Cation-Controlled Enantioselective and Diastereoselective Synthesis of Indolines: An Autoinductive Phase-Transfer Initiated 5-endo-trig Process

Article abstract of DOI:10.1021/jacs.5b08834

A catalytic enantioselective approach to the synthesis of indolines bearing two asymmetric centers, one of which is all-carbon and quaternary, is described. This reaction proceeds with high levels of diastereoselectivity (>20:1) and high levels of enantioselectivity (up to 99.5:0.5 er) in the presence of CsOH·H₂O and a quinine-derived ammonium salt. The reaction most likely proceeds via a delocalized 2-aza-pentadienyl anion that cyclizes either by a suprafacial electrocyclic mechanism, or through a kinetically controlled 5-endo-trig Mannich process. Density functional theory calculations are used to probe these two mechanistic pathways and lead to the conclusion that a nonpericyclic mechanism is most probable. The base-catalyzed interconversion of diastereoisomeric indolines in the presence of certain quaternary ammonium catalysts is observed; this may be rationalized as a cycloreversion-cyclization process. Mechanistic investigations have demonstrated that the reaction is initiated via a Makosza-like interfacial process, and kinetic analysis has shown that the reaction possesses a significant induction period consistent with autoinduction. A zwitterionic quinine-derived entity generated by deprotonation of an ammonium salt with the anionic reaction product is identified as a key catalytic species and the role that protonation plays in the enantioselective process outlined. We also propose that the reaction subsequently occurs entirely within the organic phase. Consequently, the reaction may be better described as a phase-transfer-initiated rather than a phase-transfer-catalyzed process; this observation may have implications for mechanistic pathways followed by other phase-transfer-mediated reactions.

Full text of DOI:10.1021/jacs.5b08834

Subscriber access provided by EPFL | Scientific Information and Libraries
Article
Cation-controlled enantio- and diastereoselective synthesis of
indolines: an autoinductive phase–transfer initiated 5-endo-trig process
Krishna Sharma, Jamie R. Wolstenhulme, Phillip P. Painter, David Yeo, Francisca
Grande-Carmona, Craig P. Johnston, Dean J Tantillo, and Martin D. Smith
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.5b08834 • Publication Date (Web): 23 Sep 2015
Downloaded from http://pubs.acs.org on September 24, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Cation-controlled enantio- and diastereoselective
synthesis of indolines: an autoinductive phase–transfer initiated
5-endo-trig process
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Krishna Sharma,1 Jamie R. Wolstenhulme,1 Phillip P. Painter,2 David Yeo,1 Francisca Grande-Carmona,1
Craig P. Johnston,1 Dean J. Tantillo2* and Martin D. Smith1*
1. Chemistry Research Laboratory, University of Oxford, 12 Mansfield Road, Oxford OX1 3TA, UK.
2. Department of Chemistry, University of California, Davis, California 95616, USA.
ABSTRACT: A catalytic enantioselective approach to the synthesis of indolines bearing two asymmetric centers, one of which is all-carbon
and quaternary, is described. This reaction proceeds with high levels of diastereoselectivity (>20:1) and high levels of enantioselectivity (up
to 99.5:0.5 e.r.) in the presence of CsOH·H2O and a quinine-derived ammonium salt. The reaction most likely proceeds via a delocalized 2-
aza-pentadienyl anion that cyclizes either by a suprafacial electrocyclic mechanism, or through a kinetically controlled 5-endo-trig Mannich
process. Density functional theory calculations are used to probe these two mechanistic pathways and lead to the conclusion that a non-
pericyclic mechanism is most probable. The base-catalyzed interconversion of diastereoisomeric indolines in the presence of certain quater-
nary ammonium catalysts is observed; this may be rationalized as a cycloreversion-cyclization process. Mechanistic investigations have
demonstrated that the reaction is initiated via a Mąkosza-like interfacial process, and kinetic analysis has shown that the reaction possesses a
significant induction period consistent with autoinduction. A zwitterionic quinine-derived entity generated by deprotonation of an ammoni-
um salt with the anionic reaction product is identified as a key catalytic species and the role that protonation plays in the enantioselective pro-
cess outlined. We also propose that the reaction subsequently occurs entirely within the organic phase. Consequently, the reaction may be
better described as a phase–transfer initiated rather than a phase–transfer catalyzed process; this observation may have implications for mech-
anistic pathways followed by other phase-transfer mediated reactions.
a general process that would permit the synthesis of stereochemi-
INTRODUCTION
cally more complex materials.6,7 Herein we describe an extension of
Asymmetric phase-transfer catalysis is a powerful technique that
this approach to generate indolines bearing two asymmetric cen-
enables a wide range of transformations under mild conditions,
ters, one of which is quaternary and all-carbon, and through density
often using inexpensive and environmentally benign reagents.1
functional theory (DFT) calculations, probe whether or not the
Since its invention, a wide range of asymmetric carbon–carbon
reaction is pericyclic.
bond forming reactions have been disclosed.2 We have focused on
extending the application of phase-transfer catalysis to new reaction
RESULTS AND DISCUSSION
A suitable substrate to probe the potential for diastereoselective
manifolds and cascade processes.3 With this in mind, we conceived
an approach to the asymmetric catalysis of electrocyclic reactions
cyclization was synthesized in three simple and scalable steps,
based upon using a chiral counterion to achieve π-face selectivity in
(Scheme 1).
anionic cyclization manifolds.4 This ideal led to the development
of a method to generate indolines with a single asymmetric center5
but also posed a series of questions – notably whether the reaction
was likely to be pericyclic, and also whether the transformation was
Scheme 1
Ph
Ph
F
(i)
i
(ii), (iii)
99%
i
CO2Pr
NO2
CO2Pr
45%
NO2
N
Ph
Catalytic enantioselective cation-directed 5-endo-trig cyclization
1
2
3
R3
R3
+
i
R4
N
(i) PhCH2CO2 Pr, NaH, DMF, 0˚C to RT (ii) H2, Pd/C, MeOH
(iii) MgSO4, PhCHO, toluene, rt.
CN
R2
CN
R1
R1
base
5-endo-trig
N
H
R2
N
up to 99.5:0.5 e.r.; > 20:1 d.r.
Nucleophilic aromatic substitution of o-fluoronitrobenzene 1 with
isopropyl 2-phenylacetate in the presence of NaH in DMF afforded
2 in 45% yield. Quantitative reduction with hydrogen and palladi-
um on carbon afforded an aniline intermediate, which was treated
with benzaldehyde in the presence of magnesium sulfate to afford
imine 3 in 99% isolated yield. With this intermediate in hand, we
probed its reactivity and selectivity under basic conditions (Table
CN
N
N
Ar
CN
Ar
Electrocyclization
5-endo-trig
Figure 1: Outline of current study
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 11
1). The system was unreactive with potassium carbonate or cesium 30˚C (>99:1 e.r. and >20:1 d.r.) at the expense of conversion
1
2
3
4
5
6
7
8
carbonate in toluene in the presence of tetrabutylammonium chlo-
ride (TBAC). In contrast, ester 3 cyclized smoothly in 24h with
potassium tert-butoxide in THF, yielding indolines 4 and 5 in a
ratio of 2.5:1 respectively (entry 3). We lowered the reaction tem-
perature to 0˚C and after 1h observed complete consumption of 3
to afford 4 (>20:1 d.r., 4:5, entry 4).8
(10%). Para- and ortho-substituted imines were well tolerated and
generally cyclized with excellent enantioselectivity (up to 99:1 e.r.
for 12) and with good diastereoselectivity. Substrates bearing me-
ta-substituents on the imine arene underwent cyclization but were
not as enantioselective (10, 71:29 e.r. and 11, 88:12 e.r.).
Table 2. Diastereo- and enantioselective synthesis of indolines
Table 1. Exploration of base-catalyzed synthesis of indolines
H
Ph
+
R4N
9
Ph
Ph
catalyst 6
Ph
Ph
i
i
CO2Pr
CO2Pr
H
N
Cl-
CO2iPr
Ph
CO2iPr
Ph
CO2iPr
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CsOH•H2O
toluene
24 h, -15 ˚C
base
solvent
24 h, rt
Ph
R2
R2
N
N
OH
N
N
H
4
N
H
5
H
N
6
3
Conversion d.r.
e.r.
