Palladium-catalyzed oxidative allylic C-H silylation

Article abstract of DOI:10.1021/ol200445b

Palladium-catalyzed allylic C-H silylation was performed with use of hexamethyldisilane as the silyl source. These C-H functionalization reactions occur only in the presence of hypervalent iodine reagents or other strong oxidants and proceed with excellen

Full text of DOI:10.1021/ol200445b

ORGANIC
LETTERS
2011
Vol. 13, No. 7
1888–1891
Palladium-Catalyzed Oxidative Allylic
C-H Silylation
Johanna M. Larsson, Tony S. N. Zhao, and Kalman J. Szabo*
ꢀ ꢀ
ꢀ
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University,
SE-10691, Stockholm, Sweden
kalman@organ.su.se
Received February 18, 2011
ABSTRACT
Palladium-catalyzed allylic C-H silylation was performed with use of hexamethyldisilane as the silyl source. These C-H functionalization
reactions occur only in the presence of hypervalent iodine reagents or other strong oxidants and proceed with excellent regioselectivity, providing
the linear allylic isomer of the allylsilane products. In demonstrating the first oxidative allylic C-H silylation of alkenes, this study marks an
important advance for the catalytic C-H functionalization method.
Catalytic C-H functionalization provides a simple,
atom-economical alternative to some of the classical func-
tional group transformation based procedures.1-11 Cur-
rently, further development of this potentially very
powerful method faces two important challenges: (i) find-
ing selective functionalization techniques for the transfor-
mation of only a single C-H bond in the substrate to avoid
formation of complex isomeric mixtures of products and
(ii) extending the synthetic scope of the C-H bond activa-
tion based methods to create a wide variety of C-X
bonds.
ofalkenesand areneswithmaingroup reagentsisemerging
as a powerful tool for the synthesis of organometallic
reagents.3,10,11 For example, C-H borylation has become
a useful approach for obtaining organoboronate reagents
from alkenes and arenes.3 Catalytic C-H silylation, how-
ever, is much less developed than the corresponding
borylation2,10 processes, probably because of the lower
reactivity of commonly employed organometallic silyl
sources compared to the borane sources.
Previously reported C-H silylation reactions include
aromatic,12-14 benzylic,15 vinylic,16-18 and alkyl19,20 sub-
strates. However, in contrast to allylic C-H borylation,21-24
catalytic allylic C-H silylation has remained almost entirely
undeveloped. To date, catalytic allylic silylations have entirely
relied on the conversion of preinstalled functional
Although the most developed areas involve replacing
C-H with C-C and C-O bonds,4-9 the functionalization
(1) Dyker, G. Handbook of C-H Transformations: Applications in
Organic Synthesis; Wiley-VCH: Weinheim, Germany, 2005.
(2) Hartwig, J. Organotransition Metal Chemistry; University Science
Books: Sausalito, CA, 2010.
(3) Mkhalid, I. A. I.; Murphy, J. M.; Barnard, J. H.; Marder, T. B.;
Hartwig, J. F. Chem. Rev. 2010, 110, 890.
(12) Tsukada, N.; Hartwig, J. F. J. Am. Chem. Soc. 2005, 127, 5022.
(13) Lu, B.; Falck, J. R. Angew. Chem., Int. Ed. 2008, 47, 7508.
(14) Ihara, H.; Suginome, M. J. Am. Chem. Soc. 2009, 131, 7502.
(15) Kakiuchi, F.; Tsuchiya, K.; Matsumoto, M.; Mizushima, B.;
Chatani, N. J. Am. Chem. Soc. 2004, 126, 12792.
(4) Alberico, D.; Scott, M. E.; Lautens, M. Chem. Rev. 2007, 107,
174.
(16) LaPointe, A. M.; Rix, F. C.; Brookhart, M. J. Am. Chem. Soc.
1997, 119, 906.
(17) Lu, B.; Falck, J. R. J. Org. Chem. 2010, 75, 1701.
(18) Jiang, Y.; Blacque, O.; Fox, T.; Frech, C. M.; Berke, H. Chem.;
Eur. J. 2009, 15, 2121.
(5) Stuart, D. R.; Fagnou, K. Science 2007, 316, 1172.
(6) Colby, D. A.; Bergman, R. G.; Ellman, J. A. Chem. Rev. 2010,
110, 624.
(7) Lewis, J. C.; Bergman, R.; Ellman, J. A. Acc. Chem. Res. 2008, 41,
1013.
(19) Sadow, A. D.; Tilley, T. D. Angew. Chem., Int. Ed. 2003, 42, 803.
(20) Sadow, A. D.; Tilley, T. D. J. Am. Chem. Soc. 2005, 127, 643.
(21) Caballero, A.; Sabo-Etienne, S. Organometallics 2007, 26, 1191.
(8) Sehnal, P.; Taylor, R. J. K.; Fairlamb, I. J. S. Chem. Rev. 2010,
110, 824.
(9) Lyons, T. W.; Sanford, M. S. Chem. Rev. 2010, 110, 1147.
(10) Marciniec, B. Coord. Chem. Rev. 2005, 249, 2374.
(11) Jazzar, R.; Hitce, J.; Renaudat, A.; Sofack-Kreutzer, J.; Baudoin,
O. Chem.;Eur. J. 2010, 16, 2654.
ꢀ
(22) Olsson, V. J.; Szabo, K. J. Angew. Chem., Int. Ed. 2007, 46, 6891.
ꢀ
(23) Olsson, V. J.; Szabo, K. J. J. Org. Chem. 2009, 74, 7715.
(24) Kondoh, A.; Jamison, T. F. Chem. Commun. 2010, 46, 907.
r
10.1021/ol200445b
Published on Web 03/08/2011
2011 American Chemical Society

Table 1. Oxidative Silylation of Allylic C-H Bondsa
Scheme 1. Synthesis of Allylsilanes by Catalytic C-H Silylation
groups.25-32 We have found only a single reaction reported
by Ishikawa and co-workers,33 which is based on palladium-
catalyzed C-H activation of a single substrate using a highly
reactive disilacyclobutene derivative as the silyl source. How-
ever, the poor accessibility of disilacyclobutenes and the low
chemoselectivity of this process limits the practical use of the
method.
Our efforts were directed toward the development of
palladium-catalyzed allylic C-H silylation using readily
accessible starting materials, including commercially avail-
able hexamethyldisilane as the silyl source. We have now
found that such reactions can be performed in the presence
of hypervalent iodine reagents34 (orotherstrongoxidants),
which have been successfully employed in many recently
reported C-H functionalization processes.8,9,35-37 For
example, we have shown that palladium-catalyzed allylic
C-H acyloxylation36 and vinylic C-H borylation37 can be
performed with iodine(III) reagents. In a typical reaction
(Scheme 1) the alkene (1), hexamethyldisilane (2), and the
appropriate oxidant (3) were reacted in the presence of
palladium catalyst 4 at 80 °C for 18 h.
These reaction conditions are relatively mild compared
to the conditions for the aromatic12-14 or benzylic15 C-H
silylation reactions, and because of the oxidative condi-
tions sacrificial hydrogen acceptors (such as norbornene)
are not needed. A high level of functional group tolerance
was achieved even in the presence of strong oxidants (such
as 3a,b); ester (1a-c), amide (1d,e,g), and benzyl (1a,d)
functional groups remained unchanged. The regioselectivity
a Unlesss otherwise stated, a mixture of 1 (0.2 mmol), 2 (0.4 mmol),
the oxidant, and the palladiumcatalyst 4 (5 mol %) was stirred for 18 h at
80 °C. b Method A: 3a (0.4 mmol) was used as oxidant in DME (0.5 mL).
