Homogenous Pd-catalyzed asymmetric hydrogenation of unprotected indoles: Scope and mechanistic studies

Article abstract of DOI:10.1021/ja502020b

An efficient palladium-catalyzed asymmetric hydrogenation of a variety of unprotected indoles has been developed that gives up to 98% ee using a strong Br?nsted acid as the activator. This methodology was applied in the facile synthesis of biologically active products containing a chiral indoline skeleton. The mechanism of Pd-catalyzed asymmetric hydrogenation was investigated as well. Isotope-labeling reactions and ESI-HRMS proved that an iminium salt formed by protonation of the C=C bond of indoles was the significant intermediate in this reaction. The important proposed active catalytic Pd-H species was observed with ¹H NMR spectroscopy. It was found that proton exchange between the Pd-H active species and solvent trifluoroethanol (TFE) did not occur, although this proton exchange had been previously observed between metal hydrides and alcoholic solvents. Density functional theory calculations were also carried out to give further insight into the mechanism of Pd-catalyzed asymmetric hydrogenation of indoles. This combination of experimental and theoretical studies suggests that Pd-catalyzed hydrogenation goes through a stepwise outer-sphere and ionic hydrogenation mechanism. The activation of hydrogen gas is a heterolytic process assisted by trifluoroacetate of Pd complex via a six-membered-ring transition state. The reaction proceeds well in polar solvent TFE owing to its ability to stabilize the ionic intermediates in the Pd-H generation step. The strong Br?nsted acid activator can remarkably decrease the energy barrier for both Pd-H generation and hydrogenation. The high enantioselectivity arises from a hydrogen-bonding interaction between N-H of the iminium salt and oxygen of the coordinated trifluoroacetate in the eight-membered-ring transition state for hydride transfer, while the active chiral Pd complex is a typical bifunctional catalyst, effecting both the hydrogenation and hydrogen-bonding interaction between the iminium salt and the coordinated trifluoroacetate of Pd complex. Notably, the Pd-catalyzed asymmetric hydrogenation is relatively tolerant to oxygen, acid, and water.

Full text of DOI:10.1021/ja502020b

Article
pubs.acs.org/JACS
Homogenous Pd-Catalyzed Asymmetric Hydrogenation of
Unprotected Indoles: Scope and Mechanistic Studies
Ying Duan,‡,† Lu Li,#,† Mu-Wang Chen,‡ Chang-Bin Yu,‡ Hong-Jun Fan,*,‡ and Yong-Gui Zhou*,‡
‡State Key Laboratory of Molecular Reaction Dynamics, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian
116023, China
#State Key Laboratory of Fine Chemicals, Institute of Coal Chemical Engineering, School of Chemical Engineering, Dalian University
of Technology, Dalian 116024, China
S
* Supporting Information
ABSTRACT: An efficient palladium-catalyzed asymmetric hy-
drogenation of a variety of unprotected indoles has been
developed that gives up to 98% ee using a strong Brønsted acid
as the activator. This methodology was applied in the facile
synthesis of biologically active products containing a chiral
indoline skeleton. The mechanism of Pd-catalyzed asymmetric
hydrogenation was investigated as well. Isotope-labeling reactions
and ESI-HRMS proved that an iminium salt formed by protonation of the CC bond of indoles was the significant intermediate
in this reaction. The important proposed active catalytic Pd−H species was observed with 1H NMR spectroscopy. It was found
that proton exchange between the Pd−H active species and solvent trifluoroethanol (TFE) did not occur, although this proton
exchange had been previously observed between metal hydrides and alcoholic solvents. Density functional theory calculations
were also carried out to give further insight into the mechanism of Pd-catalyzed asymmetric hydrogenation of indoles. This
combination of experimental and theoretical studies suggests that Pd-catalyzed hydrogenation goes through a stepwise outer-
sphere and ionic hydrogenation mechanism. The activation of hydrogen gas is a heterolytic process assisted by trifluoroacetate of
Pd complex via a six-membered-ring transition state. The reaction proceeds well in polar solvent TFE owing to its ability to
stabilize the ionic intermediates in the Pd−H generation step. The strong Brønsted acid activator can remarkably decrease the
energy barrier for both Pd−H generation and hydrogenation. The high enantioselectivity arises from a hydrogen-bonding
interaction between N−H of the iminium salt and oxygen of the coordinated trifluoroacetate in the eight-membered-ring
transition state for hydride transfer, while the active chiral Pd complex is a typical bifunctional catalyst, effecting both the
hydrogenation and hydrogen-bonding interaction between the iminium salt and the coordinated trifluoroacetate of Pd complex.
Notably, the Pd-catalyzed asymmetric hydrogenation is relatively tolerant to oxygen, acid, and water.
and Agbossou−Niedercorn7 then found that a complex of Rh
INTRODUCTION
■
and PinPhos or WalPhos as catalyst could also promote
Chiral indolines are a significant class of alkaloids found in
numerous bioactive compounds, such as pharmaceuticals,
asymmetric hydrogenation of N-protected indoles. The Ir/N,P-
ligand system was a desirable candidate as well, yet it gave
herbicides, and insecticides.1 These privileged motifs are key
barely satisfactory yields.8 There was virtually no development
building blocks in the discovery of drugs. Although the demand
in the hydrogenation of simple indoles until our group
for enantiomerically pure indolines is growing as time goes on,
introduced the strategy of Brønsted acid activation (Scheme
the methods to reach them are still limited. Because indoles
have inherent stability and it is difficult to modify their aromatic
1).9 In that work, a powerful Pd catalyst system was selected
due to its inertness to strong Brønsted acids. Subsequently,
properties, the journey from indoles to indolines has a long way
dehydration-triggered asymmetric hydrogenation10 and consec-
to go. From the perspective of atom economy, asymmetric
utive Brønsted acid/Pd-complex-promoted tandem reactions11
hydrogenation is the most direct and important way to achieve
this process.2 As a huge family of heteroaromatic compounds,
were also reported. Moreover, 3-(toluenesulfonamidoalkyl)-
indoles were synthesized and hydrogenated to form chiral
indoles have been shown to have important applications in
organic synthesis, but they exhibited inertness in asymmetric
indolines.12,13
Chiral Pd-based catalysts have achieved great successes in
hydrogenation until 2000.
asymmetric hydrogenation of a large range of substrates,14 such
as activated imines,15 enamines,16 functional ketones,17
olefins,18 and also heteroaromatic compounds.9−12,19 However,
In 2000, Ito, Kuwano, and co-workers3 reported the first
example of asymmetric hydrogenation of N-protected indoles
using a Rh/Ph-TRAP complex as catalyst. Thereafter, Rh/Ph-
TRAP4 and Ru/Ph-TRAP5 were found to be useful in
asymmetric hydrogenation of indoles with an electron-with-
drawing group on the nitrogen atom. The groups of Minnaard6
Received: February 26, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 1. Strong Brønsted Acid Activation Strategy for
Asymmetric Hydrogenation of Unprotected Indoles
Table 1. Optimization of Conditions for Asymmetric
a
Hydrogenation of 2-Substituted Indole 1a
conv
b
c
entry ligand
solvent
acid
(%)
ee (%)
1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L1
L2
L3
L4
L5
L6
L7
L8
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
−
TfOH
<5
61
−
2
54 (R)
62 (R)
69 (R)
8 (R)
3
PhSO3H
TsOH·H2O
TFA
>95
>95
>95
>95
44
4
5
6
L-CSA
71 (R)
66 (R)
25 (R)
4 (R)
7
D-CSA
PhCO2H
salicylic acid
L-CSA
8
<5
there are also some issues to be noted in these reactions. First,
trifluoroethanol (TFE) has always been the best solvent, but
that was challenged in the hydrogenation of simple indoles and
pyrroles. Second, the detailed mechanism of Pd-catalyzed
asymmetric hydrogenation remains to be elucidated, in terms of
the fashion of hydrogen gas activation or the dissociation of Pd
catalyst from the product. This question is relatively unex-
plored, in contrast to Ir-,20 Ru-,21 or Rh22-catalyzed asymmetric
hydrogenations. In this article, we extend the scope of Pd-
catalyzed asymmetric hydrogenation of unprotected indoles
and gain insight into the mechanism of Pd-catalyzed
asymmetric hydrogenation by a combination of experimental
and theoretical studies.
9
35
10
11
12
13
14
15
16
17
18
19
20
21
22
23
MeOH
19
40 (R)
57 (R)
−
toluene
L-CSA
57
THF
L-CSA
<5
DCM
L-CSA
89
73 (R)
82 (R)
85 (R)
80 (R)
86 (R)
81 (R)
84 (R)
84 (R)
91 (R)
80 (R)
−
DCM/TFE (2:1)
DCM/TFE (1:1)
DCM/TFE (1:2)
DCM/TFE (1:1)
DCM/TFE (1:1)
DCM/TFE (1:1)
DCM/TFE (1:1)
DCM/TFE (1:1)
DCM/TFE (1:1)
DCM/TFE (1:1)
L-CSA
>95
>95
>95
>95
>95
>95
>95
>95
49
L-CSA
L-CSA
L-CSA
L-CSA
L-CSA
L-CSA
L-CSA
L-CSA
RESULTS AND DISCUSSION
■
L-CSA
<5
Asymmetric Hydrogenation of Simple 2-Substituted
Indoles. 2-Methylindole was chosen as the model substrate to
assay the best reaction conditions for asymmetric hydro-
genation, and Pd(OCOCF3)2/(R)-SegPhos complex was
applied to begin the investigation. As shown in Table 1,
hydrogenation does not occur in the absence of Brønsted acid
(entry 1). To achieve full conversion and high enantio-
selectivity requires a stoichiometric amount of Brønsted acid to
activate 2-methylindole. First, the effects of Brønsted acids on
the reactivity and enantioselectivity were examined. Among the
strong Brønsted acids, PhSO3H, TsOH·H2O, and L-camphor-
sulfonic acid (L-CSA) gave full conversions and moderate
enantiomeric excess (ee) values (Table 1, entries 3, 4, and 6); L-
CSA gave full conversion and the highest enantioselectivity
(entry 6, 71% ee). In contrast, D-CSA with the opposite
configuration gave product with 44% conversion and somewhat
lower ee (entry 7, 66% ee). These varied activities and ee values
may ascribed to the different matching ability of the chiral
ligand and the two CSAs. However, with weak Brønsted acids,
such as benzoic acid and salicylic acid, either low turnover or
low enantioselectivity was observed. Therefore, it was
concluded that L-CSA was the best activator to promote this
asymmetric hydrogenation, on the basis of the data in Table 1.
Second, the effect of solvent was evaluated, and it was found
that solvent influenced both conversions and ee values
dramatically (Table 1, entries 6, 10−13). Fortunately, the
combination of equal volumes of DCM and TFE resulted in full
conversion and higher ee (Table 1, entries 15 vs 14 and 16).
Finally, examination of various bisphosphine ligands showed
that (R)-H8-BINAP promoted the ee value to 91% (Table 1,
entries 17−23). After screening a wide range of parameters for
indole 1a, we determined the optimized conditions: Pd-
(OCOCF3)2/(R)-H8-BINAP/L-CSA/rt/TFE:DCM = 1:1.
a
Conditions: indole 1a (0.25 mmol), acid (0.25 mmol), Pd-
(OCOCF3)2 (2 mol%), ligand (2.4 mol%), solvent (3.0 mL), 24 h,
b
c
1
rt. Determined by H NMR. Determined by HPLC.
With the optimized conditions in hand, we investigated the
range of substrates (Table 2). Those substrates bearing primary
or secondary alkyl groups could all give excellent ee values and
moderate to excellent yields. The length of 2-alkyl substituent
as well as the size of the cyclic alkyl substituent had little effect
on the enantioselectivities (Table 2, entries 1−5). 2-Phenethyl-
and 2-benzylindole could also give prominent ee values, 93%
and 95%, respectively (Table 2, entries 6 and 7). The steric
effect of benzene ring bearing on benzyl displayed little concern
with activities and enantioselectivities (Table 2, entries 8−11).
γ-Ester group was tolerated and high ee was obtained with the
slight modification of the reaction conditions (Table 2, entry
12), but hydroxyl was not so fortunate (Table 2, entry 13).
However, the presence of group at C5 of indole, such as 5-
fluoro and 5-methyl, decreased the ee values to 88% and 84%,
respectively (Table 2, entries 14 and 15). Unfortunately, 2-
carbalkoxy and 2-styryl group inhibited the reaction, and the
B
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 2. Pd-Catalyzed Asymmetric Hydrogenation of 2-
Substituted Indoles 1
Scheme 2. Asymmetric Hydrogenation of 2,3-Disubstituted
Indoles 3 via a Dynamic Kinetic Resolution Process
a
b
c
entry
R1
R2
yield (%)
ee (%)
1
H
Me
88 (2a)
82 (2b)
89 (2c)
90 (2d)
85 (2e)
89 (2f)
99 (2g)
84 (2h)
95 (2i)
82 (2j)
78 (2k)
98 (2l)
<5
91 (R)
93 (R)
92 (R)
95 (S)
95 (S)
93 (R)
95 (R)
94 (R)
94 (R)
93 (R)
96 (R)
84 (R)
−
2
H
n-Bu
3
H
n-pentyl
4
H
Cy
5
H
cyclopentyl
phenethyl
Bn
6
H
7
H
8
H
2-MeC6H4CH2
3-MeC6H4CH2
4-MeC6H4CH2
1-naphthyl-CH2
CH2CH2COOEt
CH2CH2CH2OH
Me
9
H
10
11
H
H
d
12
H
13
14
15
16
17
18
19
20
H
5-F
5-Me
H
84 (2m)
81 (2n)
<5
88 (R)
84 (R)
−
−
−
(2:1) gave better results than the others (Table 3, entry 3 vs
entries 1 and 2), yielding only the cis-2,3-disubstituted indoline
Me
COOEt
H
styryl (E)
CF3
<5
Table 3. Optimization for Asymmetric Hydrogenation of
a
H
<5
2,3-Disubstituted Indole 3a
H
t-Bu
<5
−
−
H
Ph
<5
a
Conditions: 1 (0.25 mmol), L-CSA (0.25 mmol), Pd(OCOCF3)2 (2
mol%), (R)-H8-BINAP (2.4 mol%), DCM/TFE (3 mL, 1:1), 24 h, rt.
b
c
d
Isolated yield. Determined by HPLC. 50 °C.
b
c
entry
solvent
acid
yield (%)
ee (%)
double bond of the styryl group was intact (Table 2, entries 16
and 17). Meanwhile, sterically more demanding groups (tert-
butyl) and electron-withdrawing goups (trifluoromethyl group)
were problematic with low activity (Table 2, entries 18 and 19).
Disappointingly, almost no reaction was observed when 2-
phenylindole was subjected to the standard conditions (Table
2, entry 20).
Asymmetric Hydrogenation of Simple 2,3-Disubsti-
tuted Indoles. Encouraged by successful asymmetric hydro-
genation of unprotected 2-substituted indoles, we next turned
our attention to asymmetric hydrogenation of 2,3-disubstituted
indoles. Unlike the results with the simple 2-substituted indoles,
two contiguous chiral centers will be obtained in the
hydrogenation of 2,3-disubstituted indoles. These two chiral
carbons are produced in different steps: asymmetric proto-
nation and asymmetric hydrogenation. Initially, the chiral
center at the 3-position is generated from protonation of indole
3. Subsequently, selective hydrogenation of one of the isomers
gives the hydrogenation product 4 (Scheme 2). Indeed, a
classical dynamic kinetic asymmetric transformation process is
involved in this reaction.
1
2
3
4
5
6
7
DCM/TFE (1:2)
DCM/TFE (1:1)
DCM/TFE (2:1)
DCM/TFE (2:1)
DCM/TFE (2:1)
DCM/TFE (2:1)
DCM/TFE (2:1)
DCM/TFE (2:1)
TsOH·H2O
TsOH·H2O
TsOH·H2O
L-CSA
77
98
84
95
85
75
80
91
86
87
88
85
86
81
83
93
D-CSA
MeSO3H
EtSO3H
d
8
TsOH·H2O
a
Conditions: indole 3a (0.25 mmol), acid (0.25 mmol), Pd-
(OCOCF3)2 (2 mol%), (R)-BINAP (2.4 mol%), solvent (3.0 mL),
24 h, 50 °C. Isolated yield. Determined by HPLC. (R)-H8-BINAP
b
c
d
was used as the ligand.
isomer. A survey of sulfonic acids indicated that p-
toluenesulfonic acid (TsOH·H2O) was the best choice (Table
3, entry 3 vs entries 4−7). The ee value was enhanced to 93%
(Table 3, entry 8) when asymmetric hydrogenation was
performed with (R)-H8-BINAP as ligand. Therefore, the
optimized reaction conditions are as follows: Pd(OCOCF3)2/
(R)-H8-BINAP/TsOH·H2O/DCM:TFE = 2:1 at 50 °C.
Under the optimized conditions, a variety of 2,3-disustituted
indoles were subjected to asymmetric hydrogenation, and
excellent enantioselectivities and diastereoselectivities were
obtained (Table 4). 3-Benzyl-substituted indoles were slightly
inferior to 3-alkyl-substituted indoles in terms of ee for this
asymmetric hydrogenation (Table 4, entries 1−3 vs entries 4
and 5). The introduction of electron-donating and electron-
withdrawing groups on the phenyl ring of 3-benzyl was found
to not affect the reaction significantly (Table 4, entry 2 vs entry
3). To ensure the high enantioselectivity of indolines 4, a slight
modification was made, such as reducing the hydrogen pressure
Accordingly, to obtain enantiomerically pure products, the
protonation step should be much faster than the hydrogenation
step (k1 ≫ k2). A higher reaction temperature should accelerate
the protonation step, and a lower pressure of hydrogen gas
should decrease the rate of hydrogenation of the iminium
intermediate. Based on the theoretical analysis, a higher
temperature as well as a lower pressure of hydrogen gas
might guarantee the success of this process.
Thus, the reaction temperature of hydrogenation was
increased to 50 °C for further reaction optimization. Initial
solvent screening showed that the mixed solvent DCM/TFE
C
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
tate coordinated on the Pd catalyst. These two points are
crucial for Pd-catalyzed asymmetric hydrogenation of un-
protected indoles (for details, see the Mechanistic Study
section).
Table 4. Asymmetric Hydrogenation of 2,3-Disubstituted
Indoles 3
a
Product Elaboration. To demonstrate the utility of Pd-
catalyzed asymmetric hydrogenation of the unprotected
indoles, a direct and efficient method was developed to
synthesize the enantiopure indoline skeleton derivative 8, an
neuroleptic agent in many types of antidepressants.23 As shown
in Scheme 4, asymmetric hydrogenation of 2,3-disubstituted
b
c
entry
R1
R2/R3
yield (%)
ee (%)
1
2
3
4
5
6
H
Me/Bn
93 (4a)
94 (4b)
97 (4c)
81 (4d)
89 (4e)
93 (4f)
91 (4g)
96 (4h)
84 (4i)
93 (4j)
92 (4k)
95 (4l)
87 (4m)
82 (4n)
98 (4o)
87 (4p)
83 (4q)
<5
91
91
91
94
94
94
H
Me/4-MeOC6H4CH2
Me/4-FC6H4CH2
Me/CyCH2
H
Scheme 4. Synthesis of Biologically Active Product 8
H
H
Me/i-PrCH2
phenethyl/Bn
−(CH2)4−
−(CH2)5−
Me/Me
H
d
7
H
91 (R,R)
d
8
H
90 (R,R)
de
,
9
H
92 (R,R)
10
11
12
13
14
15
16
17
18
5-F
5-F
7-Me
7-Me
7-Me
7-Me
7-Me
7-Me
H
Me/Bn
87
92
Me/CyCH2
Me/Bn
97
Me/CyCH2
96
Me/2-MeC6H4CH2
Me/3-MeC6H4CH2
Me/4-MeC6H4CH2
−(CH2)4−
98
97
97
de
,
96 (R,R)
−
Me/CO2Et
a
Conditions: indole 3 (0.25 mmol), TsOH·H2O (0.25 mmol),
Pd(OCOCF3)2 (2 mol%), (R)-H8-BINAP (2.4 mol%), DCM/TFE
b
c
indole 3g gave the key intermediate hexahydrocarbazole 4g
with 88% ee. Subsequent protection with 2-chloroacetyl
chloride and a simple recrystallization afforded the intermediate
7 with 99% ee. Finally, N-alkylation gave the biologically active
compound 8 in high overall yields and up to 99.9% ee.
(3.0 mL, 2:1), 24 h, 50 °C. Isolated yield. Determined by HPLC.
d
e
Using L-CSA and solvent (DCM/TFE 1:1) at rt. 50 °C and H2 (300
psi).
(Table 4, entries 9 and 17). We found that ring-fused substrates
could also give satisfying results (Table 4, entries 7−9 and 17),
and indoles bearing a 5-F substituent displayed slightly lower ee
than other indoles (Table 4, entries 10 and 11). 7-Methyl-
substituted 2,3-disubstituted indoles showed remarkable
enantioselectivities (96−98% ee, Table 4, entries 12−17).
Unfortunately, 3-carbalkoxy-2-methylindole 3r could not be
hydrogenated under the standard conditions (Table 4, entry
18).
Asymmetric Hydrogenation of N-Protected Indoles.
N-Protected indoles were also subjected to hydrogenation
under the above standard conditions (Scheme 3). Unfortu-
nately, it was found that these substrates were not good
partners for Pd catalyst system. The low enantioselectivity
might be due to that fact that the protecting groups on the N-
atom prevent the generation of an iminium intermediate and/
or inhibit hydrogen-bonding interaction with the trifluoroace-
MECHANISTIC STUDY
■
Iminium Intermediate. Iminium and vinylogous iminium
have been found to be significant intermediates in numerous
organic reactions,24 and they are also cucial for the success of
the asymmetric hydrogenation of unprotected indoles.
Further confirmation of the presence of the iminium in the
hydrogenation reaction is required; therefore, iminium salt was
conveniently prepared with the equivalent 2-methylindole and
TsOH·H2O in CDCl3 after stirring 5 min at room temperature.
1H NMR investigation indicated that the signal of the hydrogen
at the 3-position became broad and shifted upfield, and the N−
H peak disappeared rapidly. When TsOH·H2O was exchanged
with D2O before being applied to this system, all the peaks
1
from the H NMR study in CDCl3 remained the same as
before, except for the partial disappearance of the 3-H of
iminium (Figure 1). Subsequent electrospray ionization mass
spectroscopic analysis of a similar solution showed peaks at m/
z+ 132.0833 and m/z− 171.0083, which agreed with the ion pair
of the iminium intermediate (eq 1). In this way, the iminium
salt generated before hydrogenation was identified.
Scheme 3. Asymmetric Hydrogenation of N-Protected
Indoles
D
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
The first step was generation of the iminium intermediate, and
then this active iminium intermediate was hydrogenated to
afford the desired product. When the asymmetric hydro-
1
genation was run in deuterated TFE, H NMR analysis of the
crude products showed that two deuterium atoms were
incorporated at the 3-position of 2-methylindoline (eq 2),
which suggested a fast, reversible process of protonation and
deprotonation. Thus, two deuterium atoms were incorporated
at the 3-position before hydrogenation. When 2-methylindole
was subjected to hydrogenation with D2, 2-deuterio-2-
methylindoline with 92% incorporation was obtained without
deuterium at the 3-position (eq 3). The above experimental
results suggested that the simple unprotected indoles can be
activated by a Brønsted acid to form iminium in situ, which can
then be hydrogenated by the chiral Pd catalyst.
Active Pd Hydride Species. Metal hydride active species
have been involved not only in hydrogenation but also in many
other organic reactions as the key intermediate. However,
several factors impede the thorough study of the metal hydride
intermediates.25 Generally, the metal hydrides are generated in
the reaction in only trace amounts, and the metal hydrides are
unstable in many conditions. Therefore, it is challenging to
isolate them and determine their conformations or properties.26
Given the importance of metal hydrides in such reactions,
especially in hydrogenation reactions, here we present the most
likely paths to generate Pd−H in the Pd-catalyzed asymmetric
hydrogenation, based on our experience.
1
Figure 1. H NMR study of TsOH·H2O or TsOD·D2O as well as 2-
methylindole in CDCl3.
Isotopic labeling experiments using deuterated solvent and
deuterium gas also affirmed the above conclusion (eqs 2 and 3).
As can be seen from Scheme 5 and Figure 2, in path A, first,
protonation of one CF3COO− ligand by strong acid TsOH
gives INT A1, and then the CF3COOH ligand is detached from
the Pd catalyst to generate an unoccupied site, INT A2. H2
combines with this active center (INT A3) and is activated with
Scheme 5. Proposed Pathways for Reaction of Hydrogen Gas and Pd Catalyst To Generate Pd−H Species
E
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
CF3COO−. The transition states marked with * in Figure 2 are
for the ligand coordination or detachment steps. Harvey has
recently suggested that such processes correspond to diffusion
control,28 with free energy barriers of ca. 4.5 kcal/mol, and are
not expected to be competitive with the rate-determining
barriers. The key transition state TS A is shown in Figure 3. It is
Figure 2. Computed reaction energy profiles for path A (with the help
of TsOH) and path B (strong acid is not involved) to generate
palladium hydride in TFE. Relative energies in kcal/mol are given.
assistance from Pd and the coordinated CF3COO− ligand
(TS A). The last step is the ligand exchange (or proton
exchange) between CF3COO− and the coordinated
CF3COOH (INT A4) to offer the Pd−H species 10. In path
B, one CF3COO− ligand is directly detached from the Pd
catalyst to generate an unoccupied site, INT A2. The following
steps are the same as in path A. In path C, H2 coordinates to Pd
catalyst first (INT C1) and then is activated by Pd and a
CF3COO− ligand. Pd−H species 10 is generated with the
remaining coordinated CF3COOH. Path D is similar to path C,
except that CF3COO− is on the apical position in INT D1,
while H2 is on the apical position in INT C1. H2 is activated by
Pd and a CF3COO− ligand. In path E, H2 is activated by the Pd
catalyst through oxidative addition, followed by reductive
elimination of CF3COOH to generate the Pd−H species 10. In
path F, the first step is similar to that in path B, but H2 is
activated by the unsaturated Pd catalyst center through
oxidative addition. Finally, the Pd−H species is generated
after release of CF3COOH. Paths A, B, and F are ionic
pathways, while paths C, D, and E are not. Paths C and D may
suffer due to the five-coordinated TS C, INT C1, and INT D1,
which are not common for transition metals with d8
configuration. Paths E and F may suffer due to the oxidative
addition, since H2 may not be able to oxidize Pd(II) to Pd(IV).
DFT calculations were carried out to elucidate which is the
most facile path,27 using TFE as the solvent (B3LYP; for
details, see Supporting Information). First, the rate-determining
barrier of path C is 32.9 kcal/mol. The energies of INT D1
(30.3 kcal/mol), INT E1 (46.6 kcal/mol), and INT F1 (44.8
kcal/mol) were found to be too high. Therefore, paths C−F
were ruled out. All rate-determining barriers mentioned in this
section were relative to the most stable structure prior to the
rate-determining step, which means INT A2 for path A and
reactant Pd catalyst 9 for all other paths. The reaction profiles
of paths A and B are shown in Figure 2. We found that path A
is preferred, with the rate-determining barrier of 19.2 kcal/mol.
Protonation of the Pd catalyst by TsOH requires 3.8 kcal/mol.
Releasing the CF3COOH ligand gains 8.0 kcal/mol and offers
an unoccupied site on Pd ion, which facilitates the H2
coordination and activation. The H2 activation step is the
rate-determining step. The barrier of path B is higher than that
of path A by 13.8 kcal/mol(ΔGTS B − ΔGTS A), which arises
from the thermodynamic stability difference between TsO− and
Figure 3. Computed structure of TS A. Bond lengths are given in Å.
an earlier transition state (H−H = 0.94 Å in TS A) with square-
planar structure as one would expect for d8 transition metals.
The transition state (TS C) of the non-ionic path C is shown in
Figure 4. It has a square-pyramid structure, with the weakly
coordinated CF3COO− (Pd−O = 2.52 Å) on the apical
position.
1
Toward lifting the veil of active Pd−H, H NMR was at the
top of our list. As to the equilibrium between Pd catalyst and
Pd−H species (Scheme 6), the low concentration of Pd−H
met our expectation, but also made it difficult to study. Under
Figure 4. Computed structure of TS C. Bond lengths are given in Å.
F
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
1
(Scheme 7, entry 1 vs entry 3). The results of these
experiments suggest that chiral the Pd catalyst has a relatively
impassive behavior to air and water. To explain this, we
proposed that the Pd−H generated from Pd(OCOCF3)2 under
H2 atmosphere could be oxidized to hydrogen peroxide or
water and Pd(II) by oxygen in the presence of trifluoroacetic
acid (Scheme 8), which has been reported by Sigman and
Scheme 6. Generation and H NMR Spectrum of Pd−H
Scheme 8. Proposed Reasons for the Stability of Pd Catalyst
to O2
rigidly controlled conditions, fortunately, some proofs were
1
obtained. A diagnostic virtual triplet in the H NMR spectrum
at −9.22 ppm (J = 84.4 Hz) (Scheme 6) and the chemical shift
were consistent with most of the metal hydrides reported.26
Since th coupling constant was between those of hydrogen
coupling with cis-phosphorus (J = 5−16 Hz) and hydrogen
coupling with trans-phosphorus (J ≈ 160 Hz), it was inferred
that the hydrogen in Pd−H couples with phosphorus with an
angle between the cis and trans positions. This could be
ascribed to equilibrium between the reversible reductive
elimination of TFA and some other intermediates (Scheme
6). Also, an interesting phenomenon was observed in studying
Pd−H species. When the chiral palladium catalyst [Pd-
(OCOCF3)2(R)-BINAP] was put in the TFE, the solution
was yellow. After being subjected to hydrogen gas, the solution
turned from yellow to red, and deep red about 15 min later.
This color does not fade over time or upon subjecting the
solution to oxygen.
Stahl.25f,g,i−k But it was not ruled out that the Pd−H complex is
simply unreactive toward O2. Actually, the formed Pd(II)
species is our hydrogenation catalyst precursor, so the Pd−H
can be regenerated. Therefore, the Pd catalyst is relatively
insensitive to air, acid, and water.
Roles of Solvent Trifluoroethanol. Fluorinated alcohols,
especially hexafluoroisopropanol (HFIP) and TFE, are used as
solvents, cosolvents, and additives in many reactions, given
their unique properties.29 These properties, which made
fluorinated alcohols a range of robust solvents, were usually
considered to be high hydrogen bond donor ability, low
nucleophilicity, high ionizing power, ability to solvate water,
and so on. TFE, as one of the inexpensive fluorinated alcohols,
has been commonly applied in synthetic chemistry.
Recently, asymmetric hydrogenation has gained much
interest because of its superb ability to construct chiral
compounds. Our group contributed many examples to this
thriving area. When we reviewed these hydrogenation reactions,
an interesting phenomenon was noticed: excellent enantio-
selectivity and high activity are generally obtained in TFE in the
Pd-catalyzed asymmetric hydrogenations. According to the
literature and our experiences, two conclusions might be drawn:
(1) TFE participates in the formation of some key
intermediates,30 or (2) TFE stabilizes these intermediates. As
can be seen from Scheme 9, TFE participated in the exchange
Robustness of Pd Catalyst to Air and Water. In the
process of running experiments, we found that the chiral Pd-
catalyzed asymmetric hydrogenation is relatively robust to air,
acid, and water. In order to study this issue further, several
controlled experiments were designed, as shown in Scheme 7
Scheme 7. Controlled Experiments on the Influence of Air
and H2O
Scheme 9. Exchange of the Counterion of Pd−H Complex
and TFE
(Table 1, entry 18 is equal to Scheme 7, entry 1). First, we
performed the hydrogenation in the atmosphere without the
Schlenk technique (Scheme 7, entry 2); i.e., the catalyst was
prepared in the air and the substrate, Brønsted acid, solvent,
and the catalyst were also mixed in the air. The experimental
results confirmed our deduction that the hydrogenation went
on well with similar activity and enantioselectivity (Scheme 7,
entry 1 vs entry 2). Second, we added water to the reaction
solution (Scheme 7, entry 3). No difference was noticed until 4
equiv of water relative to substrate was added to the reaction
of the counterion of the Pd catalyst before or after the
formation of Pd−H complex. Since CF3CH2O− is more
nucleophilic than CF3COO−, this exchange seemed reasonable.
But when we subjected the Pd catalyst to H2 in TFE, followed
by in situ IR, we found there is barely a peak at 1790 cm−1,
which is the characteristic peak of the carbonyl group of
CF3COOH.
The role of TFE has also been studied theoretically. First we
studied whether TFE was involved in the reaction directly by
replacing one CF3COO− with CF3CH2O−. We only considered
G
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 10. Possible TFE-Participating Pathway for Reaction of H2 and Pd Catalyst To Generate Pd−H Species
path A because it is much more facile than the other paths. The
new path is path G, as shown in Scheme 10. Replacing
CF3COO− with CF3CH2O− was found to be endothermic (ΔG
= 7.8 kcal/mol), which is consistent with the fact that
CF3COOH has not been detected experimentally. The rate-
determining barrier (relative to INT A2, which is the most
stable structure prior to the rate-determining step) is calculated
to be 29.9 kcal/mol, higher than that of the corresponding non-
TFE-participating path by more than 10 kcal/mol. Therefore,
the theoretical results do not support that TFE participates in
the Pd−H formation reaction directly.
Since we ruled out the TFE-participating paths, we began to
accept the hypothesis that TFE stabilizes some key
intermediates or transition states. This could be true since
TFE is a highly polar solvent (ε = 26.5) which is able to
stabilize the ions in ionic paths such as path A, and it is
inevitable that the barrier of path A is higher in a less polar
solvent. To test this hypothesis, we got the reaction profiles for
path A in CH2Cl2 (ε = 8.93) for comparison (Figure 5). As
Scheme 11. Possible Interaction of the TFE and Pd−H
Species
phenomenon was also observed in many other metal hydride
species, such as Ru−H and Ni−H.31a,c,32 It is reported that
Ru−H and Ni−H could exchange proton with alcohol solvent
(Scheme 12), but this exchange was not observed in the indole
hydrogenation because there was no deuteron on the 2-C when
d3-TFE was subjected to hydrogenation (see eq 2).
Scheme 12. Possible Proton Exchange of the TFE and Pd−H
Species
Proposed Catalytic Cycle of Indole Hydrogenation.
The catalytic mechanisms for hydrogenation reactions can be
classified as (i) inner-sphere or outer-sphere mechanisms on
the basis of the interaction between substrate and catalyst, (ii)
ionic and non-ionic mechanisms according to the charge
separation in the intermediates and transition states, as well as
(iii) stepwise and associated mechanisms on the basis of the
nature of the reaction pathway. In our search for the most
plausible pathway for the indole hydrogenation, we have
evaluated four paths including both inner-sphere and outer-
sphere, ionic and non-ionic, stepwise and associated paths. As
can be seen from Schemes 13−15 as well as Figures 6 and 7,
path I (Scheme 13, Figure 7), first indole 1a is protonated by
TsOH to give iminium α, and then iminium α combines with
Pd−H to generate INT 11b. The next step is the hydride
transfer from Pd−H to α to generate INT 11c, which
subsequently separates into product 2a and INT A2 and then
follows path A to get the active catalyst Pd−H. This is an outer-
sphere and ionic hydrogenation pathway. Alternatively, 1a can
also be protonated by CF3COOH (generated in the Pd−H
formation step) instead of TsOH.
Figure 5. Computed reaction energy profiles for path A to generate
palladium hydride in TFE and CH2Cl2. Relative free energies in kcal/
mol are given.
expected, we found that in CH2Cl2 the rate-determining barrier
of path A is 24.2 kcal/mol (relative to Pd catalyst 9), which is
TSA
CH2Cl2
TSA
CF3CH2OH
9.2 kcal/mol higher than in TFE (ΔG
− ΔG
).
The ΔG of all other ionic intermediates also increases by 8−10
kcal/mol. On the basis of these calculations, we proposed that
the unique properties of TFE and other fluorinated alcohols
stem from their high polarity, which is able to stabilize the ionic
intermediates during the catalytic reaction. Generally, the
heterolytic splitting of a bond is very difficult and is usually
facilitated by the strong solvation effect of the ions. For path A,
the barrier is 89.4 kcal/mol in the gas phase, 24.2 kcal/mol in
CH2Cl2, and 19.2 kcal/mol in TFE, which makes it facile under
room temperature and high H2 pressure.
Path II (Scheme 14), protonation of coordinated CF3COO−
by TsOH to give INT 12a, followed by associated proton and
hydride transfer to give 2a and INT A2, involves a six-
membered-ring transition state. This is an outer-sphere and
non-ionic pathway. Path III is similar to path II except that
CF3COOH coordinates to Pd by CO instead of the hydroxyl
group. Accordingly, the associated hydrogenation has an eight-
membered-ring transition state.
As a strong hydrogen bond donor,15a,31 the hydrogen atom
of TFE can interact with the Pd−H (Scheme 11). This
H
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 13. Proposed Mechanism for the Asymmetric
Hydrogenation of Indoles
Scheme 15. Inner-Sphere Pathway for Indole Asymmetric
Hydrogenation
Scheme 14. Associated Pathways II and III for Indole
Hydrogenation
the coordinated iminium. Path IV is an inner-sphere and non-
ionic pathway.
All four pathways mentioned above have been studied
computationally, and the reaction profiles in TFE are shown in
Figure 6. Path I was found to be the most plausible one.
Figure 6. Computed reaction energy profiles for all paths to achieve
asymmetric hydrogenation of unprotected indoles in TFE. Relative
energies in kcal/mol are given.
Protonation of 1a by TsOH is thermodynamically favorable
(ΔG = −5.8 kcal/mol), and the rate-determining step is the
hydride transfer. As discussed above, all rate-determining
barriers were relative to the most stable structure prior to the
rate-determining step, which means α for path I, INT 13a for
path III, and 10 for the other paths. The barriers for formation
of (R)-2a and (S)-2a are 18.4 and 22.7 kcal/mol, respectively
(Figure 7). The lower barrier for formation of (R)-2a is in good
agreement with the high enantioselectivity observed exper-
imentally. In CH2Cl2, both barriers are slightly lower (17.3
kcal/mol for (R)-2a and 21.6 kcal/mol for (S)-2a) than in TFE.
This is because the protonation of 1a by TsOH is already quite
facile in CH2Cl2, while TFE overstabilizes the protonated
intermediates and lowers its activity.
In path IV (Scheme 15), Pd−H addition to the CC bond
of 1a generates INT 14 first, and then the product 2a and
INT A2 are generated by proton transfer from CF3COOH to
To confirm that the observed enantioselectivity is not
method-dependent, the two key transition states leading to R
I
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
and there is actually an eight-membered-ring transition state.
Therefore, these transition states obey the “three-point
interaction model”: the N−H···O hydrogen bond, the Pd···
H···C bond, and the steric interaction between the ligand and
the substrate. As shown in Figure 8, the steric interactions come
mainly from CH3 of the substrate and H(Ph) of the chiral
ligand (rH···H = 2.50 Å in TS 11-1-R and 2.43 Å in TS 11-1-S),
as well as CH2 of the substrate and Ph of the chiral ligand (the
distance between H(CH2) and center of Ph is 3.15 Å for
TS 11-1-R and 2.96 Å for TS 11-1-S). We do see a larger steric
interaction for TS 11-1-S. In summary, despite the bond
interaction and the steric interaction that exist for common
stepwise mechanisms, here the hydrogen bond acts as the third
point, and is expected to be crucial for the high enantio-
selectivity. This deduction is also demonstrated by the results of
hydrogenation of N-protected indoles, in which either low
activity or poor enantioselectivity was obtained due to the
absence of the hydrogen bond between N−H of iminium and
oxygen of coordinated trifluoroacetate (Scheme 3).
Figure 7. Computed reaction energy profiles for path I to achieve
asymmetric hydrogenation of unprotected indoles in TFE. Relative
free energies in kcal/mol are given.
Paths II and III feature associated hydrogenation, which is
usually thought to be found only for asymmetric reactions. The
protonation costs 10.1 kcal/mol for path II and −3.5 kcal/mol
for path III. The rate-determining steps for paths II and III are
the associated hydrogenation steps, with the barrier of 59.2
kcal/mol (relative to 10) for path II and 57.9 kcal/mol (relative
to INT 13a) for path III (transition-state structures are shown
in Figures 9 and 10). Our calculations suggest that both
and S isomers were reoptimized using the M06-2X functional.
We found that TS-R is lower in energy (ΔG in CF3CH2OH)
than TS-S by 3.5 kcal/mol, which is close to the 4.3 kcal/mol
found by the B3LYP functional.
The indole hydrogenation described by this work does not
occur without the presence of the strong acids such as L-CSA,
TsOH, and MeSO3H, etc., in contrast to the asymmetric
hydrogenation of activated imines catalyzed by Pd complex-
es.15e Using TsOH as the example, we have shown that the
strong acid is involved in both the Pd−H generation and the
hydrogenation. Without strong acid, indole can be protonated
by the trifluoroacetic acid generated in the Pd−H formation
step (ΔG = 8.0 kcal/mol). The barriers for formation of (R)-2a
and (S)-2a are 26.4 and 30.7 kcal/mol, respectively, both are
much higher than those in path I.
Theoretically, a “three-point interaction model” 33 was
usually used to explain the high enantioselectivity. According
to this model, in order to have enantioselectivity, there must be
at least three interactions between the catalyst and the substrate
in the transition state, and at least one of those interactions
must be stabilizing. Based on this model, in general one would
expect that an associate pathway is a better mechanism for the
enantioselectivity, and the three points are one steric
interaction and two bonding interactions. To understand why
path I, a stepwise path, has very good enantioselectivity, we
show the transition-states structures (TS 11-1-R and TS 11-1-
S) for hydride transfer in Figure 8. Interestingly, we found that
both transition states have a strong hydrogen bond between
N−H of iminium and oxygen of the coordinated CF3COO−,
Figure 9. Computed structure of TS 12. Bond lengths are given in Å.
associated pathways degraded to stepwise pathways, where
proton transfer to indole is followed by hydride transfer to
iminium. Compared with path I, the hydride transfer is the
same, while the proton transfer is not competitive because
indole is more easily protonated (costs −5.8 kcal/mol) than the
coordinated CF3COO−. For path IV, the barrier for Pd−H
addition is 25.0 kcal/mol. However, the following proton
transfer is much more difficult (energy barrier is 47.0 kcal/mol
relative to 10); therefore, this pathway is not preferred.
Generally, Pd-catalyzed asymmetric hydrogenation includes
some important features: (1) Activation of hydrogen gas is a
heterolytic process assisted by coordinated trifluoroacetate. (2)
This is an ionic hydrogenation process. The hydride is from the
hydrogen gas, while the proton is from a Brønsted acid. (3)
Figure 8. Computed structures of TS 11-1 (R and S). Bond lengths
are given in Å.
J
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
observed experimentally was reproduced by the calculation, and
is proposed to arise from the hydrogen-bonding interaction
between N−H of iminium salt and oxygen of the coordinated
trifluoroacetate ligand in the eight-membered-ring transition
state for hydride transfer. So, active chiral Pd complex is a
typical bifunctional catalyst, leading to asymmetric hydro-
genation and hydrogen-bonding interaction between coordi-
nated trifluoroacetate ligand of Pd catalyst and N−H of
iminium salt. The above mechanistic study may provide some
useful hints for chiral catalysis design and extend the scope of
Pd-catalyzed asymmetric hydrogenation.
ASSOCIATED CONTENT
* Supporting Information
■
S
Complete experimental procedures and characterization data
for the prepared compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
fanhj@dicp.ac.cn
■
Figure 10. Computed structure of TS 13. Bond lengths are given in Å.
ygzhou@dicp.ac.cn
The mechanism is an outer-sphere hydrogenation mechanism.
(4) The catalyst is bifunctional, leading to asymmetric
hydrogenation and hydrogen-bonding interaction between
trifluoroacetate of the Pd catalyst and iminium salt in the
eight-membered-ring transition state for hydride transfer. (5)
The highly polar solvent TFE stabilizes the ionic intermediates
in active Pd−H species generation.
Author Contributions
†Y.D. and L.L. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was supported by by National Natural Science
Foundation of China (21125208, 21032003) and National
Basic Research Program of China (2010CB833300). This paper
is dedicated to Prof. Li-Xin Dai on the occasion of his 90th
birthday.
CONCLUSION
■
In conclusion, we report an extensive substrate scope for
asymmetric hydrogenation of indoles, including 2-substituted
and 2,3-disubstituted indoles. In this process, an equivalent of
strong Brønsted acid was necessary to complete the reaction,
giving up to 98% ee. This methodology was also applied in
product elaboration for the synthesis of the biologically active
indoline skeleton. Notably, a drawback is the low activity and/
or enantioselectivity obtained for the 2-aryl-substituted and
electron-withdrawing group located at the 2-position of the
indoles. Given the compatibility of a Pd catalyst in this reaction,
we also probed the properties of the Pd catalyst and the
mechanism of indole hydrogenation by combining experimen-
tal studies and DFT calculations. Iminium salt was demon-
strated to be the significant active intermediate in this reaction
through the isotope-labeled reactions and ESI-HRMS. The
strong acid protonates the CC double bond of indole much
easier, and therefore can remarkably decrease the energy
REFERENCES
■
(1) (a) Southon, I. W.; Buckingham, J. Dictionary of Alkaloids;
Chapman and Hall: New York, 1989. (b) Neuss, N.; Neuss, M. N. In
The Alkaloids; Brossi, A., Suffness, M., Eds.; Academic Press: San
Diego, 1990; p 229. (c) Gueritte, F.; Fahy, J. In Anticancer Agents from
Natural Products; Cragg, G. M., Kingstom, D. G. I., Newman, D. J.,
Eds.; CRC Press: Boca Raton, FL, 2005; p 123. (d) Modern Alkaloids;
Fattorusso, E., Taglialatela-Scafati, O., Eds.; Wiley-VCH: Weinheim,
2008, and references therein.
(2) For recent reviews on asymmetric hydrogenation of aromatic
compounds, see: (a) Zhao, D.; Glorius, F. Angew. Chem., Int. Ed. 2013,
52, 9616. (b) Wang, D.-S.; Chen, Q.-A.; Lu, S.-M.; Zhou, Y.-G. Chem.
Rev. 2012, 112, 2557. (c) Tang, W.; Zhang, X. Chem. Rev. 2003, 103,
3029. (d) Kuwano, R. Heterocycles 2008, 76, 909. (e) Zhou, Y.-G. Acc.
Chem. Res. 2007, 40, 1357. (f) Lu, S.-M.; Han, X.-W.; Zhou, Y.-G.
Chin. J. Org. Chem. 2005, 25, 634. (g) Glorius, F. Org. Biomol. Chem.
2005, 3, 4171. For recent examples, see: (h) Ye, Z.-S.; Guo, R.-N.;
Cai, X.-F.; Chen, M.-W.; Shi, L.; Zhou, Y.-G. Angew. Chem., Int. Ed.
2013, 52, 3685. (i) Iimuro, A.; Yamaji, K.; Kandula, S.; Nagano, T.;
Kita, Y.; Mashima, K. Angew. Chem., Int. Ed. 2013, 52, 2046. (j) Ye, Z.-
S.; Chen, M.-W.; Chen, Q.-A.; Shi, L.; Duan, Y.; Zhou, Y.-G. Angew.
Chem., Int. Ed. 2012, 51, 10181. (k) Urban, S.; Beiring, B.; Ortega, N.;
Paul, D.; Glorius, F. J. Am. Chem. Soc. 2012, 134, 15241. (l) Kuwano,
R.; Morioka, R.; Kashiwabara, M.; Kameyama, N. Angew. Chem., Int.
Ed. 2012, 51, 4136. (m) Shi, L.; Ye, Z.-S.; Cao, L.-L.; Guo, R.-N.; Hu,
Y.; Zhou, Y.-G. Angew. Chem., Int. Ed. 2012, 51, 8286. (n) Wang, T.;
Zhuo, L.-G.; Li, Z. W.; Chen, F.; Ding, Z.; He, Y.; Fan, Q.-H.; Xiang,
J.; Yu, Z.-X.; Chan, A. S. C. J. Am. Chem. Soc. 2011, 133, 9878.
(o) Urban, S.; Ortega, N.; Glorius, F. Angew. Chem., Int. Ed. 2011, 50,
3803. (p) Ortega, N.; Urban, S.; Beiring, B.; Glorius, F. Angew. Chem.,
1
barrier. According to the H NMR spectrum of the Pd−H
species, which was considered to be the active species, proton
exchange between the Pd−H active species and solvent TFE
could not occur. We have shown that in TFE the Pd−H active
catalyst is generated by an ionic mechanism in which H2 was
split by Pd and trifluoroacetate via a six-membered-ring
transition state. The reaction is only facile at room temperature
with the help of the strong Brønsted acid. Highly polar solvent
such as TFE is able to stabilize the ionic intermediates, making
it better than the other solvents. We also propose that the Pd-
catalyzed indole hydrogenation of unprotected indoles under-
goes a stepwise, ionic, and outer-sphere hydrogenation
mechanism,20,21 which features protonation of indole by a
strong Brønsted acid and hydride transfer from active Pd−H to
a protonated indole intermediate. The high enantioselectivity
K
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Int. Ed. 2012, 51, 1710. (q) Ortega, N.; Tang, D.-T. D.; Urban, S.;
Zhao, D.; Glorius, F. Angew. Chem., Int. Ed. 2013, 52, 9500.
(3) (a) Kuwano, R.; Sato, K.; Kurokawa, T.; Karube, D.; Ito, Y. J. Am.
Chem. Soc. 2000, 122, 7614. (b) Kuwano, R.; Kashiwabara, M.; Sato,
K.; Ito, T.; Kaneda, K.; Ito, Y. Tetrahedron: Asymmetry 2006, 17, 521.
(4) Kuwano, R.; Kaneda, K.; Ito, T.; Sato, K.; Kurokawa, T.; Ito, Y.
Org. Lett. 2004, 6, 2213.
(5) Kuwano, R.; Kashiwabara, M. Org. Lett. 2006, 8, 2653.
(6) Mrsic, N.; Jerphagnon, T.; Minnaard, A. J.; Feringa, B. L.; de
Vries, J. G. Tetrahedron: Asymmetry 2010, 21, 7.
(20) (a) Zhu, Y.; Burgess, K. Acc. Chem. Res. 2012, 45, 1623.
