Palladium-catalyzed enantioselective allylic alkylations through C-H activation

Article abstract of DOI:10.1002/anie.201207870

A new ligand class: The title reaction was made possible by the discovery of a new class of phosphoramidite ligands. A variety of sterically and electronically diverse allylarenes undergo reaction with 2-acetyl-1-tetralones to form quaternary carbon stereocenters. This is a conceptually and mechanistically distinct strategy from traditional methods for the synthesis of enantioenriched allylic substitution products. 2,6-DMBQ=2,6- dimethylbenzoquinone. Copyright

Full text of DOI:10.1002/anie.201207870

Angewandte
Chemie
DOI: 10.1002/anie.201207870
Synthetic Methods
Palladium-Catalyzed Enantioselective Allylic Alkylations through
C–H Activation**
Barry M. Trost,* David A. Thaisrivongs, and Etienne J. Donckele
Metal-catalyzed asymmetric allylic alkylation (AAA) reac-
tions, particularly those employing palladium,[1] copper,[2]
iridium,[3] or molybdenum,[4] have found broad utility in the
construction of a diverse array of enantioenriched products
from achiral starting materials. Irrespective of the catalyst,
however, one inviolable requirement for nearly all substrates
that undergo these transformations is the presence of an
allylic leaving group.[5] Such functionality permits an appro-
priate transition metal catalyst to transform one allylic
substituent into another through a redox-neutral event
(Scheme 1). One way to increase the efficiency and chemo-
selectivity of these reactions would be to break this paradigm
ity issues that arise from carrying these reactive functional
groups through a synthetic sequence by replacing them with
À
relatively inert C H bonds.
Stoichiometric palladium-mediated allylic alkylations of
À
alkenes that proceed through insertion into allylic C H bonds
are well known,[6] but until recently, reports of metal-
catalyzed allylic alkylations that did not require leaving
groups were exceedingly rare.[7] In 2008, Shi and co-workers
reported that Pd(OAc)2, in the presence of 1,2-bis(benzylsul-
finyl)ethane[8] and benzoquinone, catalyzes both the intra-
À
molecular allylic alkylation of activated and unactivated C H
bonds with b-dicarbonyl compounds and the corresponding
[9]
À
intermolecular allylic alkylation of activated C H bonds.
This latter finding was corroborated in a simultaneous
disclosure by White and Young, who reported that Pd(OAc)2,
in the presence of 1,2-bis(phenylsulfinyl)ethane and 2,6-
dimethylbenzoquinone, catalyzes the analogous intermolec-
À
ular allylic C–H alkylation of activated C H bonds with
methyl nitroacetate.[10] These authors later expanded the
scope of their work to include a similar intermolecular allylic
[11]
À
C–H alkylation of unactivated C H bonds.
There are,
however, no reports of stoichiometric or catalytic allylic
alkylations which proceed through C–H activation that are
asymmetric. Herein we describe the first examples of such
catalytic enantioselective allylic alkylations. These reactions
are additionally noteworthy, because they stereoselectively
generate quaternary carbon stereocenters, one of the most
challenging feats in enantioselective catalysis,[12] and they
employ prochiral nucleophiles, a class of substrates that chiral
allylic alkylation catalysts often struggle to enantiodiscrimi-
nate.[13]
Scheme 1. Comparison of traditional and oxidative palladium-catalyzed
allylic alkylations.
