Palladium-catalyzed enantioselective allylic alkylation of thiocarboxylate ions: Asymmetri synthesis of allylic thioesters and memory effect/dynamic kinetic resolution of allylic esters

Article abstract of DOI:10.1021/jo049756x

The palladium-catalyzed allylic alkylation of KSAc and KSBz with racemic cyclic and acyclic allylic esters by using N,N'-(1R,2R)-1,2-cyclohexandiylbis[2-(diphenylphosphino)-benzamide] as ligand frequently gave the corresponding allylic thioesters with high ee values and yields. The reaction of the cyclic allylic carbonates with KSAc in the presence of H20 was accompanied by a partial palladium-catalyzed enantioselective "hydrolysis" of the substrates with formation of the corresponding enantioenriched allylic alcohols. The degree of the "hydrolysis" was strongly dependent on the solvent and the thiocarboxylate ion. Highly selective kinetic resolutions (KRs) were observed in the palladium-catalyzed reaction of the racemic cyclohexenyl and cycloheptenyl acetates with KSAc. While the KR of the cyclohexenyl acetate is characterized by a selectivity factor S = 72 ± 19, that of the cycloheptenyl acetate afforded (R)-cycloheptenyl acetate of ≥ 99% ee in 48% yield and (S)-cycloheptenyl thioacetate of 98% ee in 50% yield. The palladium-catalyzed reaction of the racemic cyclopentenyl acetate with KSAc showed a strong "memory effect" (ME), that is, both enantiomers reacted with different enantioselectivities. The ME was probed by studying the palladium-catalyzed reactions of both the matched acetate of ≥ 99% ee and the mismatched acetate of ≥ 99% ee with KSAc. The acetates not only reacted with different enantioselectivities and rates but also suffered an unexpected and concomitant palladium-catalyzed racemization in the presence of the chiral ligand. This led in the case of the mismatched acetate to a temporary dynamic kinetic resolution (DKR) that featured a racemization of the mismatched acetate by the chiral catalyst. Studies of the palladium-catalyzed reaction of the racemic cyclopentenyl acetate, carbonate, and naphthoate with KSAc in the presence of the chiral ligand also showed the ME to be strongly dependent on the nucleofuge. This also allowed the synthesis of (S)-cyclopentenyl thioacetate of 92% ee in high yield from the racemic cyclopentenyl naphthoate.

Full text of DOI:10.1021/jo049756x

P a lla d iu m -Ca ta lyzed En a n tioselective Allylic Alk yla tion of
Th ioca r boxyla te Ion s: Asym m etr ic Syn th esis of Allylic Th ioester s
a n d Mem or y Effect/Dyn a m ic Kin etic Resolu tion of Allylic Ester s
Bernhard J . Lu¨ssem and Hans-J oachim Gais*
Institut fu¨r Organische Chemie der Rheinisch-Westfa¨lischen Technischen Hochschule (RWTH) Aachen,
Prof.-Pirlet-Str. 1, 52056 Aachen, Germany
gais@rwth-aachen.de
Received February 11, 2004
The palladium-catalyzed allylic alkylation of KSAc and KSBz with racemic cyclic and acyclic allylic
esters by using N,N′-(1R,2R)-1,2-cyclohexandiylbis[2-(diphenylphosphino)-benzamide] as ligand
frequently gave the corresponding allylic thioesters with high ee values and yields. The reaction of
the cyclic allylic carbonates with KSAc in the presence of H2O was accompanied by a partial
palladium-catalyzed enantioselective “hydrolysis” of the substrates with formation of the corre-
sponding enantioenriched allylic alcohols. The degree of the “hydrolysis” was strongly dependent
on the solvent and the thiocarboxylate ion. Highly selective kinetic resolutions (KRs) were observed
in the palladium-catalyzed reaction of the racemic cyclohexenyl and cycloheptenyl acetates with
KSAc. While the KR of the cyclohexenyl acetate is characterized by a selectivity factor S )
72 ( 19, that of the cycloheptenyl acetate afforded (R)-cycloheptenyl acetate of g99% ee in 48%
yield and (S)-cycloheptenyl thioacetate of 98% ee in 50% yield. The palladium-catalyzed reaction
of the racemic cyclopentenyl acetate with KSAc showed a strong “memory effect” (ME), that is,
both enantiomers reacted with different enantioselectivities. The ME was probed by studying the
palladium-catalyzed reactions of both the matched acetate of g99% ee and the mismatched acetate
of g99% ee with KSAc. The acetates not only reacted with different enantioselectivities and rates
but also suffered an unexpected and concomitant palladium-catalyzed racemization in the presence
of the chiral ligand. This led in the case of the mismatched acetate to a temporary dynamic kinetic
resolution (DKR) that featured a racemization of the mismatched acetate by the chiral catalyst.
