Highly enantioselective trapping of zwitterionic intermediates by imines

Article abstract of DOI:10.1038/nchem.1406

Reactions with the unstable and highly reactive zwitterionic intermediates generated in processes catalysed by transition metals are providing new opportunities for molecular constructions. Insertion reactions involve the collapse of zwitterionic intermediates, but trapping them would allow structural elaborations that are not currently available. To synthesize complex molecules in this manner, reactive electrophiles can be used to trap the zwitterionic intermediates. Here, we describe the use of imines, activated by chiral organocatalysts, and a highly efficient integrated rhodium and chiral Brnsted acid co-catalysed process to trap zwitterionic intermediates that have been proposed previously to undergo a formal C-H insertion reaction, allowing us to obtain polyfunctionalized indole and oxindole derivatives in a single step with excellent diastereoselectivity and enantioselectivity.

Full text of DOI:10.1038/nchem.1406

ARTICLES
PUBLISHED ONLINE: 29 JULY 2012 | DOI: 10.1038/NCHEM.1406
Highly enantioselective trapping of zwitterionic
intermediates by imines
Huang Qiu†, Ming Li†, Li-Qin Jiang, Feng-Ping Lv, Li Zan, Chang-Wei Zhai, Michael P. Doyle and
Wen-Hao Hu
*
Reactions with the unstable and highly reactive zwitterionic intermediates generated in processes catalysed by transition
metals are providing new opportunities for molecular constructions. Insertion reactions involve the collapse of zwitterionic
intermediates, but trapping them would allow structural elaborations that are not currently available. To synthesize
complex molecules in this manner, reactive electrophiles can be used to trap the zwitterionic intermediates. Here, we
describe the use of imines, activated by chiral organocatalysts, and a highly efficient integrated rhodium and chiral
Brønsted acid co-catalysed process to trap zwitterionic intermediates that have been proposed previously to undergo a
formal C–H insertion reaction, allowing us to obtain polyfunctionalized indole and oxindole derivatives in a single step with
excellent diastereoselectivity and enantioselectivity.
arbenoids generated by transition metals and diazo com-
It has been proposed that the reaction of a carbenoid with a
pounds are among the most active and useful intermediates highly nucleophilic indole substrate could also generate a proposed
C
in organic chemistry1. Owing to their highly Lewis acidic zwitterionic intermediate with a similar iminium fragment (Fig. 1c,
character, they undergo numerous electrophilic transformations, intermediate III)12. Several catalytic systems have been reported that
including X–H insertions2, cyclopropanations3 and cycloaddition can be used for the C–H functionalization of indole with diazo com-
reactions4. Ylide species formed from carbenoids and nucleophiles pounds derived from this intermediate13,14. We posited that this
such as amines, alcohols and mercaptans are very useful for carrying intermediate could also be trapped by an imine, and therefore
out carbene transfer reactions, and collectively have become an effi- decided to develop a novel three-component reaction for the effi-
cient tool for constructing C–X bonds5.
cient construction of indole derivatives.
C–H functionalization is an effective approach for the direct and
To control the reaction selectivity (chemo-, diastereo- and enan-
selective replacement of C–H bonds with new bonds (such as C–C, tioselectivity) of these potential transformations, the design of a
C–O and C–N)6. These transformations have broad potential in proper catalytic system is crucial. From previous studies it is clear
synthetic reactions because unactivated C–H bonds are the simplest that a chiral rhodium complex is the most widely used type of catalyst
and most common structural motifs to appear in naturally occurring in diazo decompositions15. Numerous types of chiral catalytic systems
organic molecules7. Transition metal-catalysed reactions of diazo have been developed to create enantiomerically enriched products16,17
.
compounds represent one of the most efficient means for bringing Based on previous experience we were aware that trapping active
about C–H bond functionalization8.