by developing an oxidative method to perform analogous
AAA reactions with unfunctionalized alkenes. Specifically,
the use of an allylic hydrogen atom as a leaving group would
not only eliminate the need to install the allylic esters,
carbonates, carbamates, halides, or phosphates necessary for
traditional AAA reactions, but also avert the chemoselectiv-
Given that the desired reactivity had been demonstrated,
the focus of efforts to render allylic C–H alkylations
enantioselective turned immediately to ligand design. Chiral
phosphorus-containing compounds are the most successful
class of ligands utilized in asymmetric catalysis.[14] At the
outset of our program, however, there were claims about the
unsuitable nature of phosphine ligands under the oxidative
conditions necessary for palladium-catalyzed allylic C–H
activation.[15] Nevertheless, the ubiquitous use of chiral
phosphines and our success with such ligands for palladium-
catalyzed AAA reactions led us to examine their use in allylic
C–H alkylations, and we subsequently demonstrated that
PPh3, a common phosphine ligand, does indeed promote
palladium-catalyzed allylic alkylation through C–H activation
in a wide array of contexts.[16] Buoyed by this success, we
evaluated many known chiral mono- and bidentate phospho-
rus ligand classes for analogous enantioselective reactions; all
such investigations were fruitless.[17]
[*] Prof. B. M. Trost, D. A. Thaisrivongs, E. J. Donckele
Department of Chemistry, Stanford University
Stanford CA 94305-5080 (USA)
E-mail: bmtrost@stanford.edu
Homepage: http://www.stanford.edu/group/bmtrost
[**] We thank the National Science Foundation for their support of our
programs and Johnson Matthey for generously providing us with
palladium salts. D.A.T. acknowledges support from a Stanford
Graduate Fellowship (The William K. Bowes, Jr. Foundation) and an
Eli Lilly and Company Graduate Fellowship. E.J.D. acknowledges an
Explo’ra Sup grant from CPE Lyon.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201207870.
Angew. Chem. Int. Ed. 2013, 52, 1523 –1526
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1523

.
Angewandte
Communications
Ultimately, we were guided by our groupꢀs previous
efforts in the design of ligands for asymmetric palladium-
catalyzed trimethylenemethane cycloadditions[18] to the dis-
covery of a novel class of pyroglutamic acid derived[19]
phosphoramidite ligands that promote palladium-catalyzed
allylic C–H alkylations. The preparation of this novel class of
chiral phosphoramidite ligands begins with pyroglutamic acid,
both enantiomers of which are readily commercially avail-
able. In a representative synthesis, esterification of l-pyro-
glutamic acid (1) with 1-naphthalene methanol using N,N’-
dicyclohexylcarbodiimide (DCC) in the presence of catalytic
4-dimethylaminopyridine (DMAP) followed by carbamate
formation with di-tert-butyl dicarbonate (Boc2O) affords
pyrrolidine 2 in 63% overall yield (Scheme 2). Chemoselec-
trifluoroacetic acid (TFA) liberates the unprotected pyrroli-
dine 5 in 74% yield; compound 5 undergoes base-promoted
coupling with a slight excess of chlorophosphite 6 to afford
representative ligand 7 in 54% yield. Importantly, ligands
prepared with the enantiomer of 6 were inferior both with
respect to reactivity and stereoselectivity. One noteworthy
feature of this synthesis with particular relevance to ligand
optimization is its flexibility. At the respective stages of the
modular sequence, different ester, aryl, and diol pieces could
be introduced, thereby allowing rapid access to a diversity of
ligand structures in the service of structure–activity relation-
ship studies. Use of (S)-BINOL with the (S,S)-pyrrolidine
proved to be the matched case.
Experiments designed to evaluate a series of ligands in
this class employed equimolar amounts of 2-acetyl-1-tetra-
lone (8), allylbenzene, 2,6-dimethylbenzoquinone (2,6-
DMBQ), and Et3N in the presence of Pd(OAc)2 (5.0 mol%)
and the ligand (7.5 mol%) in THF at 608C for 24 h. When the
pyrrolidine subunit is substituted with a phenyl ring (10),
allylic C–H alkylation product 9 is obtained in 83% con-
version and 72% ee (Scheme 3).[20] Substitution on this
aromatic ring is not well tolerated. A reaction with the
corresponding para-biphenyl ligand 11 gave 9 in only 14%
conversion and 30% ee, and the meta-biphenyl analogue 12
gave no desired product at all. The meta-terphenyl ligand 13
yielded 9 in 70% ee, but only at 10% conversion. Further
increasing the steric bulk is deleterious, as the reaction with
Scheme 2. Synthesis of chiral phosphoramidite ligands based on
pyroglutamic acid.
tive semireduction of the lactam carbonyl with lithium
triethylborohydride and then treatment with methanol and
catalytic amounts of para-toluenesulfonic acid monohydrate
provides the corresponding aminal 3 in 37% yield. Copper(I)
bromide dimethyl sulfide mediated addition of phenylmag-
nesium bromide in the presence of boron trifluoride diethyl
etherate gives the desired trans-substituted pyrrolidine 4 in
83% yield. Notably, in every such reaction we have per-
formed, the diastereocontrol of the cuprate addition has been
complete. Cleavage of the tert-butyl carboxyl group with
Scheme 3. Palladium-catalyzed enantioselective allylic alkylations
through C–H activation with selected phosphoramidite ligands.
