Mechanistic insight into the regioselective palladation of indole derivatives: Tetranuclear indolyl palladacycles with high C2-Pd or C3-Pd bond selectivity

Article abstract of DOI:10.1021/om300284t

This research provides a straightforward understanding of the regioselective palladation of indole derivatives by capturing both C2-Pd and C3-Pd intermediates of N-phenylindole and N-methylindole. Those tetranuclear indolyl palladacycles suggest that the regioselectivity of palladation depends on the N-substituted protective groups of indoles as well as the acidity of the reaction medium.

Full text of DOI:10.1021/om300284t

Communication
pubs.acs.org/Organometallics
Mechanistic Insight into the Regioselective Palladation of Indole
Derivatives: Tetranuclear Indolyl Palladacycles with High C2−Pd
or C3−Pd Bond Selectivity
Yang Li,† Wen-Hua Wang,† Ke-Han He,† and Zhang-Jie Shi*,†,‡
†Beijing National Laboratory of Molecular Sciences (BNLMS) and Key Laboratory of Bioorganic Chemistry and Molecular
Engineering of Ministry of Education, College of Chemistry and Molecular Engineering, Peking University, Beijing,
People's Republic of China
‡State Key Laboratory of Organometallic Chemistry, SIOC, CAS, Shanghai 200032, People's Republic of China
S
* Supporting Information
ABSTRACT: This research provides a straightforward under-
standing of the regioselective palladation of indole derivatives
by capturing both C2−Pd and C3−Pd intermediates of
N-phenylindole and N-methylindole. Those tetranuclear indolyl
palladacycles suggest that the regioselectivity of palladation
depends on the N-substituted protective groups of indoles as
well as the acidity of the reaction medium.
s a ubiquitous heterocyclic structural unit, the indolyl
scaffold exists in many natural products and pharmaceut-
studies. Herein, our research gives an understanding of the
regioselective palladation of indole derivatives by capturing
both C2−Pd and C3−Pd intermediates of N-phenylindole and
N-methylindole.
A
ical agents, and various methods have been developed to
functionalize indole derivatives.1 Direct functionalization of the
indolyl core by C−H transformation provides highly efficient
access to complicated indolyl analogues.2 However, challenges
still remain ,because both the chemo- and regioselectivities
must be well controlled by avoiding the formation of intractable
isomeric mixtures. Among the transition-metal-catalyzed direct
functionalizations of indolyl derivatives, catalytic systems with
Pd have shown excellent catalytic ability and regioselectivity.3−6
Particularly, in some cases, completely controllable C2 or C3
selectivity from the same substrate was observed by simple
tuning the reaction parameters.7
Some mechanistic studies were investigated to understand
the regioselectivity of the indolyl functionalization.7b,8 During
this research, Sames and his co-workers have carried out
comprehensive investigations on N-methylindolyl derivatives as
model substrates.7b To provide a rationale for the observed C2
regioselectivity, they presented an electrophilic palladation
pathway, accompanied by a 1,2-migration. After that, DeBoef
and his co-workers also described a systematic concerted
metalation deprotonation (CMD) process to approach the C2
regioselectivity by using the N-(2-(trimethylsilyl)ethoxy)methyl
(SEM) indolyl derivatives as model substrates. However, no
intermediates were isolated during these detailed mechanistic
In accord with the reported PdII-catalyzed highly regiose-
lective functionalization of indoles, it seems that basic
conditions favor C3 selectivity.5 In contrast, C2 selectivity is
mostly observed under acidic conditions.6 On the basis of these
investigations, N-phenylindole (1a) was selected for our initial
investigation due to its rigidity, proper electron density, and
relatively lower reactivity. These characteristics of 1a may help
to stabilize the desired intermediates. The unsubstituted C2
and C3 positions of 1a provided potential active sites to
investigate the regioselectivity. To obtain the C3 palladation
product, sodium pivalate was chosen as a proper base because
of the following factors. First, in some cases of direct
functionalization of indoles, the promoted efficiency was
observed in the presence of pivalyl ligand.5b,7c,8a,b,9 Second,
the bulky pivalyl ligand may play a key role in the stabilization
of Pd−indole species. Third, the lipophilic pivalyl ligand might
facilitate the precipitation of Pd−indole intermediates from a
polar solvent.
Received: April 8, 2012
Published: June 6, 2012
© 2012 American Chemical Society
4397
dx.doi.org/10.1021/om300284t | Organometallics 2012, 31, 4397−4400

Organometallics
Communication
Treatment of 1a with [Pd(OAc)2]3 (2a) in the presence of
PivONa in MeOH at room temperature generated, as we
predicted, the C3 palladation compound 10 as the major
product in 58% isolated yield, accompanied by 7−9 as minor
products (Scheme 1). The C3 palladation could be attributed
to the high electron density at the C3 position of the indolyl
derivatives. Meanwhile, no C2 palladation compound was ob-
served. Gratifyingly, X-ray-quality crystals of 10 were obtained
by slow volatilization of its DCM and acetone solution at 5 °C.
Simultaneously, crystals of 7−9 were also obtained from the
mother liquor of the above reaction mixture.
Scheme 1. Regioselective Electrophilic Palladation at the C3
a
Position of N-Phenylindole
Compound 10 is a novel tetranuclear palladacycle with 1a as
bridges (Figure 1). Pd2 shows a typical σ sp2 Pd2−C3 bond. In
comparison with Pd2, Pd1 engages in M−π interactions with
the C2C3 moiety. This PdII−π (C2C3) interaction pres-
ents a η2 coordination with the deprotonated C3 atom and
proton-bearing C2 atom. Meanwhile, C2−H is kept untouched
during the electrophilic palladation of N-phenylindole. To the
best of our knowledge, a structurally characterized PdII−indole
complex including indole with unsubstituted C2 and C3
positions has not been reported.10,11 In addition, the co-
existence of σ and π Pd−C bonding models in M−indole
complexes was first observed in compound 10.12
On the basis of the X-ray structure characterization of
compounds 7−10, the pathway to C3 palladation of 1 is pro-
posed as follows (Figure 1): (1) 2a [Pd3(L1)6] (L1 = OAc−)
initially undergoes a ligand exchange process to form the new
trinuclear palladacycle 7 [Pd3(L2)6] (L2 = PivO−); (2) the
tetranuclear palladacycle 8 [Pd4(L2)6(L3)2] (L2 = PivO−, L3 =
MeO−) is derived from reorganization of the trinuclear
palladacycle 7 in the presence of MeOH; (3) two carboxylate
ligands of tetranuclear palladacycle 8 [Pd4(L2)6(L3)2] disso-
ciate, followed by the electrophilic substitution of 8 with 1a
at the C3 position. Then a proper base facilitated the
deprotonation of indole derivatives to generate the sp2 C3−
Pd species [Pd4(L1)2(L2)2(L3)2(indole)2] (10; L1 = L2 = PivO−,
L3 = MeO−, indole = N-phenylindole). The formation of C3
palladation compound 9 is suggested by a similar pathway, in
which an incomplete ligand exchange is involved.
Subsequently, we focused our efforts to obtain Pd−indole
complexes with metalation at the C2 position. After many trials
a
Gray spheres represent t-Bu.
Figure 1. Crystal structures of 7−10 and the possible pathway to 10 through C3 palladation of N-phenylindole (L1 = OAc, L2 = PivO, L3 = MeO−).
Disordered atoms, some hydrogen atoms, and the solvent molecule are omitted for clarity. The crystal structure of 2a is based on CCDC 262708.13
4398
dx.doi.org/10.1021/om300284t | Organometallics 2012, 31, 4397−4400

Organometallics
Communication
Scheme 2. pH-Dependent Highly Regioselective Electrophilic Palladation at the C2 or C3 Position of N-Methylindole and X-ray
a
Crystal Structures of Compounds 12 and 13
a
Disordered atoms and some hydrogen atoms are omitted for clarity.
under different pH conditions, we failed to isolate our expected
indolyl C-2 palladacycles. Interestingly, the unexpected C3
palladation compound 11 was observed and isolated in 34%
yield when compound 1a was treated with 2a in the presence of
PivOH at 45 °C (Scheme 1). Its structure shows a similarity
with compound 10 by an analysis of the single-crystal structure
(see the Supporting Information). The difference is that the
indole molecules are bridged by the chlorides. At the same
time, no C2 palladation compound was detected.
Thus, we failed to detect the expected C2−Pd indolyl
palladacycles under acidic conditions. Possibly, this result is due
to the electronic characteristics of the phenyl group, which
induces the lower reactivity of N-phenylindole (1a). The steric
hindrance of the N-phenyl group might be another factor
inhibiting the C2 palladation. To prove the regioselectivity
based on the pH of the reaction medium in many reactions
by the Pd intermediates, we changed the substrate from
N-phenylindole (1a) to N-methylindole (1b), due to its high
reactivity.2
By treatment of 1b with 2a in the presence of PivONa as a
base in MeOH, C3 indolyl palladacycle 12 was obtained as a
dark red solid smoothly in 34% isolated yield (Scheme 2).