intermediates such as ammonium ylides and oxonium ylides with
A recent theoretical study9 confirmed the widely held belief that C–H different electrophiles gave only negligible enantioselectivity when
bond insertion occurs in a single step via a three-centred transition state using a chiral rhodium catalyst18. To achieve asymmetric catalysis, a
(Fig. 1a, intermediate I) and that the rhodium catalyst does not directly chiral co-catalyst19 that could contribute to the creation of a strong
interact with the C–H bond undergoing insertion10. Among the various chiral environment and also activate the reaction component and
types of C–H insertion reactions in which diazo compounds partici- suppress other side reaction pathways was investigated. Recently,
pate, Doyle et al. proposed the existence of a zwitterionic intermediate our group reported rhodium and chiral phosphoric acid (PPA) co-
(Fig. 1b, intermediate II) during the insertion of a carbenoid generated catalysed asymmetric multicomponent reactions20,21. We proved
by N-aryl diazoamide into an aromatic ring11. This zwitterionic inter- that the combination of a rhodium catalyst and a PPA organo-catalyst
mediate is electronically favourable, because the positive charge of the is compatible with diazo decomposition chemistry22,23. Furthermore,
intermediate is stabilized by the electron-rich oxindole with the this is a promising strategy in the development of new and valuable
iminium component. In addition, the negative charge is stabilized by reactions as it is possible that unprecedented transformations may
the carbenoid component and usually undergoes rapid proton transfer not be achievable using one catalyst alone24.
to generate C–H-functionalized products (Fig. 1b, path I).
We envisioned that chiral PPA-activated imines might be appro-
As part of our ongoing interest in exploring novel reactions priate electrophiles25,26 to be used to trap zwitterionic intermediates
involving the trapping of active intermediates to efficiently construct and achieve high diastereoselectivity and enantioselectivity. If
polyfunctional molecules, we considered developing a new process true, this would permit facile access to both polyfunctional indole
to trap the proposed zwitterionic intermediates using electrophiles and oxindole derivatives with contiguous chiral centres. Specifically,
such as activated imines (Fig. 1b, path II). Demonstrating this trap- it would facilitate the construction of two C–C bonds in a single-
ping process would provide experimental evidence for the existence step transformation, which remains challenging using existing meth-
of the proposed zwitterionic intermediates, and also an efficient odologies27. Furthermore, the net transformation would be a comp-
approach to oxindole derivatives.
lement to existing asymmetric synthetic methodologies involving
Shanghai Engineering Research Centre of Molecular Therapeutics and New Drug Development, East China Normal University, Shanghai, 200062, China,
†These authors contributed equally to this work. e-mail: whu@chem.ecnu.edu.cn
*
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.
1

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1406
ARTICLES
–
[Rh]
a
N2
b
1, 2 proton
transfer
A
F
E
[Rh]
N
[Rh]
+
B
H
–
O[Rh]
O
O
+
+
D
N
Path I
N
N
O
A
Intermediate II
Ar2
Ar1
A
D
*
Path II
N
O
B
C
F
H
H
O
H
Ar1
*
N
P
O
This chemistry
C
[Rh]
O
N
*
O
E
Ar2
Intermediate I
c
–
N2
[Rh]
–
[Rh]
Ph
O
Ph
Ph
COOMe
Ph
COOMe
1, 2 proton
transfer
COOMe
[Rh]
OMe
A
D
H
+
[Rh]
+
B
C
C
F
Path I
+
+
N
N
N
E
N
B
Intermediate III
Ph
NHAr1
MeO2C
*
Path II
*
O
Ar2
O
H
Ar1
P
This chemistry
O
N
*
O
N
Ar2
Figure 1 | Concept for trapping zwitterionic intermediates with imines. a, Proposed three-centred intermediate (I) for C–H bond insertion. b, Two pathways
for the transformation based on zwitterionic intermediates. Zwitterionic intermediate II, generated by N-aryl diazoamides in the presence of a rhodium
catalyst, can undergo 1–2 proton transfer to form C–H insertion products (path I). The envisioned path II would form oxindole derivatives by trapping the
zwitterionic intermediate by an imine with the activation of chiral PPA. c, Extension of the zwitterionic intermediate trapping process to a three-component
reaction. Pathways are also shown for the transformation derived from a similar zwitterionic intermediate (III) generated by methyl phenyldiazoacetate
and an indole in the presence of a rhodium catalyst.
C–H functionalization of indoles and oxindoles, the derivatives of reaction that 5aa was not derived from compound A and imine
which are widely present in bioactive natural products28,29
.
2a in the presence of Rh2(OAc)4, which proved that the zwitterionic
intermediate trapping process worked.