Method B: 3b (0.4 mmol) and 3c (0.2 mmol) were used as oxidants in
THF (0.5 mL) in the presence of 6 (0.2 mmol). c E/Z ratio determined by
1H NMR from the crude reaction mixture. d Isolated yield of both
isomers. e THF was used as a solvent. f 6 (0.1 mmol) was added. g At
60 °C.
of the silylation reaction was excellent with only linear
isomers forming. A high stereoselectivity was achieved for
most substrates. Typically, the trans isomer was formed as
the major (entries 1-6) or the only product (entries 7 and
10). However, for sulfone 1f and sulfonamide 1g the
selectivity was changed and the cis isomer became the
major product (entries 8 and 9). With the exception of 5f
and 5g, the trans and cis allylsilanes could easily be
separated by column chromatography and isolated as
single isomers.
In the absence of alkene substrate (1) iodane 3a and
disilane 2 underwent degradation even without palladium
catalyst. When the reaction was conducted under standard
catalytic conditions, but without oxidants(3), formation of
allylsilane products could not be observed. We found that
hypervalent iodine reagent 3a proved efficient for most of
(25) Suginome, M.; Ito, Y. Chem. Rev. 2000, 100, 3221.
(26) Horn, K. A. Chem. Rev. 1995, 95, 1317.
(27) Tsuji, Y.; Funato, M.; Ozawa, M.; Ogiyama, H.; Kajita, S.;
Kawamura, T. J. Org. Chem. 1996, 61, 5779.
ꢀ
ꢀ
(28) Macsari, I.; Hupe, E.; Szabo, K. J. J. Org. Chem. 1999, 64, 9547.
(29) Suginome, M.; Ohmura, T.; Miyake, Y.; Mitani, S.; Ito, Y.;
Murakami, M. J. Am. Chem. Soc. 2003, 125, 11174.
(30) Hayashi, S.; Hirano, K.; Yorimitsu, H.; Oshima, K. J. Am.
Chem. Soc. 2007, 129, 12650.
(31) Moser, R.; Nishikata, T.; Lipshutz, B. H. Org. Lett. 2010, 12, 28.
ꢀ
(32) Selander, N.; Paasch, J. R.; Szabo, K. J. J. Am. Chem. Soc. 2011,
133, 409.
(33) Naka, A.; Hayashi, M.; Okazaki, S.; Ishikawa, M. Organome-
tallics 1994, 13, 4994.
(34) Zhdankin, V. V.; Stang, P. J. Chem. Rev. 2008, 108, 5299.
(35) Dick, A. R.; Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc. 2004,
126, 2300.
€
(36) Pilarski, L. T.; Selander, N.; Bose, D.; Szabo, K. J. Org. Lett.
ꢀ
2009, 11, 5518.
ꢀ
(37) Selander, N.; Willy, B.; Szabo, K. J. Angew. Chem., Int. Ed. 2010,
49, 4051.
Org. Lett., Vol. 13, No. 7, 2011
1889

the substrates, providing good yields and selectivities.
Other iodine(III) reagents, such as PhI(OAc)2, PhI-
(OCOCF3)2, or PhI(OBz)2, could also be applied, but the
use of these oxidants resulted in lower yields of the silylated
product. Iodonium reagent 3a could be replaced by per-
oxide 3b. However, to achieve as high yields and selectiv-
ities with 3b as with 3a, additives such as benzoquinone,
BQ (3c) and 4-nitrobenzoic acid (6) alsohad to be used (cf.,
entries 1 and 2). BQ (3c) may both act as a co-oxidant38,39
and facilitate40 the coordination of the alkene to the
palladium catalyst. Inclusion of 6 is required to ensure
protic conditions necessary for the redox process of BQ.
Although the 3b,c oxidation system is less powerful than 3a,
it doesnot produce PhI as a byproduct, and therefore Heck-
type41 side reactions could be suppressed. These side reac-
tions were particularly problematic for C-H silylation of
sulfones 1f,g, for which formation of the Heck-type product
with PhI decreased the yield of silanes 5f,g (Table 1).
