Contemporaneous dual catalysis by coupling highly transient nucleophilic and electrophilic intermediates generated in situ

Article abstract of DOI:10.1021/ja110501v

We report herein a new process, which we call contemporaneous dual catalysis, that selectively couples two highly reactive catalytic intermediates while overcoming competing trapping by stoichiometric reactive species also present in the reaction. The rea

Full text of DOI:10.1021/ja110501v

COMMUNICATION
pubs.acs.org/JACS
Contemporaneous Dual Catalysis by Coupling Highly Transient
Nucleophilic and Electrophilic Intermediates Generated in Situ
Barry M. Trost* and Xinjun Luan
Department of Chemistry, Stanford University, Stanford, California 94305-5080, United States
S Supporting Information
b
Scheme 1. (a) Conventional Catalysis; (b) Typical Dual
Catalysis; (c) Contemporaneous Dual Catalysis
ABSTRACT: We report herein a new process, which we
call contemporaneous dual catalysis, that selectively couples
two highly reactive catalytic intermediates while overcoming
competing trapping by stoichiometric reactive species also
present in the reaction. The reaction proceeds via the
convergence of a vanadium-catalyzed propargylic rearrange-
ment and a palladium-catalyzed allylic alkylation. It gener-
ates a variety of R-allylated R,β-unsaturated ketones, which
are not readily accessible by other means. Notably, this dual
catalysis is achieved using low catalyst loadings (1.0 mol %
[Pd], 1.5 mol % [V]) and gives good to excellent yields (up
to 98%) of the desired products.
he development of new catalytic methods has been an area of
T
intense research in both industrial and academic laboratories
over the last century. During this time, thousands of catalytic
reactions have been invented,1,2 driven in part by the growing
demands of the pharmaceutical, agrochemical, petrochemical,
flavors, fragrances, and materials industries for rapid access to
complex chemical structures. As the global demand for fine
chemicals continues to grow, minimization of the environmental
impact of these chemical manufacturing processes has become
increasingly important. To facilitate the continuous advance-
ment of our society and sustainability of our planet, it is vital to
design and implement new modes of catalysis that enable pre-
viously inaccessible chemical processes while minimizing the
waste generated and energy required for their production.
In general, catalytic transformations rely on the principle of
capturing the small amount of catalyst-activated intermediate with a
coupling partner present in the reaction. To date, the majority of
catalytic reactions have been developed by activating a substrate (I)
with a substoichiometric amount of a catalyst and trapping the
resulting activated species with a second reactant (II), which is
generally present in stoichiometric or superstoichiometric quantities
(Scheme 1a).3 A significant drawback of this approach is that if the
second reagent II requires activation for the reaction to occur, it is
done so stoichiometrically. Recently, dual catalysis has been a popu-
lar target for many research groups.4 This strategy employs the
selective activation of two substrates (I and II) by two different cata-
lysts (catalystI and catalystII) to generate I0 and II0, which are sub-
sequently coupled to afford I0-II0 with concomitant regeneration of
both catalysts (Scheme 1b).5 One advantage of this approach is that
it allows reactions to proceed without the addition of stoichiometric
activating reagents. Furthermore, these reactions have been shown to
have novel reactivity, generating products that are difficult or
impossible to obtain by the traditional one-catalyst and/or one-
reaction approach.
Despite their advantages, the number of dual-catalyzed reac-
tions that have been developed is dramatically less than those
where a single catalyst is employed. One factor that contributes
to this disparity is that the in situ-generated reactive intermediates
are not necessarily chemoselective. They can be intercepted with
stoichiometric amounts of reactant (I or II) prior to the desired
reaction. In theory, if the relative rate constants (k, k0, and k00) for the
trapping of catalytic species I0 with either catalytic species II0 or
substrate I or II are similar, the activated intermediate I0 is more likely
to react with I or II because they are present in higher concentrations
than the relatively small amount of II0. Therefore, an important
goal is the development of contemporaneous dual catalysis, where the
two intermediates I0 and II0, formed irreversibly in situ, selectively react
with each other in the presence of large excesses of competing reagents
(Scheme 1c). To conquer this challenge, it is critical to control the
relative rates of formation of each reactive intermediate (such that k.
k0 and k . k00), thereby enhancing their affinity toward each other
relative to the starting materials. While this approach is attractive in
theory, its practical implementation remains elusive.