Studies of the palladium-catalyzed reaction of the racemic cyclopentenyl acetate, carbonate, and
naphthoate with KSAc in the presence of the chiral ligand also showed the ME to be strongly
dependent on the nucleofuge. This also allowed the synthesis of (S)-cyclopentenyl thioacetate of
92% ee in high yield from the racemic cyclopentenyl naphthoate.
In tr od u ction
synthetically important allylic sulfur compounds.8 It was
found that both the alkylation of sulfinate ions and
2-pyrimidinethiol (external nucleophiles) with racemic
allylic esters1a,b,d,e,2 and the mechanistically related O,S-
rearrangement of racemic allylic O-allylic sulfinates6c and
thiocarbamates6a,b (internal nucleophiles) (Scheme 1)
provide excellent methods for the asymmetric synthesis
of allylic sulfur compounds including sulfones, carbam-
ates, and pyrimidyl-substituted sulfides.
The palladium(0)-catalyzed enantioselective allylic
alkylation of sulfur nucleophiles1-3a and the allylic O,S-
rearrangement4-6 have been intensively studied in recent
years.7 The main incentive for these studies was the
development of a catalytic asymmetric synthesis of
* To whom correspondence should be addressed. Fax: (+49) 241-
8092665.
(1) (a) Eichelmann, H.; Gais, H.-J . Tetrahedron: Asymmetry 1995,
6, 643. (b) Gais, H.-J .; Eichelmann, H.; Spalthoff, N.; Gerhards, F.;
Frank, M.; Raabe, G. Tetrahedron: Asymmetry 1998, 9, 235. (c) Frank,
M.; Gais, H.-J . Tetrahedron: Asymmetry 1998, 9, 3353. (d) Gais, H.-
J .; Spalthoff, N.; J agusch, T.; Frank, M.; Raabe, G. Tetrahedron Lett.
2000, 41, 3809. (e) Gais, H.-J .; J agusch, T.; Spalthoff, N.; Gerhards,
F.; Frank, M.; Raabe, G. Chem. Eur. J . 2003, 9, 4202.
(2) (a) Trost, B. M.; Organ, M. G.; O’Doherty, G. A. J . Am. Chem.
Soc. 1995, 117, 9662. (b) Trost, M.; Krische, M. J .; Radinov, R.; Zanoni,
G. J . Am. Chem. Soc. 1996, 118, 6297. (c) Trost, B. M.; Krueger, A. C.;
Bunt, R. C.; Zambrano, J . J . Am. Chem. Soc. 1996, 118, 6520. (d) Trost,
B. M.; Radinov R. J . Am. Chem. Soc. 1997, 119, 5962.
(3) (a) Sinou, D.; Divekar, S.; Safi, M.; Soufiaoui, M. Sulfur Lett.
1999, 22, 125. (b) Divekar, S.; Safi, M.; Soufiaou, M.; Sinou, D.
Tetrahedron 1999, 55, 4369.
(6) (a) Bo¨hme, A.; Gais, H.-J . Tetrahedron: Asymmetry 1999, 10,
2511. (b) Gais, H.-J .; Bo¨hme, A. J . Org. Chem. 2002, 67, 1153. (c)
J agusch, T.; Gais, H.-J .; Bondarev, O. J . Org. Chem. 2004, 69, 2731.
