Catalytic allylic C-H acetoxylation and benzoyloxylation via suggested (η3-Allyl)palladium(IV) intermediates

Article abstract of DOI:10.1021/ol9023369

"Chemical Equation Presented" Palladium-catalyzed allylic acetoxylations and benzoyloxylations were carried out using iodonium salts. The reactions proceed under mild conditions with high reglo- and stereoselectivity. The catalysis can be performed under

Full text of DOI:10.1021/ol9023369

ORGANIC
LETTERS
2009
Vol. 11, No. 23
5518-5521
Catalytic Allylic C-H Acetoxylation and
Benzoyloxylation via Suggested
(η3-Allyl)palladium(IV) Intermediates
Lukasz T. Pilarski, Nicklas Selander, Dietrich Bo¨se, and Ka´lma´n J. Szabo´*
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm UniVersity,
SE-10691, Stockhom, Sweden
kalman@organ.su.se
Received October 9, 2009
ABSTRACT
Palladium-catalyzed allylic acetoxylations and benzoyloxylations were carried out using iodonium salts. The reactions proceed under mild
conditions with high regio- and stereoselectivity. The catalysis can be performed under both acidic and nonacidic conditions without use of
BQ or other external oxidants and activator ligands. Deuterium labeling experiments clearly show that the catalytic reaction proceeds through
(η3-allyl)palladium intermediates. A stoichiometric study with one of the catalysts provided evidence for the formation of a Pd(IV) species.
Palladium-catalyzed allylic acetoxylation is one of the
most efficient procedures for C-H functionalization of
alkenes.1,2 Åkermark,2a,b Ba¨ckvall,2c,d White,2e,f and
others2g,h have presented many excellent applications for
preparation of allylic acetates via palladium-catalyzed ac-
etoxylation of the allylic C-H bond in various alkenes.
In the vast majority of these studies, benzoquinone (BQ)
was used as a crucial additive. According to in-depth
mechanistic studies, BQ plays a dual role in the acetoxy-
lation reactions: (i) reoxidation of Pd(0) to Pd(II) to
maintain the catalytic cycle and (ii) activation of the Pd(II)
intermediate for selective nucleophilic attack.3 Although
these two important features render BQ an almost indis-
pensible component in allylic acetoxylation reactions,1c
there are a couple of disadvantages as well. For example,
the oxidation potential of BQ is pH dependent, as
protonation is involved in the redox process generating
hydroquinone. As a result, in most catalytic applications
acidic conditions have to be applied, which reduce the
nucleophilicity of the acetate component.2
* To whom correspondence should be addressed (www.organ.su.se/ks).
(1) (a) Tsuji, J. Palladium Reagents and Catalysts. New PerspectiVes
for the 21st Century; Wiley: Chichester, 2004. (b) Dick, A. R.; Sanford,
M. S. Tetrahedron 2006, 62, 2440. (c) Popp, B. V.; Stahl, S. S. Top.
In this paper, we disclose our results of a conceptually
new method for allylic C-H acetoxylation/benzoyloxylation
reactions, which can be performed without BQ (eq 1, Table
1). Our basic idea is replacement of the traditional Pd(0)/
Pd(II) catalytic cycle with a Pd(II)/Pd(IV) cycle.4 Iodonium
salts 3a,b are applied as oxidants for allylic C-H function-
alization of alkenes 2a-f in the presence of palladium
catalysts 1a or 1b (eq 1).
Organomet. Chem. 2007, 22, 149
.
(2) (a) Hansson, S.; Heumann, A.; Rein, T.; Åkermark, B. J. Org. Chem.
1990, 55, 975. (b) Åkermark, B.; Larsson, E. M.; Oslob, J. D. J. Org.
Chem. 1994, 59, 5729. (c) Grennberg, H.; Ba¨ckvall, J.-E. Chem.sEur. J.
1998, 4, 1083. (d) Grennberg, H.; Simon, V.; Ba¨ckvall, J.-E. J. Chem. Soc.
Chem. Commun. 1994, 265. (e) Chen, M. S.; White, M. C. J. Am. Chem.
Soc. 2004, 126, 1346. (f) Delcamp, J. H.; White, M. C. J. Am. Chem. Soc.
2006, 128, 15076. (g) McMurry, J. E.; Kocovsky, P. Tetrahedron Lett. 1984,
25, 4187. (h) Tsuji, J.; Sakai, K.; Nagashima, H.; Shimuzu, I. Tetrahedron
Lett. 1981, 22, 131
.
10.1021/ol9023369 CCC: $xxxx
Published on Web 11/09/2009
 2009 American Chemical Society

for 18 h. Palladium acetate 1a and pincer complex 1b
proved to be equally efficient catalysts for model substrate
2a (cf. entries 1 and 2). For this reason, and because of
its commercial availability, the majority of the reactions
were performed using 1a. However, the fact that 1b readily
catalyzed the process provided a handle for the exploration
of some important mechanistic aspects of the reaction
(vide infra). Most of the acetoxylation reactions (entries
1, 2, 5, and 7-10) were performed in acetic acid with 3a
in the presence of KOAc.
Table 1. Pd-Catalyzed C-H Acetoxylation and
Benzoyloxylation of Alkenes with Iodonium Saltsa
The benzoyloxylation reactions (entries 4 and 6) were
conducted with 3b as oxidant in the presence of LiOBz
in MeCN solvent. Thus, the catalytic C-H benzyloxyla-
tion could be performed under slightly basic conditions,
which is not possible with BQ as oxidant (vide supra).
We found that the acetoxylation reaction can also be
performed under similar nonacidic conditions (entry 3).
The catalytic reactions proceeded smoothly with both
terminal (2a) and internal (2b-f) alkenes. In the case of
carboxylates 2a-d we did not observe isomerization
reported for the classical Pd(0)/Pd(II) reactions, in which
BQ pentyl nitrite or CuCl2 were used as oxidants.2h Our
conditions gave exclusively one regio- and stereoisomer
of each product. The only exception was the acetoxylation
of 2e in which the cis and trans products were formed in
a 1:1 ratio. In the case of 2b-d, the alkyl chain length
did not influence the selectivity of the reaction. Acetoxy-
lation of cyclohexene (2f) was performed with high
regioselectivity without formation of the vinyl-substituted
product (entry 10).
We were particularly interested in exploring two basic
mechanistic aspects of the catalytic process: (1) the
mechanism of the nucleophilic attack and (2) the oxidation
state of the palladium under the reaction conditions. The
nucleophilic attack of the C-H acetoxylation reaction can
proceed by two different mechanisms.2c,d One possibility
is the formation of an (η3-allyl)palladium (such as 5, eq 2)
followed by nucleophilic attack of the acetate. The other
possibility is an acetoxy-palladation (through intermediates
such as 6, eq 3), followed by ꢀ-hydride elimination to give
the final product. To differentiate between these two mech-
anisms, we carried out an isotope labeling study with
monodeuterated cyclohexene (2f-1d). Using this substrate
two different (η3-allyl)palladium complexes (5a and 5b, eq
2) may be expected to form, each of which could react with
the nucleophile at either of the terminal positions. This would
give three allylic acetoxy products, 4h-1d, 4h-2d, and 4h-
3d, in an expected 1:2:1 ratio. Alternatively, if the palladium
complex undergoes acetoxy-palladation, two insertion com-
plexes may form: 6a and 6b (eq 3). Subsequent ꢀ-hydride
elimination would give two allylic acetates 4h-1d and 4h-
2d in a 1:1 ratio. In this case, formation of 4h-3d (in which
a Alkene 2 (0.3 mmol), iodonium salt 3 (0.6 mmol), catalyst 1 (5 mol
%), and KOAc (0.3 mmol) or LiOBz (0.3 mmol) were dissolved/suspended
in the indicated solvent (1 mL), and the reaction mixture was stirred at 40
°C for 18 h. KOAc and LiOBz were used with oxidants 3a and 3b,
respectively. b Isolated yield. c Cis and trans isomers formed in a 1:1 ratio.
Under our conditions, alkene 2 (0.3 mmol), iodonium
salt 3 (0.6 mmol), catalyst 1 (5 mol %), and the appropriate
carboxylate salt base were reacted in AcOH or MeCN.
Potassium acetate was used with PhI(OAc)2 (3a), while
lithium benzoate was used with PhI(OBz)2 (3b). The
reactions were conducted under mild conditions (40 °C)
(3) (a) Ba¨ckvall, J. E.; Bystro¨m, S. E.; Nordberg, R. E. J. Org. Chem.
1984, 49, 4619. (b) Ba¨ckvall, J.-E.; Gogoll, A. Tetrahedron Lett. 1988, 29,
2243. (c) Szabo´, K. J. Organometallics 1998, 17, 1677.
(4) (a) Canty, A. J.; Rodemann, T.; Ryan, J. H. AdV. Organomet. Chem.,
Vol. 55, Elsevier, New York, 2008, p. 279. (b) Canty, A. J. Dalton Trans.
2009, DOI: 10.1039/b914080h. (c) Deprez, N. R.; Sanford, M. S. Inorg.
Chem. 2007, 46, 1924. (d) Stowers, K. J.; Sanford, M. S. Org. Lett. 2009,
11, 4584. (e) Dick, A. R.; Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc.
2004, 126, 2300. (f) Daugulis, O.; Zaitsev, V. D. Angew. Chem., Int. Ed.
2005, 44, 4046. (g) Tong, X.; Beller, M.; Tse, M. K. J. Am. Chem. Soc.
2007, 129, 4906. (h) Mun˜iz, K.; Ho¨velmann, C. H.; Streuff, J. J. Am. Chem.
Soc. 2008, 130, 763. (i) Liu, G.; Stahl, S. S. J. Am. Chem. Soc. 2006, 128,
7179. (j) Bressy, C.; Alberico, D.; Lautens, M. J. Am. Chem. Soc. 2005,
127, 13148. (k) Alexanian, E. J.; Lee, C.; Sorensen, E. J. J. Am. Chem.
Soc. 2005, 127, 7690. (l) Aydin, J.; Larsson, J. M.; Selander, N.; Szabo´,
K. J. Org. Lett. 2009, 11, 2852. (m) Merritt, E. A.; Olofsson, B. Angew.
Chem., Int. Ed. 2009, DOI: 10.1002/anie.200904689.
(5) (a) Lagunas, M.-C.; Gossage, R. A.; Spek, A. L.; van Koten, G.
Organometallics 1998, 17, 731. (b) Canty, A. J.; Rodemann, T.; Skelton,
B. W.; White, A. H. Organometallics 2006, 25, 3996. (c) Dick, A. R.;
Kampf, J. W.; Sanford, M. S. J. Am. Chem. Soc. 2005, 127, 12790. (d)
Racowski, J. M.; Dick, A. R.; Sanford, M. S. J. Am. Chem. Soc. 2009,
131, 10974.
(6) Stang, P. J.; Boehshar, M.; Wingert, H.; Kitamura, T. J. Am. Chem.
Soc. 1988, 110, 3272.
Org. Lett., Vol. 11, No. 23, 2009
5519

the deuterium and the acetate group are separated by three
carbon atoms) is not expected to occur.
the reaction was monitored by 1H NMR spectroscopy (Figure
1). The signals for the CH2 and CH3 protons of the side arms
Figure 1.
Oxidation of the Pd(II) central atom of 1b. The 1H NMR
shifts are given in ppm values.
of 1b appear as singlets resonating at 4.04 and 3.01 ppm,
respectively.5a,b This arises from the fast inversion of the
nitrogen atoms of the side arms, lending a time-averaged
C
2V symmetry to the complex. Addition of 3 equiv of 3a to
the CDCl3 solution of 1b led to a color change from pale
yellow to amber. Notable changes in the 1H NMR spectrum
were (i) the splitting of the CH3 signals into two distinct
sharp singlets resonating at 2.88 and 3.24 ppm and (ii) the
emergence of two doublets (J ) 14 Hz) at 4.46 and 4.38
ppm. (We also observed the formation of iodobenzene,
indicating removal of the acetate groups from 3a.) Canty
and co-workers reported similar characteristic changes in the
1H NMR spectrum for the reaction of NCN complex 1b and
an ethynyl iodonium salt.4a,b,5b A reasonable explanation for
the observed changes in the NMR signals is the expansion
of the palladium coordination sphere from tetracoordinated
(1b) to hexacoordinated (7) species, due to oxidation of the
palladium atom from Pd(II) to Pd(IV). The change of the
coordination number hinders the ring inversion process in
the side arms and reduces the NMR time-scale averaged
symmetry of the complex from C2V to Cs. Consequently, the
chemical environment of the two CH2 side arm protons will
be different, giving rise to a geminal coupling pattern.
Similarly, the chemical environments of the CH3 groups will
also become different, and therefore, these groups resonate
as two distinct singlets. These results suggest that under the
applied catalytic conditions the palladium atom of the catalyst
may be oxidized to Pd(IV) by the iodonium salts 3.
On the basis of these results, we propose the catalytic cycle
shown in Figure 2. The suggested initial step is the oxidative
addition of the Pd(II) complex with the iodonium salt (3a
or 3b) to give Pd(IV) complex 8. The coordination of alkene
2 affords the complex 9 and subsequent deprotonation
generates the (η3-allyl)palladium complex 10, in accordance
with the results of our deuterium labeling study (eqs 2-4).
Alternatively, the (η3-allyl)palladium(II) complex may be
formed first in the case of 1a, followed by oxidation of the
palladium by the iodonium salt. However, an (η3-allyl)pal-
ladium(II) complex cannot be formed from 1b directly, as
in 1b only a single coordination site is available on the metal
center. We assume that formation of (η3-allyl)palladium(IV)
complex 10 from 9 is thermodynamically favorable with the
efficient π-donor allyl moiety stabilizing the electron-poor
palladium(IV) center.
When we performed the catalytic acetoxylation of 2f-1d
under our standard conditions (entry 10), we obtained three
deuterated products (4h-1d, 4h-2d, and 4h-3d, eq 4) in a
2
1:2:1 ratio, as determined by H NMR spectroscopy.
Thus, we conclude that the acetoxylation of 2f involves
an (η3-allyl)palladium intermediate, and very probably the
rest of the substrates (2a-e) react via analogous intermedi-
ates. Interestingly, previous studies on the Pd(0)/Pd(II)-based
catalytic acetoxylation using BQ also showed evidence of
(η3-allyl)palladium intermediates.2c,d
A further important mechanistic question concerns the
oxidation state of palladium. It is known that iodonium salts
are able to oxidize Pd(II) species to Pd(IV).4,5 Sanford and
co-workers have shown that benzoyloxy iodonium salts (e.g.,
3a,b) readily oxidize Pd(II) complexes to form stable Pd(IV)
species.5c,d
This process provided the evidence for the intermediacy
of Pd(IV) species in directed C-H acetoxylation of het-
eroaromatics and related substrates.4e Furthermore, van Koten
and co-workers demonstrated that NCN pincer complex 1b
reacts with PhICl2 to form a Pd(IV) species, which was
observed by 1H NMR spectroscopy.5a In addition, Canty and
co-workers have reported that pincer complex 1b can
undergo oxidative addition with an alkynyl iodonium salt to
give the corresponding Pd(IV) species.5b
Similarly, we attempted to generate Pd(IV) intermediates
starting from pincer complex 1b, which we found to be as
active a catalyst for our acetoxylation protocol as 1a (cf.,
entries 1 and 2). In our investigation, NCN complex 1b was
reacted with iodonium salt 3a in CDCl3 and the progress of
5520
Org. Lett., Vol. 11, No. 23, 2009

3b). When the catalytic reaction of 2a and acetoxy iodonium
salt 3a (entry 3) was repeated with LiOBz as an additive in
place of KOAc, a mixture of 4a and 4b formed. This
indicates that the carboxylate groups may undergo exchange
either in the iodonium salt6 (i.e., 3a is converted to 3b) or
in complexes 8-10. New synthetic applications based on
this principle are underway in our laboratory.
In summary, we have presented a new catalytic allylic
C-H functionalization process and evidence for the forma-
tion of (η3-allyl)palladium(IV) intermediates. The reactions
are suitable for selective acetoxylation and benzoyloxylation
of various terminal and internal alkenes using iodonium salts
as oxidants. This catalytic process is regio- and stereoselec-
tive, and it can be performed under mild conditions. Unlike
some of the Pd(0)/Pd(II) acetoxylation methods,2h the present
procedure does not induce rearrangement of the reagents or
products. Thus, the presented procedure is complementary
to the existing Pd(0)/Pd(II) methods, when rearrangements
or isomerizations have to be avoided or application of
nonacidic conditions is required. The presented process is
also very interesting mechanistically. As far as we know,
this is the first catalytic process in which an (η3-allyl)pal-
ladium(IV) complex is suggested as the key intermediate in
iodonium salt mediated oxidation reactions. Thus, the
presented procedure significantly broadens the scope of
palladium-catalyzed processes proceeding via a presumed
Pd(II)/Pd(IV) redox cycle.4,5
Figure 2. Proposed catalytic cycle for the allylic C-H acetoxyla-
tion/benzoyloxylation reaction.
In the final step of the cycle, reductive elimination gives
products 4 and regenerates the Pd(II) catalyst. This does not
require external activation, as the driving force is the
reduction of Pd(IV) to Pd(II). In contrast, reactions via the
Pd(0)/Pd(II) catalytic cycle require the presence of an
activator ligand. The (η3-allyl)palladium(II)acetate and chlo-
ride complexes are known to be stable for reductive
elimination or external nucleophilic attack, unless an activator
ligand (such as BQ) is present in the reaction mixture.3
Significantly, the reductive elimination from the (η3-al-
lyl)palladium(IV) complex (such as 10) appears to be
inherently facile, obviating the need for BQ or other activator
ligands. The regioselectivity of the nucleophilic attack is the
same as the regioselectivity in the corresponding Pd(0)/Pd(II)
process.2h
Acknowledgment. A postdoctoral stipend by the Carl
Tryggers Stiftelse for L.T.P. and the support of the Swedish
Research Council is greatly acknowledged.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for 4a-h as well as details
of the mechanistic studies are provided. This material is
available free of charge via the Internet at http://pubs.acs.org.
The carboxylate nucleophile of complex 10 does not
necessarily arise from the employed iodonium salts (3a or
OL9023369
Org. Lett., Vol. 11, No. 23, 2009
5521