A central theme of the research in our laboratories is the
development of transition-metal-catalyzed reactions.6 As part of
these studies, we have demonstrated that electrophilic π-al-
lyl-palladium intermediate 1 can be generated from a variety
of alkenes containing allylic leaving groups in the presence of a
palladium(0) catalyst (Figure 1a). Upon its formation, 1 can
react with a wide range of nucleophilic coupling partners,
Received: November 22, 2010
Published: January 18, 2011
r
2011 American Chemical Society
1706
dx.doi.org/10.1021/ja110501v J. Am. Chem. Soc. 2011, 133, 1706–1709
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 2. Catalytic Transformations Promoted by
Vanadium and/or Palladium Catalysts
Figure 1. Catalytic formation of (a) π-allyl-palladium 1 and (b)
vanadium-allenoate 2.
(cycle A0). Intermediate 1 also has the potential to undergo
nucleophilic attack by the propargylic alcohol to generate O-allylated
alcohol 5 (cycle B0). The challenge, therefore, is to find conditions
under which the selective reaction between intermediates 1 and 2 to
afford the desired enone 6 proceeds faster than the alternative
processes involving stoichiometric partners that produce the unde-
sired 4 and/or 5. In order to accomplish this, we set out to identify
both the ligands and the V and Pd catalyst ratio ([V]/[Pd]) that
would selectively promote cycles A and B rather than A0 and B0. The
daunting challenge of identifying two catalysts that are compatible and
chemoselective is a critical prerequisite for this type of dual catalysis. In
addition, the proposed dual catalysis would result in the formation of
R-allyl-substituted R,β-unsaturated ketone products 6, which are not
readily accessible by other means.
To test this new dual catalysis concept, we began by choosing
propargyl alcohol 7 and allyl carbonate 8 as the coupling partners
along with a catalyst combination of OdV(OSiPh3)3 and Pd2-
(dba)3 CHCl3 (dba = trans,trans-dibenzylideneacetone) com-
3
Figure 2. Design blueprint for the contemporaneous dual catalysis.
plexed with the phosphine ligand 1,1-bis(diphenylphosphino)-
methane (DPPM).9 As both vanadium and palladium catalysts
have the potential to react with substrates 7and/or 8independent of
the other catalyst,10,11 three parallel reactions were conducted under
identical conditions using [V], [Pd], and both [Pd] and [V] as cata-
lyst(s), respectively. When only the vanadium catalyst was em-
ployed, R,β-unsaturated ketone 9,12 the Meyer-Schuster rearran-
gement product, was isolated in 94% yield with complete recovery of
allylic carbonate 8 (Scheme 2a). When the palladium complex was
used as the sole catalyst, O-allylated alcohol 1013 was isolated in 85%
yield as the only product (Scheme 2b). Much to our delight, when
both the vanadium and palladium catalysts were employed simulta-
neously, the desired allylated enone 11 was isolated in high yield.14
These preliminary studies revealed that the successful execution of
our contemporaneous dual catalysis gives R-allylated enone 11
with excellent yield (98%) and good stereoselectivity (E:Z = 5:1)
(Scheme 2c) by outcompeting both of the two side reactions
(Meyer-Schuster rearrangement and O-allylation). We hypothe-
sized that enone 11 might result from rearrangement of 10 by the
vanadium catalysis in a tandem catalytic process. However, when 10
was subjected to the reaction conditions with either [V] or [V]/[Pd]
catalyst(s), 11 was not observed. This control reaction indicated that
contemporaneous dual catalysis had occurred and that the reaction
did not proceed through two stepwise transformations.
including alcohols, amines, enolates, and malonates, to generate
new allylic C-O, C-N, or C-C bonds.7 In a similar pursuit, we
have demonstrated that oxyvanadium catalysts can perform a 1,3-
transposition of propargylic alcohols to generate nucleophilic
vanadium-allenoate intermediate 2 (Figure 1b). We have
shown that 2 can be intercepted in an intermolecular fashion
by aldehydes or imines rather than simple protonation, which
affords the Meyer-Schuster rearrangement product.8 In our
efforts to develop a contemporaneous dual catalysis process and
to expand the utility of the vanadium-catalyzed rearrangement,
we wondered whether the catalytic generation of the nucleophilic
vanadium-allenoate 2 could be coupled with the catalytic
generation of the electrophilic π-allyl-palladium complex 1 to
generate a 1,4-diene product faster than trapping of either 1 or 2
with the propargylic alcohol could occur via nucleophilic attack
or protonation, respectively (see below).
A detailed description of the proposal is shown in Figure 2. To
enter cycle A, the vanadium catalyst undergoes transesterification
with the propargylic alcohol (ROH) to generate vanadium ester
3. The ester then undergoes a 1,3-transposition to generate
vanadium-allenoate 2. Subsequently, in a dual catalysis mani-
fold, the reactive allenoate intermediate 2 can be intercepted by
the π-allyl-palladium intermediate 1 that is generated in the
parallel cycle B to give the desired product 6. Alternatively,
intermediate 2 could undergo protonation to generate enone 4
To further test the validity of our method involving two
parallel catalytic cycles, we performed a study that varied the
1707
dx.doi.org/10.1021/ja110501v |J. Am. Chem. Soc. 2011, 133, 1706–1709

Journal of the American Chemical Society
COMMUNICATION
Table 1. Studies of [V]/[Pd] Ratio Variations in Dual Catalysis
Table 2. Survey of Allyl Carbonate Scope in Dual Catalysis
[V]
[Pd]
(mol %)a
Conv.
(%)b
entry
(mol %)
9:10:11c
E:Z for 11c
1
5.0
1.0
5.0
1.5
1.5
5.0
5.0
1.0
1.0
1.0
100
93
3:1:47
1:12:17
30:1:16
5:1:65
1:0:5
3:1
3:1
6:1
6:1
7:1
2
3
96
4
5d
62
100
a Pd2(dba)3 CHCl3/DPPM = 1:2.4. b Determined by H NMR spec-
1
3
troscopy by observing consumption of 7. c Determined by H NMR
1
spectroscopy. d Reaction was performed by using 1.2 equiv of 7 at 80 °C
for 16 h to give a 98% isolated yield of 11.
[V]/[Pd] ratio of the reaction shown in Scheme 2c. The results
of these studies are summarized in Table 1. As expected, the dual
catalysis was extremely sensitive to the [V]/[Pd] ratio, and the
proper ratio was critical for the promotion of cycles A and B
relative to A0 and B0. In general, when an excess of the palladium
catalyst with respect to the vanadium catalyst was used, formation
of the O-allylated alcohol product 10 began to compete with that
of product 11 (entry 2). Similarly, when vanadium was in excess
relative to palladium, the generation of the Meyer-Schuster
rearrangement product 9 occurred at an elevated rate (entry 3). In
the optimized reaction, more than 1.0 equiv of propargyl alcohol 7
was required in order to obtain complete consumption of the allyl
carbonate 8 (entry 5). In this way, this reaction proceeded smoothly
to give rise to an excellent isolated yield of 11 (98%) with very good
E:Z stereoselectivity (7:1).
From the viewpoint of organic synthesis, the R-alkylation of
carbonyl compounds is one of the most fundamental carbon-
carbon bond-forming reactions. Conventionally, R-alkylated
carbonyl compounds are obtained by deprotonation of an R-
proton followed by addition to an electrophile, such as an alkyl
halide. However, these reactions are often limited to carbonyl-
containing compounds that can be regioselectively deprotonated
by a base or reacted with an amine to afford the metal enolates or
enamines, respectively.15 Since it is not possible to generate
allenoates by deprotonation of R,β-unsaturated carbonyl com-
pounds, the R-alkylations of R,β-unsaturated carbonyl compounds
are rare.16 Excitingly, the dual-catalytic method offers a new dis-
connection and allows access to these synthetically valuable
R-alkylated R,β-unsaturated carbonyl compounds in one step
from a propargyl alcohol and an allyl carbonate, both of which are
readily available. Importantly, our attempts to trap the vana-
dium-allenoate 2 with a variety of allyl halides were unsuccessful
(no amount of allylated product was observed). We suggest that
the allylating agent in the Pd-catalyzed reaction, being positively
charged, enhances its ability to intercept the vanadium enolate.
Having identified conditions for the vanadium- and palladium-
catalyzed coupling of propargylic alcohols with allylic carbonates,
we sought to examine the scope of this new transformation.