Result and discussion
To control the reaction selectivity effectively, various chiral PPAs
To validate our hypothesis, we slowly added a solution of a-methyl- were screened to activate imine 2a to achieve the best stereo-
N-methyl-N-phenyl diazoacetamide (1a) in CH2Cl2 (DCM) into selectivity. We first carried out catalytic asymmetric reactions
a mixture of imine 2a and a catalytic amount of Rh2(OAc)4 using chiral PPA 7a (10 mol%) and determined that the desired
(2 mol%) in DCM. The desired product 5aa was obtained at 65% product 5aa was obtained with the same diastereomeric ratio
yield with good diastereoselectivity (89:11) (Table 1, entry 1). A (d.r.) (89:11) and 51% enantiomeric excess (e.e.) (Table 1,
formal intramolecular C–H insertion process was observed to entry 2). A control experiment for the reaction in the absence of
compete with the trapping process, as C–H insertion compound Rh2(OAc)4 resulted in total recovery of the starting materials,
A (Fig. 1b) was also isolated. We determined through a control indicating that PPA 7a alone was unable to catalyse the reaction.
Table 1 | Catalyst screening and optimization of reaction conditions.
Cl
R
Rh2(OAc)4
N2
(S)-7a: R = Ph
(S)-7f: R = 9-phenanthryl
(S)-7g: R = 4-CF3Ph
(S)-7h: R = 4-MeOPh
R
*
N
catalyst 7
N
N
H
O
O
(S)-7b: R = SiPh3
(S)-7c: R = 3,5-(CF3)2Ph
R
O
O
+
P
O
Solvent
*
OH
Cl
N
temperature
(S)-7d: R = 1,1'-biphenyl-4yl (S)-7i: R = 3,5-Cl2Ph
(S)-7e: R = 2-naphthyl
R
1a: R = Me
1b: R = Et
1c: R = H
2a
5aa: R = Me
5ba: R = Et
5ca: R = H
Catalyst (S)-7
Entry
1
Co-catalyst Solvent T (8C) Yield (%)*
(d.r., syn:anti)†
e.e. (%)‡ Entry
1
Co-catalyst Solvent T (8C) Yield (%)*
(d.r., syn:anti)†
65 (90:10)
e.e. (%)‡
1
1a
–
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
DCM
25
25
25
25
25
25
25
25
25
65 (89:11)
62 (89:11)
70 (96:4)
65 (87:13)
60 (89:11)
55 (87:13)
58 (88:12)
73 (96:4)
55 (92:8)
0
51
10
11
1a (S)-7i
1a (S)-7b
1a (S)-7b
1a (S)-7b
1a (S)-7b
1a (S)-7b
1a (S)-7b
1b (S)-7b
1c (S)-7b
DCM
25
62
5 0
37
87
88
90
91
2
3
4
5
6
7
8
9
1a (S)-7a
1a (S)-7b
1a (S)-7c
1a (S)-7d
1a (S)-7e
1a (S)-7f
1a (S)-7g
1a (S)-7h
To lu en e 2 5
5 2 ( 93 :7 )
68 (95:5)
35 (96:4)
70 (96:4)
73 (97:3)
75 (97:3)
77 (98:2)
70 (90:10)
81
12
13
14
15
16
17
18
DCE
25
54
48
50
48
80
70
THF
25
DCM
DCM
DCM
DCM
DCM
0
210
220
220
220
96
83
Reaction conditions: Rh2(OAc)4/1/2a/7 ¼ 0.002/0.2/0.1/0.01 mmol, 1 in 1 ml solvent was added to a solution of 2a, 7 and Rh2(OAc)4 in 1 ml solvent under an argon atmosphere for 1 h. *Isolated yield.
†Determined either by 1H NMR spectroscopy or high-performance liquid chromatography (HPLC) analysis of the crude reaction mixture. ‡Determined by chiral HPLC analysis.
2
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1406
ARTICLES
Table 2 | Enantioselective reaction of diazoacetamides with various imines.
Ar2
R1
N2
R2
N
Ar1
Rh2(OAc)4 (2 mol%)
N
H
R3
R1
(S)-7b (10 mol%)
N
Ar2
O
Ar1
+
O
N
R3
DCM, –20 °C
R2
1
2
5
Entry Product
Yield
(%)*
d.r.
e.e.
Entry Product
Yield
(%)*
d.r.
e.e.
