Palladium-catalyzed oxidative rearrangement of diaryl alkenyl carbinols to β,β-diaryl α,β-unsaturated ketones

Article abstract of DOI:10.1021/ol201177t

An unusual oxidative palladium-catalyzed rearrangement of diaryl alkenyl carbinols to β,β-diaryl α,β-unsaturated ketones is described. The geometry of the alkene product is not determined by the electronic nature of the aryl substitutents but rather is determined by substitution pattern on the aryl rings. The reaction proceeds in good yields, utilizes oxygen at atmospheric pressure as the terminal oxidant, and tolerates a variety of functional groups on the aryl rings.

Full text of DOI:10.1021/ol201177t

ORGANIC
LETTERS
2011
Vol. 13, No. 14
3648–3651
Palladium-Catalyzed Oxidative
Rearrangement of Diaryl Alkenyl
Carbinols to β,β-Diaryl
r,β-Unsaturated Ketones
David Rosa and Arturo Orellana*
Department of Chemistry, York University, 440 Chemistry Building,
4700 Keele Street, Toronto, ON, Canada M3J 1P3
aorellan@yorku.ca
Received May 14, 2011
ABSTRACT
An unusual oxidative palladium-catalyzed rearrangement of diaryl alkenyl carbinols to β,β-diaryl R,β-unsaturated ketones is described. The
geometry of the alkene product is not determined by the electronic nature of the aryl substitutents but rather is determined by substitution pattern
on the aryl rings. The reaction proceeds in good yields, utilizes oxygen at atmospheric pressure as the terminal oxidant, and tolerates a variety of
functional groups on the aryl rings.
Tremendous advances in transition metal catalysis have
revolutionized synthetic organic chemistry. In particular,
palladium catalyzed reactions have found widespread use,
and progress in this area continues unabated.1 We have
been interested in studying the palladium-catalyzed chem-
istry of tertiary alcohols2 proceeding through a carbonÀ
carbon bond cleavage step.3 In earlier reports,4 we have
documented carbonÀcarbon bond forming reactions of
palladium homoenolates.5 In these instances, the forma-
tion of the homoenolate moiety occurred via the known
palladium-catalyzed and strain-driven cleavage or ring-
expansion of cyclopropanols and isopropenyl-tert-cyclo-
butanols, respectively. Withaninterest indevelopinga new
entry into palladium homoenolates, we began studying
tertiary allylic alcohols that do not benefit from ring strain
as a driving force. Specifically, we envisioned that diaryl
alkenyl carbinols6,7 could give rise to a palladium homo-
enolate through electrophilic activation of the alkenyl
function followed by an aryl 1,2-migration.8 The resulting
palladium homoenolate could then participate in further
reactions.
(1) Reviews: (a) Metal-Catalyzed Cross-Coupling Reactions; de
Meijere, A., Diederich, F., Eds.; Wiley-VCH: Weinheim, Germany, 2004.
(b) Nicolaou, K. C.; Bulger, P. G.; Sarlah, D. Angew. Chem., Int. Ed. 2005,
45, 4442.
(2) For a review of palladium-catalyzed reactions of alcohols, see:
Muzart, J. Tetrahedron 2005, 61, 9423.
(3) For reviews on palladium-catalyzed reactions proceeding
through β-carboelimination, see: (a) Seiser, T.; Cramer, N. Org. Biomol.
Chem. 2009, 7, 2835. (b) Satoh, T.; Miura, M. Top. Organomet. Chem
2007, 24, 61. (c) Satoh, T; Miura, M. Top. Organomet. Chem 2005, 14, 1.
(d) Kuwajima, I.; Nakamura, E. In Topics in Current Chemistry; Springer-
Verlag: Berlin, Germany, 1990; Vol. 155, p 1. (e) Kuwajima, I.; Nakamura,
E. In Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon: Oxford, U.K., 1991; Vol. 2, p 441. (f) Kuwajima, I. Pure Appl.
Chem. 1988, 60, 115.
(4) (a) Schweinitz, A.; Chtchemelinine, A.; Orellana, A. Org. Lett.
2011, 13, 232. (b) Rosa, D.; Orellana, A. Org. Lett. 2011, 13, 110.
(5) For a good entry into the palladium homoenolate literature, see:
(a) Molander, G. A.; Jean-Gerard, L. J. Org. Chem. 2009, 74, 1297. For
a concise overview of homoenolate chemistry see: (b) Lettan, R. B.;
Galliford, C. V.; Woodward, C. C.; Scheidt., K. A. J. Am. Chem. Soc.
2009, 131, 8805.
(6) Diaryl alkenyl carbinols were readily prepared by either addition
of alkenyllithiums or Grignard reagents to the corresponding benzo-
phenones, or addition of aryllithiums to the corresponding unsaturated
ketones. See Supporting Information for details.
(7) This substrate class has not been explored extensively. For the
synthesis of epoxides through palladium-catalyzed carboetherification,
see: Hayashi, S; Yorimitsu, H.; Oshima, K. J. Am. Chem. Soc. 2009, 131,
2052.
(8) For selected examples of aryl 1,2-migration in transition-metal-
catalyzed reactions, see: (a) Shen, H.-C.; Pal, S.; Lian, J.-J.; Liu, R.-S. J.
Am. Chem. Soc. 2003, 125, 15762. (b) Dudnik, A. S.; Grevorgyan, V.
Angew. Chem., Int. Ed. 2007, 46, 5195.
r
10.1021/ol201177t
Published on Web 06/15/2011
2011 American Chemical Society

In initial experiments, treatment of tertiary allylic alco-
hol 1 with stoichiometric amounts of palladium(II) salts did
not promotethedesired aryl1,2-shift, but ratherresultedin
the formation of indene 2 and β,β-diaryl R,β-unsaturated
ketone 3 (eq 1). The formation of arylated indene 2 is likely
the result of ionization of the allylic alcohol and intramo-
lecular FriedelÀCrafts reaction.9 The formation of unsa-
turated ketone 3,10 in which a reorganization of the carbon
framework associated with the allylic alcohols from a
branched to a linear arrangement of carbon atoms has
occurred, was unexpected and is not readily rationalized.
We note in passing that 1,1-diaryl styrene systems of this
type are useful synthetic intermediates, and that their
hydrogenation leads to 1,1-diaryl fragments11 that are
valuable in drug discovery. More importantly, the unusual
reactivity observed prompted us to study this reaction
further.
solvent mixture significantly diminished the generation of
benzophenone.
Table 1. Reaction Optimizationa,g
time
(h)
yield (%)d
(3:4)e
solventb
Pd sourcec
base
1
MeCNf
DMF
NMP
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMA
DMAf,h
PdCl2
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
K2CO3
20
20
20
20
10
10
24
10
10
10
48
10
10
10
10
16 (100:0)
26 (85:15)
35 (72:28)
76 (81:19)
68 (70:30)
68 (78:22)
39 (67:33)
80 (69:31)
74 (72:28)
90 (72:28)
26 (50:50)
80 (51:49)
77 (82:18)
80 (88:12)
80 (98:2)
2
PdCl2
3
PdCl2
4
PdCl2
5
PdBr2
6
PdCl2MeCN2
PdCl2(PPh3)2
PdCl2
7
8
9
PdCl2
Na2CO3
NaHCO3
Ag2CO3
NaOAc
KOtBu
CsOAc
10
11
12
13
14
15
PdCl2
PdCl2
PdCl2
PdCl2
PdCl2
PdCl2
CsOAc
We reasoned that the inclusion of a stoichiometric
amount of inorganic base would prevent formation of aryl
indenes. Under these conditions the reaction provided a
mixture of ketone 3 and benzophenone, the latter product
likely arising through palladium-catalyzed β-carboelimina-
tion of the isopropenyl fragment. Further optimization
focused on using catalytic amounts of palladium in the
presence of a co-oxidant (Table 1). It was quickly established
that common oxidants (Cu and Ag salts, DDQ) did not
ameliorate the reaction. The use of molecular oxygen as the
terminal oxidant12 was explored using a variety of solvents
with limited success. However, the use of dimethylacetamide
(DMA)13 resulted in a significantly increased yield. A screen
of palladium(II) sources and bases revealed PdCl2 and
CsOAc to be optimal. Finally, the use of DMA/MeCN
a All reactions were conducted on a 0.11 mmol scale at a 0.1 M
concentration using 20 mol % of the palladium catalyst and 1.1 equiv of
base. b All reactions were conducted at 100 °C unless otherwise noted.
c Pd(OAc)2 and PdTFA2 did not yield any unsaturated ketone. d Combined
isolated yield of benzophenone and unsaturated ketone. e Ratio determined
by quantitative 1H NMR of crude reaction mixtures. f Reactions conducted
at 80 °C. g A catalyst loading study revealed that as little as 1 mol % of
PdCl2 could be used. However, the reaction required 72 h to reach
completion. For convenience a catalyst loading of 20 mol % was used on
all subsequent reactions. h A 3:1 ratio of DMA/MeCN was used.
Next, we conducted a series of experiments aimed at
expanding the scope of this reaction (Table 2). Not surpris-
ingly, the carbinol derived from benzophenone provided
the expected product in good yield (entry 1). A substrate
bearing a phenyl and an electron-rich aryl ring provided a
1:1 mixture of products in good yield (entry 2). Substitut-
ing the phenyl ring with an electron-poor aryl ring did not
affect the product distribution (entry 3). Likewise, the use
of a substrate bearing two ortho-substituted aryl rings also
resulted in a 1:1 mixture of products (entry 4). Taken
together, these results suggest that the process is insensitive
to the electronic nature of the aryl ring. In contrast, the use
of a substrate bearing only one ortho-substituted aryl ring
resulted in the formation of a single product (entry 5).14
A series of substrates bearing one ortho-susbtituted aryl
ring were subjected to the optimized reaction conditions to
explore the generality of this selectivity (Figure 1). In all
cases studied only the product having a cis-relationship
between the ortho-substituted ring and acyl moiety was
(9) Smith, C. D.; Rosoch, G.; Mui, L.; Batey, R. J. Org. Chem. 2010,
75, 4716.
(10) The structure of ketone 3 could be readily assigned based on
calculation of the chemical shift for the vinyl proton and characterizaion
of its hydrogenation product. See Supporting Information for full
details.
(11) For a leading reference on the enantioselective hydrogenation of
1,1-diaryl alkenes, see: Wang, X.; Guram, A.; Caille, S.; Hu, J.; Preston,
J. P.; Ronk, M.; Walker, S. Org. Lett. 2011, 13, 1881.
(12) Reviews: (a) Gligorich, K. M.; Sigman, M. S. Chem. Commun.
2009, 3854. (b) Stahl, S. S. Science 2005, 309, 1824. (c) Stahl, S. Angew.
Chem., Int. Ed. 2003, 43, 3400.
(13) (a) Mitsudome, T.; Mizumoto, K.; Mizugaki, T.; Jitsukawa, K.;
Kaneda, K. Angew. Chem., Int. Ed. 2010, 49, 1238. (b) Mitsudome, T.;
Umetani, T.; Nozaka, N.; Mori, K.; Mizugaki, T.; Ebitami, K.; Kaneda,
K. Angew. Chem., Int. Ed. 2006, 68, 5236.
(14) The double bond geometry of all products arising from unsym-
metrical carbinols was determined using nuclear Overhauser difference
(NOE) experiments. See Supporting Information for details.
Org. Lett., Vol. 13, No. 14, 2011
3649

