Carbocyclization of unsaturated thioesters under palladium catalysis

Article abstract of DOI:10.1039/c3sc50486g

An intramolecular thioester-olefin cross-coupling reaction for the preparation of cyclic ketone structures from unsaturated thioesters is described. This method capitalizes on the unique reactivity of thioesters with low-valent palladium catalysts and copper(i) salts to form acyl-metal species. Trapping of such intermediates with alkene functional groups generates the corresponding exo-methylene cycloalkanone products. Unsaturated thioester substrates, which can be accessed using a suite of modern synthetic methods, can now be regarded as precursors to complex carbocyclic derivatives.

Full text of DOI:10.1039/c3sc50486g

Chemical Science
View Article Online
View Journal
EDGE ARTICLE
Carbocyclization of unsaturated thioesters under
palladium catalysis†
Cite this: DOI: 10.1039/c3sc50486g
*
Arun P. Thottumkara, Toshiki Kurokawa‡ and J. Du Bois
An intramolecular thioester–olefin cross-coupling reaction for the preparation of cyclic ketone structures
from unsaturated thioesters is described. This method capitalizes on the unique reactivity of thioesters
with low-valent palladium catalysts and copper(I) salts to form acyl-metal species. Trapping of such
intermediates with alkene functional groups generates the corresponding exo-methylene cycloalkanone
products. Unsaturated thioester substrates, which can be accessed using a suite of modern synthetic
methods, can now be regarded as precursors to complex carbocyclic derivatives.
Received 20th February 2013
Accepted 9th April 2013
DOI: 10.1039/c3sc50486g
www.rsc.org/chemicalscience
The modern synthetic chemist, while empowered with an
Although intramolecular olen hydroacylation of unsatu-
impressive arsenal of reaction technologies that enable rapid rated aldehydes has been studied extensively, the preparation of
assembly of acyclic molecules, is confronted with the rich cycloalkanone products through analogous coupling reactions
challenges posed by architecturally complex cyclic structural of carboxylic acid derivatives and alkenes (or alkynes) nds
motifs. In principle, stereodened linear molecules can func- considerably less precedent.2,3 In our hands, attempts to
tion as precursors to carbocyclic derivatives. Yet few methods, promote cyclization of d,3-unsaturated acid chlorides under
among which alkene metathesis1 is arguably foremost, make palladium catalysis afforded none of the desired cyclic products
possible the conjoining of two ends of a chain to form ring (eqn (1)).4,5 These results prompted us to consider alternative
products, a limitation that has frustrated our own efforts in starting materials, which included thioester derivatives. The
synthesis. Motivated by the difficulties of preparing in conver- decision to investigate such substrates (i.e., 1, Table 1) was
gent fashion a number of heteroatom-rich polycyclic targets, we based on the growing body of literature highlighting the facility
have sought to identify new types of carbocyclization processes of thioesters to engage in metal-catalyzed transformations.6
from functionalized acyclic molecules bearing common ‘end’ Pioneering work by Fukuyama et al.7 and Liebeskind et al.8 has
groups. Our efforts in this area have resulted in the develop- demonstrated that thioesters may be directly converted to the
ment of an intramolecular coupling reaction of unsaturated corresponding aldehydes and ketones under palladium catal-
thioesters (Fig. 1A). This method operates under the action of a ysis. These conditions oen take advantage of thiophilic metal
catalytic palladium source, copper thiophene-2-carboxylate, and co-factors both to activate the thioester C–S bond for oxidative
zinc formate, the latter of which is critical for achieving efficient addition and to sequester the liberated thiolate anion. Privi-
turnover numbers. We expect this process to be of general utility leged among the many available thiophilic metal additives
in synthesis and hope that our ndings embolden future available are the copper(I) salts made prominent by Liebeskind.
studies to identify new catalytic processes that transform In a series of reports, Liebeskind has demonstrated the superior
complex linear molecules into carbocyclic ring products.
ability of select copper(I) carboxylates, in particular copper(I)
thiophene-2-carboxylate (CuTC), in concert with palladium salts
to facilitate acyl-palladium formation.7 These studies provided
the backdrop for our own investigations.
Fig. 1 A palladium-catalyzed thioester carbocylization reaction.
Thioester 1 (Table 1) was selected as a convenient starting
material for exploratory reaction development. Initial attempts
to promote cyclization of this material successfully afforded
enone 2, albeit at low conversion. Extensive screening of palla-
dium sources,9 ligands, solvents, and temperature had little
Department of Chemistry, Stanford University, Stanford, CA 94305, USA. E-mail:
jdubois@stanford.edu
† Electronic supplementary information (ESI) available: Experimental details and
1H NMR spectra for all new compounds. See DOI: 10.1039/c3sc50486g
‡ Current address: Eisai Pharmaceuticals, Tokyo, Japan.
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci.

View Article Online
Chemical Science
Edge Article
Table 1 Evaluation of reaction conditions for catalytic thioester carbocyclization
similar enhancements in turnover number (entries 13 and 14).
At this time, we speculate that zinc formate may retard the rate
of Pd aggregate and/or Pd black formation.
Examination of our optimized protocol with thioester deriva-
tives appearing in Table 2 serves to illustrate the scope and utility
of the cyclization reaction. The critical role of zinc formate in this
cyclization process is evident from entries 2, 3, 5, 6, 9 and 10;
Entrya
Ligand
Additive
CuTC equiv.
Conversionb reactions conducted in the absence of zinc formate furnish <30%
of the desired product (product % conversion in parentheses). By
1
2
3
4
5
6
7
8
P(2-furyl)3
P(OMe)3
P(OMe)3
P(OEt)3
—
—
—
—
—
Et3N
3.2
3.2
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
1.6
0
25c
contrast, our optimized conditions perform with a range of
disparate examples, affording both ve and six-membered car-
bocycles in moderate to excellent yield. As noted in related
30
30
30
10
20
15
30
10
—
80
0
P(OiPr)3
P(OMe)3
P(OMe)3
P(OMe)3
P(OMe)3
P(OMe)3
P(OMe)3
P(OMe)3
P(OMe)3
P(OMe)3
iPr2NEt
Zn(OAc)2
Zn(CO3)
Zn(OTf)2
Zn(O2CH)2
Zn(O2CH)2
NaO2CH
Cu2O/HCO2H
Table 2 Carbocycle synthesis via thioester–olefin cross-coupling
9
10
11
12
13
14
1.6
1.6
35d
5e
a
Reactions were performed with 5 mol% Pd(OAc)2, 1.6 equiv. CuTC, 1.0
b
equiv. ligand, and 1.0 equiv. additive unless otherwise noted. Percent
conversion estimated by integration of the unpuried 1H NMR spectrum.
c
e
d
10 mol% P(2-furyl)3. Reaction performed with 2.0 equiv. NaO2CH.
Reaction performed with 1.0 equiv. Cu2O and 2.0 equiv. HCO2H.
effect on improving the reaction. Use of phosphine ligands,
such as P(2-furyl)3, in catalytic amounts necessitated that a large
excess of CuTC also be employed (entry 1). Increasing ligand
loading did not provide for a compensatory reduction in CuTC
equivalents. The overall amount of CuTC, however, could be
lowered with no evident loss in product conversion by replacing
the triarylphosphine ligand with a trialkylphosphite, speci-
cally P(OMe)3 (entries 2 and 3). The effectiveness of P(OMe)3 in
this process may be due, in part, to its inuence on the partial
dissolution of CuTC, which appears to be largely insoluble in
THF. It is also possible that the phosphite ligand creates a more
sterically accessible and electrophilic coordination environ-
ment for alkene binding.8d In this regard, use of other phosphite
ligands such as P(OiPr)3 shows a pronounced diminution in
product formation (entry 5, Table 1).
Having selected P(OMe)3 as the optimal ligand/additive, we
were challenged with making additional modications to the
protocol that would substantially improve catalyst turnover
numbers. The inuence of tertiary amine additives on reaction
performance was examined with the expectation that a nitrogen
base would facilitate Pd(0) regeneration (entries 6 and 7); these
conditions gave disappointing results. In considering that both
Liebeskind10 and Fukuyama7b have utilized zinc compounds to
effect cross-coupling reactions with thioorganic substrates, we
questioned whether a zinc additive would promote turnover
(entries 8–11). Of the zinc(II) salts tested, the formate complex
(entry 11) was unique, with production of 2 increasing by over
40% as compared to all other reaction conditions examined.
