Catalyzed catalysis using carbophilic Lewis acidic gold and Lewis basic palladium: Synthesis of substituted butenolides and isocoumarins

Article abstract of DOI:10.1021/ja9068497

(Figure Presented) A new strategy for gold and palladium dual-catalytic reactivity and turnover, called catalyzed catalysis, enhanced the synthetic usefulness of vinylgold intermediates by providing dual-catalytic carbon-carbon cross-coupling as an alternative to protodemetalation. This protocol enabled the synthesis of substituted butenolides and isocoumarins from allyl esters. Kinetic and spectroscopic experiments support a mechanism in which the Lewis acidic gold complex catalyzes both an initial rearrangement step and a subsequent Lewis basic palladium oxidative-addition step.

Full text of DOI:10.1021/ja9068497

Published on Web 11/24/2009
Catalyzed Catalysis Using Carbophilic Lewis Acidic Gold and Lewis Basic
Palladium: Synthesis of Substituted Butenolides and Isocoumarins
Yili Shi, Katrina E. Roth, Stephen D. Ramgren, and Suzanne A. Blum*
Department of Chemistry, UniVersity of California, IrVine, California 92697-2025
Received August 12, 2009; E-mail: blums@uci.edu
Scheme 1. Proposed Mechanism
Reactions catalyzed by two different metals at the same time,
or dual-catalyzed reactions, can access reactivity and selectivity
not available to single-metal systems while offering the attractive
environmental and cost benefits of substoichiometric use of both
metals.1-5 Considerably less is known about strategies for designing
dual-catalyzed reactions than about their single-metal counterparts.1
This paucity of predictive power arises in part from questions
regarding the compatibility of the two metals with each other and
the additional complexity of initiating the catalytic turnover of both
metals. In this work, we have developed a valuable new strategy,
catalyzed catalysis, to achieve turnover of both metals and novel
reactivity. Specifically, we employ a carbophilic6 Lewis acidic Au
catalyst to catalyze the cross-coupling reactivity of a second Lewis
basic Pd catalyst in order to functionalize vinylgold intermediates
arising from intramolecular substrate rearrangements. In the past
decade, vinylgold intermediates have been indicated in the rear-
rangement of numerous heteroatom-containing substrates;6,7 however,
their reactivity has been dominated by protodemetalation,6,7 with
electrophilic trapping as a notable alternative.8-10 Our Au/Pd cross-
coupling reaction enhances the synthetic usefulness of these intermedi-
ates by providing dual-catalytic C-C bond formation as an alternative
to protodemetalation.11 We have showcased this reaction in the
synthesis of substituted butenolides and isocoumarins, structural motifs
present in numerous biologically active compounds.12,13
Table 1. Scope of Au/Pd-Catalyzed Butenolide Rearrangement
We anticipated that the Au-catalyzed rearrangement of allenoate 1
would produce allyl oxonium 2, in analogy to the stoichiometric
formation of the ethyloxonium ion reported by Hammond14 (Scheme
1). We hypothesized that this rearrangement15,16 would enhance the
reactivity of the ester toward deallylation17 by Pd(0) because of the
increased rate of oxidative addition of Pd(0) toward allylic substrates
bearing cationic leaving groups.18 Subsequent transmetalation between
neutral vinylgold complex 3 and π-allyl Pd complex 411 followed by
C-C bond-forming reductive elimination19 would complete both
metals’ catalytic cycles and form substituted butenolide 5. In essence,
the Au therefore would catalyze both the initial rearrangement step
and the subsequent Pd oxidative addition step, first by lowering the
energy of the allene antibonding orbital in 1 and then by redistributing
this electron deficiency through the substrate’s rearrangement via
lowering of the energy of the allyl-oxygen antibonding orbital in
oxonium 2, thus lowering the barrier for oxidative addition by the
second Pd catalyst (Scheme 1).18 In this proposed pathway, Au
catalytically creates a different species with increased reactivity toward
catalytic oxidative addition of Pd. The term catalyzed catalysis is
applied to convey this mechanistic process.20
a Using 10 mol % PPh3AuCl/AgOTf and 10 mol % Pd2dba3.
product 5g (92%). The robustness of the Brønsted acid-sensitive
tert-butyldimethylsilyl protecting group toward the reaction condi-
tions showcased the selective reactivity of the carbophilic Lewis
acid Au while providing a handle for subsequent orthogonal
reactivity. Similarly, the dual-catalyzed reaction conditions avoided
potential side reactions from oxidative addition into the aryl bromide
bond (5e, 75%). Substitution of the allyl moiety (R3) at the internal
position was also tolerated [5h (78%) and 5i (85%)] despite the
increased acid sensitivity of the starting allenoates 1h and 1i toward
Lewis acid-catalyzed ionization.
To probe the intermediacy of π-allyl palladium complex 4 and
vinylgold complex 3 in this reaction, a crossover experiment
between allenoates 1b and 1h was examined.22 Specifically, 5 mol
% PPh3AuCl/AgOTf and 5 mol % Pd2dba3 was added to 0.5 equiv
of 1b and 0.5 equiv of 1h in a single reaction vessel (eq 1). The
crude reaction mixture was analyzed by 1H NMR spectroscopy and
mass spectrometry (MS), which revealed that products 5b and 5h
and crossover products 5a and 5i were produced in a 1:1:0.8:0.8
ratio. The observed crossover is consistent with the intermediacy
of 3 and 4 followed by intermolecular transmetalation and reductive
elimination.19 The unequal ratio of crossover to noncrossover
products could arise from a small difference in the rates of reactivity
between the differentially substituted starting materials.
In order to explore this hypothesis, allenoate 1a was treated with
5 mol % PPh3AuCl/AgOTf and 5 mol % Pd2dba3 in CD2Cl2 (Table
1). As predicted, this reaction produced butenolide 5a (97%) and
required both Au and Pd.21 The reaction proceeded in high
conversion with a variety of substituents at the R1, R2, and R3
positions (Table 1, 1.5-24 h). For example, geminal substitution
on the allenyl moiety was tolerated in the formation of spirocyclic
9
18022 J. AM. CHEM. SOC. 2009, 131, 18022–18023
10.1021/ja9068497  2009 American Chemical Society

C O M M U N I C A T I O N S
83%, 98% respectively).27 This result demonstrated that the
vinylgold intermediate need not be previously reported as observ-
able26 in order to serve as an efficient cross-coupling partner.
In conclusion, we have established a fundamental pattern of dual-
catalytic cross-coupling reactivity and turnover with carbophilic
Lewis acidic Au catalysts and Lewis basic Pd catalysts. More
broadly, this C-C cross-coupling28 method represents a practical
alternative to protodemetalation for elaborating the large number
of recently reported organogold oxonium intermediates.6,7
Acknowledgment. We thank the ACS-PRF, the NSF (a CAREER
Award), and the University of California Irvine for funding, the
Allergan Foundation for a fellowship to Y.S., Dr. John Greaves for
help with MS, and Mr. Stephen Canham for HPLC assistance.
A series of 1H NMR spectroscopic kinetic experiments revealed
a first-order kinetic dependence on the concentration of Pd2dba3,
confirming the role of Pd in the reaction (at 5 mol % PPh3AuOTf
and 0.10 M 1b, kobs ) 0.0014, 0.0016, and 0.0026 M-1 s-1 at 5,
7.5, and 10 mol % Pd, respectively).23 Over this loading range,
the reaction displayed zeroth-order substrate saturation kinetics.24
While the presence of Au was required for the observed product
formation, loadings of over 5 mol % PPh3AuOTf resulted in
decreased conversion, consistent with catalyst decomposition prior
to the completion of the reaction at higher Au loadings.
Supporting Information Available: Experimental procedures,
kinetic data, and compound characterization data. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Sammis, G. M.; Danjo, H.; Jacobsen, E. N. J. Am. Chem. Soc. 2004, 126, 9928.
(2) Podder, S.; Choudhury, J.; Roy, U. K.; Roy, S. J. Org. Chem. 2007, 72, 3100.
(3) Shi, Y.; Peterson, S. M.; Haberaecker, W. W., III; Blum, S. A. J. Am.
Chem. Soc. 2008, 130, 2168.
We next examined which step was rate-determining. We
discovered that reactant 1b and rearrangement product 2b reach a
rapid equilibrium strongly favoring 2b (>95% by 1H NMR
spectroscopy) (eq 2) Specifically, addition of the precatalyst
(4) For a Ag/Pd cross-coupling reaction limited to carboxylic acids, see: Ma,
S.; Shi, Z. J. Org. Chem. 1998, 63, 6387. For a Pd-only reaction, see: Ma,
S.; Yu, Z. J. Org. Chem. 2003, 68, 6149.
(5) Chinchilla, R.; Na´jera, C. Chem. ReV. 2007, 107, 874.
(6) Gorin, D. J.; Sherry, B. D.; Toste, F. D. Chem. ReV. 2008, 108, 3351.
(7) Hashmi, A. S. K. Chem. ReV. 2007, 107, 3180.
(8) Dube´, P.; Toste, F. D. J. Am. Chem. Soc. 2006, 128, 12062.
(9) Nakamura, I.; Sato, T.; Terada, M.; Yamamoto, Y. Org. Lett. 2007, 9, 4081.
(10) Luzung, M. R.; Mauleo´n, P.; Toste, F. D. J. Am. Chem. Soc. 2007, 129, 12402.
(11) We previously reported a series of Au/Pd cross-coupling reactions that were
stoichiometric in Au and catalytic in Pd: Shi, Y.; Ramgren, S. D.; Blum,
S. A. Organometallics 2009, 28, 1275. While the current manuscript was
under revision, another example of a Au/Pd cross-coupling reaction that is
also stoichiometric in Au and catalytic in Pd was reported: Hashmi, A. S. K.;
Lothschu¨tz, C.; Do¨pp, R.; Rudolph, M.; Ramamurthi, T. D.; Rominger, F.
Angew. Chem., Int. Ed. 2009, 48, 1.
Pd(PPh3)4 did not initiate cross-coupling reactivity but instead
quickly regenerated the starting material, presumably by trapping
PPh3AuOTf as the catalytically inactive [(PPh3)2Au]+, as was
observed when 1 equiv of PPh3 was added to 1 equiv of 2b (>95%
regeneration of 1b within 20 min). This is an early example of an
experimentally detected Au-catalyzed substrate rearrangement in
which the equilibrium and the back reaction could be observed and
kinetically probed. This result suggested that a low energy barrier