We have previously shown that alkenes undergo palla-
dium-catalyzed C-H acyloxylation in the presence of
hypervalent iodine reagents.36 Thisreaction couldcompete
with the silylationunder ourreaction conditions. Choiceof
solvent and oxidant was important to decrease or even
completely avoid the formation of the acyloxylation pro-
ducts. Pd(OAc)2 (4a) proved to be an efficient catalyst for
most of the reactions. However, for some substrates the
chemoselectivity was low. For example, the transforma-
tion of amide 1e to 5e was accompanied by considerable
amounts of acyloxylation products. The chemoselectivity
of the reaction could be shifted to formation of allylsilane
5e by employing palladacycle 4b as catalyst (entry 7). The
same catalyst also gave high yield and selectivity for C-H
silylation of 1a using the 3b,c oxidant system (entry 3).
C-H silylation of allylbenzene derivative 1h under our
typical conditions (at 80 °C) using 4a gave a complex
reaction mixture and a low yield of silylated product 5h.
However, we have found that the reactivity could be
increased and the formation of Heck-type and other by-
products reduced by decreasing the temperature to 60 °C
and using selenium-based42 palladium catalyst 4c. The
substrate scope of the present C-H silylation method is
limited to terminal alkenes with electron-withdrawing
functional groups.
Figure 1. Plausible catalytic cycle (Ar = p-NO2-C6H4).
We36,37 and others9,43-46 have shown that hypervalent
iodine reagents easily oxidize Pd(II). The catalytic cycle is
proposed to start with oxidation of the Pd(II) catalyst
resulting in complex 7. After coordination of the alkene
substrate (1), complex 8 may undergo internal deprotona-
tion to form allylpalladium complex 9.
Although allylpalladium(IV) species are not commonly
invoked as catalytic intermediates, Canty and co-workers
have reported synthesis and characterization of such com-
plexes.47 Furthermore, the mechanistic studies reported by
Liu40 and by our group36 indicated that allylpalladium(IV)
intermediates may occur in palladium-catalyzed allylic
C-H functionalization reactions performed in the pre-
sence of hypervalent iodines. The deprotonation of 8 may
proceed via a so-called concerted metalation deprotona-
tion (CMD) mechanism, which was proposed by Fagnou
and co-workers for aromatic C-H activation reactions.48
According to the CMD mechanism the deprotonation of
the substrate occurs by an acyloxy ion (acetate, benzoate,
etc.).
It should be noted that the oxidative addition and the
allyl formation step could also happen in reverse order,
and the order of these two processes can also be sub-
strate dependent. Complex 9 may undergo two different
reactions. Transmetalation with disilane 2 would provide
10 and complex 11, which subsequently would undergo
reductive elimination, affording the allylsilane product (5).
Although the mechanistic details have yet to be eluci-
dated, the observed regioselectivity is consistent with an
allylpalladium mechanism. It is well documented that
palladium complexes with an unsymmetrically substituted
allyl moiety undergo attack at the least substituted termi-
nus, affording the linear allylic regioisomers.2,27,28 On the
basis of this and our previous experience with oxidative
C-H borylation37 and C-H acetoxylation36 reactions, we
suggest a plausible catalytic cycle in Figure 1.
(43) Lagunas, M. C.; Gossage, R. A.; Spek, A. L.; van Koten, G.
Organometallics 1998, 17, 731.
(44) Canty, A. J.; Rodemann, T.; Skelton, B. W.; White, A. H.
Organometallics 2006, 25, 3996.
(45) Racowski, J. M.; Dick, A. R.; Sanford, M. S. J. Am. Chem. Soc.
2009, 131, 10974.
(46) Powers, D. C.; Geibel, M. A. L.; Klein, J. E. M. N.; Ritter, T.
€
(38) Grennberg, H.; Simon, V.; Backvall, J.-E. J. Chem. Soc., Chem.
Commun. 1994, 265.
