 127.9, 132.4, 132.8,
obtained together with p-tolyldisulfide (50%). H NMR (CDCl3,
300 MHz): 1.43 (d, J ) 6.9 Hz, 3H), 3.17-3.22 (m, 1H), 3.64 (s,
3H), 3.73 (s, 3H), 4.52 (d, J ) 9.1 Hz, 1H), 6.60-6. (m, 4H),
6.96-6.98 (m, 5H), 7.00-7.22 (m, 3H). 13C NMR (CDCl3, 75
MHz): 17.4, 21.1, 46.2, 55.3, 79.9, 107.7, 114.3, 118.7, 123.1,
124.2, 127.2, 127.5, 129.2, 133.4, 137.0, 139.0, 150.0. [R]D20 -71.9
(c ) 0.8, CHCl3).
20
136.9, 147.9, 149.4. [R]D -35.6 (c ) 0.5, CHCl3).
Acknowledgment. Financial support of this work by the
Ministerio de Educacio´n y Ciencia (Grants No. CTQ2006-06741/
BQU and CTQ2007-61462) is gratefully acknowledged. A.P., V.M.,
S.C., and S.M. express their thanks for predoctoral fellowships.
(2S,3S)-1-(3,4-Dimethoxyphenyl)-3-methyl-2-phenylindoline
(trans-5o). By means of the process described above at rt, trans-
5o was obtained as a yellow oil in 83% yield from amine anti-3o
1
in a step-by-step process. H NMR (CDCl3, 300 MHz): 1.35 (d, J
Supporting Information Available: Experimental procedures,
spectroscopic data for synthesis of imines 2d, 2h, 2j, 2l, 2m
and the compounds 3a-n, and NMR and X-ray data used to
support the π,π-stacking interactions. This material is available
free of charge via the Internet at http://pubs.acs.org.
) 6.7 Hz, 3H), 3.19-3.12 (m, 1H), 3.67 (s, 3H), 3.33 (s, 3H),
4.58 (d, J ) 8.9 Hz, 1H), 6.62-6.73 (m, 4H), 7.04-7.21 (m, 6H),
7.33 (d, J ) 9.2 Hz, 2H). 13C NMR (CDCl3, 75 MHz): 17.4, 21.1,
46.2, 55.3, 56.3, 107.7, 114.3, 118.7, 123.1, 124.2, 126.3, 126.7
20
(2C), 127.2, 127.5, 129.2, 133.4, 135.3, 137.0, 139.0, 150.0. [R]D
-88.9 (c ) 0.6, CHCl3).
JA8096527
9
J. AM. CHEM. SOC. VOL. 131, NO. 26, 2009 9441