A three-component palladium-catalyzed oxidative C-C coupling reaction: A domino process in two dimensions

Article abstract of DOI:10.1002/anie.201300833

Easy as ABC. or CBA! Indoles can be selectively functionalized twice by a catalytic oxidative three-component coupling of arylboronates and β-ketoesters. The difference in the rates (see scheme) of the competing processes are amplified under the conditions of the three-component reaction. Copyright

Full text of DOI:10.1002/anie.201300833

.
Angewandte
Communications
DOI: 10.1002/anie.201300833
Multicomponent Reactions
ꢀ
A Three-Component Palladium-Catalyzed Oxidative C C Coupling
Reaction: A Domino Process in Two Dimensions**
Roshan Y. Nimje, Mikko V. Leskinen, and Petri M. Pihko*
In multicomponent reactions (MCRs), three or more reac-
tants combine in a single chemical step to give a product that
contains nearly all the atoms of the individual reactants.
Compared with multistep reaction routes, MCRs are highly
attractive in terms of step economy.[1,2] Typical MCRs proceed
in a linear domino mode: in the first stage, the components A
and B (Scheme 1a) give rise to a reactive intermediate (AB)
which then reacts with a third component (C), and so on, until
the sequence is terminated. Very often, the first steps are
reversible, and only the terminating steps are irreversible
(type I or type II MCRs).[3,4]
In contrast, if three components A, B, and C could react
with each other in any order, such three-component domino
reactions (3CR) could proceed through an alternative two-
dimensional split domino process (Scheme 1b). In such
a process, there are two kinetic alternatives to give the final
ABC product, via the AB or the BC intermediates. If the steps
are irreversible, the efficiency of the process could be
compromised if one of the pathways leads to a dead end
(see the box in Scheme 1b). Alternatively, unwanted homo-
coupling or heterocoupling reactions between the compo-
nents could also jeopardize the projected 3CR process. Given
these constraints, it is not surprising that nearly all known
MCRs and domino reactions proceed in a strict linear fashion,
with the functionalities of the individual components deter-
mining the order of events.[5]
Scheme 1. The difference between a) linear and b) two-dimensional
multicomponent sequences. Bz=benzoyl, TFA=trifluoroacetate
Herein, we describe a successful two-dimensional 3CR
that proceeds through a mild Pd-catalyzed oxidative coupling
between b-ketoesters (component A), indoles (component
B), and aryl boronates (component C; Scheme 1c). This 3CR
could theoretically proceed via both AB and BC intermedi-
ates, and we demonstrate herein that both pathways are
viable. To the best of our knowledge, catalytic oxidative 3CRs
have not been reported in the literature.[6] The 3CR generates
densely functionalized 2,3-disubstituted indoles directly from
the three components.[7]
[*] Dr. R. Y. Nimje, M. V. Leskinen, Prof. Dr. P. M. Pihko
Department of Chemistry and NanoScience Center
University of Jyvꢀskylꢀ
We hypothesized that the reaction conditions of our
previously reported[8] cross-dehydrogenative coupling reac-
tion between b-ketoesters and indoles could also enable other
oxidative coupling reactions at the free 2-position of the
indole. Two-component oxidative couplings between indoles
at the 2-position and arylboronates have previously been
reported, by the groups of Shi,[9a] Zhang,[9b] and Studer.[10]
However, direct oxidative 2,3-difunctionalization reactions of
indoles have not been developed, and we were concerned that
owing to the increased steric requirements of such a process
a CuII co-catalyst might be required,[9a,b] and this could
compromise the regioselectivity for the b-ketoester.[8a] How-
ever, we anticipated that the use of the more-electrophilic PdII
precursor Pd(TFA)2, might overcome these problems, as this
catalyst was shown to be superior to other PdII sources in the
cross-dehydrogenative coupling reactions.[8a]
P. O. B. 35, 40014 JYU (Finland)
E-mail: petri.pihko@jyu.fi
Homepage: http://tinyurl.com/pihkogroup
[**] We thank the Academy of Finland (projects 138854 and 139046),
Tekes, Orion Pharma, Hormos Medical, and AB Enzymes for
financial support. We also thank Mirja Lahtiperꢀ and Esa Haapa-
niemi for assistance with mass spectrometry and NMR spectros-
copy, respectively. Antti Neuvonen and Dr. K.-T. Yip are acknowl-
edged for assistance with the X-ray and kinetic analysis, respectively.