J. Am. Chem. Soc. 2009, 131, 17050.
(39) Popp, B. V.; Stahl, S. S. Top. Organomet. Chem. 2007, 22, 149.
(40) Yin, G.; Wu, Y.; Liu, G. J. Am. Chem. Soc. 2010, 132, 11978.
(47) Byers, P. K.; Canty, A. J. J. Chem. Soc., Chem. Commun. 1988,
639.
ꢀ
(41) Aydin, J.; Larsson, J. M.; Selander, N.; Szabo, K. J. Org. Lett.
2009, 11, 2852.
(42) Yao, Q.; Kinney, E. P.; Zheng, C. Org. Lett. 2004, 6, 2997.
(48) (a) Gorelsky, S. I.; Lapointe, D.; Fagnou, K. J. Am. Chem. Soc.
2008, 130, 10848. (b) Sun, H.-Y.; Gorelsky, S. I.; Stuart, D. R.; Campeau,
L.-C.; Fagnou, K. J. Org. Chem. 2010, 75, 8180.
1890
Org. Lett., Vol. 13, No. 7, 2011

prepared 12 via an alternative synthetic route and attempted
to perform the silylation using 12 instead of 1a (Scheme 2). In
this process we could not observe formation of allylsilane 5a
at all. Thus, even if acyloxylated byproducts are formed
under our reaction conditions (by reductive elimination
from 9) these types of compounds are not intermediates in
the present allylic C-H silylation reaction.
Scheme 2. The Reaction Cannot Be Carried out As a Tandem
Sequence via Intermediate Formation of Acyloxy Compound
12
In the case of using a mixture of 3b and 3c as oxidant the
reaction may follow the above Pd(II)/Pd(IV) pathway;
however, we cannot discount a classical Pd(0)/Pd(II) redox
cycle. Mechanistic studies including DFT modeling are
underway to explore the mechanistic aspects of this novel
allylic C-H silylation mechanism.
In summary, we have presented the first oxidative allylic
C-H silylation reaction. Our procedure is suitable for the
completely regioselective and highly stereoselective synth-
esis of functionalized allylsilanes from alkenes. Notable
features include the strongly oxidative conditions and that,
unusually, palladium catalyzes the C-H silylation process.2
Thus, our method constitutes an important new C-H
functionalization approach and complements existing meth-
ods for the synthesis of allylsilanes.
Alternatively, 9 may undergo reductive elimination pro-
viding an acyloxylated product (such as12, Scheme 2). Our
previous studies have shown that allyl-Pd(II) acetate com-
plexes are relatively stable to reductive elimination.28
However, after transmetalation with disilanes they under-
go very fast reductive elimination to give allylsilanes.28 We
believe that the same difference in reactivity applies for the
corresponding allyl-Pd(IV) complexes, such as 9. A possi-
ble reason for the formation of both cis and trans forms of
the allylsilane products 5 could be an η3-η1-η3 isomeriza-
tion of the allyl moiety in 9.2,49
An alternative to the above mechanism (Figure 1) could be
that palladium performs a tandem C-H acyloxylation36-
allylic substitution27,28 sequence in the transformation of the
alkenes to allylsilanes. We have also considered this possibi-
lity (Scheme 2) for transformation of 1a to 5a. Thus, we
attempted to perform C-H acyloxylation under our stan-
dard reaction conditions (entry 1) in the absence of 2.
However, this reaction did not give any acyloxy product
12. To double check this alternative mechanistic pathway, we
Acknowledgment. This work was supported by the
Swedish Research Council. We thank Dr. N. Selander,
Department of Organic Chemistry, Stockholm University,
for help with the synthesis of complex 4c.
Supporting Information Available. Detailed experimen-
tal procedures and characterization and 1H and 13C
NMR spectra of the products. This material is available
free of charge via the Internet at http://pubs.acs.org.
ꢀ
(49) Solin, N.; Szabo, K. J. Organometallics 2001, 20, 5464.
Org. Lett., Vol. 13, No. 7, 2011
1891