(syn:anti)†
(%)‡
(syn:anti)†
(%)‡
MeO
MeO
OMe
Cl
Cl
N
H
O
1
77
98:2
96
8
77
99:1
45
N
H
O
N
5ba
N
5bh
Br
N
H
2
3
4
5
84
75
57
65
99:1
99:1
99:1
99:1
97
97
95
96
9
57
99:1
98:2
93:7
98:2
65
90
89
88
N
H
O
N
O
5bb
5bc
5bd
5be
N
5bi
5ab
5cb
Br
Br
N
H
10
81
N
H
O
N
O
N
Br
CF3
N
H
11§
57
H
N
H
O
N
O
N
Br
N
H
Br
12
70
N
H
O
N
O
N
5bj
Cl
Br
6
78
99:1
97
13
67
98:2
91
N
H
N
H
O
O
N
N
5bf
Bn
5bk
Br
Br
Br
N
H
7
68
99:1
90
14
80
98:2
98
N
O
Br
H
O
N
5bg
N
5bl
Reaction conditions: 1/2/7b/Rh2(OAc)4 ¼ 0.2/0.1/0.01/0.002 mmol, 1 in 1 ml CH2Cl2 was added to a solution of 2, 7b and Rh2(OAc)4 in 1 ml CH2Cl2 at 220 8C for 1 h under an argon atmosphere and stirred at
220 8C for another 1 h. *Isolated yield. †Determined either by 1H-NMR spectroscopy or HPLC analysis of the crude reaction mixture. ‡Determined by chiral HPLC analysis. §THF was used as solvent and the
reaction temperature was 25 8C.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.
3

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1406
ARTICLES
Table 3 | Enantioselective reaction of a-aryl-a-diazoacetates and indoles with various imines.
R1
1
NH2Ar
Ar2
MeO2C
Rh2(OAc)4 (1 mol%)
N2
(S)-7b (2 mol%)
R2
Ar2
R2
N
Ar1
+
+
R1
CO2Me
N
R3
N
4Å MS
R3
toluene, –10 °C
4
3
2
6
Entry Product
Yield
(%)*
d.r.
e.e.
Entry Product
Yield
(%)*
d.r.
e.e.
(anti:syn )†
.20:1
(%)‡
(anti:syn )†
(%)‡
MeO
HN
HN
MeO2C
MeO2C
1
93
99
9
76
.20:1
.20:1
94
96
N
N
6b
6j
Br
S
HN
HN
MeO2C
MeO2C
2
76
.20:1
97
10
95
N
N
6k
CF3
6c
Ph
HN
HN
MeO2C
MeO2C
3
4
5
6
85
58
75
98
.20:1
.20:1
.20:1
.20:1
67
93
99
94
11
12
13
14
50
72
75
85
.20:1
.20:1
.20:1
.20:1
97
98
95
97
N
N
6l
6d
HN
HN
MeO2C
MeO2C
MeO
Br
N
N
6e
6m
6n
6o
HN
HN
MeO2C
MeO2C
Br
N
N
Br
6f
HN
Cl
HN
MeO2C
MeO2C
N
N
6g
OMe
HN
MeO2C
HN
MeO2C
7
85
.20:1
95
15
80
.20:1
96
N
6p
Bn
N
6h
Br
HN
MeO2C
HN
MeO2C
8
98
.20:1
84
16
75
.20:1
86
N
Ph
N
6q
6i
Reaction conditions: 2/3/4/7b/Rh2(OAc)4 ¼ 0.3/0.3/0.25/0.005/0.0025 mmol, 3 in 2 ml toluene was added to a solution of 4Å MS (0.1 g), 2, 4, 7b and Rh2(OAc)4 in 2 ml toluene under an argon
atmosphere at 210 8C for 2 h and stirred at 210 8C for 6 h. *Isolated yield. †Determined by 1H-NMR spectroscopy of the crude reaction mixture. ‡Determined by chiral HPLC analysis.
4
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1406
ARTICLES
(2b) (Supplementary Table S1). PPA 7b was also found to be the
optimal co-catalyst. The best result was obtained when the reaction
was conducted at 210 8C in toluene with 4Å MS as an additive,
resulting in 6a in 94% yield, .20:1 d.r. and 97% e.e. Through a sep-
arate experiment, we found that the addition of 4Å MS was crucial
to enantioselective control and that the e.e. decreased dramatically
without 4Å MS (Supplementary Table S1).