Crystals of 12 suitable for X-ray analysis were also obtained by
slow volatilization of its DCM and acetone solution at 5 °C.
X-ray crystallography of 12 shows a tetranuclear palladacycle
similar to compound 10. No characteristic peak of a C2 indolyl
palladacycle (a single peak at about 6.2 ppm assigned to
C3−H) was observed in the remaining unidentified complexes
and the original reaction mixture.
We further tried to obtain the Pd−indole complex with
palladation at the C2 position with 1b under acidic conditions.
First, the reaction was conducted in AcOH; compound 1b
was quickly consumed, accompanied by the precipitation of
Pd black. After many trials, we found that, by treatment of 1b
with 2a in PivOH at 45 °C in the presence of NaCl, a distinct
red-brown solid could be isolated in 15% yield. To our delight,
the structure of this red-brown solid was clearly identified as the
C2 indolyl palladacycle 13 by X-ray crystal analysis (Scheme 2).
No characteristic peak of a Pd−indole C3 complex (a single
peak at about 9.2 ppm assigned to C2−H) was observed in the
1H NMR spectrum of the remaining unidentified complexes.
Under acidic conditions with N-methylindole as a substrate,
complete inversion of the regioselectivity from C3 to C2
palladation was observed.
In the crystal structure of complex 13, Pd1 presents a typical
σ sp2 C2−Pd bond. Pd2 is engaged in T-shaped bonding to the
C3 position. The Pd2−π(C3) interaction presents a η1
coordination with the proton-bearing C3 atom. The Pd2−H3
distance (1.89 Å) and Pd2−C3−H3 angle (66.5°) indicate
some degree of η2 Pd2−H3−C3 agostic-type bonding, which is
a meager compensation for the loss of Pd−C2 π bonding.14
Furthermore, comparison of 12 and 13 indicates the
following. (1) Different T-shaped PdII−π interactions were
discovered in 12 and 13. The C2C3 bond length (1.41(3) Å
in 12) in a η2 PdII−π(C2C3) pattern is obviously elongated
as compared with the observed value (1.38(2) Å in 13) in a η1
PdII−π(C3) pattern, and both of them are longer than that in
free indole (1.365(9) Å).10 This may be a consequence of
PdII−π interactions giving weak C2C3 bonds. (2) Pd2−C3
distances are shorter than Pd2−C2 distances in 12 and 13,
which demonstrates that Pd2 atoms are engaged more with C3
than with C2. This effect probably results from the first contact
of Pd with indole occurring at the “nucleophilic” C3 position.
The productive C−H activation on the indole nucleus likely
benefits from the T-shaped Pd−π interactions. (3) The square-
planar coordination structures of Pd ions are achieved with
4399
dx.doi.org/10.1021/om300284t | Organometallics 2012, 31, 4397−4400

Organometallics
Communication
Chem. 2007, 72, 1476. (e) Bellina, F.; Cauteruccio, S.; Rossi, R. Eur. J.
Org. Chem. 2006, 2006, 1379. (f) Jiao, L.; Bach, T. J. Am. Chem. Soc.
2011, 133, 12990.
bridging carboxyl and methoxy ligands in 12 and bridging
carboxyl and chlorine in 13, respectively. In addition, the
methoxy bridges two Pd ions which are connected to the same
indole in compound 12, whereas the chloride bridges two Pd
ions which are connected with different indole molecules in 13.
(4) Compound 12 shows a structure characterized by Ci
symmetry, which has a tetranuclear palladacycle scaffold with
two trans N-methylindole groups as bridging units. 13 shows a
structure characterized by C2 symmetry, which has a
tetranuclear palladacycle scaffold with two cis N-methylindoles
as bridging units. These studies are consistent with the previous
reports on direct functionalization of N-methylindoles via Pd
catalysis.
In summary, to unveil the mechanism of the regioselectivity
in the direct functionalization of indole and its derivatives,
several novel tetranuclear indolyl palladacycles were synthe-
sized and structurally determined by X-ray crystallography.