Ph
Ph
Ph
Entry Solvent
Base
Catalyst
i
[%]
(4:5) (major)
CO2Pr
i
i
CO2Pr
CO2Pr
1
2
3
4*
5
6*
7
toluene 33% K2CO3 (aq.) Bu4N+Cl-
none
none
>95
>95
>95
>95
>95
none
60
-
-
-
-
-
-
-
-
-
-
-
N
N
N
H
H
H
toluene
THF
THF
THF
Cs2CO3
tBuOK
tBuOK
tBuOK
KOH
Bu4N+Cl-
none
Cl
Br
4. >20:1 d.r.
97:3 e.r. (anti)
74% yield
8. 20:1 d.r.
97:3 e.r. (anti)
70% yield
2.5:1
>20:1
<1:20
2:1
<1:20
-
<1:20
6:1
>20:1 96:4
>20:1 97:3
7. 4.2:1 d.r.
98:2 e.r. (anti)
75% yield
none
Bu4N+Cl-
none
Ph
Ph
Ph
i
i
i
CO2Pr
CO2Pr
CO2Pr
THF
N
N
N
THF
toluene
toluene
KOH
KOH
KOH
KOH
CsOH·H2O
CsOH·H2O
Bu4N+Cl-
none
Bu4N+Cl-
N-BnCD+Cl-
N-BnCD+Cl-
N-BnCD+Cl-
H
H
H
8
9
Cl
NO2
Cl
9. 17:1 d.r.
97:3 e.r. (anti)
90% yield
10. 6.9:1 d.r.
71:29 e.r. (anti)
72% yield
11. 3.4:1 d.r.
88:12 e.r. (anti)
81% yield
10* toluene
11* toluene
12§ toluene
>95
>95
>95
96:4
Ph
Ph
Ph
i
i
i
CO2Pr
CO2Pr
CO2Pr
N
N
N
H
H
H
Conditions: 10 mol% catalyst, 2 eq. base. d.r. established by exami-
Br
NO2
Me
1
nation of crude H NMR spectra; e.r. established by chiral station-
12. 10:1 d.r.
99:1 e.r. (anti)
84% yield
13. 7.5:1 d.r.
97:3 e.r. (anti)
60% yield
14. 4.1:1 d.r.
97:3 e.r. (anti)
85% yield
ary phase HPLC. *reaction temperature was 0˚C. §reaction tem-
perature was –15˚C. N-BnCD+Cl- = N-benzylcinchonidinium
chloride 6.
Ph
Ph
Ph
i
i
CO2Pr
O
CO2Pr
N
N
N
H
This result was in contrast to the outcome of a similar reaction in
the presence of TBAC, which gave complete conversion but with a
complete reversal of selectivity, generating indolines 4 and 5 in a
ratio of <1:20, respectively (entry 5). This observation indicated
the possibility that under certain reaction conditions interconver-
sion of the two isomers was occurring. Broadly similar results were
observed with solid KOH9 in THF (entries 6 and 7), and this
demonstrated the potential for a highly diastereoselective cycliza-
tion to afford indolines bearing two stereocenters. However, we
rationalized that we were unlikely to attain an enantioselective
transformation unless a switch to a less polar solvent could be
achieved. Gratifyingly, solid KOH in toluene in the presence of
TBAC led to 60% conversion and a <1:20 ratio of indolines (4:5,
entry 9). Consequently, we focused on the use of this solvent and
base combination to examine the potential for an asymmetric trans-
formation. Using these conditions in the presence of N-
benzylcinchonidinium chloride (entry 10) gave complete conver-
sion, a 6:1 ratio of diastereoisomers and high levels of enantioselec-
tivity (96:4 e.r., major diastereoisomer). Reducing the temperature
to –15˚C and a switch to solid CsOH·H2O gave 97:3 e.r. and >
20:1 d.r. (entry 12).
H
H
F
Cl
15. 12:1 d.r.
90:10 e.r. (anti)
64% yield
16. 3.0:1 d.r.
98:2 e.r. (anti)
89% yield
17
Conditions: 50 mg substrate, 2.5 mL toluene, 2 eq. CsOH·H2O, 10
mol% catalyst; reaction time: 24 h at –15˚C. d.r. established by
examination of crude 1H NMR spectra; e.r. established by chiral
stationary phase HPLC. Yields are for isolated material.
BENZONITRILE SUBSTRATES
We observed that some ester-containing substrates such as 9 un-
derwent uncontrollable oxidation in the reaction mixture or on
standing to generate indoles such as 17 (Table 2). Whilst this un-
desired reactivity could possibly be controlled through further
functionalization, we realized that this significantly compromised
the potential utility of these building blocks, and decided that a
change of substrate might offer an alternative stability profile. Con-
sequently, we decided to investigate an alternative electron-
withdrawing group, and focused on changing the isopropyl ester for
a nitrile functional group. These substrates can be made by a strate-
gically similar process outlined previously for the ester substrates
(Scheme 2). An SNAr reaction between a series of substituted fluo-
ronitrobenzenes and phenylacetonitrile was mediated by aqueous
NaOH in the presence of stoichiometric tetrabutylammonium
hydrogen sulfate.10 Reduction of the nitro group to the aniline gen-
With these optimized conditions in-hand, we examined the reac-
tion of a range of substrates, which cyclized smoothly with good to
excellent diastereo- and enantioselectivity. The simple phenyl sub-
strate cyclized at –15˚C to give indoline 4 in >20:1 d.r. and 97:3
e.r.; selectivity could be augmented by running the reaction at –
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
erally occurred smoothly with zinc and ammonium chloride, and bears an N-anthracenyl group, did not lead to a significant im-
1
2
3
4
5
6
7
8
condensation with an aldehyde afforded imines 20; these were
often formed and used without formal purification.
provement in enantio- or diastereoselectivity (1:1 d.r.; 56:44 e.r.,
entry 6). We also examined a series of quininium derivatives; cata-
lyst 26 gave an improved e.r. of 89:11 and d.r. of 14:1. Varying the
nature of the N-substituent on the catalyst led to small changes in
selectivity (entries 8-10) but no improvements. Pseudoenantio-
meric catalysts 30 and 31 led to the predicted reversal of enanti-
oselectivity but no overall improvement and these were not pur-
sued further (entries 11 and 12). We focused on 26 as the opti-
mum catalyst for the transformation, and lowering the temperature
of the reaction to –50˚C and reducing catalyst loading to 5 mol%
improved both diastereo- and enantioselectivity (>20:1 d.r.; 95:5
e.r., entry 13). With these optimal conditions in hand, we probed
the generality of this transformation (Table 4).
Scheme 2
Ph
Ph
F
(i)
(ii), (iii)
CN
CN
R1
R1
R1
R2
NO2
NO2
19
N
18
20
9
-
(i) PhCH2CN, NaOH, Bu4N+HSO4 , toluene, RT (ii) Zn, NH4Cl,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
acetone/water, RT (iii) MgSO4, R2CHO, toluene, RT
Under conditions that were successful for the previous ester sys-
tem, the simple substrate 21 underwent smooth cyclization to
yield a diastereoisomeric mixture of indolines in moderate e.r.
(72:28) and d.r (3.6:1, 22:23). Lowering the temperature to –
50˚C led to an increase in d.r. to >20:1 but e.r. and conversion were
still disappointingly low (Table 3, entry 2). Consequently, we in-
vestigated a range of different base and catalyst combinations,
which demonstrated that CsOH·H2O, which facilitated complete
conversion at a lower temperature than KOH, was optimum.
Table 4. Diastereo- and enantioselective synthesis of indolines
(i)
R2
H
O
OMe
N
Ph
MgSO4, toluene
Ph
H
N
Cl-
CN
R2
R1
CN
R1
Ph
(ii) catalyst 26
CsOH•H2O
toluene, 48 h, –50 ˚C
N
H
OH
NH2
26
Table 3. Optimization of benzonitrile substrates
Ph
Ph
Ph
+
R4N
Ph
X-
CN
CN
CN
Ph
Ph
CN
Ph
CN
Ph
CN
Ph
N
N
N
H
base
solvent
16 h
H
H
N
N
H
22
N
H
23
Me
22. >20:1 d.r.
94:6 e.r. (anti)
75% yield
32. >20:1 d.r.
90:10 e.r. (anti)
84% yield
33. >20:1 d.r.
95:5 e.r. (anti)
86% yield
21
Temp.
(˚C)
d.r.
e.r.
Ph
Ph
Ph
Entry
Base
Catalyst
(22:23) (major)
CN
CN
CN
O
O
N
H
1*
2*
3
4
5
6
7
8
9
10
11
12
13§
KOH
KOH
6
6
6
6
24
25
26
27
28
29
30
31
26
0
3.6:1
>20:1
4:1
4:1
3:1
72:28
78:22
74:26
81:19
50:50
56:44
89:11
89:11
85:15
90:10
80:20
75:25
95:5
N
H
F3C
N
H
–50
–50
–30
–30
–30
–30
–30
–30
–30
–30
–30
–50
OMe
34. >20:1 d.r.
97:3 e.r. (anti)
81% yield
35. >20:1 d.r.
99:1 e.r. (anti)
91% yield
36. >20:1 d.r.
95:5 e.r. (anti)
83% yield
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
CsOH·H2O
Ph
Ph
CN
CN
1:1
N
H
F3C
N
H
14:1
14:1
10:1
16:1
4:1
37. >20:1 d.r.
97:3 e.r. (anti)
34% yield
38. >20:1 d.r.
99.5:0.5 e.r. (anti)
37
78% yield
Ph
Ph
Ph
CN
CN
CN
5:1
>20:1
Me
Me
N
H
N
H
Ph
N
H
Me
H
X-
39. > 20:1 d.r.
84:16 e.r. (anti)
30% yield
40. > 20:1 d.r.
94:6 e.r. (anti)
50% yield
41. > 20:1 d.r.
85:15 e.r. (anti)
70% yield
6. R1=H, R2=H, R3=Ph, X=Cl
H
OH
X- 24. R1=H, R2=allyl, R3=Ph, X=Br
R1
25. R1=H, R2=H, R3=anthracenyl, X=Cl
26. R1=OMe, R2=H, R3=Ph, X=Cl
27. R1=OMe, R2=H, R3=p-BrC6H4, X=Br
28. R1=OMe, R2=H, R3=p-(OMe)C6H4, X=Br
29. R1=OiPr, R2=H, R3=Ph, X=Br
N
H
N
Ph
Ph
R3
H
Ph
N
R3
OR2
CN
Me
CN
CN
30. R3 = Ph, X=Cl
31. R3 = p-(CF3)C6H4, X=Cl
N
N
H
O
N
H
N
H
Br
NO2
42. >20:1 d.r.
78:22 e.r. (anti)
76% yield
43. >20:1 d.r.
78:22 e.r. (anti)
64% yield
44. >20:1 d.r.
99:1 e.r. (anti)
89% yield
Conditions: 0.1 mmol imine, 1.3 mL toluene, 0.2 mmol base, 10
mol% catalyst; reaction time: 2 – 48 h. d.r. established by examina-
tion of the crude 1H NMR spectra; e.r. established by chiral station-
ary phase HPLC. *1 mmol of base used. §5 mol% catalyst used.
Conditions: (i) aniline (1.0 eq.), aldehyde (1.1 eq.), MgSO4 (5 eq.)
(ii) 2.0 eq. CsOH·H2O, 5 mol% catalyst; reaction time: 48 h at –
50˚C. d.r. established by examination of the crude H NMR spec-
tra; e.r. established by chiral stationary phase HPLC. Yields are for
isolated material.
1
Under these conditions, using N-benzyl cinchonidinium chloride
as catalyst afforded indolines in a moderately improved d.r. and e.r
(4:1 d.r.; 81:19 e.r, entry 4). Protecting the free hydroxyl group as
an allyl ether (catalyst 24) gave entirely racemic material, con-
sistent with previous observations on the importance of Brønsted
acidic groups in phase-transfer catalysis.11 Use of catalyst 25, which
Although intermediate imines could be isolated and purified, in
practice we used imines without purification and hence all yields
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
are quoted over two steps from the corresponding anilines. In all
Figure 2: Catalyst dependent diastereoselectivity in ester- and ni-
trile-substituted benzhydryl systems. a10 mol% catalyst b 1 eq. KOH
1
2
3
4
5
6
7
8
cases the reaction is completely diastereoselective, affording the
product in which the C-3 nitrile group is syn- to the C-2 substituent
on the indoline. Nominally increasing the size of the aryl substitu-
ent (to 2-naphthyl 32) led to a decrease in enantioselectivity (to
90:10 e.r.), although substitution on the 4-position of the parent
arene (as in 33) leads to high e.r. and d.r. Electron-rich arenes such
as 2-furyl (34 and 35), and para-methoxyphenyl 36 were well
tolerated, giving selectivities of up to 99:1 e.r. The relative stereo-
chemistry of cyclohexyl substituted indoline 37 was confirmed by
X-ray crystallography, illustrating the syn- nature of the nitrile to
the C-2 substituent. Both electron withdrawing and donating sub-
stituents are tolerated on the indolinyl arene without compromis-
ing selectivity; the trifluoromethyl substituted indoline bearing a
cyclohexyl group in the 2-position 38 can be generated in 99.5:0.5
e.r. Alternative alkyl substituents proved to be more challenging but
indolines bearing a tert-butyl group (39, 84:16 e.r.), an unsaturat-
ed sidechain (40, 95:5 e.r.) or a cyclopentyl group (41, 85:15 e.r.)
can also be accommodated. In the ester-substituted system out-
lined earlier, electronegative substituents on the imine led to cy-
clization with very high levels of selectivity; this is in marked con-
trast to this system where electron-withdrawing groups such as 4-
bromo 42 and 4-nitro 43 yield significantly lower levels of enanti-
oselectivity (78:22 e.r.). Alkyl substitution on the indoline (such as
in 44) is permitted without compromizing selectivity (99:1 e.r.).
1
c 1 eq. CsOH·H2O. d.r. established by examination of the crude H
NMR spectra; e.r. established by chiral stationary phase HPLC.
Substrate 3 cyclizes upon treatment with solid KOH in toluene in
the presence of tetra-N-butylammonium hydrogensulfate
(TBAHS) to give indoline 5 in 78% yield as a single diastereoiso-
mer. The stereochemical course of this reaction was probed
through X-ray crystallography and nOe experiments, which are
consistent with the C-2 arene and the C-3 ester being on opposite
sides of the indoline nucleus (Figure 2A). Treatment of substrate 3
with N-benzylcinchonidinium chloride in toluene in the presence
of CsOH·H2O at –15˚C also led to the generation of a single indo-
line 4 with very high levels of diastereo- and enantioselectivity.
However, nOe analysis demonstrated that this was the opposite
diastereoisomer to that generated using an achiral phase-transfer
catalyst (in which the C-2 arene and the C-3 ester were on the same
side of the indoline core). Intriguingly, treatment of this material 4
(at 99:1 e.r. and d.r. >20:1) with tetra-N-butylammonium hydro-
gen sulfate and KOH in toluene at room temperature led to smooth
interconversion to the opposite diastereoisomer 5 (and complete
racemization). Similarly, we probed the variation in diastereoselec-
tivity for the nitrile system (Figure 2B). This demonstrated that a
single diastereoisomer is generated with the quinine-derived cata-
lyst 26 (at –50˚C), but that a mixture of diastereoisomers (in
which the opposite diastereoisomer predominates) is produced in
the presence of a simple alkylammonium salt
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
IS DIASTEREOSELECTIVITY A CONSEQUENCE OF
REVERSIBILTY UNDER CERTAIN CONDITIONS?
We believe that this is only consistent with reversible ring closing
and ring opening in the presence of TBAHS to ultimately yield the
thermodynamically most stable product. It follows that diastere-
omer 4 is kinetically favored under these reaction conditions and
we propose that the cinchoninium salt leads to kinetic and effec-
tively irreversible cyclization to indoline 4. To probe this we fol-
The observation that diastereoselectivity in the reaction of ester 3
varied with reaction temperature suggested to us that equilibration
between the two diastereoisomers could be occurring under certain
conditions. In order to probe this behavior, we examined the reac-
tions of substrates 3 and 21 in the presence of chiral and achiral
phase-transfer catalysts (Figure 2).
1
lowed the reaction with TBAHS and KOH by H NMR spectros-
[A]
Ph
Ph
copy. This confirmed that a low concentration of indoline 4 was
present throughout the reaction, before being ultimately converted
to indoline 5 over time. This is consistent with the proposed cy-
clization-cycloreversion mechanism, but does not explain why dif-
ferent diastereoisomers are produced with different catalysts. Dif-
ferences in reactivity between structurally distinct ammonium salts
in base-mediated phase-transfer reactions have been noted previ-
ously. In a comprehensive study, Denmark and co-workers
demonstrated that ammonium charge accessibility and catalyst
cross-sectional area were key factors in catalyst reactivity.12 To un-
derstand the reaction mechanism further we decided to (i) explore
the kinetic profile of the reaction of 21 with different phase-transfer
catalysts, and (ii) perform quantum chemical calculations to probe
reaction mechanisms and the relative thermodynamic stabilities of
diastereoisomeric products such as 22 and 23.
catalyst 6
i
CO2Pr
i
CO2Pr
CsOH•H2O
toluene
16 h, –15 ˚C
N
Ph
N
Ph
H
3
4. >20:1 d.r.
97:3 e.r.
5
Bu
Bu
Bu
Bu
Bu
KOHa,b
toluene
24 h, rt
KOHa,b
N
N
toluene
24 h, rt
Bu
Bu
Bu
H
HSO4
HSO4
Cl-
Ph
i
H
N
CO2Pr
N
Ph
OH
H
N
6
5. >20:1 d.r.
rac
[B]
Ph
Ph
catalyst 26
CN
Ph
CN
CsOH•H2O
toluene
16 h, –50 ˚C
N
N
Ph
H
MECHANISM: KINETIC STUDIES
21
22. >20:1 d.r.
94:6 e.r.
Our initial focus was on following the progress of the reaction with
time; we found it to be technically challenging to follow the pro-
gress of the reaction under the optimized reactions conditions (at –
50˚C), and the reaction is complete in fewer than 5 mins at room
temperature. Consequently we manipulated the reaction condi-
tions (through dilution and a reduction in the amount of base) to
enable the reaction to proceed to complete conversion at room
temperature over a period of approximately one hour (Figure 3).13
We were able to demonstrate that the reaction proceeds, albeit
22
Bu
Bu
Bu
Bu
Bu
Bu
CsOH•H2Oa,c
CsOH•H2Oa,c
N
N
Bu
toluene
24 h, rt
toluene
24 h, rt
Bu
H
HSO4
HSO4
Cl-
OMe
Ph
H
N
CN
Ph
OH
N
H
N
26
23. 3:2 d.r.
rac
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
slowly, in the absence of a phase-transfer catalyst, suggesting that dium anion of 4-methyl indoline derivative 45 (>20:1 d.r., 82:18
1
2
3
4
5
6
7
8
the reaction is likely to be interfacial in nature with deprotonation
occurring on the surface boundary between the solid base and tolu-
ene solution. The reaction was also mostly invariant with stirring
rate until below 200 rpm, when the reaction rate dropped signifi-
cantly, suggesting that this is when the reaction is intrinsically rate
limited by mass transfer.14 To establish a kinetic profile of the reac-
tion, aliquots were removed from the reaction mixture, filtered
though a prepacked pipette containing silica gel, eluted with a hex-
ane/IPA mixture and directly analyzed by HPLC; this enabled a
plot of % conversion to indoline vs. time to be produced (Figure 3).
The reaction kinetic profile is clearly sigmoidal in shape and pos-
sesses a significant induction period that is superficially suggestive
of autocatalysis or autoinduction.15 The observation of autocataly-
sis in phase-transfer reactions is relatively rare.16,17
e.r.) and added a substoichiometric amount to a solution of imine
21 in toluene at room temperature. Under these conditions we
observed essentially no cyclization to product over a 1 h timescale
and 4-methyl indoline 46 was recovered unchanged from the reac-
tion mixture (Scheme 3).
Scheme 3
Ph
Me
CN
9
Indoline recovered.
No change
in d.r. or e.r.
N
Na+
Ph
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Ph
45 (82:18 e.r., >20:1 d.r.)
Ph
Me
(30 mol %)
CN
Ph
CN
no reaction
1 h, rt
N
toluene, rt
N
H
Ph
21
46
H
Ph
OMe
Consequently, we performed a related experiment whereby the 4-
methyl indolinyl sodium salt 45 and ammonium salt 26 were pre-
mixed and then added to a stirring solution of imine 21 in toluene.
We were able to follow the conversion of imine 21 to indoline 22
by sampling rapidly during the first 10 minutes of the reaction
(Figure 4).
Ph
10 mol% cat. 26
CN
Ph
CN
Ph
H
N
Cl-
Ph
20 mol%
CsOH•H2O
toluene, rt
N
N
OH
H
N
22: > 20:1 d.r.
83:17 e.r.
97% isolated yield
21
26
Ph
100
Me
CN
90
80
70
60
50
40
30
20
10
0
N
Na+
Ph
Ph
45 (82:18 e.r., >20:1 d.r.)
Ph
Ph
Me
(30 mol %)
CN
CN
CN
N
Ph
N
Ph
N
Ph
10 mol% cat. 26
H
H
premixed
toluene, rt
21
22. > 20:1 d.r.
83:17 e.r.
97% isolated yield
46. Indoline recovered.
No change
in d.r. or e.r.
100
90
80
70
60
50
40
30
20
10
0
0
500
1000 1500 2000 2500 3000
time (s)
Figure 3: Plot of conversion to product versus time. T0 corre-
sponds to the time when base was added to the reaction mixture.
The order in catalyst was determined through variation of the cata-
lyst concentration and monitoring reaction progress under the
conditions above; this showed the order to be 0.56 (R2= 0.98). It
has been demonstrated by Dolling et al.18 that catalysts such as 26
exist as mono-deprotonated dimeric species under basic reaction
conditions. These authors suggested that the dimeric species is
significantly more soluble in the organic phase than the monomeric
derivative. This is consistent with the catalyst existing in a dimeric
form in its resting state prior to rapid and reversible dissociation to
an active monomeric form; this also implies that a process other
than dimer dissociation is responsible for the induction period.19
We also examined the reaction in the presence of a C-9 O-alkylated
catalyst to see whether the presence of a free OH group was a pre-
requisite for reactivity. This catalyst effected cyclization at a signifi-
cantly slower rate than catalyst 26, and both reactions also pos-
sessed an induction period.20 We considered that it was plausible
that the product of the cyclization reaction – an indolinyl anion –
could be accelerating the reaction through the production of a
more organic soluble base. Once generated, this material would
possess an appropriate pKa value to deprotonate the benzhydryl
position in 21 and, in concert with the quininium-derived cation,
effect asymmetric cyclization. To probe this, we generated the so-
0
500
1000
1500
2000
2500
3000
time (s)
Figure 4: Plot of conversion to product versus time. Red markers
indicate reaction progress in presence of preformed indolinyl ani-
on/ammonium salt mixture. T0 corresponds to the time when the
preformed indolinyl anion/ammonium salt mixture was added to
substrate. Blue markers are data from Figure 3 (for comparison).
Under these conditions the kinetic profile of the reaction was sig-
nificantly different and there was effectively no induction period.
This is suggestive that the reaction proceeds rapidly only when
there is sufficient concentration of the indolinyl anion paired with
the ammonium counterion in the organic phase.21 However, this
does not inform us whether the role of this anion is to deprotonate
the imine substrate 21 prior to cyclization or whether another spe-
cies is involved in this process. To investigate the deprotonation of
imine 21 we prepared the deuterated substrate 47 (90% D)22 and
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
treated this under the optimized cyclization conditions (Scheme riod to essentially zero, but the maximum rate of reaction was sig-
1
2
3
4
5
6
7
8
4). This demonstrated that there is a significant kinetic isotope
effect, consistent with deprotonation being the turnover-limiting
step in the reaction.23
nificantly slower than that observed in Figure 3. As a result, we
prestirred for a longer period of 6 h before adding substrate 21
(green circles). Under these conditions the reaction proceeds rap-
idly to complete conversion (in under 4 minutes) with identical
enantioselectivity and a maximum rate of reaction comparable to
that observed in Figures 3 and 4, but with no induction period. We
attribute the lower rate of reaction with a shorter pre-stirring period
to a lower concentration of active catalyst. We thus propose that
the overall process can be described by the catalytic cycle below
(Figure 6). We suggest that the catalyst resting state is likely to be
dimeric by virtue of our observation that its order is 0.56, consistent
with the proposal of Dolling.18 Our observation that pre-stirring
catalyst and base for a time comparable with the induction period
does not reproduce the maximal rate is suggestive that this may not
be the only route for the formation of the active catalyst. An alter-
native route involves initial deprotonation, likely by an interfacial
process, of the organic soluble imine 21 by solid cesium hydroxide
to form a cesium cyanocarbanion 48.
Scheme 4
Ph
Ph
D
10 mol% cat. 26
CN
Ph
CN
kH/kD = 4.4 ± 1.6
CsOH•H2O
(20 mol%)
toluene, rt
N
N
Ph
H
47 (90% D)
9
22. > 20:1 d.r.
83:17 e.r.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
d.r. established by examination of the crude 1H NMR spectrum; e.r.
established by chiral stationary phase HPLC.
We must reconcile this with (i) the observation that the reaction
can reverse under certain conditions, and (ii) the requirement for
an effectively irreversible reaction in the presence of catalyst 26.
This suggests that the indolinyl anion cannot be an autocatalyst as
this is incompatible with being involved in both the turnover-
limiting deprotonation and a low barrier protonation. Consequent-
ly, we proposed the indolinyl anion cyclization product could speed
up the reaction by deprotonating 26 to generate a zwitterionic
species which could accelerate a kinetically meaningful step in the
reaction sequence.24 To investigate this we treated ammonium salt
26 with a stoichiometric amount of CsOH·H2O in toluene and,
after stirring at rt for (i) 30 minutes and (ii) 6 h, we added substrate
21 and followed conversion to indoline 22 via HPLC (Figure 5).
H
H
OMe
Cl-
H
N
insoluble
catalyst
Ph
OH
N
26
base
H
H
OMe
OMe
N
soluble mono-
deprotonated dimer
(catalyst resting state)
H
N
H
N
Cl-
OMe
Ph
OH
Ph
O-
Cl-
H
N
N
Ph
OH
fast equilibrium
N
26
Ph
Ph
Ph
H
(10 mol%)
CN
Ph
CN
Ph
CN
OMe
N
CsOH•H2O (10 mol%)
toluene,
N
H
N
N
H
Ph
H
N
22
Ph
O-
21
premixed, rt, time
22. > 20:1 d.r.
83:17 e.r.
> 20:1 d.r.
83:17 e.r.
97% isolated yield
Ph
51
CN
Ph
N
21
100
90
80
70
60
50
40
30
20
10
0
H
OMe
N
Ph
Turnover
limiting
step
CN
H
N
Ph
N
Ph
OH
50
H
Catalyst
controlled
cyclization
OMe
N
Ph
H
N
CN
Ph
N
Ph
49
OH
H
OMe
PTC initiated
process
Cl-
CsCl
0
500
1000
1500
2000
2500
3000
H
N
time (s)
occurs at interfacial
boundary
Ph
OH
N
26
Figure 5: Plot of conversion to product versus time. Green circles
indicate reaction progress after 6 h prestirring; orange squares indi-
cate reaction progress after 30 mins prestirring. T0 corresponds to
the time when substrate was added to the reaction. Blue and red
markers are data from Figures 3 & 4 respectively (for comparison).
Ph
Ph
Cs
CN
CsOH•H2O(s)
CN
Ph
N
21
N
Ph
48
Figure 6: Proposed catalytic cycle (for reaction at RT, 10 mol%
catalyst, 20 mol% CsOH·H2O).
Pre-stirring for 30 minutes – a period comparable with the ob-
served induction period in Figure 3 – led to the conversion plot
above (Figure 5, orange squares). This reduced the induction pe-
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
This anion is relatively unreactive towards cyclization in the inter-
calculations28 we first examined the reaction of a malonate-derived
substrate (R1=R2=CO2Me). The ring-closure can be formulated as
either a 6π electrocyclization or a 5-endo-trig ring closure where a
malonate anion nucleophile attacks an imine electrophile, analo-
gous to a Mannich reaction.29
1
2
3
4
5
6
7
8
facial region until it undergoes ion exchange with the ammonium
salt to afford 49, whereby this substrate·ammonium complex mi-
grates to the organic phase and undergoes cyclization controlled by
the chiral cation to generate an enantioenriched indolinyl anion
paired with the chiral ammonium salt 50. This indolinyl anion can
then deprotonate the alcohol functional group on the cinchona-
derived ammonium salt, leading to the generation of the product
22 and the zwitterion 51. In order to achieve high levels of enanti-
oselectivity, the presence of a Brønsted acid in the organic phase is
essential to facilitate rapid trapping of the indolinyl anion. Zwitteri-
on 51 can then function as a Brønsted base and deprotonate an-
other molecule of imine 21 in the turnover-limiting step leading to
49 and 22 – but this time in the organic phase – and continuing
the cycle. The induction period is thus a reflection of the require-
ment to generate a sufficient concentration of species 51 in the
organic phase.25 Consequently, the reaction can be considered as an
interfacial phase-transfer reaction, but this may not be the most
accurate description of the overall process. Once initiated, we be-
lieve the reaction proceeds entirely in the homogeneous organic
phase, with the zwitterion 51 being responsible for turnover. As
such, the reaction may be better described as a phase-transfer initi-
ated and counterion-directed cyclization than a phase-transfer cata-
lyzed process. This mechanism could potentially operate in a range
of phase-transfer reactions where the pKa of an anion formed by
nucleophilic addition is able to deprotonate a pro-nucleophile and
allow the reaction to turn over. This is entirely consistent with the
principle of co-catalysis outlined by Mąkosza.26 Based on this mod-
el, we can also speculate on how differences in catalyst structure
(between simple alkylammonium salts and quinine-derived salts)
lead to the observed differences in kinetic and thermodynamic
product distribution. In the presence of a catalyst that does not
contain an alcohol functional group (such as tetrabutylammonium
hydrogen sulfate or O-allylated cinchonidinium derivative 24), we
observed (i) that the rate of formation of product 22 is significantly
lower, and (ii) the reaction leads to a different diastereoisomeric
outcome than when catalyst 26 is used. Based on our proposed
mechanism (Figure 5) we suggest that the observed difference in
diastereoselectivity arises by a process that occurs entirely in the
organic phase and hence we can neglect ion transport and interfa-
cial processes. This assumes that the ammonium salts are concen-
trated in the organic phase and hence solubility and lipophilicity are
less important than they are in processes where genuine transfer
between phases is a prerequisite for the bulk of the reaction to oc-
cur. In the absence of a proton source in toluene, we believe the
indolinyl anion could sequester the catalyst and slow turnover.