C13
C14
C22
C15
C8
C12
C11
Cl
C23
C2
C10
C9
C1
N
H
C3
O1
C21
C18
N1
C16
O
C4
C5
C20
C19
C7
N
C6
C24
N2
C17
A broad range of aryl imines, aryldiazoacetates and indoles were
then investigated under standard reaction conditions and the results
summarized in Table 3. Electron-donating imines (Table 3, entry 3)
afforded relatively poor enantioselectivity compared with electron-
withdrawing and neutral imines (Table 3, entries 1,2,4,6). Diazo
compounds with substitutions at the m- and p-positions were all tol-
erated (Table 3, entries 7–9), and a heterocyclic thiophene-derived
diazo compound was also employed in the reaction to give 5k in
high yield with .20:1 d.r. and 96% e.e. (Table 3, entry 10). It is
noteworthy that methyl trans-styryldiazoacetate, which has been
reported to induce [3 þ 2] annulation with indole4, can be
adapted to our dual catalysis system, resulting in the three-
component product being isolated at moderate yield with 97% e.e.
(Table 3, entry 11). The carbon–carbon double bond leaves poten-
tial functionalities for further chemical transformations. Also, the
substitution of indoles has little impact on the reaction. Various sub-
stitutions at the nitrogen atom and the aromatic ring all afforded
good results (Table 3, entries 12–16).
C11
(3S, 1'R)-5ba
Brl
C21
C9
C22
HN
O2
C31
C8
C23
C20
C19
C18
C17
C24
MeO2C
N1
C1
C30
C2
C7
C1
C16
C6
C5
N
C3
C4
C25
N2
Br
C15 C14
C26
C13
C12
C10
C29
(2R, 3R)-6b
C11
C27
C28
Figure 2 | Determination of absolute configurations of 5ba and 6b by
single-crystal X-ray analysis.
The evaluation of various PPAs (Table 1, entries 2–10) revealed
that there was no conspicuous electronic effect and that PPA 7b
bearing bulky triphenyl silyl substitutions was optimal, affording
the product with 96:4 d.r. and 81% e.e. (Table 1, entry 3). To
improve the enantioselectivity of the reaction, further optimization
of the reaction conditions was carried out. THF was identified as the
best solvent for enantioselectivity but the yield was not satisfactory
(Table 1, entry 13). A decrease in temperature improved the e.e.
value (Table 1, entries 14,15), and optimal conditions were obtained
when the reaction was carried out at –20 8C (Table 1, entry 16).
Replacing the a-methyl diazoacetamide 1a with a-ethyl diazoaceta-
mide 1b significantly increased the enantioselectivity to 96%
(Table 1, entry 17), and slightly decreased values of d.r. (90:10)
and e.e. (83%) were observed when using less sterically hindered
diazoacetamide 1c as a substrate (Table 1, entry 18).
Absolute configurations were determined by X-ray crystallogra-
phy of 5ba and 6b (Fig. 2). Products 5 favoured the syn-isomer
and 6 the anti-isomer.
a
TS I
R
O
O
O
P
O
O
O
O
H
P
C2
OH
R
R
H
N
N
R
Ph
Ph
ORhLn
(S)-7b, R = SiPh3
H
Having identified the optimized reaction conditions, the reaction
was extended to include various imines with co-catalyst. (S)-7b
In most cases, the reaction proceeded smoothly to give the
desired product 5 in good to high yield (up to 84%), excellent d.r.
(up to 99:1), and excellent enantioselectivity (up to 98% e.e.).
Aromatic imines with m- or p-substitutions were tolerated
(Table 2, entries 1,3–8). In general, electron-withdrawing imine
substrates (Table 2, entry 3) resulted in slightly lower enantio-
selectivity of the corresponding product, and greatly decreased
selectivity was observed when electron-donating groups were
used (Table 2, entry 8). It is worth noting that the imine derived
from cinnamaldehyde can also be used in the reaction, and
product 5bi was obtained with moderate yield with high d.r. and
moderate e.e. (Table 2, entry 9). Substitutions on diazo compounds
had little impact on the reaction, and the a-ethyl diazo compounds
resulted in slightly higher selectivity compared with less sterically
hindered a-methyl and the a-H compounds (Table 2, entries
10,11). Different substitutions at the nitrogen atom and on the aro-
matic ring in diazo compounds all yielded good results (Table 2,
entries 12–14).
b
D
N2
D (100%)
D
D (26%)
O
Rh2(OAc)4
10 mol% (S)-7b
D
N
*
N
CH2Cl2, 25oC
O
D
D
D
D
D
Quantitative yield
30% e.e.