Zinc formate, however, cannot be used in place of CuTC,
underscoring the importance of a copper additive for thioester
activation (entry 12). Other formate sources do not afford
a Reactions performed with 5 mol% Pd(OAc)2, 1.0 equiv. P(OMe)3, 1.6
equiv. CuTC, 1.0 equiv. Zn(O2CH)2, THF, 60 ꢀC for 18 h. b Isolated
yields following chromatography on silica gel. c Values in parentheses
represent estimated percent product formation (based on 1H NMR
integration) in reactions performed without Zn(O2CH)2.
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2013

View Article Online
Edge Article
Chemical Science
Table 3 Internal coordination facilitates ring formation
processes involving acyl-metal intermediates (e.g. intramolecular
hydroacylation),2 formation of 5-membered ring structures is
generally more efficient when compared to reactions that furnish
cyclohexyl products (entries 1 and 2, Table 2). Not surprisingly,
substrates possessing a geminally disubstituted center and that
are predisposed towards cyclization are particularly well-suited to
this process (cf. entries 3 and 4). Carbocyclization reactions also
perform efficiently with unsaturated thioesters bearing a range of
common functional groups, including ethers, nitriles, and oxy-
esters. The tricarbonyl derivative shown in entry 6 highlights the
uniqueness of the thioester to undergo selective activation and
subsequent cyclization.
Substitution at the a-carbon of the thioester delivers cyclic
ketone products in modest yields (entries 9 and 10, Table 2). In
these substrates, competitive decarbonylation and b-hydride
elimination appears to arrest catalyst turnover.11 We have found
that subjecting a-phenyl thioester 3 to the reaction conditions
returns almost entirely starting material and trace amounts of the
skipped diene (eqn (2)). Formation of this minor product is the
result of decarbonylation of the acyl-Pd intermediate and subse-
quent b-hydride elimination. The stability of the acyl-Pd species
and the rate of CO loss are arguably the principle competing
factors that lower catalyst turnover numbers. Future modica-
tions to our protocol that can extend the lifetime of the acyl-metal
adduct should lead to increased reaction scope and may allow for
the preparation of larger-membered carbocyclic rings.
a Reactions performed with 5 mol% Pd(OAc)2, 1.0 equiv. P(OMe)3, 1.6
equiv. CuTC, 1.0 equiv. Zn(O2CH)2, THF, 60 ꢀC for 18 h. b Isolated
yields following chromatography on silica gel. c Estimated conversion
to product based on 1H NMR integration.
in yields exceeding 60%. Entry 5 highlights the utility of the
thioester coupling method for the preparation of fused bicyclic
Consistent with our hypothesis that successful cyclization structures, a common motif in natural and synthetic materials.
necessitates the acyl-Pd intermediate be rapidly intercepted by Collectively, these results are consistent with a role for internal
the pendant nucleophile, internal alkenes do not engage as coordination of the acyl-Pd intermediate and offer a practical
coupling partners in this reaction. We conclude that the addi- solution for expanding the substrate scope of this method.
tional steric hindrance of such olens adversely inuences
We have initiated studies with alkyne-substituted thioesters
coordination to the acyl-Pd species. Spent reaction mixtures with the goal of identifying coupling partners other than
from tests with internal alkenes consist primarily of starting terminal alkenes that can engage in facile acyl-palladation. In
material, along with trace amounts of decarbonylated decom- these reactions and as an added feature of this coupling tech-
position products.
nology, the vinyl-Pd intermediate may be intercepted with a
Suppositions based on the instability of the putative acyl-Pd nucleophile to afford b-substituted enone products.14 We have
intermediate have guided substrate design. Following the work of found that treatment of methyl-substituted alkyne 5 with 10 mol
Willis and Dong,12 we have found that appropriate placement of a % Pd(OAc)2 and 3.2 equivalents of CuTC in the presence of
weakly coordinating group such as an alcohol or ether helps phenyl boronic acid affords ketone 6 in 56% yield (Fig. 2).