and equilibrium may be present in the initial rearrangement step
for numerous other recently reported Au-catalyzed reactions.6,7
The presence of substrate saturation kinetics suggested that the
catalyst resting state should be an observable intermediate with the
substrate bound to one or both of the metal catalysts. Indeed,
vinylgold complex 3 and the π-allyl palladium cations 4 with L )
dba and (PPh3)2 were identified in the reaction mixture by high-
resolution MS prior to completion of the reaction, supporting the
proposed mechanism.25 The observed equilibrium between 1b and
2b and the buildup of intermediates 3 and 4 suggested that the
intermolecular combination of 3 and 4 was the rate-determining
step under saturation conditions (likely via transmetalation and/or
reductive elimination to form 5, in analogy to other known
π-allyl-Pd alkylation reactions19).
(12) Alali, F. W.; Liu, X. S.; McLaughlin, J. L. J. Nat. Prod. 1999, 62, 504.
(13) Hill, R. A. Prog. Chem. Org. Nat. Prod. 1986, 49, 2.
(14) Liu, L.-P.; Xu, B.; Mashuta, M. S.; Hammond, G. B. J. Am. Chem. Soc.
2008, 130, 17642.
(15) Kang, J.-E.; Lee, E.-S.; Park, S.-I.; Shin, S. Tetrahedron Lett. 2005, 46, 7431.
(16) Piera, J.; Krumlinde, P.; Stru¨bing, D.; Ba¨ckvall, J.-E. Org. Lett. 2007, 9, 2235.
(17) Fu¨rstner, A.; Stelzer, F.; Szillat, H. J. Am. Chem. Soc. 2001, 123, 11863.
(18) Amatore, C.; Jutand, A.; Mensah, L.; Meyer, G.; Fiaud, J.-C.; Legros, J.-
Y. Eur. J. Org. Chem. 2006, 1185.
(19) Trost, B. M.; Van Vranken, D. L. Chem. ReV. 1996, 96, 395.
(20) Recent reports from Nakao and Hiyama and from our group provide
examples of the use of this mechanism as a unified concept for reaction
design (see ref 3 and: Nakao, Y.; Yada, A.; Ebata, S.; Hiyama, T. J. Am.
Chem. Soc. 2007, 129, 2428) A previous example of this mechanistic
phenomenon is the Sonogashira reaction (see ref 5).
(21) No conversion was observed with AgOTf or Pd2dba3 separately or with
only PPh3AuCl/AgOTf. The combination of AgOTf and Pd2dba3 catalyzed
the reaction less effectively alone than it did in the presence of PPh3AuCl
(49% at 1.5 h and 60% at 24 h, compared with 97% at 1.5 h when Au was
also present). Employing premade PPh3AuOTf and Pd2dba3 in the absence
of Ag confirmed the presence of a purely Au/Pd dual-catalyzed reaction.
Addition of PtCl2 did not result in product formation.
(22) Blum, S. A.; Tan, K. L.; Bergman, R. G. J. Org. Chem. 2003, 68, 4127.
(23) We estimate the error to be (0.0001-0.0002 M-1 s-1 on the basis of
triplicate runs at each loading level.
(24) Anslyn, E. V.; Dougherty, D. A. Modern Physical Organic Chemistry;
University Science Books: Sausalito, CA, 2006; p 396.
(25) We considered the possibility that cationic 2 rather than neutral 3 was the
cross-coupling partner with 4. In the ethyl series, dealkylation of the
oxonium ion did not occur, so we could query this point. Cationic ethyl 9
did not participate in cross-coupling, suggesting that neutral 3 rather than
cationic 2 was the more likely transmetalation partner:
Extension of the Au/Pd cross-coupling reaction to benzoate
substrates established the generality of this methodology. The Au-
only rearrangement of benzoates 6a and 6b previously was proposed
to proceed through vinylgold oxonium intermediates 7a and 7b (eq
3), followed by protodemetalation.26 Addition of 5 mol % Pd2dba3
(26) Aikawa, H.; Tago, S.; Umetsu, K.; Haginiwa, N.; Asao, N. Tetrahedron
2009, 65, 1774.
(27) Separately, PPh3AuCl/AgOTf or Pd2dba3 did not catalyze this transformation.
(28) Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005,
44, 4442.
to 6a-c in the presence of 5 mol % PPh3PAuCl/AgOTf resulted
in cross-coupling of 7a-c to produce isocoumarins 8a-c (89%,
JA9068497
9
J. AM. CHEM. SOC. VOL. 131, NO. 50, 2009 18023