d5-1a
c
TS II
O
O
O
P
O
R
O
O
O
R
R
H
P
C2
RhLn
O
OH
OMe
N
Ph
N
H
Ph
R
H
We next applied a similar strategy that used a PPA-activated
imine to trap zwitterionic intermediates that form intermolecularly
through the reaction of a diazo compound and an indole group
(Fig. 1c). Catalyst screening and optimization of conditions were
again carried out for the three-component reactions of methyl phe-
nyldiazoacetate (3a), N-methyl indole (4a) and phenylbenzylimine
(S)-7b, R = SiPh3
Figure 3 | Rationalization for the stereoselective control of the reactions.
a, Proposed transition state for the two-component reaction. b, Controlled
deuterium labelling experiment to support the ‘indirect proton transfer’
process. c, Proposed transition state for the three-component reaction.
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
5
© 2012 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY DOI: 10.1038/NCHEM.1406
ARTICLES
14. Cai, Y., Zhu, S.-F., Wang, G.-P. & Zhou, Q.-L. Iron-catalyzed C–H
fuctionalization of indoles. Adv. Synth. Catal. 353, 2939–2944 (2011).
15. Hansen, J. & Davies, H. M. L. High symmetry dirhodium(II) paddlewheel
complexes as chiral catalysts. Coord. Chem. Rev. 252, 545–555 (2008).
16. Doyle, M. P. & Forbes, D. C. Recent advances in asymmetric catalytic metal
carbene transformations. Chem. Rev. 98, 911–935 (1998).
17. Li, Z.-J. & Davies, H. M. L. Enantioselective C–C bond formation by rhodium-
catalyzed tandem ylide formation/[2,3]-sigmatropic rearrangement between
donor/acceptor carbenoids and allylic alcohols. J. Am. Chem. Soc. 132,
396–401 (2010).
18. Ji, J.-J. et al. Diastereoselectivity switch in cooperatively catalyzed three-
component reactions of an aryldiazoacetate, an alcohol, and a b,g-unsaturated
a-keto ester. J. Org. Chem. 76, 5821–5824 (2011).
19. Shao, Z.-H. & Zhang, H.-B. Combining transition metal catalysis and
organocatalysis: a broad new concept for catalysis. Chem. Soc. Rev. 38,
2745–2755 (2009).
20. Hu, W.-H. et al. Cooperative catalysis with chiral Brønsted acid–Rh2(OAc)4:
highly enantioselective three-component reactions of diazo compounds with
alcohols and imines. J. Am. Chem. Soc. 130, 7782–7783 (2008).
21. Jiang, J. et al. Diastereoselectively switchable enantioselective trapping of carbamate
ammonium ylides with imines. J. Am. Chem. Soc. 133, 8428–8431 (2011).
22. Williams, A. L. & Johnston, J. N. The Brønsted acid-catalyzed direct
Aza-Darzens synthesis of N-alkyl cis-aziridines. J. Am. Chem. Soc. 126,
1612–1613 (2004).
23. Johnston, J. N., Muchalski, H. & Troyer T. L. To protonate or alkylate?
Stereoselective Brønsted acid catalysis of C–C bond formation using
diazoalkanes. Angew. Chem. Int. Ed. 49, 2290–2298 (2010).
24. Zhong, C. & Shi, X.-D. When organocatalysis meets transition-metal catalysis.
Eur. J. Org. Chem. 2999–3025 (2010).
To rationalize the observed stereochemistry, an interaction
model proposed by Simon and Goodman30 can be used to explain
the stereoselective outcome of current imine additions catalysed
by BINOL-PPA catalysts (Fig. 3a). For example, in the (S)-7b cata-
lysed reaction of 1b and 2b, transition state I (TSI) is proposed to
afford 5bb with the observed stereochemistry. The direction of the
N-phenyl group on the energetically favoured E-imine is towards
the empty side of the catalyst pocket for the initial interaction
between the imine substrate and the catalyst. The imine is protonated
by the PPA catalyst with the imine maintaining its E configuration31.
We propose that the key interaction that defines the stereochemical
outcome of this reaction is a weak hydrogen bond between the
Lewis basic phosphoryl oxygen atom32 with the acidic C–H proton
in the zwitterionic intermediate, whereby proton transfer occurs
through the PPA in forming the product. Like the similar case
reported by Zhou and colleagues33, the PPA catalyst acts as a
proton-transfer shuttle in the reaction. As a result of these inter-
actions, chiral induction is achieved when the nucleophilic carbon
attacks the imine from the silicon face to give (3S, 1′R)-5bb.
To obtain more insight into the proposed proton transfer
pathway of the zwitterionic intermediate, we carried out a deuter-
ium isotope experiment comprising a PPA-catalysed C–D insertion
of d5-1a (Fig. 3b). A dramatic decrease in the deuterium content of
the product was observed, indicating an ‘indirect proton transfer’34
from the corresponding zwitterionic intermediate. Using the same
model, the observed stereochemistry for the three-component reac-
tion catalysed by (S)-7 can be rationalized by a similar transition
state (TSII) to give (2R,3R)-6 (Fig. 3c).
25. Uraguchi, D., Sorimachi, K. & Terada, M. Organocatalytic asymmetric aza-
Friedel–Crafts alkylation of furan. J. Am. Chem. Soc. 126, 11804–11805 (2004).
26. Akiyama, T., Itoh, J., Yokota, K. & Fuchibe, K. Enantioselective Mannich-type
reaction catalyzed by a chiral Brøsted acid. Angew. Chem. Int. Ed. 43,
1566–1568 (2004).
In summary, we have developed a highly efficient integrated
rhodium and chiral Brønsted acid co-catalysed process for the trapping
of zwitterionic intermediates by imines. Polyfunctionalized indole and
oxindole derivatives were obtained in a single-step transformation with
excellent diastereoselectivity and enantioselectivity. Further exploration
using other electrophiles to trap these intermediates, as well as inves-
tigations into other types of zwitterionic intermediate trapping pro-
cesses, are currently under way in our laboratories.
27. Tian, X., Jiang, K., Peng, J., Du, W. & Chen Y.-C. Organocatalytic stereoselective
Mannich reaction of 3-substituted oxindoles. Org. Lett. 10, 3583–3586 (2008).
28. Galliford, C. V. & Scheidt, K. A. Pyrrolidinyl-spirooxindole natural products as
inspirations for the development of potential therapeutic agents. Angew. Chem.
Int. Ed. 46, 8748–8758 (2007).
29. Zhang, D., Song, H. & Qin, Y. Total synthesis of indoline alkaloids: a
cyclopropanation strategy. Acc. Chem. Res. 44, 447–457 (2011).
30. Simon, L. & Goodman, J. M. A model for the enantioselectivity of imine reactions
catalyzed by BINOL-phosphoric acid catalysts. J. Org. Chem. 76, 1775–1788 (2011).
31. Liu, H., Dagousset, G., Masson, G., Retailleau, P. & Zhu, J.-P. Chiral Brønsted
acid-catalyzed enantioselective three-component Povarov reaction. J. Am. Chem.
Soc. 131, 4598–4599 (2009).
32. Jia, Y.-X., Zhong, J., Zhu, S.-F., Zhang, C.-M. & Zhou, Q.-L. Chiral Brøsted
acid catalyzed enantioselective Friedel–Crafts reaction of indoles and a-aryl
enamides: construction of quaternary carbon atoms. Angew. Chem. Int. Ed.
46, 5565–5567 (2007).
33. Xu, B., Zhu S.-F., Xie X.-L., Shen J.-J. & Zhou Q.-L. Asymmetric N–H insertion
reaction cooperatively catalyzed by rhodium and chiral spiro phosphoric acids.
Angew. Chem. Int. Ed. 50, 11483–11486 (2011).
Received 21 March 2012; accepted 12 June 2012;
published online 29 July 2012; corrected online 14 August 2012
References
1. Dorwald, F. Z. (eds) Metal Carbenes in Organic Synthesis (Wiley-VCH, 2007).
2. Zhu, S.-F., Cai, Y., Mao, H.-X., Xie, J.-H. & Zhou, Q.-L. Enantioselective iron-
catalysed O–H bond insertions. Nature Chem. 2, 546–551 (2010).
3. Zhu, S.-F., Xu, X., Perman, J. A. & Zhang, X. P. A general and efficient cobalt(II)-
based catalytic system for highly stereoselective cyclopropanation of alkenes with
a-cyanodiazoacetates. J. Am. Chem. Soc. 132, 12796–12799 (2010).