This has been observed in the nitroarylation of acetonitriles, which
requires a stoichiometric quantity of phase-transfer agent to pro-
ceed to completion.10 Where formation of a zwitterion is not possi-
ble (as with catalyst 24) the system could be forced into an alterna-
tive autocatalytic cycle where the indolinyl anion must deprotonate
imine 21 directly; it could be expected that this would be slower
than deprotonation of an alcohol, which is reflected in the overall
lower rate. This longer-lived indoline anion is then subject to cy-
cloreversion and racemization, consistent with the complete lack of
enantioselectivity demonstrated by 24.27
N
R2
R1
Electrocyclization
9
R1
N
R2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
R1
Ph
R2
-H+
+H+
N
H
Ph
R2
R1
N
5-endo-trig
1.44
1.50
2.16
1.36
1.33
3.01
+5.7
+19.9
1.44
1.33
1.52
1.61
1.4
4
+14.8
+5.3
1.43
1.36
1.34
1.50
2.99
[0.0]
Figure 7: Structures and relative free energies (M06-2X/6-
31+G(d,p); selected distances in Å) of stationary points involved in
ring closure of Z-imine (top) and E-imine (bottom).
The imine in the reactant can adopt either an E- or Z-geometry,
and distinct transition state structures for indoline formation were
found for both (Figure 7). Although activation barriers for ring
closure from either the E- or Z-reactant differ by less than 1 kcal
mol-1, the transition state structure for reaction of the E-imine is 5
kcal mol-1 lower than that for reaction of the Z-imine—a conse-
quence of the fact that the E-reactant is 6 kcal mol-1 lower in energy
than its Z-isomer. In both cases, visual inspection of the transition
state structure geometries does not readily reveal if they are best
described as pericyclic, so NICS (Nucleus Independent Chemical
Shift) calculations30 were used to probe aromaticity changes along
the associated reaction coordinates.31 Points at both the center of
the benzene ring and the center of the forming 5-membered ring
were examined (blue spheres in Figure 8). On the basis of the
NICS criteria, the arene in both systems becomes less aromatic as
cyclization occurs. This indicates significant delocalization of the
product anion into the benzene ring, which disrupts aromaticity,
but not reactant anion delocalization, consistent with the twisting
of reactant geometries to avoid steric clashes.
MECHANISM: QUANTUM CALCULATIONS
We decided to apply computation to address another key mecha-
nistic issue — are these reactions pericyclic or not? Using DFT
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
20$
15$
10$
5$
20$
15$
10$
5$
1
2
3
4
5
6
7
2.16
2.05
2.93
2.66
IRC$
IRC$
NICS(Forming$Ring)$
NICS(0)$
NICS(Forming)$
IRC
IRC
NICS(0)$
8
9
0$
0$
!15$
!10$
!5$
0$
5$
10$
15$
!15$
!10$
!5$
0$
5$
10$
15$
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NICS (forming ring)
NICS (forming ring)
1.65
1.65
!5$
!5$
NICS (benzene)
NICS (benzene)
!10$
!10$
Reaction Coordinate
Reaction Coordinate
Figure 8: Intrinsic reaction coordinate (blue) and NICS values (red and green; blue spheres on structures indicate points at which NICS
values were calculated) for ring-closure of Z-(left) and E-(right) imines; M06-2X/6-31+G(d,p) (forming C–C bond distances in Å).
NICS values for the forming ring indicate a decrease of aromaticity
throughout the reaction for the E-reactant (following a region
where there is little variation in NICS value), but an increase and
then decrease of aromaticity (i.e., a decrease then increase in NICS
value) along the reaction coordinate for the Z-isomer. The former
is consistent with a Mannich-type process, but the latter is what one
would expect for a pericyclic reaction where aromaticity is maximal
near the delocalized transition state structure.32 The malonate
group in the transition state structure for cyclization of the Z-
reactant is also twisted into a geometry expected for disrotation,
while that for cyclization of the E-reactant is essentially perpendicu-
lar to the benzene π-system, thereby shutting down the delocaliza-
tion expected for a pericyclic reaction. These results are consistent
with the E-isomer reacting via the Mannich-type pathway and the
Z-isomer reacting via a pericyclic reaction.33 The lack of cyclic delo-
calization during cyclization for the former is connected to the E-
imine geometry, which alters the position of the imine substituent,
providing a better geometry for π-stacking in the transition state
structure; this, however, appears to come at the cost of reducing
conjugation.34 Experimentally, the cyclization almost certainly
involves closure of E-imines.35 We subsequently examined the
cyclization of anions 52 and 57 (derived from 21 and 3 respec-
tively) in an attempt to understand why we observe interconver-
sion between different diastereoisomers (Figure 9). Starting with
cyanocarbanion 52 (Figure 10A), we identified transition state
structures for the cyclization to indolinyl anions 53 and 54. The
transition state structure (TS2) for cyclization to 54 is 2.1 kcal mol-
1
lower than that for formation of 53 (TS 1) — a consequence of
the fact that minimum 56 is lower in energy than 55. A similar
picture emerges for the cyclization of ester enolate 57, and in both
cases, transition state structures with both phenyl groups near to
each other are predicted to be lower in energy than those with dis-
tal phenyl groups. The proximity of the aryl groups in the lower
energy transition state structure is consistent with π-stacking play-
ing an important role in diastereoselection. These results are con-
sistent with the experimental observations for reactions in the pres-
ence of achiral phase-transfer catalysts, and may also reflect the
ability of chiral catalysts such as 26 to overwhelm the inherent
stereochemical preference of the cyclization.
A
B
Ph
Ph
Ph
Ph
Ph
Ph
5-endo-trig
via TS 2
5-endo-trig
via TS 1
5-endo-trig
via TS 4
5-endo-trig
via TS 3
CN
CN
Ph
CN
CO2Me
CO2Me
Ph
CO2Me
N
Ph
N
N
Ph
N
Ph
N
N
Ph
54
52
53
59
57
58
2.09
2.92
3.08
2.20
55. +2.9
TS 1. +11.0
53. +2.9
60. +2.4
TS 3. +10.9
58. +5.8
3.09
1.34
2.09
2.98
2.20
56. [0.0]
TS 2. +8.9
54. +0.1
61. [0.0]
TS 4. +9.5
59. +2.0
Figure 9: Structures and relative free energies (M06-2X/6-31+G(d,p); distances in Å) of stationary points involved in ring-closure of [A]
conformers 55 and 56 of anion 52 (from 21)and [B] conformers 60 and 61 of anion 57 (derived from methyl ester derivative of 3).
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
scholarship (to KS). We also acknowledge computational support from
We have recently described a 5-endo-trig cyclization to form in-
danes6 that has some similarities with this reaction. Although in
this case there is no obvious alternative (5-exo-trig) pathway, the
ease with which this nominally disfavored cyclization occurs under
kinetic control is noteworthy. Calculations performed for the cy-
clization of cyanocarbanion 52 demonstrate that in the minimum,
the cyano group and both arenes do not lie in a plane, so this anion
is not delocalized into the nitrogen-substituted arene. The transi-
tion states for the formation of diastereoisomers 53 and 54 have
angles of attack onto the imine of 96.2˚ (TS 1) and 98.3˚ (TS 2)
respectively. These are significantly smaller than the Bürgi-Dunitz
angle and are consistent with previous observations that approach
trajectory may not be decisive in irreversible ring closing reactions.6
the NSF’s XSEDE program.
ABBREVIATIONS
REFERENCES
1. (a) Shirakawa, S.; Maruoka, K. Angew. Chem. Int. Ed. 2013, 52,
4312–4348; (b) Ooi, T.; Maruoka, K. Angew. Chem. Int. Ed. 2007, 46,
4222–4266.
2. (a) Hiyama, T.; Mishima, T.; Sawada, H.; Nozaki, H. J. Am. Chem.
Soc. 1975, 97, 1626-1627; (b) Fiaud, J.C. Tetrahedron Lett. 1975, 16,
3495-3496; (c) Helder, R.; Hummelen, J. C.; Laane, R. W. P. M.; Wiering,
J. S.; Wynberg, H. Tetrahedron Lett. 1976, 17, 1831–1834; (d) Colonna,
S.; Fornasier, R. J. Chem. Soc. Perkin 1 1978, 371-373; (e) O’Donnell, M.