The results of varying the allylic carbonate coupling partner are
shown in Table 2. These results demonstrate that the reac-
tion conditions are general and that the products can be obtained
in moderate to excellent yields (57-98% yield) with high
a Using 1.5 mol % Pd2(dba)3 CHCl3, 3.6 mol % DPPM, and 4.5 mol %
3
OdV(OSiPh3)3. b Using 1.5 equiv of 7 and a reaction time of 48 h.
stereoselectivity (E:Z up to 12:1). More specifically, both alkyl
and aryl substituents can be incorporated at the terminal position of
the olefin (entries 2, 3, and 8), including heterocycles (entries 5 and
6). Additionally, substitution on the internal carbon of the olefin is
tolerated (entry 7), as well as substitution adjacent to the oxygen
(entries 4 and 8). Notably, the observed reactivity difference among
these allyl carbonates tested is consistent with the rate of formation
of the π-allyl-palladium intermediate; carbonates, which are known
to undergo slow oxidative addition, afforded lower yields of the
coupled products. Thus, the choice of a suitable palladium catalyst
aids in improving the efficiency of the overall reaction by tuning the
rate of π-allyl-palladium formation.
Next, we sought to expand the scope of the propargyl alcohols.
A variety of propargyl alcohols were synthesized and subjected to
the reaction conditions (Table 3). Gratifyingly, alcohol substrates
bearing various substituents, including those with electron-rich and
electron-poor aromatics, heterocycles, vinyl groups, and alkyl groups
at the propargylic position participated well in the reaction and
provided the R-alkylated enone products in moderate to excellent
yields (61-96% yield) with up to >19:1 E:Z stereoselectivity.
Additionally, a wide range of substituents with various steric and
electronic properties located on the terminus of the carbon-carbon
triple bond were accommodated.
Overall, our new contemporaneous dual catalysis methodol-
ogy has proven to be applicable to a wide variety of substrates,
including both highly functionalized allyl carbonates and pro-
pargyl alcohols. These general reaction conditions, although not
optimized except for the initial example, generally provided
promising levels of both yield and stereoselectivity. Accordingly,
R-allylated enones were always the predominant products, with
little to no amount of O-allylated product observed. As this dual-
catalytic transformation is performed by a combination of
vanadium and palladium catalysts, its efficiency and selectivity
1708
dx.doi.org/10.1021/ja110501v |J. Am. Chem. Soc. 2011, 133, 1706–1709

Journal of the American Chemical Society
COMMUNICATION
Table 3. Survey of Propargyl Alcohol Scope in Dual Catalysis
of the scope and mechanism of this contemporaneous dual
catalysis are in progress.
’ ASSOCIATED CONTENT
S
Supporting Information. Complete experimental details
b
and characterization of new products. This material is available
free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
bmtrost@stanford.edu
’ ACKNOWLEDGMENT
We thank the NSF (CHE 0948222) for financial support of
this project. X.L. is grateful for a Swiss National Science
Foundation postdoctoral fellowship.
’ REFERENCES
(1) Hartwig, J. F. Organotransition Metal Chemistry: From Bonding to
Catalysis; University Science Books: Mill Valley, CA, 2010.
(2) Organocatalysis; Reetz, M. T., List, B., Jaroch, S., Weinmann, H.,
Eds.; Springer: Berlin, 2008.
(3) Masel, R. I. Chemical Kinetics and Catalysis; Wiley: New York,
2001.
(4) (a) Sammis, G. M.; Danjo, H.; Jacobsen, E. N. J. Am. Chem. Soc.
2004, 126, 9928. (b) Ibrahem, I.; Cꢀordova, A. Angew. Chem., Int. Ed.
2006, 45, 1952. (c) Mukherjee, S.; List, B. J. Am. Chem. Soc. 2007, 129,
11336. (d) Nicewicz, D. A.; MacMillan, D. W. C. Science 2008, 322, 77.
(e) Ikeda, M.; Miyake, Y.; Nishibayashi, Y. Angew. Chem., Int. Ed. 2010,
49, 7289.
(5) (a) Lee, J. M.; Na, Y.; Han, H.; Chang, S. Chem. Soc. Rev. 2004,
33, 302. (b) Shibasaki, M.; Kanai, M.; Matsunaga, S.; Kumagai, N. Acc.
