Kinetic resolution in palladium-catalyzed asymmetric allylic alkylations by a P,O ligand system.

Article abstract of DOI:10.1021/ol0161256

[reaction: see text] Through the investigation of peptide-based phosphine oxazoline ligands, a simple P,O ligand system was developed. This system provides palladium complexes that are capable of a very high degree of kinetic resolution of 1,3-diphenylpro

Full text of DOI:10.1021/ol0161256

ORGANIC
LETTERS
2001
Vol. 3, No. 14
2237-2240
Kinetic Resolution in
Palladium-Catalyzed Asymmetric Allylic
Alkylations by a P,O Ligand System
Scott R. Gilbertson* and Ping Lan
Department of Chemistry, Washington UniVersity, One Brookings DriVe,
Campus 1134, Saint Louis Missouri, 63130-4899
srg@wuchem.wustl.edu
Received May 14, 2001
ABSTRACT
Through the investigation of peptide-based phosphine oxazoline ligands, a simple P,O ligand system was developed. This system provides
palladium complexes that are capable of a very high degree of kinetic resolution of 1,3-diphenylprop-2-enyl acetate. The isolated palladium
complex was synthesized, characterized, and determined to be an effective procatalyst.
Over the course of the past few years, we have been involved
in the development of peptide-based phosphine ligands for
asymmetric catalysis.1-4 We have developed chemistry that
allows for the introduction of phosphines into a variety of
different peptide secondary structures. Additionally, we have
used parallel synthesis in the construction of these new
ligands.5-8 Recently, a number of workers have become
interested in the use of phosphine-oxazoline ligands in
asymmetric catalysis.9-22 In the course of screening a series
of peptide-based phosphine-oxazoline ligands, we observed
significant kinetic resolution of the starting material. This
Letter reports the development of a simple P,O ligand system,
the kinetic resolution of racemic starting allyl acetate, and
attempts to discuss some of the mechanistic implications of
our observations.
Initially we synthesized a series of phosphine-oxazoline
ligands that were designed to have a â-turn secondary
structure. These ligands were based on the observation that
(11) Hashimoto, Y.; Horie, Y.; Hayashi, M.; Saigo, K. Tetrahedron Asym.
2000, 11, 2205-2210.
(12) Loiseleur, O.; Hayashi, M.; Schmees, N.; Pfaltz, A. Synthesis 1997,
1338-1345.
(1) Gilbertson, S. R.; Chen, G.; McLoughlin, M. J. Am. Chem. Soc. 1994,
116, 4481-4482.
(2) Gilbertson, S. R.; Wang, X. J. Org. Chem. 1996, 61, 434-435.
(3) Gilbertson, S. R.; Pawlick, R. V. Angew. Chem., Int. Ed. Engl. 1996,
35, 902.
(13) Ogasawara, M.; Yoshida, K.; Hayashi, T. Heterocycles 2000, 52,
195-201.
(14) Porte, A. M.; Reibenspies, J.; Burgess, K. J. Am. Chem. Soc. 1998,
120, 9180-9187.
(4) Gilbertson, S. R.; Wang, X.; Hoge, G. S.; Klug, C. A.; Schaefer, J.
Organometallics 1996, 15, 4678-4680.
(15) Sagasser, I.; Helmchen, G. Tetrahedron Lett. 1998, 39, 261-264.
(16) Spinz, J.; Helmchen, G. Tetrahedron Lett. 1993, 34, 1769.
(17) Sprinz, J.; Kiefer, M.; Helmchen, G.; Reggelin, M.; Huttner, G.;
Walter, O.; Zsolnai, L. Tetrahedron Lett. 1994, 35, 1523-1526.
(18) Gilbertson, S. R.; Fu, Z. Org. Lett. 2001, 3, 161-164.
(19) Langer, T.; Janssen, J.; Helmchen, G. Tetrahedron: Asymmetry
1996, 7, 1599-1602.
(5) Gilbertson, S. R.; Wang, X. Tetrahedron Lett. 1996, 37, 6475-6478.
(6) Gilbertson, S. R.; Chen, G.; Kao, J.; Beatty, A.; Campana, C. F. J.
Org. Chem. 1997, 62, 5557-5566.
(7) Gilbertson, S. R.; Wang, X. Tetrahedron 1999, 11609-11619.
(8) Gilbertson, S. R.; Collibee, S. E.; Agarkov, A. J. Am. Chem. Soc.
2000, 122, 6522-6523.
(20) Newman, L. M.; Williams, M. J.; McCague, R.; Potter, G. A.
Tetrahedron: Asymmetry 1996, 7, 1597-1598.
(21) Gilbertson, S. R.; Xie, D. Angew. Chem. 1999, 38, 2750-2752.
(22) Gilbertson, S. R.; Fu, Z.; Xie, D. Tetrahedron Lett. 2001, 42, 365-
368.
(9) Dawson, G. J.; Frost, C. G.; Williams, J. M. J. Tetrahedron Lett.
1993, 34, 3149.
(10) Gilbertson, S. R.; Chang, C.-W. J. Org. Chem. 1998, 63, 8424-
8431.
10.1021/ol0161256 CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/14/2001

the proline D-amino acid sequence has a strong preference
for the formation of â-turn type structures.23-26 On one side
of the presumed turn-forming residues, we positioned a
phosphine-containing amino acid while on the other side we
placed an oxazoline (3-5, Scheme 1). Our initial results with
Table 1. Catalysis of Reaction 1 with Ligands 6-8
Scheme 1
1(S):1(R)a (convn)
ligand
solvent
time
3.5 h
S valueb
ee of 2, %c
6
6
toluene
THF
1:1 (26%)
1.4:1 (45%)
1.8
17 (S)
26 (S)
4 h
6
7
7
7
8
CH3CN
toluene
THF
4 h
1.4:1 (29%)
2.8
3.7:1 (52%)
5.6
19:1 (51%)
42.0
1.3:1 (50%)
1.4
53 (S)
64 (S)
86 (S)
60 (S)
44 (S)
2 h
40 min
<5 min
2 d
CH3CN
toluene
1.3:1 (25%)
2.6
8
8
THF
CH3CN
1 d
30 min
1:1 (11%)
6.2:1 (57%)
7.0
12 (S)
52 (S)
these ligands provided moderate selectivity, with ligand 5
providing a 40% ee.
a ee of recovered starting material was determined by HPLC on Chiralcel
OJ; solvent i-PrOH/hexane ) 1/99; flow rate 1 mL/min; retention time tR(1-
(S)) ) 27.4 min, tR(1-(R)) ) 32.9 min. b S ) kfast/kslow ) ln[(1 - C/100)(1
- ee/100)]/ln[1 - C/100)(1 + ee/100)] (C ) conversion; ee ) enantiomeric
excess of the recovered substrate. c Enantiomeric excesses were determined
by HPLC on Chiralpak AD; solvent i-PrOH/hexane ) 10/90; flow rate 1
mL/min; retention time, tR(2-(R)) ) 10.0 min, tR(2-(S)) ) 13.7 min.
In an attempt to develop more selective catalysts, we
decided to study the chemistry of this system. There were a
number issues that needed to be studied in order to improve
the selectivity of this system. The first question we had to
deal with was how the transition metal was chelated to the
ligand. While these ligands were designed to be bidentate
chelators, it was not a certainty that they were performing
in such a manner. Additionally, if bidentate, we needed to
determine that the palladium was bound to the phosphine
and the oxazoline nitrogen. We fully expected that the
phosphine would bind to the palladium, and this was verified
by 31P NMR. However, it was not clear that the oxazoline
moiety would effectively compete with the amides in the
peptide backbone. To test these issues, we synthesized a
series of ligands that did not contain an oxazoline group (6-
8, Table 1).
Surprisingly these ligands performed significantly better
than the original design, with the palladium complex of
ligand 7 providing the allylation product in up to 86% ee.
Just as significant, a large difference in the rate of reaction
for the two enantiomers of the starting material was observed.
The R enantiomer of 1 reacted significantly faster than the
S enantiomer. When the reaction run in THF was stopped at
51% conversion, the ratio of enantiomers of the starting
material was 19/1 (S value of 42).27-30 Through the years
there have been reports of kinetic resolution in the palladium-
catalyzed allylation.31-34 To the best of our knowledge, the
results we observe are among the highest S values observed
for acyclic allyl acetates.35
Ligand 8 was synthesized to test the importance of the
amide functionality. While catalysis with complexes of ligand
8 proceeded with some selectivity, the reaction rates were
significantly slower than with 7.
To help address the importance of the proline and the
phenyl group R to the phosphine, ligands 9, 10, and 11 were
synthesized and tested. In this case ligand 10 proved to give
the highest selectivity, for both product formation as well
as selectivity, between enantiomers of the starting material.
(29) Ruble, J. C.; Tweddell, J.; Fu, G. C. J. Org. Chem. 1998, 63, 2794-
2795.
(30) Ruble, J. C.; Latham, H. A.; Fu, G. C. J. Am. Chem. Soc. 1997,
119, 1492-1493.
(31) Hayashi, T.; Yamamoto, A.; Ito, Y. J. Chem. Soc., Chem. Commun
1986, 1090-1092.
(32) Bourghida, M.; Widhalm, M. Tetrahedron: Asymmetry 1998, 9,
1073-1083.
(23) Imperiali, B.; Kapoor, T. M. Tetrahedron 1993, 49, 3501-3510.
(24) Imperiali, B.; Fisher, S. L. J. Org. Chem. 1992, 57, 757.
(25) Imperiali, B.; Fisher, S. L.; Moats, R. A.; Prins, T. J. J. Am. Chem.
Soc. 1992, 114, 3182-3188.
(26) Wilmont, C. M.; Thornton, J. M. J. Mol. Biol. 1988, 203, 221.
(27) Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J. J. Am. Chem.
Soc. 1982, 104, 7294-7299.
(28) Martin, V. S.; Woodard, S. S.; Katsuki, T.; Yamada, Y.; Ikeda, M.;
Sharpless, K. B. J. Am. Chem. Soc. 1981, 103, 6237-6240.
(33) Dierkes, P.; Ramdeehul, S.; Barloy, L.; De Cian, A.; Fisher, J.;
Kamer, P. C. J.; van Leeuwen, P. W. N. M.; Osborn, J. A. Angew. Chem.,
Int. Ed. 1998, 37, 3116-3118.
(34) Ramdeehul, S.; Dierkes, P.; Aguado, R.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M.; Osborn, J. A. Angew. Chem., Int. Ed. 1998, 37,
3118-3121.
(35) Longmire, J. M.; Wang, B.; Zhang, X. Tetrahedron Lett. 2000, 41,
5435-5439.
2238
Org. Lett., Vol. 3, No. 14, 2001

It is interesting that significant selectivity is observed with
the very simple ligand 10.36 The results with ligand 11 give
an indication of the importance of the phenyl group next to
the phosphine. The only difference between ligands 7 and
11 is this phenyl group. In the case of 11, the selectivity is
low with no kinetic preference for either starting enantiomer.
This appears to indicate that the principal element of control
in this system is the chiral center next to the phosphine group.
On the basis of the observed selectivities in Table 2, it
the complex generated in situ from [Pd(η3-C3H5)Cl]2 and
ligand 13 (Table 3). Only the R acetate was recovered at
Table 3. Catalysis of 1 with Ligands 12-13
1(S):1(R)a (convn)
Table 2. Catalysis of Reaction 1 with Ligands 9-11a
ligand
solvent
toluene
time
1 h
S valueb
ee of 2, %c
68 (S)
12
2.4:1 (43%)
5.0
12
12
13
13
13
THF
3.5 h
3.8:1 (56%)
4.6
1.5:1 (42%)
2.1
1:6.6 (45%)
44.0
only (R) 1 (63%)
12
61 (S)
44 (S)
88 (R)
73 (R)
62 (R)
CH3CN
toluene
THF
5 min
1 h
1(S):1(R)b (convn)
ligand
solvent
toluene
time
4 h
S valuec
ee of 2, %d
36 (R)
0.5 h
<5 min
9
1:1.6 (37%)
2.8
1:5.5 (61%)
5.1
1:2.3 (40%)
5.5
1.9:1 (38%)
4.1
7.3:1 (58%)
7.6
7.3:1 (64%)
5.4
1:1 (48%)
1:1 (60%)
1:1 (56%)
CH3CN
1:2.8 (56%)
3.3
9
9
THF
1 d
7 h
36 (R)
60 (R)
72 (S)
64 (S)
61 (S)
a ee of recovered starting material was determined by HPLC on Chiralcel
OJ; solvent i-PrOH/hexane ) 1/99; flow rate 1 mL/min; retention time tR(1-
(S)) ) 27.4 min, tR(1-(R)) ) 32.9 min. b S ) kfast/kslow ) ln[(1 - C/100)(1
- ee/100)]/ln[1 - C/100)(1 + ee/100)] (C ) conversion; ee ) enantiomeric
excess of the recovered substrate. c Enantiomeric excesses were determined
by HPLC on Chiralpak AD; solvent i-PrOH/hexane ) 10/90; flow rate 1
mL/min; retention time, tR(2-(R)) ) 10.0 min, tR(2-(S)) ) 13.7 min.
CH3CN
toluene
THF
10
10
10
3.5 h
1.6 h
0.5 h
CH3CN
56% conversion, with the product obtained in 82% ee.
An NMR study of the complex indicates that the major
conformation of the molecule is 15a (Scheme 2). The ratio
11
11
11
toluene
THF
CH3CN
10 min
<5 min
<5 min
0
8 (S)
22 (R)
a All reactions were run at -25 °C. b ee of recovered starting material
was determined by HPLC on Chiralcel OJ; solvent i-PrOH/hexane ) 1/99;
flow rate 1 mL/min; retention time tR(1-(S)) ) 27.4 min, tR(1-(R)) ) 32.9
min. c S ) kfast/kslow ) ln[(1 - C/100)(1 - ee/100)]/ln[1 - C/100)(1 +
ee/100)] (C ) conversion; ee ) enantiomeric excess of the recovered
substrate. d Enantiomeric excesses were determined by HPLC on Chiralpak
AD; solvent i-PrOH/hexane ) 10/90; flow rate 1 mL/min; retention time,
tR(2-(R)) ) 10.0 min, tR(2-(S)) ) 13.7 min.
Scheme 2
appears that the preferred groups coordinating to the pal-
ladium are the phosphine and an amide or ester carbonyl.
With this in mind, we decided to synthesize a set of ligands
where the group next to the carbonyl ligand would be chiral
as well as rigid. Ligands 12 and 13 were synthesized. Ligand
13 gave selectivity comparable to that of ligand 7, with a
very large difference in the rate of reaction between the two
enantiomers of the starting material, when the reaction is
run in toluene solvent (S value ) 44).
In an effort to sort out the origin of the observed product
selectivity and the difference in the rate of reaction between
the two enantiomers of the starting material, the π-allyl
complex of ligand 13 and palladium was synthesized and
isolated (15). When this complex was employed as a
procatalyst in THF, the extent of kinetic resolution and the
ee of the product corresponded to the reaction catalyzed by
of isomers (15a:15b) as judged by 31P NMR is 3.9:1 at room
temperature. The NMR data support a P,O complex in which
the phosphine and a carbonyl oxygen are coordinated to the
palladium. This is significant since prior ligands of this type
have been used in a 4-fold excess relative to the metal and
have been proposed to be monodentate ligands.37
(36) Okauchi, T.; Fujita, K.; Ohtaguro, T.; Ohshima, S.; Minami, T.
Tetrahedron: Asymmetry 2000, 11, 1397-1403.
Org. Lett., Vol. 3, No. 14, 2001
2239

Stoichiometric addition of dimethylmalonate in THF gives
the product that is expected from addition trans to the
phosphine in complex 15a. However, this addition proceeded
with significantly lower selectivity than the catalytic reaction,
32% ee vs 82% ee.
Scheme 3
The observed results are difficult to explain by a single
mechanism. One theory would be that addition of the
nucleophile is fast relative to the formation of the allyl
complex. If this is so, then the ratio of the products is partially
explained by the difference in rates of the formation of the
two different allyl complexes. Thus, the difference in
selectivity between the catalytic reaction and the stoichio-
metric addition of a nucleophile to an allyl complex would
then be explained as the difference between adding to the
kinetic ratio of allyl complexes vs the thermodynamic ratio.
This explanation is complicated by the observation by
Helmchen that in his system the isomerization of the allyl
complex is 50 times faster than addition of the nucleophile.17
In this reaction there have been observations of “memory
effects”. One way to describe such observations is that there
are different ratios of rates between isomerization and
nucleophilic attack. In most cases the situation Helmchen
observed is operating. The allyl complex is undergoing
isomerization significantly faster than nucleophilic addition.
In select cases this does not appear to be the case. It is in
these cases that the stereochemistry of the starting material
has significant influence on the stereochemistry of the
product.
this preference is well-known, it is not necessarily true that
the addition is exclusively trans in P,O type ligands.
A number of workers have attempted to discuss the origin
of selectivity in this reaction. A variety of plausible explana-
tions have been forwarded. It appears clear that the selectivity
in this reaction is controlled by a balance of factors. Small
changes in the ligand system or reaction conditions can not
only decrease selectivity of the product but can also cause a
different mechanistic pathway to be responsible for the
selectivity that is observed. With the ligands reported here,
we generally see good selectivity for one enantiomer of the
product when we also have significant kinetic resolution.
Clearly this is not necessary with other ligands given that in
many cases the reaction using racemic substrate can be run
to completion with excellent product selectivity.
Additionally, we have observed that when matched and
mismatched sets of allyl acetate and ligand are reacted, the
S enantiomer is obtained in both cases (Scheme 3). There
appears to be a clear preference for formation of the S
enantiomer regardless of the difference in the rate of reaction
for the two enantiomeric allyl acetates. This is the type of
preference that operates in any of the many successful
asymmetric palladium-catalyzed allylations, given that it is
possible to obtain good selectivity and high yields from
racemic allyl acetates.
Acknowledgment. This work was partially supported by
NIH R01 GM56490-04 and Washington University. We also
gratefully acknowledge the Washington University High-
Resolution NMR Facility, partially supported by NIH
1S10R02004, and the Washington University Mass Spec-
trometry Resource Center, partially supported by NI-
HRR00954, for their assistance.
In the above discussion we assume that the nucleophile
will add exclusively trans to the phosphine ligand. While
Supporting Information Available: Experimental pro-
cedures and characterization of compounds. This material
is available free of charge via the Internet at http://pubs.acs.org.
(37) Okada, Y.; Minami, T.; Sasaki, Y.; Umezu, Y.; Yamaguchi, M.
Tetrahedron Lett 1990, 31, 3905-3908.
OL0161256
2240
Org. Lett., Vol. 3, No. 14, 2001