Author contributions: M.V.L. conceived and initiated the study and
performed the kinetic experiments. R.Y.N. explored the scope of the
reaction. P.M.P. supervised the project and wrote the paper
together with the co-authors.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201300833.
4818
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4818 –4822

Angewandte
Chemie
Table 1: Screening of various aryl sources for the 3CR.[a]
Entry
1
X
t [h]
Yield [%][a]
83
20
2
3
24
24
11
59
4
48
0
5
6
48
48
0
0
7
24
0
[a] Reaction conditions: 1a (0.6 mmol, 1.5 equiv), 2a (0.4 mmol,
1 equiv), 3 (0.8 mmol, 2 equiv), Pd(TFA)2 (0.04 mmol, 10 mol%), and
tBuOOBz (1.0 mmol, 2.5 equiv) in iPrOH/AcOH 4:1 (0.5 mL).
[b] Determined by 1H NMR analysis using dibenzyl ether as an internal
standard.
Scheme 2. Oxidative domino cross-couplings with various arylboronate
and indole components. The yields of the products after purification
are reported. Reaction conditions, unless otherwise indicated:
1 (1.5 equiv), 2 (0.4 mmol, 1.0 equiv), 3 (0.8 mmol, 2.0 equiv),
tBuOOBz(2.5 equiv), Pd(TFA)2 (0.1 equiv), iPrOH/AcOH (4:1; 0.5 mL)
at 258C. [a] Dioxane/AcOH (4:1; 0.5 mL) used as a solvent. [b] 2
(3 equiv) used in two installments. [c] 1:1 mixture of atropisomers. The
combined yield of two atropisomers is given.
To our delight, when the three components b-ketoester
1a, indole 2a, and phenylboronic acid 3 (Table 1) were
subjected to our reaction conditions with 2.5 equivalents of
the oxidant (tBuOOBz), the reaction proceeded cleanly to
give the ABC product 4a (83% conversion, 76% yield upon
isolation). Further studies (Table 1) indicated that phenyl-
boronic acid was indeed the optimal coupling partner. Other
oxidants, acid additives, and solvents were also screened (see
the Supporting Information) but these generally turned out to
be inferior to the optimal conditions.
The scope of the reaction turned out to be very wide.
Scheme 2 summarizes the exploration of the scope with
different arylboronates and indoles. Both electron-donating
and electron-withdrawing groups are tolerated on the boro-
nate (products 4c–4i), as well as on indole (4k–4n). With o-
tolylboronic acid, the product 4i was obtained as a 1:1 mixture
of two atropisomers. Even compound 4j, derived from o-
isopropylphenylboronic acid, was obtained in good yield and
the two diastereomeric atropisomers could readily be sepa-
rated and characterized (see the Supporting Informa-
tion).[11,12] These experiments also attest to the high steric
tolerance of the 3CR process.
Different b-ketoesters are also tolerated (Scheme 3). 3CR
products derived from b-keto lactones (4p–4q), 7-membered
b-ketoesters (4r), fused b-ketoesters (4s), and even open-
chain b-ketoesters (4t) can be accessed with remarkable ease
by this oxidative three-component coupling reaction. Finally,
the use of a menthyl ester provides the product with moderate
but promising diastereoselectivity (4u; 3:1 d.r.), and even
more remarkably, with o-tolylboronic acid, the reaction gives
only two major products in a 1.8:1 ratio (4v; presumably the
atropisomeric products) instead of the expected statistical
mixture of four diastereomers.
These investigations underline the high tolerance of the
reaction to steric effects and suggest that some level of control
for the diastereoselectivity should also be possible. The
reaction proceeds under very mild reaction conditions, at
room temperature, with benign solvents (AcOH + iPrOH or
dioxane).
In all cases, the reactions proceeded with high regiochem-
ical fidelity: the arylboronate 3 was selectively coupled at the
2-position of indole, and the cross-dehydrogenative coupling
of b-ketoester was fully regioselective for the b’ position of
the b-ketoester and the 3-position of the indole compo-
nent.[8–10] What was more remarkable, however, was that the
dehydrogenative conditions of the reaction did not result in
unwanted coupling between the b-ketoester 1 and the
boronate 3 (A + C coupling). Indeed, even in a control
experiment with 1a and 3a but no indole component, the rate
of the formation of 8a (A + C product) was very low
(Scheme 4).
To probe whether the 3CR could proceed through both
A + B and B + C pathways, or whether one of the pathways
was dominating, we carried out the following experiments.
First, monitoring the reaction progress (Figure 1a) by
1H NMR spectroscopy indicated that the formation of 5
Angew. Chem. Int. Ed. 2013, 52, 4818 –4822
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4819

.
Angewandte
Communications
Scheme 3. Oxidative domino cross-couplings with different b-keto-
esters. Yields of pure products are reported and the X-ray crystallo-
graphic structure of 4q is shown.[15] Reaction conditions, unless
otherwise indicated: 1 (1.5 equiv), 2 (0.4 mmol, 1.0 equiv), 3
(0.8 mmol, 2.0 equiv), tBuOOBz (2.5 equiv), Pd(TFA)2 (0.1 equiv),
iPrOH/AcOH (4:1; 0.5 mL) at 258C. [a] Combined yield of diastereo-
mers.
Figure 1. a) Monitoring of the temporal progress of the three-component
A+B+C (1a+2a+3a) coupling by 1H NMR spectroscopy. b) Monitoring
of the temporal progress of the A+B (1a+2a) two-component coupling
by 1H NMR spectroscopy. c) Initial rate kinetics of the B+C (2a+3) two-
component coupling. Reaction conditions: [1a]0 =0.44m, [2a]0 =0.29m,
[3]0 =0.58m, [tBuOOBz]0 =0.73m, 10 mol% Pd(TFA)2, 4:1 [D8]dioxane/
AcOH, 300 K. In the three-component coupling, [7] could not be reliably
monitored in the early stages of the reaction (see the Supporting
Information).
Scheme 4. Control experiment to probe the possibility of the formation
of the A+C coupling product.
(A + B coupling) was slower than the formation of 6 (B + C)
under the 3CR conditions, with a more than 20-fold initial
rate difference between these two competing processes.
Each coupling step of the process was then examined
individually by two-component coupling (2CR) experiments
under otherwise identical reaction conditions. The initial
coupling reactions between 1a + 2a (A + B) and 2a + 3 (B +
C) were studied first. These experiments revealed that the
boronate indole coupling (2a + 3; B + C) proceeded at nearly
identical rates in the 3CR (5.5 mm minꢀ1) and in the 2CR
(5.1 mm minꢀ1; Figure 1c). In stark contrast, the cross-dehy-
drogenative coupling between b-ketoester 1a and indole 2a
(A + B) proceeded nearly an order of magnitude faster under
the 2CR conditions (1.5 mm minꢀ1) compared to the 3CR
conditions (0.2 mm minꢀ1; see Figure 1a). The difference in
the rates of the individual 2CRs A + B and B + C, thus
appears to be amplified in the 3CR.
The second step of the 3CR was studied next. Although
the rate of the formation of the AB product 5 was marginal
under the 3CR conditions, the AB + C control experiment
demonstrated that 5 readily reacted with 3 (C) under
4820 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4818 –4822

Angewandte
Chemie
otherwise identical reaction conditions to give 4 (ABC
product).[12] Thus, the AB + C!ABC pathway, although
slower than the A + BC pathway, was nevertheless shown to
be productive. In turn, a competition experiment with 1a (A)
and equimolar concentrations of 2a (B) and 6 (BC) indicated
that these two competing cross-dehydrogenative coupling
reactions to give 5 (AB) and 4 (ABC) proceed at nearly
identical rates.[12] These experiments confirmed that the 3CR
proceeds predominantly through the “B + C, then A” path-
way, with the dominance established by the faster B + C
coupling relative to the A + B coupling. However, both
pathways are viable routes to the product.
[1] a) “Towards the ideal synthesis”: P. A. Wender, S. T. Handy,
D. L. Wright, Chem. Ind. 1997, 765; For more recent
a
perspective, see: b) T. Gaich, P. S. Baran, J. Org. Chem. 2010,
75, 4657.
[2] For selected reviews, see: a) E. Ruijter, R. Scheffelaar, R. V. A.
Orru, Angew. Chem. 2011, 123, 6358; Angew. Chem. Int. Ed.
2011, 50, 6234; b) J. E. Biggs-Houck, A. Younai, J. T. Shaw, Curr.
Opin. Chem. Biol. 2010, 14, 371; c) G. Guillena, D. J. Ramꢀn, M.
Yus, Tetrahedron: Asymmetry 2007, 18, 693; d) A. Dçmling,
Chem. Rev. 2006, 106, 17; e) Multicomponent Reactions (Eds.: J.
Zhu, H. Bienaymꢁ), Wiley-VCH, Weinheim, 2005.
[3] A. Dçmling, I. Ugi, Angew. Chem. 2000, 112, 3300; Angew.
Chem. Int. Ed. 2000, 39, 3168.
The suppression of the A + B coupling relative to the B +
C coupling under the 3CR conditions could be due to
partitioning of the catalyst between the A + B and B + C
catalytic cycle, as both reactions are competing for the same
Pd catalyst pool. However, the reactions are also competing
for the same substrate, indole 2a (B). The sigmoidal shape of
the curve for [5] in the two-component A + B coupling is
characteristic of a consecutive reaction and suggests that
enone 7 is an intermediate (see Figure 1b).[13] The initial
delay, inherent in the consecutive nature of the A + B
coupling, gives this reaction a serious kinetic disadvantage
compared to the B + C coupling; by the time the concen-
tration of the intermediate (e.g. enone 7) reaches its peak,
most of the indole 2a (B) has already been consumed.
Consequently, the reaction will be diverted towards the A +
BC pathway, possibly with 6 (BC) competing with 2a (B) for
enone 7. Although the rate differences between the two-
component A + B and B + C experiments is small, these
differences are amplified in the three-component A + B + C
reaction because the A + B reaction is a consecutive reaction
with an induction period, whereas the B + C reaction starts to
consume indole 2a (B) without detectable delay.[14]
Such kinetic amplifications could be useful in designing
selective reactions and MCR processes, or for probing the
induction periods of reactions. For example, two competing
reactions with similar initial rates but different induction
periods should display significant rate differences in a com-
petition experiment for the same substrate, even in cases
where the intermediates themselves would escape detection.
In summary, palladium-catalyzed oxidative three-compo-
nent coupling between b-ketoesters, indoles, and arylboronic
acids proceeds under very mild reaction conditions with
excellent regioselectivity for the products. Even very hin-
dered compounds, such as atropisomeric products 4j and 4v,
can be accessed in good yields. Indole couples faster with the
arylboronic acid than the b-ketoester, and this rate difference
is significant under the three-component reaction conditions,
possibly because the coupling with b-ketoester involves an
enone intermediate. Further implications of this study on the
mechanism of both oxidative coupling reactions are being
studied and will be reported separately.
[4] L. F. Tietze, G. Brasche, K. M. Gericke, Domino Reactions in
Organic Synthesis, Wiley-VCH, Weinheim, 2006.
[5] During the preparation of this manuscript, Enders and co-
workers described an organocatalytic branched (i.e. split)
domino sequence: X. Zenge, Q. Ni, G. Raabe, D. Enders,
Angew. Chem. 2013, 125, 3050; Angew. Chem. Int. Ed. 2013, 52,
2977.
[6] Oxidative domino cross-couplings have been previously de-
scribed, but typically these reactions involve only two compo-
nents. For selected examples, see: a) S. Hajdok, J. Conrad, U.
Beifuss, J. Org. Chem. 2012, 77, 445; b) K. Hackeloer, G.
Schnakenburg, S. R. Waldvogel, Org. Lett. 2011, 13, 916; for
a rare example with three components, see: c) L. Jiangsheng, C.
Feifei, L. Zhiwei, X. Yuan, C. Chao, L. Weidong, C. Zhong, Chin.