promote cyclization of 3,z-unsaturated thioesters to give cyclo- Despite the similarity of these conditions to known methods for
hexanone products (Table 3).13 Comparison of entries 1–3 show thioester–boronic acid cross-coupling, none of the phenyl
demonstrable differences in reaction performance based on the ketone product is detected in this reaction.8a Although terminal
choice of the b-substituent group. The highest product yield is alkynes fail to cyclize,15 other internal alkynes such as 7 are
obtained for entry 1, in which the tertiary alcohol is positioned efficiently and stereospecically converted to the b,b-disubsti-
appropriately to chelate the acyl-Pd species. Thioesters bearing tuted enone products. Such compounds are difficult to prepare
weakly coordinating b-groups (e.g., acetate, silyl ether) are in isomerically pure form using alternative tactics,16 a fact that
considerably less effective as substrates (entries 2 and 3), and the serves to further underscore the utility of our method.
gem-dimethyl derivative shown in entry 4 gives no detectable
We have developed an intramolecular thioester–olen cross-
amount of the desired cyclohexenone. Other b-alcohol and ether coupling reaction as a technology for generating cycloalkanones
derivatives (entries 5 and 6) can be successfully employed in this from simple acyclic precursors. Under the action of catalytic
cyclization process, giving the desired 6-membered ring structures Pd(OAc)2, starting thioesters decorated with a number of
This journal is ª The Royal Society of Chemistry 2013
Chem. Sci.

View Article Online
Chemical Science
Edge Article
5 Cyanoformates have also been employed to generate acyl-Pd
species, see: (a) Y. Hirata, A. Yada, E. Morita, Y. Nakao,
T. Hiyama, M. Ohashi and S. Ogoshi, J. Am. Chem. Soc.,
2010, 132, 10070; (b) N. R. Rondia, S. M. Levi, J. M. Ryass,
R. A. Vaden Berg and C. J. Douglas, Org. Lett., 2011, 13, 1940.
6 For
a recent mini-review of thioester cross-coupling
chemistry, see: H. Prokopcov and C. O. Kappe, Angew.
Chem., Int. Ed., 2009, 48, 2276.
7 (a) T. Fukuyama, S. C. Lin and L. Li, J. Am. Chem. Soc., 1990,
112, 7050; (b) H. Tokuyama, S. Yokoshima, T. Yamashita and
T. Fukuyama, Tetrahedron Lett., 1998, 39, 3189.
Fig. 2 Alkyne substrates afford tetra-substituted olefin products.
8 (a) L. S. Liebeskind and J. Srogl, J. Am. Chem. Soc., 2000, 122,
11260; (b) C. Savarin, J. Srogl and L. S. Liebeskind, Org. Lett.,
2000, 2, 3229; (c) R. Wittenberg, J. Srogl, M. Egi and
L. S. Liebeskind, Org. Lett., 2003, 5, 3033; (d) H. Yang,
H. Li, R. Wittenberg, M. Egi, W. Huang and
L. S. Liebeskind, J. Am. Chem. Soc., 2007, 129, 1132; (e)
H. Yang and L. S. Liebeskind, Org. Lett., 2007, 9, 2993; (f)
J. M. Villalobos, J. Srogl and L. S. Liebeskind, J. Am. Chem.
Soc., 2007, 129, 15734; (g) Z. Zhang, M. G. Lindale and
L. S. Liebeskind, J. Am. Chem. Soc., 2011, 133, 6403.
9 Other metal sources including rhodium(I) and nickel(I) are
known to form acyl-metal species from thioesters. In our
studies, however, only palladium salts proved successful for
thioester–olen cross-coupling. For recent examples of
rhodium-catalyzed transformationsof thioesters, see: (a)
M. Arisawa, Y. Nihei, T. Suzuki and M. Yamaguchi, Org. Lett.,
2012, 14, 855; (b) M. Arisawa, T. Yamada and M. Yamaguchi,
Tetrahedron Lett., 2010, 51, 6090; (c) M. Arisawa, T. Kubota
and M. Yamaguchi, Tetrahedron Lett., 2008, 49, 1975; (d) For a
nickel-catalyzed decarbonylation of thioesters, see: E. Wenkert
and D. Chianelli, J. Chem. Soc., Chem. Commun., 1991, 627.
common functional groups can be efficiently processed to the
corresponding enones. Our ndings offer a distinct approach
for carbocycle synthesis and give impetus for further investi-
gations to identify methods that enable ring formation from
polysubstituted acyclic starting materials.