Chem. Res. 2009, 42, 1117.
a Run for 48 h. b Using 1.5 mol % Pd2(dba)3 CHCl3, 3.6 mol % DPPM,
3
and 4.5 mol % OdV(OSiPh3)3. c tert-Butyl cinnamyl carbonate was used
instead of 8. d The reaction used 1.5 equiv of propargyl alcohol and was
run at 100 °C for 48 h.
are affected not only by the catalyst loading but also by the ratio
of the two catalysts with respect to each other. To highlight the
importance of the proper [V]/[Pd] ratio, we conducted several
experiments that demonstrate its impact. For example, when
catalyst loading ([V]/[Pd]) was increased from 4.5/3.0 mol % to
5.0/5.0 mol % while maintaining other conditions constant for
entry 7 in Table 2, the yield of desired product was increased to
97% (vs 88%) without erosion of stereoselectivity. When only
the palladium catalyst loading was increased from 1.0 to 2.0 mol %
for entry 3 in Table 3, the yield of desired product was dramatically
decreased (<5 vs 67%), whereas the O-allylated product was
obtained in 83% yield. When only the vanadium catalyst loading
was decreased from 4.5 to 3.0 mol % for entry 5 in Table 3, the yield
of desired product was significantly increased to 87% (vs 78%)
without decrease of stereoselectivity. These examples demonstrate
the important nature of the relative catalyst loadings and the poten-
tial to further optimize individual reactions.
In conclusion, we have established the feasibility of coupling
two highly reactive catalytic intermediates while omitting the
competing side reactions with stoichiometric reagents. This
transformation was achieved through the use of a combination
of a vanadium-catalyzed 1,3-transposition of propargylic alcohols
and a palladium-catalyzed alkylation of allylic carbonates. To our
knowledge, this is the first example of trapping a catalytic amount
of a highly reactive metal intermediate with a second catalytically
generated metal intermediate in the presence of stoichiometric
amounts of competitive reagents. We anticipate that this new
tactic will provide access to a wide range of new chemical
processes for generating important compounds. Further studies
(6) (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395.
(b) Trost, B. M. Acc. Chem. Res. 2002, 35, 695.
(7) Trost, B. M.; Fandrick, D. R. Aldrichimica Acta 2007, 40, 59.
(8) (a) Trost, B. M.; Oi, S. J. Am. Chem. Soc. 2001, 123, 1230. (b)
Trost, B. M.; Chung, C. K. J. Am. Chem. Soc. 2006, 128, 10358.
(9) It is interesting to note the superiority of DPPM, which
traditionally has not been used as a ligand for allylic alkylation.
(10) Pauling, H.; Andrews, D. A.; Hindley, N. C. Helv. Chem. Acta
1976, 59, 1233.
(11) Haight, A. R.; Stoner, E. J.; Peterson, M. J.; Grover, V. K. J. Org.
Chem. 2003, 68, 8092.
(12) Iwasawa, N.; Maeyama, K.; Saitou, M. J. Am. Chem. Soc. 1997,
119, 1486.
ꢀ
(13) Sanz, R.; Martínez, A.; Alvarez-Gutiꢀerrez, J. M.; Rodríguez, F.
Eur. J. Org. Chem. 2006, 1383.
(14) This result represents the first example of dual catalysis where
two reactive intermediates generated irreversibly in situ react with each
other in the presence of large excesses of competing reagents. It is
important to note that the previous examples of dual catalysis have
employed a catalytically generated enamine intermediate that is formed
reversibly and is much more stable under the reaction conditions in
comparison with our case, where the protonation of the vanadium-
allenoate is much more facile and irreversible (see ref 4.)
(15) Carey, F. A.; Sundberg, R. J. Advanced Organic Chemistry, 5th
ed.; Springer: New York, 2007.
(16) (a) Frank, S. A.; Mergott, D. J.; Roush, W. R. J. Am. Chem. Soc.
2002, 124, 2404. (b) Jellerichs, B. G.; Kong, J. R.; Krische, M. J. J. Am.
Chem. Soc. 2003, 125, 7758.
1709
dx.doi.org/10.1021/ja110501v |J. Am. Chem. Soc. 2011, 133, 1706–1709