J. Chem. 2012, 30, 1699.
ꢀ
[7] Reactions that achieve the formation of two C C bonds at the
indole nucleus in a single operation are rare. For an example of
a cross-dehydrogenative coupling where the 2,3-disubstituted
products were generated in trace amounts, see: a) D. R. Stuart,
E. Villemure, K. Fagnou, J. Am. Chem. Soc. 2007, 129, 12072; for
an example of a stoichiometric 2,3-difunctionalization, see:
b) M. Ishikura, H. Kato, M. Ohnuki, Chem. Commun. 2002, 220;
c) M. Ishikura, H. Kato, Tetrahedron 2002, 58, 9827; for
examples of 2,3-difunctionalizations involving annulations, see:
d) G. Zhang, X. Huang, G. Li, L. Zhang, J. Am. Chem. Soc. 2008,
130, 1814; e) R. Liu, J. Zhang, Adv. Synth. Catal. 2011, 353, 36;
f) X. Han, H. Li, R. P. Hughes, J. Wu, Angew. Chem. 2012, 124,
10536; Angew. Chem. Int. Ed. 2012, 51, 10390; for a review of
catalytic functionalization of indoles, see: g) M. Bandini, A.
Eichholzer, Angew. Chem. 2009, 121, 9786; Angew. Chem. Int.
Ed. 2009, 48, 9608.
[8] a) M. V. Leskinen, K.-T. Yip, A. Valkonen, P. M. Pihko, J. Am.
Chem. Soc. 2012, 134, 5750; b) K.-T. Yip, R. Y. Nimje, M. V.
Leskinen, P. M. Pihko, Chem. Eur. J. 2012, 18, 12590.
[9] a) S.-D. Yang, C.-L. Sun, Z. Fang, B.-J. Li, Y.-Z. Li, Z.-J. Shi,
Angew. Chem. 2008, 120, 1495; Angew. Chem. Int. Ed. 2008, 47,
1473; for an alternative protocol with aryltrifluoroborates, see:
b) J. Zhao, Y. Zhang, K. Cheng, J. Org. Chem. 2008, 73, 7428.
[10] S. Kirchberg, R. Frçhlich, A. Studer, Angew. Chem. 2009, 121,
4299; Angew. Chem. Int. Ed. 2009, 48, 4235.
[11] For examples of atropisomers derived from indoles, see:
a) T. D. W. Claridge, J. M. Long, J. M. Brown, D. Hibbs, M. B.
Hursthouse, Tetrahedron 1997, 53, 4035; for a related benzimi-
dazole-derived atropisomer, see: b) A. Figge, H. J. Altenbach,
D. J. Brauer, P. Tielmann, Tetrahedron: Asymmetry 2002, 13, 137;
for atropisomers derived from 1,2,3-trisubstituted pyrroles, see:
c) S. Le Gac, N. Monnier-Benoit, L. D. Metoul, S. Peti, I. Jabin,
Tetrahedron: Asymmetry 2004, 15, 139; d) J. L. A. Webb, J. Org.
Chem. 1953, 18, 1413.
Received: January 30, 2013
Published online: April 2, 2013
[12] See the Supporting Information for details.
[13] In a three-component control experiment where 250 mol% of
enone 7 was used in lieu of b-ketoester 1a but no additional
oxidant was used, the reaction gave exclusively the A + B
Keywords: homogeneous catalysis · kinetics ·
multicomponent reactions · oxidative couplings · palladium
.
Angew. Chem. Int. Ed. 2013, 52, 4818 –4822
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4821

.
Angewandte
Communications
coupling product 5 (see the Supporting Information). This
experiment indicates that enone 7 may not be capable of
reoxidizing Pd0 in the catalytic cycle (in order to sustain the B +
C oxidative coupling) and suggests that the concentration of 7
controls the rate of the A + B coupling. For a previous discussion
of the mechanism of the indole-b-ketoester coupling, see
Ref. [8a].
species here (1a, 2a, 5, 6 and 7) could act as ligands for Pd. For
evidence of the involvement of indoles in the formation of 7, see
Ref. [8b] (see the Supporting Information, p. S40).
[15] CCDC 917055 (4q) contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
[14] We are aware that given the complexity of the system, other
factors are likely to play a role as well. For example, most of the
4822 www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 4818 –4822