Table 2. Effect of Aryl Substitution on Product Distributiona
a All reactions were conducted at 0.1 M concentration using 20 mol
% of PdCl2, 1.1 equiv of CsOAc, and a DMA/MeCN ratio of 3:1 (v/v).
b Isolated yields. c Determined using 1H NMR of crude reaction
mixtures.
observed, regardless of the electronic nature of the rings
(entries 1 to 14). This suggests that the geometry-determin-
ing step in the reaction is influenced by the steric demands
of the aryl groups. A variety of functionalized aryl groups
are tolerated in this reaction including alkyl- (entries 1 and
10), alkoxy- (entries 1 to 8, 10 to 12, and 14), trifluoro-
methyl- (entries 6 and 11), vinyl- (entry 7), chloro- (entries
3 and 14), fluoro- (entries 4 and 13) and bromo- (entry 12)
substituted aryl rings. The fact that chloro-, bromo-, and
vinyl substituents do not participate in any palladium-
catalyzed side reactions is particularly interesting. Further-
more, it is important to note that although the majority of
examples explored bear an ortho-methoxy-substituted aryl
group15 (entries 1À8), the nature of this substituent does
not appear to be an important factor. Indeed, substrates
bearing a variety of other functional groups at the ortho-
position, including phenyl (entry 9), methyl (entry 10), tri-
fluoromethyl (entry 11), bromo (entry 12), fluoro (entry 13),
and chloro (entry 14), display the same reactivity and
selectivity.
Figure 1. Selective product formation with substrates bearing
ortho-substituted aryl rings.
effect on the success of this reaction. Substrates bearing
isopropenyl (Tables 1 and 2, and Figure 1), vinyl, and 2-(1-
butenyl) groups provide the rearranged products in good
yields, while those bearing 1-(trans-1-hexenyl) or cyclohex-
enyl groups react sluggishly and provide intractable mix-
tures of products (Scheme 1).
At present, we cannot offer a clear mechanistic picture
forthisunusualreaction. However, wehavebegun probing
its mechanism. A number of studies have documented the
palladium-catalyzed β-carboelimination of aryl fragments
from triaryl carbinols to generate arylpalladium(II) inter-
mediates.16 In particular, the selective β-carboelimination of
ortho-substituted aryl rings from tertiary carbinols has been
ascribed to steric effects.17 We have conducted a number
of crossover experiments to explore the possibility of a
Substrates bearing heterocycles did not provide the
expected products but rather formed the ketone resulting
from carboelimination of the isopropenyl unit or gave
intractable reaction mixtures (not shown). The nature of
the alkene component in the substrate also has a profound
(15) A number of benzophenones were prepared in one step from
salicylaldehyde using the palladium-catalyzed cross-coupling with aryl
halides. See: Satoh, T.; Itaya, T.; Miura, M.; Nomura, M. Chem. Lett.
1996, 25, 823.
(16) Terao, Y.; Wakui, H.; Satoh, T.; Miura, M.; Nomura, M. J. Am.
Chem. Soc. 2001, 123, 10407.
3650
Org. Lett., Vol. 13, No. 14, 2011

Scheme 1. Effect of Alkene Variation
Scheme 2. Crossover Experiment
palladium-catalyzed β-carboelimination of the aryl or alkenyl
groups in the present reaction (Scheme 2).18 For example,
when two substrates derived from two different benzophe-
nones and bearing different alkenyl groups were subjected to
the optimized reaction conditions, only products A and B
were observed. Although some of the corresponding benzo-
phenones (not shown) could be detected, no evidence of
products C or D, resulting from an exchange of the alkenyl
moiety, or products E or F, arising from exchange of the aryl
substituents, was observed. This experiment suggests that the
mechanism of this reaction does not involve fragmentation of
the starting material into two reactive partners but rather
proceeds in an intramolecular fashion. Further studies aimed
at elucidating the mechanism of this transformation are
underway.
This reaction provides a single geometric isomer of the
products when one ortho-substituted aryl ring is incorpo-
rated in the substrate, proceeds in good yields, tolerates a
variety of functional groups on the aryl rings, and utilizes
oxygen at atmospheric pressure as the terminal oxidant.
Acknowledgment. We gratefully acknowledge the sup-
port of this work from York University and the Natural
Sciences and Engineering Research Council of Canada.
We thank Dr. Howard N. Hunter (York University) for
assistance in conducting NOE difference experiments.
In conclusion, we have discovered an unusual palla-
dium-catalyzed oxidative rearrangement of alkenyl diaryl
carbinols to β,β-disubstituted R,β-unsaturated ketones.
Supporting Information Available. Experimental pro-
cedures and compound characterization data. This ma-
terial is available free of charge via the Internet at http://
pubs.acs.org.
(17) Terao, Y.; Wakui, H.; Nomoto, M.; Satoh, T.; Miura, M.;
Nomura, M. J. Org. Chem. 2003, 68, 5236.
(18) See Supporting Information for experimental details.
Org. Lett., Vol. 13, No. 14, 2011
3651