J.; Bennett, W. D.; Wu, S. J. Am. Chem. Soc. 1989, 111, 2353–2355; (f)
O’Donnell, M. J.; Wu S.; Huffman; J. C. Tetrahedron 1994, 50, 4507–
4518; (g) Corey E. J.; Xu, F.; Noe, M. C. J. Am. Chem. Soc. 1997, 119,
12414–12415; (h) Lygo, B.; Wainwright, P.G. Tetrahedron Lett. 1997, 38,
8595-8598.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In all cases, these reactions are predicted to be close to thermoneu-
tral or slightly endergonic, which is consistent with the ease with
which the process reverses under certain conditions. This is also
consistent with the observation that high enantioselectivity re-
quires a catalyst bearing an alcohol to facilitate a low barrier and
essentially irreversible protonation event to drive the reaction for-
ward.
3. Maciver, E. E.; Knipe, P. C.; Cridland, A.; Thompson, A. L; Smith, M.
D. Chem. Sci. 2012, 3, 537-440.
4. Maciver, E. E.; Thompson, S;. Smith, M. D. Angew. Chem. Int. Ed.
2009, 48, 9979-9982.
5. For related cyclizations see: Speckamp, W. N.; Veenstra, S. J.; Dijkink,
J.; Fortgens, R.; J. Am. Chem. Soc. 1981, 103, 4643-4645; (b) Veenstra, S.
J.; Speckamp, W. N. J. Am. Chem. Soc. 1981, 103, 4645-4647; (c)
Veenstra, S. J.; Speckamp, W. N. J. Chem. Soc. Chem. Commun. 1982,
369-370; (d) Dijkink, J.; Zonjee, J.; de Jong, N. B. S.; Speckamp, W. N.
Heterocycles 1983, 20, 1255-1258;(e) Speckamp, W. N. Heterocycles
1984, 21, 211-234; (f) Veenstra, S. J.; Fortgens, H. P.; Vijn, R. J.; de Jong,
B. S.; Speckamp, W. N. Tetrahedron 1987, 43, 1147-1156; (f) Huisgen R.
Angew. Chem. Int. Ed. 1980, 19, 947-973; (g) Taylor, E. C.; Turchi, I.
Chem. Rev. 1979, 79, 181-231; (h) Grigg, R.; Gunaratne, H. Q. N. Tetra-
hedron Lett. 1983, 24, 1201-1204; (i) Grigg, R.; Gunaratne, H. Q. N.;
Kemp, J. J. Chem. Soc. Perkin Trans. 1 1984, 41-46; (j) Grigg, R.;
Gunaratne, H. Q. N.; Henderson, D.; Sridharan, V. Tetrahedron 1990, 46,
1599-1610; (k) Kanner, C. B.; Pandit, U. K. Tetrahedron 1981, 37, 3519-
3523.
CONCLUSIONS
We have demonstrated an approach to the diastereo- and enanti-
oselective synthesis of indolines bearing two stereocenters via a
Mannich-like 5-endo-trig process. Kinetic analysis has shown that
this is an autoinductive process that is merely initiated by phase-
transfer catalysis, rather than a true phase-transfer process. Once
sufficient concentration of the reactive catalyst – a quinine-derived
zwitterion – is produced, the reaction occurs entirely within the
organic liquid phase. The enantioselective reaction is facilitated by
a protonation event that necessitates the presence of an alcohol;
catalysts not bearing this functional group are kinetically less profi-
cient and nonselective. These observations may have implications
for the mechanisms of other phase-transfer mediated reactions and
the design of new phase-transfer catalysts.
6. Johnston, C. P.; Kothari, A.; Sergeieva, T.; Okovytyy, S. I.; Jackson, K.
E.; Paton, R. S.; Smith, M. D. Nature Chem. 2015, 7, 171-177.
7. Yin, X.-P.; Zeng, X.-P.; Liu, Y.-L.; Liao, F.-M.; Yu, J.-S.; Zhou, F.;
Zhou, J. Angew. Chem. Int. Ed. 2014, 53, 13740-13745.
ASSOCIATED CONTENT
Full synthetic details, 1H and 13C NMR spectra, crystallographic data in
.cif format and additional details on computations. The crystal struc-
ture data of 5, 22, 23 and 37 have been deposited in the Cambridge
Crystallographic Data Center (CCDC 1417302-1417305 respective-
ly). This material is available free of charge via the Internet at
http://pubs.acs.org.
8. This transformation is sensitive to the quality of the potassium tert-
butoxide employed; newly sublimed and anhydrous material is optimal.
9. Both KOH and CsOH·H2O are deliquescent solids and it is challeng-
ing to ascertain the amount of adventitious water that these materials ab-
sorb from the air. We did not take special precautions (sucg as a glove box)
to exclude moisture during weighing of these materials for reaction.
10. (a) Mąkosza, M.; Tomashewski, A. A. J. Org. Chem. 1995, 60,
5425-5429; (b) Mąkosza, M. Pure Appl. Chem. 1975, 43, 439-462.
11. (a) Ooi, T.; Ohara, D.; Tamura, M.; Maruoka, K. J. Am. Chem. Soc.
2004, 126, 6844–6845; (b) Ooi, T.; Ohara, D.; Fukumoto, K.; Maruoka,
K. Org. Lett. 2005, 7, 3195–3197; (c) Berkessel, A.; Guixa, M.; Schmidt,
F.; Neudörfl, J. M.; Lex, J. Chem. Eur. J. 2007, 13, 4483–4498; (d) Bernal,
P.; Fernández, R.; Lassaletta, J. M. Chem. Eur. J. 2010,16, 7714–7718; (e)
Weintraub, R. C.; Denmark, S. E. Heterocycles 2011, 82, 1527-1540; (f)
Johnson, K. M.;. Rattley, M. S.; Sladojevich, F.; Barber, D. M.; Nunez, M.
G.; Goldys, A. M.; Dixon, D. J. Org. Lett. 2012, 14, 2492–2495; (g) Li, M.;
Woods, P. A.; Smith, M. D. Chem. Sci. 2013, 4, 2907-2911; (h) Novacek,
J.; Waser, M. Eur. J. Org. Chem. 2013, 637–648; (i) Novacek, J.; Waser,
M. Eur. J. Org. Chem. 2014, 802-809; (j) Xiang, B.; Belyk, K. M.; Reamer,
R. A.; Yasuda, N. Angew. Chem. Int. Ed. 2014, 53, 8375-8378.
12. (a) Weintraub, R. C.; Gould, N. D.; Denmark, S. E. J. Am. Chem.
Soc. 2012, 134, 13415-13429; (b) Denmark, S. E.; Henle, J. J. Chem. Sci.
2015, 6, 2211-2218; (c) Gould, N. D.; Wolf, L. M.; Denmark, S. E. J. Org.
AUTHOR INFORMATION
Corresponding Author
•
•
martin.smith@chem.ox.ac.uk
djtantillo@ucdavis.edu
Author Contributions
All authors have given approval to the final version of the manuscript.
ACKNOWLEDGMENT
We are grateful to Prof John Brown FRS for helpful discussions. The
European Research Council has provided financial support under the
European Community’s Seventh Framework Programme (FP7/2007-
2013) / ERC grant agreement no. 259056. We are grateful to EPSRC
and Pfizer for CASE award (to CPJ), and the Tata foundation for a
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
Chem. 2011, 76, 4260-4336; (d) Gould, N. D.; Wolf, L. M.; Denmark, S.
E. J. Org. Chem. 2011, 76, 4337-4357.
13. We make the assumption that the reaction mechanism under these
modified conditions does not change from that operating under the opti-
mized conditions.
14. Rabinovitz, M.; Cohen, Y.; Halpern, M. Angew. Chem. Int. Ed.
1986, 25, 960-970.
D. J. Wallingford CT, 2009) with the B3LYP/6-31+G(d,p) (Becke, A. D. J.
Chem. Phys. 1993, 98, 1372-1377; Becke, A. D. J. Chem. Phys. 1993, 98,
5648-5652; Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785-789,
Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys.
Chem. 1994, 98, 11623-11627; Tirado-Rives, J.; Jorgensen, W. L J. Chem.