(7) For recent reviews on the palladium(0)-catalyzed enantioselective
allylic alkylation and 1,3-rearrangement, see: (a) Moreno-Man˜as, M.;
Pleixats, R. In Handbook of Organopalladium Chemistry for Organic
Synthesis; Negishi, E., de Meijere, A., Eds.; Wiley: New York, 2002;
Vol. 2; p 1707. (b) Acemoglu, L.; Williams, J . M. J . In Handbook of
Organopalladium Chemistry for Organic Synthesis; Negishi, E., de
Meijere, A., Eds.; Wiley: New York, 2002; Vol. II, p 1945. (c) Koc´ovsky´,
P.; Stary´, I. In Handbook of Organopalladium Chemistry for Organic
Synthesis; Negishi, E., de Meijere, A., Eds.; Wiley: New York, 2002;
Vol. II, p 2011. (d) Beletskaya, I. P.; Cheprakov, A. V. In Handbook of
Organopalladium Chemistry for Organic Synthesis; Negishi, E., de
Meijere, A., Eds.; Wiley: New York, 2002; Vol. II, p 2957.
(4) Hiroi, K.; Makino, K. Chem. Pharm. Bull. 1988, 36, 1744.
(5) Valk, J . M.; Claridge, T. D. W.; Brown, J . M. Tetrahedron:
Asymmetry 1995, 6, 2597.
10.1021/jo049756x CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/07/2004
J . Org. Chem. 2004, 69, 4041-4052
4041

Lu¨ssem and Gais
SCHEME 2. P a lla d iu m -Ca ta lyzed Asym m etr ic
SCHEME 1. Sim p lified Sch em e of
P a lla d iu m -Ca ta lyzed En a n tioselective Allylic
Syn th esis of Allylic Th ioester s
Alk yla tion , KR a n d ME of Ra cem ic Su bstr a tesa
a
L* ) chiral ligand. This scheme also applies to acyclic
substrates having a symmetrically substituted carbon skeleton.
stabilized carbanions11m has been hampered by the lack
of general methods for their asymmetric synthesis.6a,b,8
Unfortunately, the palladium-catalyzed enantioselective
allylic alkylation of thiols is limited in scope because of
its apparent confinement to aryl and heteroaryl thiols.1b,c,e
We have therefore developed an interest in the pal-
ladium-catalyzed enantioselective allylic alkylation of
thiocarboxylate ions as a surrogate for that of thiols.
Allylic thioesters should make highly useful starting
materials for the synthesis of allylic sulfides because of
their ready hydrolysis to the corresponding allylic thiols12
and the facile alkylation,12 arylation,6b,c,13 heteroaryla-
tion,6b,c,12 and further derivatization of the latter at the
S-atom.14 The literature on the feasibility of a palladium-
The use of the bisphosphane 47d,h,9 (Scheme 2) as ligand
for the palladium atom conveyed high enantioselectivities
not only to the second step of the reaction sequence, the
allylic alkylation of the sulfur nucleophile, but also to the
first step, the ionization of the racemic substrate.1b-e,10
For example, a highly efficient kinetic resolution (KR) of
racemic cyclic allylic carbonates was achieved through a
palladium-catalyzed reaction with lithium tert-butylsul-
finate and ligand 4.1e It was now of interest to see
whether the palladium-catalyzed allylic alkylation is also
applicable to the asymmetric synthesis of a further class
of synthetically valuable allylic sulfur compounds, the
allylic sulfides.8,11 A greater exploitation of the potential
of allylic sulfides for stereoselective synthesis including
the allylic alkylation of organometallics,11b,e sigmatropic
rearrangement11a,c,d,f-l and generation of chiral sulfur-
(11) For pertinent examples, see: (a) Hartley, R. C.; Warren, S.;
Richards, I. C. J . Chem. Soc., Perkin Trans. 1 1994, 507. (b) Calo, V.;
Fiandanese, V.; Nacci, A.; Scilimati, A. Tetrahderon 1994, 50, 7283.
(c) Schnur, R. C.; Corman, M. L. J . Org. Chem. 1994, 59, 2581. (d)
Ernst, B.; Gonda, J .; J eschke, R.; Nubbemeyer, U.; Oehrlein, R.; Bellus,
D. Helv. Chim. Acta 1997, 80, 876. (e) Dieter, R. K.; Sadanandan, N.;
Lois, E. Synlett 1997, 1114. (f) Fukuda, T.; Irie, R.; Katsuki, T.
Tetrahedron 1999, 55, 649. (g) Carter, D. S.; Van Vranken, D. L.
Tetrahedron Lett. 1999, 40, 1617. (h) Aggarwal, V. K.; Ferrara, M.;
Hainz, R.; Spey, S. E. Tetrahedron Lett. 1999, 40, 8923. (i) McMillen,
D. W.; Varga, N.; Reed, B. A.; King, C. J . Org. Chem. 2000, 65, 2532.
(j) He, S.; Kozmin, R. A.; Rawal, V. H. J . Am. Chem. Soc. 2000, 122,
190. (k) Shinada, T.; Fuji, T.; Ohtani, Y.; Yoshida, Y.; Ohfune, Y. Synlett
2002, 1341. (l) Zhang, X.; Qu, Z.; Ma, Z.; Shi, W.; J in, X.; Wang, J . J .
Org. Chem. 2002, 67, 5621. (m) Marr, F.; Fro¨hlich, R.; Wibbeling, B.;
Diedrich, C.; Hoppe, D. Eur. J . Org. Chem. 2002, 2970.
(12) (a) Gundermann, K.-D.; Hu¨mke, K. In Organische Schwefel-
verbindungen (Houben-Weyl); Klamann, D., Ed., Thieme:, Stuttgart,
1985; Vol. E11, p 158. (b) Page, P. C. B.; Wilkes, R. D.; Reynolds, D.
In Comprehensive Organic Functional Group Transformations; Ka-
tritzky, A. R., Meth-Cohn, O., Rees, C. W., Ley, S. V., Eds.; Pergamon:
Oxford, 1995; Vol. 2, p 113.
(13) (a) Harayama, H.; Nagahama, T.; Kozera, T.; Kimura, M.;
Fugami, K.; Tanaka, V.; Tamaru, Y. Bull. Chem. Soc. J pn. 1997, 70,
445. (b) Mann, G.; Baranano, D.; Hartwig, J . F.; Rheingold, A. L.;
Guzei, I. A. J . Am. Chem. Soc. 1998, 120, 9205. (c) Zheng, N.;
McWilliams, J . C.; Fleitz, F. J .; Armstrong, J . D., III; Volante, R. P. J .
Org. Chem. 1998, 63, 9606.
(8) For reviews, see: (a) Ogura, K. In Comprehensive Organic
Synthesis; Schreiber, S. L., Ed.; Pergamon: Oxford, 1991; Vol. 1, p 505.
(b) Krief, A. In Comprehensive Organic Synthesis; Pattenden, G., Ed.;
Pergamon: Oxford, 1991; Vol. 3, p 85. (c) Yamamoto, Y. In Compre-
hensive Organic Synthesis; Heathcock, C. H., Ed.; Pergamon: Oxford,
1991; Vol. 3, p 55. (d) Simpkins, N. S. Sulphones in Organic Synthesis;
Pergamon: Oxford, 1993. (e) Braun, M. In Stereoselective Synthesis
(Houben-Weyl); Helmchen, G., Hoffmann, R. W., Mulzer, J ., Schau-
mann, E., Eds.; Thieme: Stuttgart, 1995; Vol. E21b, p 1713. (f) Pyne,
S. In Stereoselective Synthesis (Houben-Weyl); Helmchen, G., Hoffmann,
R. W., Mulzer, J ., Schaumann, E., Eds.; Thieme: Stuttgart, 1995; Vol.
E21b, p 2068. (g) Mikołjczk, M.; Dabrowwicz, J .; Kiełasin´ski, P. Chiral
Sulfur Reagents; CRC Press: Boca Raton, 1997. (h) Katritzky, A. R.;
Piffl, M.; Lang, H.; Anders, E. Chem. