Catalytic Direct Dehydrogenative Cross-Couplings of C-H (Pro)Nucleophiles and Allylic Alcohols without an Additional Oxidant

Article abstract of DOI:10.1021/cs501495g

Transition-metal-catalyzed cross-coupling reactions between sp²-hybridized C atoms are of prime importance in both target and diversity oriented synthesis. Ideal cross-coupling reactions would neither require any leaving groups nor stoichiometr

Full text of DOI:10.1021/cs501495g

Research Article
pubs.acs.org/acscatalysis
Catalytic Direct Dehydrogenative Cross-Couplings of C−H
(Pro)Nucleophiles and Allylic Alcohols without an Additional Oxidant
Marcel Weiss and Rene Peters*
́
Institut fur Organische Chemie, Universitat Stuttgart, Pfaffenwaldring 55, D-70569 Stuttgart, Germany
̈
̈
S
* Supporting Information
ABSTRACT: Transition-metal-catalyzed cross-coupling reactions
between sp2-hybridized C atoms are of prime importance in both
target and diversity oriented synthesis. Ideal cross-coupling reactions
would neither require any leaving groups nor stoichiometric reagents.
In this article, we report the first direct dehydrogenative cross-
couplings between aromatic C−H bonds (in most cases using indole
substrates) and allylic alcohols, which do not require an additional
classical stoichiometric oxidizing agent and provide β-arylketones as
value-added products. Ruthenocene- or ferrocene-based bismetalla-
cycles, in which either Pd(II) or Pt(II) are the catalytically active
centers, were found to be particularly efficient catalysts. Control experiments suggest that the bismetallacycles initially transform
the allylic alcohols into vinylketones, which then alkylate the aromatic substrate in the presence of the catalyst. The fact that the
dehydrogenative coupling does not require a classical stoichiometric oxidizing agent is explained either by protonolysis of a
metallacyclic M(II)-H intermediate or by a mechanism in which an excess of the allylic alcohol substrate serves as a sacrificial
hydrogen acceptor. The title reaction is supported by cocatalytic amounts of Ni(OAc)2. In preliminary studies, it was observed
that the title reaction can as well be applied to prochiral CH-acidic pronucleophiles such as α-cyanoacetates, representing the first
examples for direct enantioselective β-ketoalkylations via allylic alcohols in the absence of an additional oxidant.
KEYWORDS: bimetallic catalysts, bispalladacycles, α-cyanoacetate, dehydrogenative cross-coupling, direct cross-coupling, ferrocene,
indole, β-ketoalkylations, oxidant-free, platinacycle, ruthenocene
sufficient reactivity. To turn over the catalytic cycle, an excess of
Cu(OAc)2 (2.1 equiv) was required as an oxidant.
INTRODUCTION
■
Cross coupling reactions between aromatic substrates and
olefins provide an indispensable tool to construct C−C bonds
for both target- and diversity-oriented synthetic applications.1
The most well-known reaction of this type is the classical
Mizoroki−Heck reaction, which is widely utilized in organic
synthesis because of its high level of generality.2 Heck type
reactions between arylhalides and allylic alcohols have been
intensively investigated in the past to allow for a synthetic
access of β-arylketones, avoiding highly electrophilic alkylating
reagents such as alkyl halides or vinylketones, which may cause
severe safety issues as a result of their toxicity and
carcinogenicity.3 The β-arylketone products are of high
synthetic value because the versatile ketone moiety offers a
number of possibilities for further functionalization.
A disadvantage of classical Heck type reactions is the need
for aromatic halides Ar-X as coupling partners and the resulting
generation of HX as a side product that needs to be trapped by
stoichiometric amounts of a base. The atom economy is thus
not ideal because large amounts of salt waste are generated.4,5
Recently, an activation strategy has been reported in which
leaving groups X on the aromatic rings could be omitted in the
generation of β-arylketones.6 Aromatic C−H bonds could be
utilized for Rh-catalyzed dehydrogenative cross-couplings with
allylic alcohols.6a A pyrimidin-2-yl substituent on the aromatic
substrate was used as an ortho-directing group to allow for a
Ideally, dehydrogenative cross-couplings would allow for the
direct construction of C−C bonds without the need for
stoichiometric reagents, generating H2 as the only side
product.7 Currently, the number of examples for oxidant-free
dehydrogenative C−C coupling reactions via H2 evolution is, in
general, still very much limited, which has been explained by a
thermodynamically unfavorable situation.7 In this context,
independent studies by the groups of Wu and Wang, who
used Ru or Co and Mn catalysts, respectively, need to be
particularly mentioned.8,9
Herein, we report the first direct dehydrogenative cross
coupling reactions of aromatic substrates Ar−H with allylic
substrates, in which there is no need for an additional classical
stoichiometric oxidizing agent. In addition ortho-directing
groups are also not needed for a sufficient reactivity. The title
reaction using allylic alcohol substrates has been accomplished
by trimetallic catalytic systems, which make use of the group 10
metals {either [Pt(II)/Pd(II)/(Ni(II)] or [Pd(II)2/Ni(II)]}.10
The precatalysts studied are either the metallocene-based
bisimidazoline11 bispalladacycles12 [FBIP-Cl]2 (1)13,14 and
Received: September 30, 2014
Revised: November 27, 2014
Published: December 8, 2014
© 2014 American Chemical Society
310
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316

ACS Catalysis
Research Article
15
a
[RuBIP-Cl]2 (2) (Figure 1) or the related ferrocene-based
Table 1. Preliminary Investigations
heterobimetallic pallada-/platinacycle [FBIPP-Cl]2 (3).16 Ni-
conv.
b
yield
c
no. precat.
X
additive (equiv)
NaOAc (0.2)
solvent
(%)
(%)
1
1
1
1
1
1
1
1
1
1
1
2
3
2.5
2.5
2.5
2.5
2.5
2.5
0
DCE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
TFE
92
97
65
66
44
33
86
50
5
2
3
54
4
[nBu4N]OAc (0.2)
Ni(OAc)2 (0.2)
Pd(OAc)2 (0.2)
Ni(OAc)2 (0.2)
Ni(OAc)2 (0.1)
Ni(OAc)2 (0.2)
Ni(OAc)2 (0.2)
Ni(OAc)2 (0.2)
Ni(OAc)2 (0.2)
36
5
100
100
9
Figure 1. Metallocene-based bis-metallacycle precatalysts relevant for
this study.
6
7
8
2.5
1.25
0.5
2.5
2.5
100
99
78
79
61
83
85
9
10
11
12
96
(OAc)2 acts as a cocatalyst. Preliminary studies showcase that
the discovered reactivity can also be applied to CH-acidic
pronucleophiles in enantioselective alkylations.
87
93
a
b
Ratio 4a/5A = 1:3. Conversion of 4a determined by 1H NMR using
c
1
an internal standard. Yield of 6aA determined by H NMR using an
internal standard.
RESULTS
■
Development of the Title Reaction. Because the dimeric
metallocene-based bismetallacycles 1−3 usually show nearly no
activity in catalytic applications as a result of the quite inert
chloride bridges, which hamper an efficient coordination of
neutral substrates, typically, these complexes need to be
activated by chloride removal, which results in the formation
of monomeric complexes.13b,14f,16 This is readily achieved in
quantitative yields by treatment with various different silver
salts according to our previously published protocols.13a,b,14−16
Because of the prominent role of indoles17 as bioactive
compounds, in our exploratory trials, we investigated the
coupling of 1-octen-3-ol (5A) with N-methylindole (4a, Table
1). With 2.5 mol % of 1 (activated by AgOTs/MeCN)13 in 1,2-
dichloroethane (DCE) at 70 °C, the β-heteroarylketone 6aA
was formed with promising efficiency in the absence of an
additional stoichiometric oxidant (entry 1).18
Substrate Scope. The reaction conditions shown in Table
1/entry 11 were subsequently applied to different indoles and
allylic alcohols (Table 2). The ruthenocene-based bispallada-
cycle 2 was selected to prepare different β-arylketones 6
because it is the most robust of the three investigated
precatalysts. The yields of the purified products 6 were usually
1
close to the yields determined by H NMR via an internal
Table 2. Application of the Optimized Conditions to Various
Indole and Allylic Alcohol Substrates
a
In 2,2,2-trifluoroethanol (TFE) the reaction outcome was
similar but with a slightly increased conversion (entry 2).
Different Brønsted basic acetate salts, such as NaOAc or
[nBu4N]OAc decreased the productivity of the indole
alkylation (entries 3−4). Lewis acidic acetate salts usually also
did not improve the reaction outcome, with the exception of
Ni(OAc)2. In the presence of 20 mol % of Ni(OAc)2, full
conversion was observed, and the product was obtained in a
yield of 86% (entry 5). In contrast, the homologous Pd(OAc)2
decreased the reaction efficiency (entry 6). Ni(OAc)2 (20 mol
%) gave only traces of product in the absence of the
bispalladacycle (entry 7).19 Reducing the Ni(OAc)2 quantity
from 20 to 10 mol % in the presence of the bispalladacycle still
allowed for a good product yield (entry 8). Reducing the
amount of the bispalladacycle precatalyst [FBIP-Cl]2 1 by a
factor of 2 (1.25 mol %) gave a similar result (79% yield, entry
9), whereas with 0.5 mol % of precatalyst, the reaction was
found to be less effective (61% yield, entry 10). A comparison
of the ferrocene bispalladacycle 1 with the ruthenocene
bispalladacycle 2 (entry 11) and the mixed ferrocene
pallada-/platinacycle 3 (entry 12) exhibited quite similar
productivities in the model reaction.
yield
b
no.
6
R1
R2
R3
Y
(%)
1
2
6aA
6bA
6cA
6dA
6eA
6fA
nPent
nPent
nPent
nPent
nPent
nPent
nPent
nPent
Me
Et
H
H
H
H
H
H
H
Me
H
H
H
H
H
83 (83)
71 (69)
81 (75)
66 (58)
33 (32)
60 (57)
54 (50)
90 (90)
61 (60)
72 (65)
78 (73)
89 (89)
H
3
Bn
H
4
3-F3CC6H4−CH2
4-BrC6H4−CH2
H
5
H
6
H
H
7
6gA
6hA
6aB
6aC
6aD
6aE
Me
Me
OBn
H
8
9
(CH2)2Ph Me
H
10
11
12
nPr
Me
Me
Me
H
Et
H
cyclo-Hex
H
a
b
Ratio 4/5 = 1:3. Yield of 6 determined by 1H NMR using an
internal standard (in parentheses: yield of isolated 6 after column
chromatography).
311
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316

ACS Catalysis
Research Article
standard, mainly depending on the sometimes limited product
stabilities under air atmosphere or on silica gel. In terms of the
indoles’ N-substituents R2, different alkyl moieties, such as
methyl (entry 1), ethyl (entry 2), benzyl (entry 3),20 or m-
(trifluoromethyl)benzyl (entry 4), were well tolerated. Notably,
a bromide substituent also was possible on a 4-Br-benzyl
residue to afford product 6eA in moderate yield (entry 5).
This latter result indicates that substrates prone to oxidative
addition pathways do not need to be categorically excluded
with the bispalladacycle 2. Noteworthy also is the observation
that a N-protecting group on the indole substrate is not strictly
required because useful yields were also obtained with the
parent indole (entry 6).
To learn more about the possible utility of oxidant-free direct
dehydrogenative couplings with allylic alcohols, a few non-
indole C-nucleophiles were studied (Scheme 2). The reaction
Scheme 2. Application of the β-Ketoalkylation to Other
a
(Pro)Nucleophiles
With an alkoxy group Y on the indole core, which is a
prominent substituent for indoles of pharmaceutical reve-
lance,17,21 the targeted cross coupling product was also formed
in useful yield (entry 7). Substituents R3 at the indole 2-
position were also accepted as displayed in entry 8. Product
6hA with R3 = Me was formed in high yield, showing that
polysubstituted indoles are accessible via the title reaction.
Regarding the allylic alcohol substrates 5, various α,β-
unbranched alkyl substituents R1 were investigated, and all
allowed for similar results (entries 9−11),22 but an α-branched
cyclohexyl residue R1 was also well tolerated (entry 12). If the
indole 3-position is blocked by a substituent, the methodology
can also be applied to functionalize the 2-position by a
dehydrogenative cross-coupling (Scheme 1a).
a
1
Yields were determined by H NMR using an internal standard (in
parentheses: yield of isolated products after column chromatography).
can also be used for electron-rich aromatic substrates other than
indoles, such as 1,3,5-trimethoxybenzene, which provided the
isolated product 10 in a yield of 73%.23 An overalkylation was
again largely avoided.
Scheme 1. (a) C−C Bond Formation Using the Indole 2-
Position and (b) Sequential Functionalization of the 3- and
2-Positions
Gratifyingly, the tandem reaction could also be applied to
CH-acidic α-cyanoacetate pronucleophiles. In that case,
diglyme is a superior solvent, and catalyst activation with silver
heptafluorobutyrate resulted in an enhanced reactivity and
allowed for a quantitative formation of 11a within 12 h at 70
°C. These preliminary studies show for the first time that
enantiocontrol is, in principle, also possible for oxidant-free
dehydrogenative couplings. Compound 11b with a sterically
demanding ester group was formed in good yield and with a
promising enantiomeric ratio of 84:16, making use of a dynamic
kinetic resolution of racemic starting material.
1
The precatalysts 1 and 2 (identified by H NMR and ESI-
MS; the activated palladacycles often act as a trap for chloride
traces upon reisolation, see e.g. ref 14i) could both be reisolated
in good yields (∼80%) by a simple filtration of the reaction
mixture over silica gel. This shows that the catalyst is relatively
stable under the reaction conditions. In addition, the
enantioselectivity obtained for 11 indicates that the pallada-
cycles used do not lose their structural identity in the catalytic
cycle and keep their coordination shell bearing the chiral ligand
at the metal centers. Optimization of the enantioselectivity and
the investigation of the scope of applicable substrates for
catalytic asymmetric reactions are currently ongoing in our
group.
Control Experiments. A direct dehydrogenative cross-
coupling employing an allylic alcohol substrate should involve
the generation of a metal hydride species at one point of the
catalytic cycle, either by a Heck-type reaction pathway6 or by
oxidation of the allylic alcohol to the corresponding enone via a
β-H-elimination.24,4,5 To investigate the fate of the allylic
alcohol 5A, the standard reaction conditions were applied to
the allylic alcohol in the absence of the nucleophile (Table 3).
Using 2.5 mol % of 1 activated by 10 mol % of AgOTs,
It was also found that this latent reactivity of the indole 2-
position causes the formation of small amounts of side products
during the above-described syntheses of products 6 by
formation of the corresponding double alkylation products
(typically up to 4%). This reactivity can also be of preparative
use for sequential double dehydrogenative cross-couplings
(Scheme 1b). In the first step, product 6aA was prepared from
model substrate 4a and 5A under standard conditions. The
second step made use of a different allylic alcohol (5D) in
larger excess and formed diketone 9 in a yield of 64%.
312
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316

ACS Catalysis
Research Article
hydride elimination, might be trapped by protonolysis with
trifluoroethanol (or the allylic alcohol in an aprotic solvent) to
form the corresponding Pd(II)-alkoxide intermediate, such as
16 (Scheme 3).27−29 Nucleophilic attack of the indole 4 at the
Table 3. Control Experiment in the Absence of a
Nucleophile
Scheme 3. Possible Simplified Mechanism with Release of
H2
a
a
no.
X
Y
yield 12A (%)
yield 13A (%)
1
2
3
2.5
2.5
0
10
0
12
6
55
40
0
10
0
a
1
Yield determined by H NMR using an internal standard.
vinylketone 12A and ethylketone 13A had been formed in
yields of 12% and 55%, respectively, after a reaction time of 24
h (entry 1). Repetition of this experiment in the absence of
Ni(OAc)2 provided similar results (entry 2). In contrast, using
10 mol % of Ni(OAc)2 in the absence of the bismetallacycle,
the formation of neither 12A nor 13A was observed (entry 3).
The major catalyst for an oxidation of the allylic alcohol is thus
the bismetallacycle, and Ni(OAc)2 should support this event.
These results suggest that the allylic alcohol 5 is first oxidized
into a vinylketone 12, which is the active electrophile in the
coupling with a nucleophilic indole. Oxidation of the allylic
alcohol 5 to the vinylketone 12 via a β-hydride elimination and
a subsequent hydrogenation of the vinylketone to an
ethylketone 13 by insertion of the CC double bond into a
Pd−H species, followed by protonolysis, would also be an
explanation why an excess of allylic alcohol (3 equiv) was
usually required for high reaction yields. To confirm the
assumption that vinylketones 12 are involved as the electro-
philic component, indole 4a was directly treated with
vinylketone 12A at a reaction temperature of only 35 °C
under otherwise unchanged reaction conditions (Table 4).
activated enone could then form the coupling product. Release
of TFE and coordination of another allylic alcohol substrate
molecule would close the catalytic cycle.
1
Analysis of the crude product mixtures by GC-MS and H
NMR revealed the generation of saturated alcohols 20 in yields
ranging from 0 to 78%, based on 1 equiv of the indole
substrates 4 for the reactions shown in Table 2 (yields of 0−
26% based on allylic alcohols 5). Compound 20 should be
formed by hydrogenation of the allylic alcohol substrates 5. A
possible catalytic cycle to rationalize this result is depicted in
Scheme 4. This alternative scenario might be an additional
explanation why a relatively large excess of the allylic alcohol is
required for high reaction efficiency.
a
Table 4. Control Experiment Using a Vinylketone Substrate
The bismetallacycles were also compared with structurally
related monopallada-30 or -platinacycles11c 21 and 22,
respectively (Scheme 5). With increased catalyst loadings to
achieve the same amount of a noble metal present as with 1−3,
only poor to low activity was noticed under the standard
reaction conditions.
b
no.
t (min)
yield 6aA (%)
1
2
3
4
1
5
24
48
64
In addition, the use of Pd(OAc)2 and (MeCN)2PdCl2 was
examined. Under typical reaction conditions (in the presence of
20 mol % of Ni(OAc)2) using 5 and 10 mol % of
(MeCN)2PdCl2, 6aA was again formed in poor yields (10
and 25%, respectively). Reactions with Pd(OAc)2 instead
resulted in low reproducibility and, in most cases, provided
varying product yields in a range of 30−40%. Addition of 10
mol % of AgOTs as a cocatalyst did not change this outcome.
These data indicate that bismetallacycles with M = Pd(II) or
Pt(II) offer a distinct reactivity advantage in the title reaction as
compared with simple Pd(II) salts. A bimetallic precoordina-
tion of the allylic alcohol substrate might, for example, facilitate
the vinylketone formation. Similar reactivity differences have
already been noticed for rearrangements of allylic imidates and
were explained by an increased binding affinity of the substrates
due to bimetallic precoordination.16
10
c
180
87
a
b
1
Ratio 4:5 = 1:3. Yield determined by H NMR using an internal
standard. In addition, the corresponding double alkylation product
was formed in 13% yield.
c
Despite the much lower reaction temperature the reaction
product 6aA was formed very rapidly.25 After 1 and 10 min,
24% and 64%, respectively, of 6aA were already detected
(entries 1 and 3).26
DISCUSSION
■
A Pd−H species such as 15 or 15′, which is assumed to be
formed by oxidation of the allylic alcohol substrate 5 via β-
313
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316

ACS Catalysis
Research Article
intermediates. The latter subsequently rapidly react with an
indole substrate in the presence of the catalyst. The fact that
there is no need for an additional stoichiometric oxidizing agent
might be explained by protonolysis of a M(II)−H intermediate
by either a protic solvent or the substrate OH groups. Because
significant amounts of fully saturated alcohols are also formed,
the allylic alcohol substrates might probably also act as
sacrificial H2 acceptors. The reactions are supported by
cocatalytic amounts of Ni(OAc)2, but the precise role of the
Ni salt is not clear yet. Preliminary studies have also shown that
the reactivity of the catalytic system can be transferred to
prochiral racemic CH-acidic pronucleophiles, such as α-
cyanoacetates, providing the tandem reaction products in
enantioenriched form. These are the first catalytic asymmetric
examples for enantioselective direct dehydrogenative couplings
via allylic alcohol electrophiles without the need for an
additional classical oxidant. These findings pave the way for
future investigations in asymmetric catalysis.
Scheme 4. Alternative Simplified Mechanism in Which
Allylic Alcohol 5 Also Serves As a Sacrificial Hydrogen
Acceptor
EXPERIMENTAL SECTION
■
General Considerations. All reactions were performed in
oven-dried (150 °C) glassware under a positive pressure of
nitrogen (about 0.2 bar). For all reactions, liquids and solutions
were added via syringes and septa. For catalysis, all glassware
used (also for catalyst activation and preparation of stock
solutions) was washed intensively with demineralized water to
remove traces of chloride. N,N-Dimethylformamide was stored
in crown-capped bottles under argon over 4 Å molecular sieves.
Acetonitrile and dichloromethane were purified by distillation
and subsequently by a solvent purification system. 2,2,2-
Trifluoroethanol was dried and degassed prior to use. For
workup procedures and column chromatography, technical
grade solvents (petrol ether and ethyl acetate) were purified by
distillation prior to use. Solvents were mostly removed at a
heating bath temperature of 40 °C and 600−10 mbar pressure
by rotary evaporation. Nonvolatile compounds were dried in
vacuo at 0.1 mbar.
Scheme 5. Control Experiments with Ferrocene
Monopalladacycle 21 and Monoplatinacycle 22
If not stated otherwise, yields refer to chromatographically
purified compounds and are calculated in mole percent of the
used starting material. Catalytic reactions were carried out in a
parallel synthesizer at 400 rpm. For thin layer chromatography,
silica gel plates were used. Visualization occurred by
fluorescence quenching under UV light, by staining with
KMnO4/NaOH followed by heating (230 °C), or both.
Purification by flash chromatography was performed on silica
gel 0.040−0.063 mm using a forced flow of eluent at 0.2−0.4
bar pressure.
General Procedure for the Catalytic Direct Dehydro-
genative Cross-Couplings. The respective chloride-bridged
precatalyst dimer 1-3 (1.00 equiv) and silver tosylate (4.00
equiv) were dissolved in MeCN (0.2 mL/mg) under nitrogen
atmosphere. The mixture was stirred overnight at room
temperature and subsequently filtered through Celite. The
filtercake was extracted with MeCN until the organic solution
was colorless. The solvent was removed by a steady stream of
nitrogen and, finally, by high vacuum. A stock solution of the
activated catalyst was prepared by dissolving the solid in dry,
degassed TFE (40 mmol/L).
The observation that Ni(OAc)2 improves the reaction
efficiency might have various reasons, and the precise role of
Ni(II) is not yet clear. In the catalytic cycle shown in Scheme 3,
protonolysis of the M-H species is assumed to be a critical
event. We hypothesize that Ni(OAc)2 might facilitate the
protonolysis step, for example, via an acetate-bridged M/Ni−H
complex.28 Such a Ni−H species might also be responsible for
hydrogenation of the allylic alcohol.
CONCLUSION
■
In conclusion, we have reported the first direct dehydrogenative
cross-couplings of aromatic substrates with allylic alcohols to
form β-arylketones without the need for an additional classical
oxidant. These reactions do not require an ortho-directing
group. Metallocene-based bispalladacycles or mixed pallada-/
platinacycles were identified as competent catalysts and are
significantly more active than related monopallada- or -platina-
cycles. Control experiments suggest that the coupling initially
proceeds via conversion of the allylic alcohols into vinylketone
This stock solution was then used to add the activated
catalyst (prepared from 2.5 mol % of precatalyst) to the
nucleophile (0.16 mmol, 1.00 equiv), the allylic alcohol 5 (0.48
mmol, 3.00 equiv), and Ni(OAc)2 (0.03 mmol 20 mol %)
under a nitrogen atmosphere. The reaction tube was sealed and
314
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316

ACS Catalysis
Research Article
Liu, B.; Li, X.-B.; Gao, X.-W.; Lei, T.; Wu, C.-J.; Li, Z.-J.; Tung, C.-H.;
Wu, L.-Z. Org. Lett. 2014, 16, 1988−1991. (c) He, R.; Huang, Z.-T.;
Zheng, Q.-Y.; Wang, C. Angew. Chem., Int. Ed. 2014, 53, 4950−4953.
(9) (a) For a related C-P bond formation, see: Zhou, A.-X.; Mao, L.-
L.; Wang, G.-W.; Yang, S.-D. Chem. Commun. 2014, 50, 8529−8532.
(b) For a related S-S bond formation, see: Li, X.-B.; Li, Z.-J.; Gao, Y.-
J.; Meng, Q.-Y.; Yu, S.; Weiss, R. G.; Tung, C.-H.; Wu, L.-Z. Angew.
Chem., Int. Ed. 2014, 53, 2085−2089.
the reaction mixture was shaken at 400 rpm for 6 h at 70 °C.
The reaction mixture was afterward suspended in petrol ether/
ethyl acetate (10:1) and subsequently purified by filtration over
silica gel.
ASSOCIATED CONTENT
■
S
* Supporting Information
The following files are available free of charge on the ACS
(10) Selected reviews about bimetallic catalysis: (a) Bratko, I.;
Publications website at DOI: 10.1021/cs501495g.
́
Gomez, M. Dalton Trans. 2013, 42, 10664−10681. (b) Shibasaki, M.;
Kanai, M.; Matsunaga, S.; Kumagai, N. Acc. Chem. Res. 2009, 42,
1117−1127. (c) Shibasaki, M.; Matsunaga, S. Chem. Soc. Rev. 2006, 35,
269−279. (d) van den Beuken, E. K.; Feringa, B. L. Tetrahedron 1998,
54, 12985−13011.
Experimental procedures, characterization data for all
new products (PDF)
Crystallographic information file for C23H27NO (CIF)
Crystallographic information file for C27H26O4 (CIF)
(11) Review about the application of imidazolines in asymmetric
catalysis: (a) Liu, H.; Du, D.-M. Adv. Synth. Catal. 2009, 351, 489−
519. Selected recent applications of bisimidazoline ligands not yet
covered in the review: (b) Nakamura, S.; Hyodo, K.; Nakamura, Y.;
Shibata, N.; Toru, T. Adv. Synth. Catal. 2008, 350, 1443−1448.
(c) Huang, H.; Peters, R. Angew. Chem., Int. Ed. 2009, 48, 604−606.
(d) Liu, H.; Du, D.-M. Adv. Synth. Catal. 2010, 352, 1113−1118.
(e) Cheng, L.; Dong, J.; You, J.; Gao, G.; Lan, J. Chem.Eur. J. 2010,
16, 6761−6765. (f) Ohara, M.; Nakamura, S.; Shibata, N. Adv. Synth.
Catal. 2011, 353, 3285−3289. (g) Hyodo, K.; Nakamura, S.; Tsuji, K.;
Ogawa, T.; Funahashi, Y.; Shibata, N. Adv. Synth. Catal. 2011, 353,
3385−3390. (h) Hyodo, K.; Nakamura, S.; Shibata, N. Angew. Chem.,
Int. Ed. 2012, 51, 10337−10341. (i) Hyodo, K.; Kondo, M.;
Funahashi, Y.; Nakamura, S. Chem.Eur. J. 2013, 19, 4128−4134.
(j) Nakamura, S.; Hyodo, K.; Nakamura, M.; Nakane, D.; Masuda, H.
Chem.Eur. J. 2013, 19, 7304−7309.
AUTHOR INFORMATION
Corresponding Author
*E-mail: rene.peters@oc.uni-stuttgart.de.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
This work was financially supported by the Deutsche
Forschungsgemeinschaft (PE 818/6-1). The Fonds der
■
Chemischen Industrie (FCI) and the Landesgraduiertenforder-
̈
ung Baden-Wurttemberg are kindly acknowledged for Ph.D.
̈
fellowships to M.W. We thank Dr. Wolfgang Frey for the X-ray
crystal structure analyses of 6cA and 10.
(12) (a) Review on palladacycles: Dupont, J., Pfeffer, M., Eds.;
Palladacycles; Wiley-VCH: Weinheim, Germany, 2008. (b) Review
about ferrocene palladacycles in asymmetric catalysis: Richards, C. J.
In Chiral Ferrocenes in Asymmetric Catalysis; Dai, L.-X., Hou, X.-L.,
Eds.; Wiley-VCH: Weinheim, 2010; p 337.
(13) (a) Jautze, S.; Seiler, P.; Peters, R. Angew. Chem., Int. Ed. 2007,
46, 1260−1264. (b) Jautze, S.; Seiler, P.; Peters. Chem.Eur. J. 2008,
14, 1430−1444. (c) Jautze, S.; Diethelm, S.; Frey, W.; Peters, R.
Organometallics 2009, 28, 2001−2004.
REFERENCES
■
(1) (a) Torborg, C.; Beller, M. Adv. Synth. Catal. 2009, 351, 3027−
3043. (b) Wu, X.-F.; Anbarasan, P.; Neumann, H.; Beller, M. Angew.
Chem., Int. Ed. 2010, 49, 9047−9050.
(2) (a) Knowles, J. P.; Whiting, A. Org. Biomol. Chem. 2007, 5, 31−
44. (b) Heravi, M. M.; Hashemi, E.; Ghobadi, N. Curr. Org. Chem.
2013, 17, 2192−2224.
(14) Examples for catalytic applications: (a) Jautze, S.; Peters, R.
Angew. Chem., Int. Ed. 2008, 47, 9284−9288. (b) Weber, M.; Jautze, S.;
Frey, W.; Peters, R. J. Am. Chem. Soc. 2010, 132, 12222−12225.
(c) Weber, M.; Jautze, S.; Frey, W.; Peters, R. Chem.Eur. J. 2012, 18,
14792−14804. (d) Weber, M.; Frey, W.; Peters, R. Adv. Synth. Catal.
2012, 354, 1443−1449. (e) Weber, M.; Frey, W.; Peters, R. Angew.
Chem., Int. Ed. 2013, 52, 13223−13227. (f) Weber, M.; Klein, J. E. M.
N.; Miehlich, B.; Frey, W.; Peters, R. Organometallics 2013, 32, 5810−
5817. (g) Weber, M.; Frey, W.; Peters, R. Chem.Eur. J. 2013, 19,
8342−8351. (h) Eitel, S. H.; Jautze, S.; Frey, W.; Peters, R. Chem. Sci.
2013, 4, 2218−2233. (i) Sudheendran, M.; Eitel, S. H.; Naumann, S.;
Buchmeiser, M. R.; Peters, R. New J. Chem. 2014, 38, 5597−5607.
(15) Hellmuth, T.; Rieckhoff, S.; Weiss, M.; Dorst, K.; Frey, W.;
Peters, R. ACS Catal. 2014, 4, 1850−1858.
(3) Selected reviews: (a) Muzart, J. Tetrahedron 2005, 61, 4179−
4212. (b) Beletskaya, I. P.; Cheprakov, A. V. Chem. Rev. 2000, 100,
3009−3066.
(4) Arylhalides can be replaced by aryldiazonium salts in Pd-catalyzed
couplings with allylic alcohols: (a) Werner, E. W.; Mei, T.-S.; Burckle,
A. J.; Sigman, M. S. Science 2012, 338, 1455−1458. (b) for a
computational study, see: Dang, Y.; Qu, S.; Wang, Z.-X.; Wang, X. J.
Am. Chem. Soc. 2014, 136, 986−998.
(5) (a) For a Pd-catalyzed oxidative cross-coupling of allylic alcohols
with N-tosylhydrazones, see also: Chen, H.; Huang, L.; Fu, W.; Liu,
X.; Jiang, H. Chem.Eur. J. 2012, 18, 10497−10500. (b) For a Pd-/
amine-catalyzed multiple relay catalysis towards chiral building blocks,
see also: Deiana, L.; Jiang, Y.; Palo-Nieto, C.; Afewerki, S.; Incerti-
́
Pradillos, C. A.; Verho, O.; Tai, C.-W.; Johnston, E. V.; Cordova, A.
(16) Weiss, M.; Frey, W.; Peters, R. Organometallics 2012, 31, 6365−
6372.
Angew. Chem., Int. Ed. 2014, 53, 3447−3451. and cited references (c)
For a sequential allylic alcohol oxidation/1,4-addition using gold-
palladium nanoclusters, see also: Yoo, W.-J.; Miyamura, H.;
Kobayashi, S. J. Am. Chem. Soc. 2011, 133, 3095−3103. (d) For an
oxidative Heck reaction using alkinyl-substituted anilines and allylic
alcohols, see: Wang, Q.; Huang, L.; Wu, X.; Jiang, H. Org. Lett. 2013,
15, 5940−5943.
(17) Selected reviews about indoles: (a) Edwankar, C. R.; Edwankar,
R. V.; Namjoshi, O. A.; Rallapalli, S. K.; Yang, J.; Cook, J. M. Curr.
Opin. Drug Discovery Dev. 2009, 12, 752−771. (b) Somei, M.; Yamada,
F. Nat. Prod. Rep. 2005, 22, 73−103. (c) Gribble, G. W. J. Chem. Soc.,
Perkin Trans. 1 2000, 1045−1075. (d) Sundberg, R. J. Indoles;
Academic Press: San Diego, 1996.
(6) (a) Shi, Z.; Boultadakis-Arapinis, M.; Glorius, F. Chem. Commun.
2013, 49, 6489−6491. (b) For an alternative oxidative method using
in situ generated indoles, see: Wang, Q.; Huang, L.; Wu, X.; Jiang, H.