These Pd−indole complexes provide a straightforward under-
standing of the regioselective palladation. Our studies also
suggest that the regioselectivity of palladation depends on not
only the N-substituted protective groups of indoles but also the
pH of the medium. These observations are beneficial for an
understanding of previously developed direct functionalization
of indole derivatives via Pd catalysis. Meanwhile, it is also
important to design new transformations toward the direct
C−H functionalization of indolyl derivatives. Hopefully, this
investigation also offers helpful insights into the regio- and
chemoselectivity of electrophilic metalation of indolyl com-
pounds with other metals and the electrophilic activation of
C−H bonds of a normal aromatic system.
(4) For palladium-catalyzed direct C3 functionalizations, see:
(a) Akita, Y.; Itagaki, Y.; Takizawa, S.; Ohta, A. Chem. Pharm. Bull.
1989, 37, 1477. (b) Lane, B. S.; Brown, M. A.; Sames, D. J. Am. Chem.
Soc. 2005, 127, 8050. (c) Djakovitch, L.; Dufaud, V.; Zaidi, R. Adv.
Synth. Catal. 2006, 348, 715. (d) Djakovitch, L.; Rouge, P.; Zaidi, R.
Catal. Commun. 2007, 8, 1561. (e) Villar, A.; Kaspar, L. T.; Cornella,
J.; Lu, P.; Larrosa, I. Org. Lett. 2009, 11, 5506. (f) Li, Y.; Wang, W.-H.;
Yang, S.-D.; Li, B.-J.; Feng, C.; Shi, Z.-J. Chem. Commun. 2010, 46,
4553.
(5) For palladium-catalyzed direct C3 functionalizations under basic
conditions, see: (a) Bandini, M.; Melloni, A.; Umani-Ronchi, A. Org.
Lett. 2004, 6, 3199. (b) Stuart, D. R.; Fagnou, K. Science 2007, 316,
1172. (c) Zhang, Z.; Hu, Z.; Yu, Z.; Lei, P.; Chi, H.; Wang, Y.; He, R.
Tetrahedron Lett. 2007, 48, 2415. (d) Bellina, F.; Benelli, F.; Rossi, R.
J. Org. Chem. 2008, 73, 5529. (e) Cusati, G.; Djakovitch, L. Tetrahedron
Lett. 2008, 49, 2499. (f) Gu, Y.; Wang, X.-M. Tetrahedron Lett. 2009,
50, 763. (g) Ackermann, L.; Barfusser, S. Synlett 2009, 808. (h) Yan,
G.; Kuang, C.; Zhang, Y.; Wang, J. Org. Lett. 2010, 12, 1052. (i) Choy,
P. Y.; Lau, C. P.; Kwong, F. Y. J. Org. Chem. 2011, 76, 80. (j) Wang, Z.;
Li, K.; Zhao, D.; Lan, J.; You, J. Angew. Chem., Int. Ed. 2011, 50, 5365.
(k) Yamaguchi, A. D.; Mandal, D.; Yamaguchi, J.; Itami, K. Chem. Lett.
2011, 40, 555.
(6) For palladium-catalyzed direct C2 functionalizations under acidic
conditions, see: (a) Deprez, N. R.; Kalyani, D.; Krause, A.; Sanford, M.
S. J. Am. Chem. Soc. 2006, 128, 4972. (b) Dwight, T. A.; Rue, N. R.;
Charyk, D.; Josselyn, R.; DeBoef, B. Org. Lett. 2007, 9, 3137.
(c) Lebrasseur, N.; Larrosa, I. J. Am. Chem. Soc. 2008, 130, 2926.
(d) Zhao, J.; Zhang, Y.; Cheng, K. J. Org. Chem. 2008, 73, 7428.
(e) Yang, S.-D.; Sun, C.-L.; Fang, Z.; Li, B.-J.; Li, Y.- Z.; Shi, Z.-J.
Angew. Chem., Int. Ed. 2009, 47, 1473. (f) Zhou, J.; Hu, P.; Zhang, M.;
Huang, S.; Wang, M.; Su, W. Chem. Eur. J. 2010, 16, 5876. (g) Liang,
Z.; Yao, B.; Zhang, Y. Org. Lett. 2010, 12, 3185. (h) Yang, L.; Zhao, L.;
Li, C.-J. Chem. Commun. 2010, 46, 4184.
(7) For transition-metal-catalyzed direct regioselective C2 and C3
functionalizations, see: (a) Grimster, N. P.; Gauntlett, C.; Godfrey, C.