1524
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1523 –1526

Angewandte
Chemie
the 3,5-di-tert-butylphenyl ligand 14 gave no desired product.
Neither the 1-naphthyl (15) or 2-naphthyl (16) ligands offered
improvement in terms of conversion or enantioselectivity, and
replacement of the aromatic ring with cyclohexane (17)
provided 9 in only 29% conversion and 25% ee.
The effect of the ester substituent of 10 on both the
reactivity and enantioselectivity of the allylic C–H alkylation
was next examined (Scheme 3). Replacing the methyl with
either a phenyl (18) or 2-naphthyl (19) group gave inactive
catalysts, presumably because these sterically demanding
groups inhibit the C–H activation event. However, substitu-
tion with a benzyl group (20) yielded 9 in 43% conversion and
85% ee. The reaction conversion increased to 100% with the
corresponding benzylic 2-naphthyl ligand 21, and the product
was obtained in 55% ee. With a benzylic 1-naphthyl ester (7),
the reaction gave 9 in 54% conversion and 89% ee.
Although the palladium-catalyzed allylic C–H alkylation
could be conducted with high enantioselectivity, incomplete
conversion was a significant problem, and it was attributed at
least in part to catalyst decomposition. An extensive screen of
11 solvents from methylene chloride through to DMSO and
DMF showed THF to be superior. At 408C, ee values as high
as 97% but only 53% conversion were observed, while at
808C complete conversion occurred but the ee dropped to
76%. Varying concentration and substrate ratios beyond 1:1
showed no beneficial effect. Through such optimization
studies, it was found that when the ligand loading was
increased to 10 mol%, the temperature decreased to 508C,
and the catalyst dosed in two portions at the beginning of the
reaction and after three hours, the desired product 9 could be
isolated in 89% yield and 85% ee (Scheme 4).
Conducting the reaction with the corresponding allyl-
para-biphenyl electrophile gave 22 in 70% yield and 67% ee,
while the para-fluoro analogue provided 23 in 75% yield and
83% ee. In a demonstration of the mildness of the reaction
conditions, 24 was obtained in 84% yield and 71% ee despite
the presence of an aldehyde. The corresponding methyl ester
and dimethyl amide gave 25 in 83% yield and 69% ee and 26
in 85% yield and 74% ee, respectively. Though the para-
trifluoromethyl derivative provided 27 in 76% yield and 66%
ee, the corresponding nitrile electrophile fails to react,
presumably because palladium coordination to the nitrogen
lone pair is inhibitory.
Highly electron-rich aromatic rings, such as that present in
29, disfavor the C–H activation event, but moderately
electron-rich substrates are tolerated, as 30 was isolated in
61% yield and 79% ee. When the para-methoxy group is
replaced with a meta-methoxy group, the strong p-donating
nature of the substituent is overwhelmed by its s-withdrawing
ability, and the correspondingly electron-deficient ring par-
ticipates in the desired reaction to provide 31 in 63% yield
and 79% ee. The meta-methyl substrate gave 32 in 77% yield
and 79% ee, and the corresponding 3,5-difluoro substrate
provided 33 in 59% yield and 65% ee. Steric bulk in the
ortho-position relative to that undergoing C–H activation is
generally not tolerated, as the failure to obtain 34 shows, but
2-allylnaphthalene reacted to give 35 in 92% yield and 71%
ee. Electron-rich and electron-poor 2-acetyl-1-tetralones also
participate in the allylic C–H alkylation: reaction with 2-
Scheme 4. Substrate scope of palladium-catalyzed enantioselective
allylic alkylations through C–H alkylation. Catalyst added in two
portions: 2.5 mol% Pd(OAc)2 and 5 mol% 7 at the beginning of the
reaction and 2.5 mol% Pd(OAc)2 and 5 mol% 7 after 3 h; reactions
performed on a 0.100 mmol scale. The absolute stereochemistry of the
products was assigned by analogy to 9, the configuration of which was
determined to be S by a comparison of its optical rotation to that
reported in a previous synthesis.[21] See the Supporting Information for
details.
acetyl-6-methoxy-1-tetralone provided 36 in 82% yield and
78% ee and 2-acetyl-7-nitro-1-tetralone analogously gave 37
in 91% yield and 78% ee. Notably, when the reaction was
conducted in the absence of a phosphoramidite ligand, no
desired product was observed, thus highlighting the essential
role of the ligand in both enabling the desired reactivity and
controlling the stereochemical course of the reaction.