Theor. Comput. 2008, 4, 297-306) and M06-2X/6-31+G(d,p) (Zhao, Y.;
Truhlar, D. Theor Chem. Acc. 2008, 120, 215-241.) methods were used
for geometry optimizations of minima and transition state structures. Tran-
sition state structures were also reoptimized using CAM-B3LYP/6-
31+G(d,p) (Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004,
1-3, 51-57) and the SMD solvation model (Marenich, A. V.; Cramer, C. J.;
Truhlar, D. G. J. Phys. Chem. B 2009, 113, 6378-6396) with THF as the
solvent to address issues associated with dispersion and solvation, respec-
tively (see Supporting Information for details). Stationary points were
characterized via frequency analysis and connected to their respective
minima via Intrinsic Reaction Coordinate (IRC) calculations (Gonzalez,
C.; Schlegel, H. B. J. Chem. Phys. 1991, 95, 5853-5860; Gonzalez, C.;
Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154-2161; C. Lee, C.; Yang, W.;
Parr, R. G. Phys. Rev. B 1988, 37, 785-789; Fukui, K. Acc Chem. Res.
1981, 14, 363-368). Energies reported herein are M06-2X/6-31+G(d,p)
Gibbs free energies at 298.5 K in the gas phase, unless otherwise noted.
29. For asymmetric Mannich reactions that proceed under phase-
transfer catalysis, see: (a) Ooi, T.; Kameda, M.; Fujii, J.; Maruoka, K. Org.
Lett. 2004, 6, 2397-2399; (b) Poulsen, T. B.; Alemparte, C.; Saaby, S.;
Bella, M.; Jørgensen, K. A. Angew. Chem. Int. Ed. 2005, 44, 2896-2899;
(c) Okada, A.; Shibuguchi, T.; Oshima, T.; Masu, H.; Yamaguchi, K.; Shi-
basaki, M. Angew. Chem. Int. Ed. 2005, 44, 4564-4567; (d) Fini, F.; Ber-
nardi, L.; Herrera, R. P.; Pettersen, D.; Ricci, A.; Sgarzani, V. Adv. Synth.
Catal. 2006, 348, 2043-2046; (e) Niess, B.; Jørgensen, K. A. Chem. Com-
mun. 2007, 1620-1622; (f) Shibuguchi, T.; Mihara, H.; Kuramochi, A.;
Ohshima, T.; Shibasaki, M. Chem. Asian J. 2007, 2, 794-801.
1
2
3
4
5
6
7
8
15. Bissette, A. J.; Fletcher, S. P. Angew. Chem. Int. Ed. 2013, 52,
12800-12826.
16. (a) Starks, C. M.; Owens, R. M. J. Am. Chem. Soc. 1973, 95, 3613-
3617; (b) Halpern, M.; Cohen, Y.; Sasson, Y.; Rabinovitz, M. Nouv. J.
Chim. 1984, 8, 443-444; (c) Buhse, T.; Nagarajan, R.; Lavabre, D.; Mi-
cheau, J. C. J. Phys. Chem. A 1997, 101, 3910–3917; (d) Buhse, T.; Lava-
bre, D.; Nagarajan, R.; Micheau, J. C. J. Phys. Chem. A 1998, 102, 10552–
10559; (e) A. J. Bissette, B. Odell and S. P. Fletcher. Nature Commun.
2014, 5, 4607.
17. Zhao, Q.; Sun, J.; Liu, B.; He, J. Chem. Eng. J. 2015, 262, 756-765.
18. (a) Dolling, U. H. Davis P., Grabowski E. J. J. J. Am. Chem. Soc.
1984, 106, 446–447; (b) Hughes, D. L.; Dolling U.-H.; Ryan, K. M.;
Schoenewaldt, E. F.; Grabowski E. J. J. J. Org. Chem. 1987, 52, 4745-
4752.
19. For an autocatalytic phase-transfer reaction in which the reaction
product increases the solubility of the catalyst in the organic phase see:
Glatzer, H. J.; Doraiswamy, L. K. Chem. Eng. Sci. 2000, 55, 5149-5160.
20. For full details see Supporting Information.
21. A corrollary of this is that the configuration and enantiopurity of the
indoline ‘catalyst’ component should not affect the enantioselectivity of the
overall process. This was confirmed by performing the reaction with 10
mol% chiral ammonium salt 26 in the presence of a substoichiometric
amount of racemic indolinyl salt 45 which resulted in identical selectivity
(83:17 e.r., d.r. >20:1).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22. Halpern, M.; Feldman, D.; Sasson, Y.; Rabinovitz, M. Angew. Chem.
Int. Ed. 1984, 23, 54-55.
23. This observation is the average of three determinations.
24. This process is best described as autoinduction: Blackmond, D. G. An-
gew. Chem. 2009, 121, 392 –396.
25. (a) Viout, P. J. Mol. Cat. 1981, 10, 231-240; (b) Nelson, K. V.; Ben-
jamin, I. J. of Phys. Chem. C. 2010, 114, 1154-1163; (c) Nelson, K. V.;
Benjamin, I. J. Phys. Chem. C. 2011, 115, 2290-2296.
26. (a) Maķosza, M.; Lasek, W. Tetrahedron 1991, 47, 2843-2850; (b)
Maķosza, M.; Chesnokov, A. Tetrahedron 2000, 56, 3553; (c) Maķosza,
M.; Chesnokov, A. Tetrahedron 2002, 58, 7295; (d) Maķosza, M.; Fedo-
30. Nucleus Independent Chemical Shift (NICS) calculations
(Schleyer, P. v. R.; Maerker, C.; Dransfeld, A.; Jiao, H.; Hommes, N. J. R. v.
E. J. Am. Chem. Soc. 1996, 118, 6317-6318; Chen, Z.; Wannere, C. S.;
Corminboeuf, C.; Puchata, R.; Schleyer, P. v. R. Chem. Rev. 2005, 105,
3842-3888) were used to probe the aromaticity of the transition states; a
dummy atom was placed at the geometric center of the benzene ring and a
second dummy atom was placed in the center of the forming 5-membered
ring (this dummy resided at the end of a 2.15 Å line segment connected to
the centroid of the benzene ring and bisecting the carbon-carbon bond of
the benzene ring and bearing the two substituents). Images of molecular
structures were created using Ball & Stick (Ball & Stick 4.0a12, Muller, N.;
Faulk, A. Johannes Kepler University Linz 2004).
31. For a related approach see: Knipe, P. C.; Gredicak, M.; Cernijenko,
A.; Paton, R. S.; Smith, M. D. Chem. Eur. J. 2014, 20, 3005-3009.
32. Jiao, H.; Schleyer, P. v. R. J. Phys. Org. Chem. 1998, 11, 655-662.
33. The IUPAC definition of an electrocyclization is ‘a molecular rear-
rangement that involves the formation of a σ-bond between the termini of a
fully conjugated linear π-electron system and a decrease by one in the
number of π-bonds’. Although this definition is consistent with the trans-
formation of 21 to 22, the lack of both aromaticity and disrotation in the
transition structure are key features that indicate the reaction is better
described as nonpericyclic.
ryński, M. Arkivoc, 2006 (iv) 7-17; (e) Lygo, B.; Beynon, C.; Lumley, C.;
McLeod, M. C.; Wade, C. E. Tetrahedron Lett. 2009, 50, 3363–3365.
27. This does not preclude the Brønsted acidic group in catalyst 26 play-
ing a role in the enantioselectivity of the cyclization.
28. Gaussian09 (Revision B.01 Frisch, M. J.; Trucks, G. W.; Schlegel, H.
B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone,
V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.;
Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.;
Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.;
Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J.
A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.;
Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.;
Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich,
S.; Daniels, A. D.; Farkas; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox,
34. For a discussion of substituent effects on a related reaction see: Gil-
more, K.; Manoharan, M.; I-Chia Wu, J.; Schleyer, P. v. R.; Alabugin, I. V. J.
Am. Chem. Soc., 2012, 134, 10584-10594.
35. We did not observe any evidence suggestive of a change in imine ge-
ometry during the reaction.
ACS Paragon Plus Environment

Page 11 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
R3
R3
+
R4N
CN
R2
CN
R1
R1
base
5-endo-trig
N
H
R2
N
up to 99.5:0.5 e.r.; > 20:1 d.r.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment