Use of Tricyclohexylphosphine to Control Regiochemistry in Palladium-catalyzed Allylic Alkylation

Article abstract of DOI:10.1021/ol9911268

(Matrix Presented) Tricyclohexylphosphine, Cy₃P, in conjuction with [(C₃H₅)PdCl]₂ catalyzes allylic alkylation of terminal allylic acetates with high regioselectivity toward branched products, which is in contra

Full text of DOI:10.1021/ol9911268

ORGANIC
LETTERS
1999
Vol. 1, No. 12
1969-1971
Use of Tricyclohexylphosphine To
Control Regiochemistry in
Palladium-Catalyzed Allylic Alkylation
A. John Blacker,† Matthew L. Clarke,‡ Michael S. Loft,§ and
,‡
Jonathan M. J. Williams*
Zeneca, Huddersfield Works, Leeds Road, Huddersfield, HD2 1FF, U.K., Department
of Chemistry, UniVersity of Bath, ClaVerton Down, Bath, BA2 7AY, U.K., and
Glaxo-Wellcome, DC1, Temple Hill, Dartford, DA1 5AH, U.K.
j.m.j.williams@bath.ac.uk
Received October 6, 1999
ABSTRACT
Tricyclohexylphosphine, Cy3P, in conjuction with [(C H )PdCl]2 catalyzes allylic alkylation of terminal allylic acetates with high regioselectivity
3
5
toward branched products, which is in contrast to most other palladium catalysts. Other ligands, even those of similar basicity or bulkiness,
did not show the same regioselectivity. Alkylation of 1,4-diacetoxy-2-butene gave predominantly branched products which had not been
observed previously.
Transition metal catalyzed allylic substitution is an important
process for organic synthesis.1 One of the most challenging
problems in allylic substitution reactions is to control the
regioselectivity when the reaction proceeds through an
unsymmetrical intermediate. For instance, when a mono-
substituted allylic acetate is used as substrate, palladium
catalysts generally give a mixture of isomers with a strong
tendency to form the linear product, although a few ligands
have been shown to reverse this regiochemistry for certain
substrates.2,3
proportion of branched products than palladium. We have
recently been interested in platinum-catalyzed allylic alky-
lation.11,12 One of the striking findings from our study was
the discovery that if the combination [(C3H5)PtCl]4/Cy3P was
used as catalyst, excellent regioselectivity toward branched
products was obtained in the alkylation of but-2-enyl (B/L
) 15:1) and hex-2-enyl (B/L ) 10:1) acetates. A Pd catalyst
(4) (a) Takeuchi, R.; Kashio, M. J. Am. Chem. Soc. 1998, 120, 8647.
(b) Janssen, J. P.; Helmchen, G. Tetrahedron Lett. 1997, 38, 8025. (c) B.
Bartels and G. Helmchen, J. Chem. Soc. Chem Commun. 1999, 741.
(5) Trost, B. M.; Hachiya, I. J. Am. Chem. Soc. 1998, 120, 1104. See
also: Glorius, F.; Pfaltz, A. Org. Lett. 1999, 1, 141.
(6) Zhang, S.; Mitsudo, T.; Kondo T.; Watanabe, Y. J. Organomet. Chem.
1993, 450, 197.
(7) Evans, P. A.; Nelson, J. D J. Am. Chem. Soc. 1998, 120, 5581.
(8) (a) Trost, B. M.; Tometzki, G. B.; Hung, M. H. J. Am. Chem. Soc.
1987, 109, 2176. (b) Lloyd-Jones, G. C.; Pfaltz, A. Angew. Chem., Int. Ed.
Engl. 1995, 34, 462 and references therein.
(9) Kurosawa, H. J. Chem. Soc., Dalton Trans. 1979, 939.
(10) Brown, J. M.; MacIntyre, J. E. J. Chem. Soc., Perkin Trans. 2 1985,
961.
Attempts to obtain branched products have focused on the
use of alternative metals as catalysts, and Ir,4 Mo,5 Ru,6 Rh,7
W,8 and Pt9-11 catalysts have all been shown to give a higher
† Zeneca.
‡ University of Bath.
§ Glaxo-Wellcome.
(1) For reviews, see: (a) Frost, C. G.; Howarth, J.; Williams, J. M. J.
Tetrahedron: Asymmetry 1992, 3, 1089. (b) Trost, B. M.; Van Vranken,
D. L. Chem. ReV. 1996, 96, 395.
(2) Hayashi, T.; Kawatsura, M.; Uozumi, Y. J. Chem. Soc., Chem.
Commun. 1997, 561.
(3) Pretot, R.; Pfaltz, A. Angew. Chem., Int. Ed. Engl. 1998, 37, 323.
See also: Trost, B. M.; Toste, F. D. J. Am. Chem. Soc, 1998, 120, 9074
and references therein.
(11) Blacker, A. J.; Clarke, M. L.; Humphries, M. E.; Loft, M. S.;
Williams, J. M. J. Chem. A. Eur. J. 1999, in press.
(12) Blacker, A. J.; Clarke, M. L.; Loft, M. S.; Williams, J. M. J.. J.
Chem. Soc., Chem. Commun. 1999, 913.
10.1021/ol9911268 CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/23/1999

to that of the Cy3P/Pd-catalyzed reactions. Tri-o-tolylphos-
phine is one of the most bulky phosphines (θ ) 194°, pKa
) 3.08) and also gives predominantly linear products, again
in contrast to the tricyclohexylphosphine system. Tris(4-
methoxyphenyl)- and tris(4-fluorophenyl)phosphines are
informative as they have a similar cone angle (θ ) 145°)
but different basicities (pKa’s ) 4.57 and 1.97, respectively.14
It can be concluded that there is no direct trend between
basicity or cone angle and the proportion of branched
products observed for this particular reaction. The formation
of predominantly linear products with a wide range of ligands
has been observed for other substrates by other workers.16
It was suggested to us that the results obtained may be
due to a strong “memory effect” 17 when Cy3P was used as
ligand (the product is the same regioisomer as the starting
material). To test this theory, the regioisomeric allylic
acetates (4 and 5) were tested (Scheme 2). The regioselec-
Table 1. Effect of Ligand on Regioselectivity of Allylic
Alkylation of But-2-enyl Acetate (1) with Sodium
Dimethylmalonate
liganda
2/3b
E/Z for 3c
Cy3Pd
11.5:1
15.4:1
1:2.0
2.1:1
1:3.9
1:1.5
1:1.2
1:1.3
1:1.1
1:1.4
N/D
N/D
N/D
4.6:1
5.5:1
5.5:1
5.7:1
5.1:1
4.7:1
6.4:1
f
Cy3P[(C3H5)PtCl]4 cat.d
Ph3Pe
f
f
Ph3P(Ph3P)2Pt-stilbene cat.d
[2,4,6-(CH3O)3C6H2]3P
tri-o-tolylphosphine
(p-CH3O-C6H4)3P
(p-FC6H4)3P
Cy2PCH2CH2PCy2
dppe
a All reaction run using 2.5 mol % of [(C3H5)PdCl]2 catalyst, 1.5 equiv
of NaCH(CO2Me)2 nucleophile, THF solvent at 20 °C unless stated. In all
cases conversion was 100%, as determined by GC. b Determined by GC
and checked by 1H NMR. c Determined by GC. d Taken from ref 11. e Taken
from ref 9. f N/D ) not determined.
Scheme 2. Tricyclohexylphosphine/Palladium Catalysts Give
Mainly Linear Products When Linear Allylic Acetates Are
Employed
in combination with this ligand was tested for comparison
and to our surprise also gave good regioselectivity.
As palladium is generally the preferred choice of metal
for allylic alkylation, we thought it important to study this
ligand effect further. In this Letter we show how we have
attempted to understand the nature of the ligand effect and
utilized it in the preparation of compounds in isomerically
pure form.
The first point we needed to address was whether any other
ligands would show this effect. Toward this end, ligand
effects on regiochemistry in the alkylation of but-2-enyl
acetate (1) with sodium dimethylmalonate were studied
(selected results are shown in Table 1 and Scheme 1).
tivity was altered (from 11.5:1 to 1.3:1 for 4 and from 3.1:1
to 1:8.8 for 5), and much greater proportions of linear
products (from linear starting material) were observed. This
is good evidence that a memory effect is associated with
this ligand.
We have also looked at the effect of Cy3P on Pd-catalyzed
allylic alkylation on commercially available cis-1,4-diac-
etoxy-2-butene (8). Allylic substitution on derivatives of
diacetate 8 have previously yielded linear isomers.18 When
compound 8 is alkylated with 1.02 equiv of sodium di-
methylmalonate using triphenylphosphine as ligand (Table
2), predominantly linear products are observed (in addition
to small amounts of byproducts).
Scheme 1
Switching the ligand to Cy3P gives a dramatic reversal of
regiochemistry (B/L ) 3.4:1). The branched isomer can be
isolated pure by column chromatography albeit in only 39%
yield due to the similar Rf of branched and linear products.
Because the starting material was linear, this result does not
fit in with the memory effect explanation.
Tricyclohexylphosphine is a bulky and strongly electron
donating phopshine. This is quantified by its large cone angle
(θ ) 170°)13 and the high pKa value of its conjugate acid
(pKa ) 9.70). This can be compared with PPh3 (θ ) 145°,
pKa ) 2.73).14 The ligands in Table 1 were chosen because
they share some characteristics with Cy3P and can therefore
be used to ascertain whether the regioselectivity observed is
a property of a general class of ligands. Tri (2,4,6-trimethoxy-
phenyl)phosphine15 is both more bulky and basic (θ ) 184°,
pKa ) 11.02) than Cy3P but gives regioselectivity opposite
When 11 is alkylated using palladium catalysts, a mixture
of inseparable isomers (12 and 13) is produced with a strong
(16) Akermark, B.; Zetterberg, K.; Hansson, S.; Krakenberger, B.;
Vitagliano, A. J. Organomet. Chem. 1987, 335, 133. Kranenburg, M.;
Kamer, P. C. J.; Van Leeuwen, P. W. N. M. Eur. J. Inorg. Chem. 1998,
25.
(13) Tolman, C. A. Chem. ReV. 1977, 77, 313.
(14) Rahman, M. M.; Liu, H.; Eriks, K.; Prock, A.; Giering, W. P.
Organometallics 1989, 8, 1.
(15) Wada, M.; Higashizaki, S. J. Chem. Soc., Chem. Commun. 1984,
482.
(17) Lloyd-Jones, G. C.; Stephen, S. C. Chem. Eur. J. 1998, 4, 2539.
Trost, B. M.; Bunt, R. C. J. Am. Chem. Soc. 1996, 118, 235.
(18) Tanigura, Y.; Nishimura, K.; Kawasaki, A.; Murahashi, S. Tetra-
hedron. Lett. 1982, 23, 5549. Genet, J. P.; Balabane, M.; Backvall, J. E.;
Nystrom, J. E. Tetrahedron Lett. 1983, 24, 2745.
1970
Org. Lett., Vol. 1, No. 12, 1999

suggest that the intermediate allyl complexes formed direct
the nucleophile to the allylic carbon atom which was
previously substituted by the acetate leaving group. While
we have no definitive explanantion for this, the work of
Mealli and Musco and their co-workers may be significant.21
They described the NMR spectroscopic properties of [(PPh3)2-
Pt(allyl)]X (X ) -BF4, -OAc) as being identical (at low
temperature where static η3-allyl complexes are present). For
the compounds [(Cy3P)2Pt(allyl)]X (X ) -BF4, -OAc), they
observed significantly different spectral properties, on chang-
ing the counterion. This was proposed to be evidence that
the acetate leaving group was still intimately involved with
the cationic palladium allyl complex. It seems feasable that
the reactions described here proceed through similar allyl
complexes in which the acetate ion is not a simple “spectator”
leaving group. The use of a chloride-free catalyst, [(Cy3P)2-
Pd(allyl)]BF4, affords results very similar to those identified
in Table 1, affording a 14:1 ratio of 2:3. Thus, the presence
of chloride is not solely responsible for the effects that we
have observed.
Table 2. Effect of Ligand on Regioselectivity in Allylic
Alkylation of Acetate 8
ligand
Ph3P
Cy3P
9/10
E/Z for 10
8/92
77/23
1.8:1
2.2:1
preference for 12 (for most ligands 12/13 ) ca. 10:1).11,19,20
When Cy3P is used as ligand, product 12 was obtained as a
single regioisomer (Scheme 3).
In conclusion, allylic alkylation using tricyclohexylphos-
phine as ligand could give insight on the mechanism of
palladium-catalyzed allylic substitution, which has recently
been shown to be more complex than first thought a few
years ago. Furthermore, the use of this cheap and com-
mercially available ligand may allow retention of the
regiochemistry of allylic acetates in the nucleophilic substitu-
tion products and should be of synthetic use.
Scheme 3
Acknowledgment. We are grateful to the LINK asym-
metric synthesis program, Glaxo-Wellcome, and Zeneca for
financial support. We also wish to thank Dr. Guy-Lloyd
Jones for a fruitful discussion.
Tricyclohexylphosphine, which was unique among the
ligands we screened, gives unusual regiochemistry in Pd-
catalyzed allylic alkylations that proceed through unsym-
metrical intermediates. Most of the results we have obtained
Supporting Information Available: Representative pro-
cedures and analytical data for compounds 2, 9, and 12. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(19) Hayashi, T.; Yamamoto, A.; Hagihara, T. J. Org. Chem. 1986, 51,
723.
(20) Togni, A. Tetrahedron: Asymmetry 1991, 2, 683.
(21) Carfagna, C.; Galarina, R.; Linn, K.; Lopez, J. A.; Mealli, C.; Musco,
A. Organometallics 1993, 12, 3019 and references therein.
OL9911268
Org. Lett., Vol. 1, No. 12, 1999
1971