R. A.; Gaunt, M. J. Angew. Chem., Int. Ed. 2005, 44, 3125. (b) Lane, B.
S.; Brown, M. A.; Sames, D. J. Am. Chem. Soc. 2005, 127, 8050.
(c) Stuart, D. R.; Villemure, E.; Fagnou, K. J. Am. Chem. Soc. 2007,
129, 12072. (d) Phipps, R. J.; Grimster, N. P.; Gaunt, M. J. J. Am.
Chem. Soc. 2008, 130, 8172. (e) Joucla, L.; Batail, N.; Djakovitch, L.
Adv. Synth. Catal. 2010, 352, 2929. (f) Shi, Z.; Cui, Y.; Jiao, N. Org.
Lett. 2010, 12, 2908. (g) Liang, Z.; Zhao, J.; Zhang, Y. J. Org . Chem.
2010, 75, 170.
ASSOCIATED CONTENT
* Supporting Information
■
S
Text, figures, tables, and CIF files giving experimental details,
characterization data, and X-ray crystal structure data. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Tel and fax: (+86) 10-6276-0890. E-mail: zshi@pku.edu.cn.
■
(8) (a) Potavathri, S.; Pereira, K. C.; Gorelsky, S. I.; Pike, A.; LeBris,
Notes
́
A. P.; DeBoef, B J. Am. Chem. Soc. 2010, 132, 14676. (b) Liegault, B.;
The authors declare no competing financial interest.
Petrov, I.; Gorelsky, S. I.; Fagnou, K. J. Org. Chem. 2010, 75, 1047.
(c) Meir, R; Kozuch, S.; Uhe, A.; Shaik, S. Chem. Eur. J. 2011, 17,
7623.
ACKNOWLEDGMENTS
■
Support of this work by the “973” Project from the MOST of
China (2009CB825300) and the NSFC (Nos. 20902006,
21002001) is gratefully acknowledged.
(9) (a) Wang, X.; Lane, B. S.; Sames, D. J. Am. Chem. Soc. 2005, 127,
4996.
(10) Onitsuka, K.; Yamamoto, M.; Suzuki, S.; Takahashi, S.
Organometallics 2002, 21, 581.
(11) (a) Shimazaki, Y.; Yokoyama, H.; Yamauchi, O. Angew. Chem.,
Int. Ed. 1999, 38, 2401. (b) Conway, B.; Hevia, E.; Kennedy, A. R.;
Mulvey, R. E. Chem. Commun. 2007, 2864.
(12) Shimazaki, Y.; Yajima, T.; Takani, M.; Yamauchi, O. Coord.
Chem. Rev. 2009, 479.
REFERENCES
■
(1) For selected reviews, see: (a) Cacchi, S.; Fabrizi, G. Chem. Rev.
2005, 105, 2873; 2011, 111, PR215 (update 1). (b) Kruger, K.;
̈
Tillack, A.; Beller, M. Adv. Synth. Catal. 2008, 350, 2153. (c) Bandini,
M.; Eichholzer, A. Angew. Chem., Int. Ed. 2009, 48, 9608. (d) Bartoli,
G.; Bencivenni, G.; Dalpozzo, R. Chem. Soc. Rev. 2010, 39, 4449.
(2) For reviews of direct functionalization of indolyl core by C−H
activation, see: (a) Djakovitch, L. Adv. Synth. Catal. 2009, 351, 673.
(b) Beck, E. M.; Gaunt, M. J. Top. Curr. Chem. 2010, 292, 85.
(3) For palladium-catalyzed direct C2 functionalizations, see:
(a) Akita, Y.; Inoue, A.; Yamamoto, K.; Ohta, A.; Kurihara, T.;
Shimizu, M. Heterocycles 1985, 23, 2327. (b) Lane, B. S.; Sames, D.
Org. Lett. 2004, 6, 2897. (c) Toure, B. B.; Lane, B. S.; Sames, D. Org.
Lett. 2006, 8, 1979. (d) Wang, X.; Gribkov, D. V.; Sames, D. J. Org.
(13) Bakhmutov, V. I.; Berry, J. F.; Cotton, F. A.; Ibragimov, S.;
Murillo, C. A. Dalton Trans. 2005, 1989.
(14) Catellani, M.; Mealli, C.; Motti, E.; Paoli, P.; Perez-Carreno, E.;
̃
Pregosin, P. S. J. Am. Chem. Soc. 2002, 124, 4336.
4400
dx.doi.org/10.1021/om300284t | Organometallics 2012, 31, 4397−4400