In conclusion, we report the first examples of catalytic
enantioselective allylic C–H alkylations, a transformation to
date unknown with any metal and one that now provides
a complementary approach to traditional methods for the
synthesis of enantioenriched allylic substitution products.
This achievement was enabled by the discovery of a new class
of phosphoramidite ligands, which promote both the palla-
dium-catalyzed C–H activation and the subsequent alkylation
event. Investigations into other C–H activation reactions
employing this unique ligand class are ongoing.
Angew. Chem. Int. Ed. 2013, 52, 1523 –1526
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1525

.
Angewandte
Communications
[7] For a singular example of a palladium(0)-catalyzed allylic C – H
alkylation of propene and 1-butene under reducing conditions,
see: L. S. Hegedus, T. Hayashi, W. H. Darlington, J. Am. Chem.
Soc. 1987, 100, 7747 – 7748. For a non-asymmetric copper- and
cobalt-catalyzed cross-dehydrogenative-coupling reaction of
Experimental Section
General procedure for the palladium-catalyzed enantioselective
allylic alkylation through C–H activation (1.00 mmol scale): A
reaction vial equipped with a stir bar was charged with the Pd(OAc)2
(5.60 mg, 0.030 mmol) and 7 (32.3 mg, 0.050 mmol). To a second
reaction vial equipped with a stir bar was added 2-acetyl-1-tetralone
(8, 188 mg, 1.00 mmol), 2,6-dimethylbenzoquinone (136 mg,
À
allylic C H bonds and activated methylenes, see: Z. Li, C.-J.
Li, J. Am. Chem. Soc. 2006, 128, 56 – 57.
[8] M. S. Chen, M. C. White, J. Am. Chem. Soc. 2004, 126, 1346 –
1347.
[9] S. Lin, C.-X. Song, G.-X. Cai, W.-H. Wang, Z.-J. Shi, J. Am.
Chem. Soc. 2008, 130, 12901 – 12903.
[10] A. J. Young, M. C. White, J. Am. Chem. Soc. 2008, 130, 14090 –
14091.
[11] A. J. Young, M. C. White, Angew. Chem. 2011, 123, 6956 – 6959;
Angew. Chem. Int. Ed. 2011, 50, 6824 – 6827.
[12] a) J. Christoffers, A. Mann, Angew. Chem. 2001, 113, 4725 –
4732; Angew. Chem. Int. Ed. 2001, 40, 4591 – 4597; b) I.
Denissova, L. Barriault, Tetrahedron 2003, 59, 10105 – 10146;
c) P. G. Cozzi, R. Hilgraf, N. Zimmermann, Eur. J. Org. Chem.
2007, 5969 – 5994; d) M. Bella, T. Gasperi, Synthesis 2009, 1583 –
1614.
[13] B. M. Trost, M. R. Machacek, A. Aponick, Acc. Chem. Res. 2006,
39, 747 – 760.
[14] Phosphorus Ligands in Asymmetric Catalysis: Synthesis and
Applications (Ed.: A. Bçrner), Wiley-VCH, New York, 2008.
[15] D. J. Covell, M. C. White, Angew. Chem. 2008, 120, 6548 – 6551;
Angew. Chem. Int. Ed. 2008, 47, 6448 – 6451.
[16] a) B. M. Trost, M. M. Hansmann, D. A. Thaisrivongs, Angew.
Chem. 2012, 124, 5034 – 5037; Angew. Chem. Int. Ed. 2012, 51,
4950 – 4953; b) B. M. Trost, D. A. Thaisrivongs, Angew. Chem.
2012, 124, 11690 – 11694; Angew. Chem. Int. Ed. 2012, 51, 11522 –
11526.
[17] For a representative sample of ligands surveyed, see the
Supporting Information.
[18] B. M. Trost, J. P. Stambuli, S. M. Silverman, U. Schwçrer, J. Am.
Chem. Soc. 2006, 128, 13328 – 13329.