4. Lian, Y.-J. & Davies, H. M. L. Rhodium-catalyzed [3þ2] annulation of indoles.
J. Am. Chem. Soc. 132, 440–441 (2010).
34. Liang Y., Zhou H.-L. & Yu Z.-X. Why is copper (I) complex more competent
than dirhodium (II) complex in catalytic asymmetric O–H insertion reactions?
A computational study of the metal carbenoid O–H insertion into water. J. Am.
Chem. Soc. 131, 17783–17785 (2009).
5. Padwa, A. & Hornbuckle, S. F. Ylide formation from the reaction of carbenes and
carbenoids with heteroatom lone pairs. Chem. Rev. 91, 263–309 (1991).
6. Godula, K. & Sames, D. C–H bond functionalization in complex organic
synthesis. Science 312, 67–72 (2006).
7. Wang, D.-H., Engle, K. M., Shi, B.-F. & Yu, J.-Q. Ligand-enabled reactivity
and selectivity in a synthetically versatile aryl C–H olefination. Science 327,
315–319 (2010).
Acknowledgements
The authors acknowledge support from the National Science Foundation of China (nos
20932003 and 21125209 to W.-H.H.), the MOST of China (2011CB808600) and from
Shanghai (10XD1401700). M.P.D. thanks the National Institutes of Health (GM 46503)
and National Science Foundation (CHE-0748121) for financial support.
8. Davies, H. M. L. & Manning, J. R. Catalytic C–H functionalization by metal
carbenoid and nitrenoid insertion. Nature 451, 417–424 (2008).
9. Nakamura, E., Yoshikai, N. & Yamanaka, M. Mechanism of C–H bond
activation/C–C bond formation reaction between diazo compound and alkane
catalyzed by dirhodium tetracarboxylate. J. Am. Chem. Soc. 124, 7181–7192 (2002).
10. Doyle, M. P., Duffy, R., Ratnikov, M. & Zhou, L. Catalytic carbene insertion into
C–H bonds. Chem. Rev. 110, 704–724 (2010).
11. Doyle, M. P., Shanklin, M. S., Pho, H. Q. & Mahapatro, S. N. Rhodium(II) acetate
and Nafion-H catalyzed decomposition of N-aryldiazoamides. Efficient
synthesis of 2(3H)-indolinones. J. Org. Chem. 53, 1017–1022 (1988).
12. Davies, H. M. L. & Hedley, S. J. Intermolecular reactions of electron-rich heterocycles
with copper and rhodium carbenoids. Chem. Soc. Rev. 36, 1109–1119 (2007).
13. DeAngelis, A., Shurtleff, V. W., Dmitrenko, O. & Fox, J. M. Rhodium(II)-
catalyzed enantioselective C–H functionalization of indoles. J. Am. Chem. Soc.
133, 1650–1653 (2011).
Author contributions
H.Q. and M.L. contributed equally to the work. H.Q. designed and performed experiments
for the three-component reaction, and M.L. designed and performed experiments for
oxindole formation. L.-Q.J. took part in the development of the initial reaction. F.-P.L.
made the organocatalysts. L.Z. and C.-W.Z. performed experiments. M.P.D. discussed,
commented on and revised the manuscript. W.-H.H. conceived and directed the project.
W.-H.H., M.L. and H.Q. wrote the paper.
Additional information
The authors declare no competing financial interests. Supplementary information
and chemical compound information accompany this paper at www.nature.com/
naturechemistry. Reprints and permission information is available online at http://www.
nature.com/reprints. Correspondence and requests for materials should be addressed
to W.-H.H.
6
NATURE CHEMISTRY | ADVANCE ONLINE PUBLICATION | www.nature.com/naturechemistry
© 2012 Macmillan Publishers Limited. All rights reserved.

ERRATUM
Highly enantioselective trapping of zwitterionic intermediates by imines
Huang Qiu, Ming Li, Li-Qin Jiang, Feng-Ping Lv, Li Zan, Chang-Wei Zhai, Michael P. Doyle and Wen-Hao Hu
Nature Chemistry 4, 733–738 (2012); published online 29 July 2012; corrected online 14 August 2012.
In the version of this Article previously published online, in Table 3, Entry 1, the structure of compound 6b was missing a Br atom label.
is has now been corrected in all versions of this Article.
© 2012 Macmillan Publishers Limited. All rights reserved.