(b) Zhou, J.; Ogle, J. W.; Fan, Y.; Banphavichit, V.; Zhu, Y.; Burgess,
K. Chem.Eur. J. 2007, 13, 7162. (c) Perry, M. C.; Cui, X. H.; Powell,
M. T.; Hou, D. R.; Reibenspies, J. H.; Burgess, K. J. Am. Chem. Soc.
2003, 125, 113. (d) O, W. W. N.; Lough, A. J.; Morris, R. H.
Organometallics 2012, 31, 2152. (e) Maeda, T.; Makino, K.; Iwasaki,
M.; Hamada, Y. Chem.Eur. J. 2010, 16, 11954. (f) Hopmann, K. H.;
Bayer, A. Organometallics 2011, 30, 2483. (g) Fan, Y. B.; Cui, X. H.;
Burgess, K.; Hall, M. B. J. Am. Chem. Soc. 2004, 126, 16688.
(h) Dobereiner, G. E.; Nova, A.; Schley, N. D.; Hazari, N.; Miller, S. J.;
Eisenstein, O.; Crabtree, R. H. J. Am. Chem. Soc. 2011, 133, 7547.
(21) (a) Sandoval, C. A.; Ohkuma, T.; Utsumi, N.; Tsutsumi, K.;
Murata, K.; Noyori, R. Chem.Asian J. 2006, 1, 102. (b) Sandoval, C.
A.; Ohkuma, T.; Muniz, K.; Noyori, R. J. Am. Chem. Soc. 2003, 125,
13490. (c) Phillips, S. D.; Fuentes, J. A.; Clarke, M. L. Chem.Eur. J.
2010, 16, 8002. (d) Kitamura, M.; Tsukamoto, M.; Bessho, Y.;
Yoshimura, M.; Kobs, U.; Widhalm, M.; Noyori, R. J. Am. Chem. Soc.
2002, 124, 6649. (e) Girard, C.; Genet, J. P.; Bulliard, M. Eur. J. Org.
Chem. 1999, 2937. (f) Ashby, M. T.; Halpern, J. J. Am. Chem. Soc.
1991, 113, 589.
(22) (a) Roucoux, A.; Thieffry, L.; Carpentier, J. F.; Devocelle, M.;
Meliet, C.; Agbossou, F.; Mortreux, A. Organometallics 1996, 15, 2440.
(b) Kless, A.; Borner, A.; Heller, D.; Selke, R. Organometallics 1997,
16, 2096. (c) Imamoto, T.; Tamura, K.; Zhang, Z.; Horiuchi, Y.;
Sugiya, M.; Yoshida, K.; Yanagisawa, A.; Gridnev, I. D. J. Am. Chem.
Soc. 2012, 134, 1754. (d) Heller, D.; Kadyrov, R.; Michalik, M.; Freier,
T.; Schmidt, U.; Krause, H. W. Tetrahedron: Asymmetry 1996, 7, 3025.
(e) Haag, D.; Runsink, J.; Scharf, H. D. Organometallics 1998, 17, 398.
(f) Gridnev, I. D.; Imamoto, T.; Hoge, G.; Kouchi, M.; Takahashi, H. J.
Am. Chem. Soc. 2008, 130, 2560. (g) Gridnev, I. D.; Higashi, N.;
Asakura, K.; Imamoto, T. J. Am. Chem. Soc. 2000, 122, 7183.
(h) Fabrello, A.; Bachelier, A.; Urrutigoity, M.; Kalck, P. Coord. Chem.
Rev. 2010, 254, 273.
(23) (a) Ettel, V.; Myska, J. Chem. Abstr. 1956, 5619. (b) Ettel, V.;
Myska, J. Collect. Czech. Chem. Commun. 1956, 21, 473. (c) Zhao, H.;
He, X. S.; Thurkauf, A.; Hoffman, D.; Kieltyka, A.; Brodbeck, R.;
Primus, R.; Wasley, J. W. F. Bioorg. Med. Chem. Lett. 2002, 12, 3111.
(d) Zhao, H.; Thurkauf, A.; He, X. S.; Hodgetts, K.; Zhang, X. Y.;
Rachwal, S.; Kover, R. X.; Hutchison, A.; Peterson, J.; Kieltyka, A.;
Brodbeck, R.; Primus, R.; Wasley, J. W. F. Bioorg. Med. Chem. Lett.
2002, 12, 3105.
(24) (a) Chen, C.-B.; Wang, X.-F.; Cao, Y.-J.; Cheng, H.-G.; Xiao,
W.-J. J. Org. Chem. 2009, 74, 3532. (b) Hinman, R. L.; Shull, E. R. J.
Org. Chem. 1961, 26, 2339. (c) Ballini, R.; Palmieri, A.; Petrini, M.;
Shaikha, R. R. Adv. Synth. Catal. 2008, 350, 129. (d) Li, Y.; Shi, F.-Q.;
He, Q.-L.; You, S.-L. Org. Lett. 2009, 11, 3182. (e) Zheng, B.-H.; Ding,
C.-H.; Hou, X.-L.; Dai, L.-X. Org. Lett. 2010, 12, 1688. (f) Jing, L.;
Wei, J.; Zhou, L.; Huang, Z.; Li, Z.; Wu, D.; Xiang, H.; Zhou, X.
Chem.Eur. J. 2010, 16, 10955.
(7) Maj, A. M.; Suisse, I.; Meliet, C.; Agbossou-Niedercorn, F.
Tetrahedron: Asymmetry 2010, 21, 2010.
(8) Baeza, A.; Pfaltz, A. Chem.Eur. J. 2010, 16, 2036.
(9) Wang, D.-S.; Chen, Q.-A.; Li, W.; Yu, C.-B.; Zhou, Y.-G.; Zhang,
X. J. Am. Chem. Soc. 2010, 132, 8909.
(10) Wang, D.-S.; Tang, J.; Zhou, Y.-G.; Chen, M.-W.; Yu, C.-B.;
Duan, Y.; Jiang, G.-F. Chem. Sci. 2011, 2, 803.
(11) Duan, Y.; Chen, M.-W.; Ye, Z.-S.; Wang, D.-S.; Chen, Q.-A.;
Zhou, Y.-G. Chem.Eur. J. 2011, 17, 7193.
(12) Duan, Y.; Chen, M.-W.; Chen, Q.-A.; Yu, C.-B.; Zhou, Y.-G.
Org. Biomol. Chem. 2012, 10, 1235.
(13) Recently, Chen and co-workers developed an asymmetric
hydrosilylation of indoles, using combined Lewis base along with
Brønsted acid activation to ensure full conversion and high
enantioselectivity: Xiao, Y.-C.; Wang, C.; Yao, Y.; Sun, J.; Chen, Y.-
C. Angew. Chem., Int. Ed. 2011, 50, 10661.
(14) Chen, Q.-A.; Ye, Z.-S.; Duan, Y.; Zhou, Y.-G. Chem. Soc. Rev.
2013, 42, 497.
(15) (a) Abe, H.; Amii, H.; Uneyama, K. Org. Lett. 2001, 3, 313.
(b) Suzuki, A.; Mae, M.; Amii, H.; Uneyama, K. J. Org. Chem. 2004,
69, 5132. (c) Wang, Y.-Q.; Zhou, Y.-G. Synlett 2006, 1189. (d) Yang,
Q.; Shang, G.; Gao, W.; Deng, J.; Zhang, X. Angew. Chem., Int. Ed.
2006, 45, 3832. (e) Wang, Y.-Q.; Lu, S.-M.; Zhou, Y.-G. J. Org. Chem.
2007, 72, 3729. (f) Wang, Y.-Q.; Yu, C.-B.; Wang, D.-W.; Wang, X.-B.;
Zhou, Y.-G. Org. Lett. 2008, 10, 2071. (g) Yu, C.-B.; Wang, D.-W.;
Zhou, Y.-G. J. Org. Chem. 2009, 74, 5633. (h) Chen, M.-W.; Duan, Y.;
Chen, Q.-A.; Wang, D.-S.; Yu, C.-B.; Zhou, Y.-G. Org. Lett. 2010, 12,
5075. (i) Zhou, X.-Y.; Bao, M.; Zhou, Y.-G. Adv. Synth. Catal. 2011,
353, 84. (j) McLaughlin, M.; Belyk, K.; Chen, C.-Y.; Linghu, X.; Pan,
J.; Qian, G.; Reamer, R. A.; Xu, Y. Org. Process Res. Dev. 2013, 17,
1052.
(16) (a) Yu, C.-B.; Gao, K.; Wang, D.-S.; Shi, L.; Zhou, Y.-G. Chem.