[19] The relative stereochemistry of these transformations was
assigned by analogy to reactions described by Wistrand and
Skrinjar. With these sets of reaction conditions, trans-substituted
pyrrolidines are always the major products. See: L.-G. Wistrand,
M. Skrinjar, Tetrahedron 1991, 47, 573 – 582. Similarly trans-
selective cuprate additions have been observed by many groups.
For examples, see: a) I. Collado, J. Ezquerra, C. Pedregal, J. Org.
Chem. 1995, 60, 5011 – 5015; b) Y. Tong, Y. M. Fobian, M. Wu,
N. D. Boyd, K. D. Moeller, J. Org. Chem. 2000, 65, 2484 – 2493;
c) R. G. Vaswani, A. R. Chamberlin, J. Org. Chem. 2008, 73,
1661 – 1681. See the Supporting Information for details.
[20] For the determination of the absolute stereochemistry of the
allylic C–H alkylation products, see the Supporting Information.
[21] R. Kuwano, K. Uchida, Y. Ito, Org. Lett. 2003, 5, 2177 – 2179.
1.00 mmol), Pd(OAc)2 (5.60 mg, 0.030 mmol), and
7 (32.3 mg,
0.050 mmol). The vials were sealed, connected with a cannula, and
evacuated and filled with Ar three times. THF (2.44 mL) was added
to the latter vial, followed by Et3N (0.139 mL, 1.00 mmol) and 4-
allyltoluene (0.132 mL, 1.00 mmol). The yellow solution was heated
to 508C and stirred for three hours, at which point THF (2.44 mL) was
added to the remaining vial and the catalyst solution added through
a cannula to the reaction, which was stirred at 508C for additional
21 h. After cooling to room temperature, the solution was concen-
trated and the crude material was purified by flash chromatography
on silica gel (CH2Cl2/Et2O = 9:1) to give the product (30, 194 mg,
59% yield, 79% ee) as a colorless oil.
Received: September 29, 2012
Revised: November 27, 2012
Published online: December 11, 2012
Keywords: allylic alkylation · C–H activation · enantioselectivity ·
.
homogeneous catalysis · palladium
[1] a) B. M. Trost, D. L. Van Vranken, Chem. Rev. 1996, 96, 395 –
422; b) B. M. Trost, M. L. Crawley, Chem. Rev. 2003, 103, 2921 –
2944; c) B. M. Trost, J. Org. Chem. 2004, 69, 5813 – 5837; d) Z.
Lu, S. Ma, Angew. Chem. 2008, 120, 264 – 303; Angew. Chem. Int.
Ed. 2008, 47, 258 – 297; e) B. M. Trost, T. Zhang, J. D. Sieber,
Chem. Sci. 2010, 1, 427 – 440.
[2] a) H. Yorimitsu, K. Oshima, Angew. Chem. 2005, 117, 4509 –
4513; Angew. Chem. Int. Ed. 2005, 44, 4435 – 4439; b) A.
Alexakis, J. E. Bꢁckvall, N. Krause, O. Pꢂmies, M. Diꢃguez,
Chem. Rev. 2008, 108, 2796 – 2823; c) C. A. Falciola, A. Alexakis,
Eur. J. Org. Chem. 2008, 3765 – 3780.
[3] a) G. Helmchen, A. Dahnz, P. Dꢄbon, M. Schelwies, R.
Wiehofen, Chem. Commun. 2007, 675 – 691; b) J. F. Hartwig,
L. M. Stanley, Acc. Chem. Res. 2010, 43, 1461 – 1475.
[4] B. M. Trost, J. R. Miller, C. M. Hoffman, Jr., J. Am. Chem. Soc.
2011, 133, 8165 – 8167, and references therein.
[5] Palladium-catalyzed AAA reactions with allene electrophiles
proceed through allene hydropalladation rather than ionization
of an allylic leaving group. For examples, see: a) B. M. Trost, C.
Jꢁkel, B. Plietker, J. Am. Chem. Soc. 2003, 125, 4438 – 4439;
b) B. M. Trost, J. Xie, J. Am. Chem. Soc. 2006, 128, 6044 – 6045.
[6] a) J. Tsuji, Acc. Chem. Res. 1969, 2, 144 – 152; b) B. M. Trost,
Tetrahedron 1977, 33, 2615 – 2649.
1526
www.angewandte.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2013, 52, 1523 –1526