Acknowledgements
We are grateful to Mr. Lorenz-Maximillian Schneider for his
assistance in the exploratory phase of this work. We thank Drs.
David Olson and Fred Menard for helpful discussions. Prof. T. K.
Vinod of Western Illinois University is acknowledged for several
discussions regarding substrate preparation. APT is the recipient
of an NSF pre-doctoral graduate fellowship. T. K. was supported as
a visiting scientist by Eisai Pharmaceuticals. We are most grateful
to Eisai, Pzer, and Stanford University for support of this work.
Notes and references
1 (a) For a recent review of Ru-catalyzed olen metathesis
reactions, see: G. C. Vougioukalakis and R. H. Grubbs, 10 J. Srogl, W. Liu, D. Marshall and L. S. Liebeskind, J. Am.
Chem. Rev., 2010, 110, 1746; (b) For a review of Mo- and W-
Chem. Soc., 1999, 121, 9449.
catalyzed olen metathesis reactions see: R. R. Schrock and 11 Liebeskind has noted decarbonylation products in cross-
A. H. Hoveyda, Angew. Chem., Int. Ed., 2003, 42, 4592.
2 For recent reviews of metal-catalyzed olen hydroacylation,
see: (a) M. C. Willis, Chem. Rev., 2010, 110, 725; (b) C.-H. Jun,
E.-A. Jo and J.-W. Park, Eur. J. Org. Chem., 2007, 1869.
3 Several examples of carbonyl hydroacylation have recently
been described; see: (a) Z. Shen, H. A. Khan and
V. M. Dong, J. Am. Chem. Soc., 2008, 130, 2916; (b)
D. H. T. Phan, B. Kim and V. M. Dong, J. Am. Chem. Soc.,
2009, 131, 15608; (c) H. A. Khan, K. G. M. Kou and
coupling reactions of peptidyl thioesters (see ref. 8d). In
this study, the use of phosphite ligands was suggested to
slow decarbonylation of the acyl-Pd species.
12 Internalchelationhasbeenutilizedtoincreasetheefficiencyof
intra- and intermolecular rhodium-catalyzed hydroacylation
´
reactions. For leading reports, see: (a) C. Gonzalez-Rodriguez
and M. C. Willis, Pure Appl. Chem., 2011, 83, 577; (b)
M. M. Coulter, P. K. Dornan and V. M. Dong, J. Am. Chem.
Soc., 2009, 131, 6932.
V. M. Dong, Chem. Sci., 2011, 2, 407; (d) For a recent 13 Substrates appearing in Table 3 were readily prepared
disclosure describing hydrocarbamoylation of terminal
alkenes, see: Y. Miyazaki, Y. Yamada, Y. Nakao and
T. Hiyama, Chem. Lett., 2012, 41, 298.
4 (a) C.-M. Andersson and A. Hallberg, J. Org. Chem., 1988, 53,
4257; (b) J. S. Brumbaugh, R. R. Whittle, M. Parvez and
through Reformatsky reaction with ethyl bromoacetate,
see: (a) G. Confalonieri, E. Marotta, F. Rama, P. Righi,
G. Rosini, R. Serra and F. Venturelli, Tetrahedron, 1994, 50,
3235; (b) M.-X. Zhao, M.-X. Wang, C.-Y. Yu, Z.-T. Huang
and G. W. J. Fleet, J. Org. Chem., 2004, 69, 997.
A. Sen, Organometallics, 1990, 9, 1735; (c) The generation 14 Boronic acid additives were also included in reactions with
and olenation of acyl-palladium species from vinyl
alkene substrates (e.g., 1, Table 1). In these cases, products
iodides has been described previously: J. M. Tour and
resulting from boronic acid incorporation were never observed.
E.-I. Negishi, J. Am. Chem. Soc., 1985, 107, 8289; (d) For 15 We speculate that copper acetylide formation is rapid and
ketone preparation through Ni-catalyzed reductive coupling interferes with the desired reaction outcome.
of acid chlorides or (2-pyridyl)thioesters, see: A. C. Wotal 16 A. B. Flynn and W. W. Ogilvie, Chem. Rev., 2007, 107,
and D. J. Weix, Org. Lett., 2012, 14, 1476. 4698.
Chem. Sci.
This journal is ª The Royal Society of Chemistry 2013