Org. Lett. 2013, 15, 5940−5943. (c) For the use of propargylic
alcohols, see: Trost, B. M.; Breder, A. Org. Lett. 2011, 13, 398−401.
(7) Overview: He, K.-H.; Li, Y. ChemSusChem 2014, 7, 2788−2790.
(8) (a) Meng, Q.-Y.; Zhong, J.-J.; Liu, Q.; Gao, X.-W.; Zhang, H.-H.;
Lei, T.; Li, Z.-J.; Feng, K.; Chen, B.; Tung, C.-H.; Wu, L.-Z. J. Am.
Chem. Soc. 2013, 135, 19052−19055. (b) Zhong, J.-J.; Meng, Q.-Y.;
(18) The regioselectivity is complementary to the alkylation of
indoles reported by Glorius (ref 6a).
(19) Although nearly all AgOTs is consumed during the catalyst
activation and AgCl removed by filtration prior to catalysis, a control
experiment has been conducted under the conditions of Table 1/entry
7, in which a mixture of AgOTs (5 mol %) and Ni(OAc)2 (20 mol %)
was used as catalyst system. This resulted in the formation of only
traces of 6aA.
315
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316

ACS Catalysis
Research Article
(20) The constitution of 6cA was confirmed by X-ray crystal
structure analysis. Supplementary crystallographic data have been
deposited with the Cambridge Crystallographic Data Centre as
deposition 1016528. This material is available free of charge via the
Internet at http://www.ccdc.cam.ac.uk/products/csd/request/.
(21) Peters, R.; Waldmeier, P.; Joncour, A. Org. Proc. Res. Develop.
2005, 9, 508−512 and cited ref. .
(22) Unfortunately, a phenyl ring was not tolerated as R1 for
currently unknown reasons.
(23) The constitution of 10 was confirmed by X-ray crystal structure
analysis. Supplementary crystallographic data have been deposited
with the Cambridge Crystallographic Data Centre as deposition
1016529. This material is available free of charge via the Internet at
http://www.ccdc.cam.ac.uk/products/csd/request/.
(24) Early studies for the isomerization of allylic alcohols into
vinylketones: (a) Alper, H.; Hachem, K. J. Org. Chem. 1980, 45, 2269−
2270. (b) Damico, R.; Logan, T. J. J. Org. Chem. 1967, 32, 2356−2358.
(c) Chin, C. S.; Park, J.; Kim, C.; Lee, S. Y.; Shin, J. H.; Kim, J. B.
Catal. Lett. 1988, 1, 203−205. (d) For an enantioselective Ru-
catalyzed isomerization of allylic alcohols into chiral aldehydes, see:
Arai, N.; Sato, K.; Azuma, K.; Ohkuma, T. Angew. Chem., Int. Ed. 2013,
52, 7500−7504.
(25) (a) Review: Saracoglu, N. In Top. Heterocycl. Chem.; Khan, M.
T. H., Ed.; Springer-Verlag: Berlin, Heidelberg, 2007; Vol. 11
(Bioactive Heterocycles V), p 1. (b) Moran, A.; Hamilton, A.; Bo,
C.; Melchiorre, P. J. Am. Chem. Soc. 2013, 135, 9091−9098.
(c) Barakat, A.; Islam, M. S.; Al Majid, A. M. A.; Al-Othman, Z. A.
Tetrahedron 2013, 69, 5185−5192.
(26) Both 2 (2.5 mol%) and Ni(OAc)2 (20 mol%) catalyze the 1,4-
addition and independently provided 6aA after 3 h in the absence of
the other catalyst in yields of 80% and 79%, respectively. In the
absence of both 2 and Ni(OAc)2, the reaction in TFE furnished 6aA in
a yield of 22%.
(27) Protonolysis of metal hydride species has also been suggested
for the oxidant-free dehydrogenative couplings reported by Wang
(Mn-H, ref 8c) and Yang (Cu-H, ref 9).
(28) Rakowski DuBois, M.; DuBois, D. L. Chem. Soc. Rev. 2009, 38,
62−72.
(29) Unfortunately, our research group has no access to suitable GC
equipment to investigate the generation of H2. One reviewer thus
suggested studying the proposed H2 formation by 1H NMR.
Unfortunately, H2 detection in trifluoroethanol solution was not
possible, since a hydrogen signal would be completely hidden by the
solvent signal, as shown for a saturated H2 solution.
(30) (a) Weiss, M. E.; Fischer, D. F.; Xin, Z.-q.; Jautze, S.; Schweizer,
W. B.; Peters, R. Angew. Chem., Int. Ed. 2006, 45, 5694−5699.
(b) Fischer, D. F.; Barakat, A.; Xin, Z.-q.; Weiss, M. E.; Peters, R.
Chem.Eur. J. 2009, 15, 8722−8741.
316
dx.doi.org/10.1021/cs501495g | ACS Catal. 2015, 5, 310−316