Commun. 2011, 47, 5052. (b) Yu, C.-B.; Gao, K.; Chen, Q.-A.; Chen,
M.-W.; Zhou, Y.-G. Tetrahedron Lett. 2012, 53, 2560.
(17) (a) Raja, R.; Thomas, J. M.; Jones, M. D.; Johnson, B. F. G.;
Vaughan, D. E. W. J. Am. Chem. Soc. 2003, 125, 14982. (b) Wang, Y.-
Q.; Lu, S.-M.; Zhou, Y.-G. Org. Lett. 2005, 7, 3235. (c) Goulioukina,
N. S.; Bondarenko, G. N.; Bogdanov, A. V.; Gavrilov, K. N.;
Beletskaya, I. P. Eur. J. Org. Chem. 2009, 510. (d) Wang, C.; Yang, G.;
Zhuang, J.; Zhang, W. Tetrahedron Lett. 2010, 51, 2044. (e) Teng, B.;
Zheng, J.; Huang, H.; Huang, P. Chin. J. Chem. 2011, 29, 1312.
(f) Zhou, X.-Y.; Wang, D.-S.; Bao, M.; Zhou, Y.-G. Tetrahedron Lett.
2011, 52, 2826.
(18) (a) Drago, D.; Pregosin, P. S. Organometallics 2002, 21, 1208.
(b) Jones, M. D.; Raja, R.; Thomas, J. M.; Johnson, B. F. G.; Lewis, D.
W.; Rouzaud, J.; Harris, K. D. M. Angew. Chem., Int. Ed. 2003, 42,
4326. (c) Tsuchiya, Y.; Hamashima, Y.; Sodeoka, M. Org. Lett. 2006, 8,
4851. (d) Monguchi, D.; Beemelmanns, C.; Hashizume, D.;
Hamashima, Y.; Sodeoka, M. J. Organomet. Chem. 2008, 693, 867.
(25) (a) Grushin, V. V. Chem. Rev. 1996, 96, 2011. (b) Hills, I. D.;
Fu, G. C. J. Am. Chem. Soc. 2004, 126, 13178. (c) Konnick, M. M.;
Gandhi, B. A.; Guzei, I. A.; Stahl, S. S. Angew. Chem., Int. Ed. 2006, 45,
2904. (d) Nielsen, R. J.; Goddard, W. A. J. Am. Chem. Soc. 2006, 128,
9651. (e) Popp, B. V.; Stahl, S. S. J. Am. Chem. Soc. 2007, 129, 4410.
(f) Konnick, M. M.; Stahl, S. S. J. Am. Chem. Soc. 2008, 130, 5753.
(g) Popp, B. V.; Stahl, S. S. Chem.Eur. J. 2009, 15, 2915.
(h) Decharin, N.; Stahl, S. S. J. Am. Chem. Soc. 2011, 133, 5732.
(i) Sigman, M. S.; Jensen, D. R. Acc. Chem. Res. 2006, 39, 221.
(j) Gligorich, K. M.; Schultz, M. J.; Sigman, M. S. J. Am. Chem. Soc.
2006, 128, 2794. (k) Schultz, M. J.; Adler, R. S.; Zierkiewicz, W.;
Privalov, T.; Sigman, M. S. J. Am. Chem. Soc. 2005, 127, 8499.
(26) (a) Leoni, P.; Sommovigo, M.; Pasquali, M.; Midollini, S.; Braga,
D.; Sabatino, P. Organometallics 1991, 10, 1038. (b) Portnoy, M.;
Frolow, F.; Milstein, D. Organometallics 1991, 10, 3960. (c) Leoni, P.;
Pasquali, M.; Sommovigo, M.; Laschi, F.; Zanello, P.; Albinati, A.;
Lianza, F.; Pregosin, P. S.; Rueegger, H. Organometallics 1993, 12,
1702. (d) Portnoy, M.; Milstein, D. Organometallics 1994, 13, 600.
(e) Arnanz, A.; Gonzal
́
ez-Arellano, C.; Juan, A.; Villaverde, G.; Corma,
chez, F. Chem. Commun. 2010, 46, 3001.
A.; Iglesias, M.; San
́
(f) Boronat, M.; Corma, A.; Gonzal
Sanchez, F. Organometallics 2010, 29, 134. (g) Wang, D.-S.; Wang, D.-
W.; Zhou, Y.-G. Synlett 2011, 947.
́
ez-Arellano, C.; Iglesias, M.;
́
(19) (a) Wang, D.-S.; Ye, Z.-S.; Chen, Q.-A.; Zhou, Y.-G.; Yu, C.-B.;
Fan, H.-J.; Duan, Y. J. Am. Chem. Soc. 2011, 133, 8866. (b) Li, C.;
Chen, J.; Fu, G.; Liu, D.; Liu, Y.; Zhang, W. Tetrahedron 2013, 69,
6839.
̌
(e) Fantasia, S.; Egbert, J. D.; Jurcík, V.; Cazin, C. S. J.; Jacobsen, H.;
Cavallo, L.; Heinekey, D. M.; Nolan, S. P. Angew. Chem., Int. Ed. 2009,
L
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
48, 5182. (f) Fulmer, G. R.; Muller, R. P.; Kemp, R. A.; Goldberg, K. I.
J. Am. Chem. Soc. 2009, 131, 1346. (g) Fulmer, G. R.; Herndon, A. N.;
Kaminsky, W.; Kemp, R. A.; Goldberg, K. I. J. Am. Chem. Soc. 2011,
133, 17713. (h) Vyas, D. J.; Larionov, E.; Besnard, C.; Guen
́ ́
ee, L.;
Mazet, C. J. Am. Chem. Soc. 2013, 135, 6830.
(27) All calculations were carried out using Density Functional
Theory as implemented in the Jaguar 7.5 suite of ab initio quantum
chemistry programs. Geometry optimizations were performed with the
B3LYP functional and the 6-31G** basis set with no symmetry
restrictions. Palladium was represented using the Los Alamos LACVP
basis. Solvation energies were evaluated by a self-consistent reaction
field (SCRF) approach based on accurate numerical solutions of the
Poisson−Boltzmann equation. The free energy in solution phase was
calculated as G(sol) = E(SCF) + ZPE − TS(gas) + Gsolv
.
(28) Fey, N.; Ridgway, B. M.; Jover, J.; McMullin, C. L.; Harvey, J. N.
Dalton Trans. 2011, 40, 11184.
(29) (a) Bonnet-Delpon, D.; Beg
́
ue,
́
J.-P.; Crousse, B. Synlett 2004,
18. (b) Borner, A.; Shuklov, I.; Dubrovina, N. Synthesis 2007, 2007,
̈
2925.
(30) (a) Eberson, L.; Hartshorn, M. P.; Persson, O.; Radner, F.
Chem. Commun. 1996, 2105. (b) Munshi, P.; Main, A. D.; Linehan, J.
C.; Tai, C.-C.; Jessop, P. G. J. Am. Chem. Soc. 2002, 124, 7963.
(31) (a) Guari, Y.; Ayllon, J. A.; Sabo-Etienne, S.; Chaudret, B.;
Hessen, B. Inorg. Chem. 1998, 37, 640. (b) Fache, F.; Piva, O. Synlett
2004, 1294. (c) Ng, S. M.; Yin, C.; Yeung, C. H.; Chan, T. C.; Lau, C.
P. Eur. J. Inorg. Chem. 2004, 2004, 1788. (d) Vuluga, D.; Legros, J.;
Crousse, B.; Slawin, A. M. Z.; Laurence, C.; Nicolet, P.; Bonnet-
Delpon, D. J. Org. Chem. 2011, 76, 1126.
(32) (a) Teunissen, H. T. Chem. Commun. 1998, 1367.
(b) Bakhmutova, E. V.; Bakhmutov, V. I.; Belkova, N. V.; Besora,
M.; Epstein, L. M.; Lledos, A.; Nikonov, G. I.; Shubina, E. S.; Tomas,
J.; Vorontsov, E. V. Chem.Eur. J. 2004, 10, 661. (c) Donghi, D.;
Beringhelli, T.; D’Alfonso, G.; Mondini, M. Chem.Eur. J. 2006, 12,
1016. (d) Dub, P. A.; Fillipov, O. A.; Silantyev, G. A.; Belkova, N. V.;
Daran, J.-C.; Epstein, L. M.; Poli, R.; Shubina, E. S. Eur. J. Inorg. Chem.
2010, 1489.
(33) (a) Easson, L. H.; Stedman, E. J. Biochem. 1933, 27, 1257.
(b) Simon, L.; Goodman, J. M. J. Am. Chem. Soc. 2008, 130, 8741.
́
M
dx.doi.org/10.1021/ja502020b | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX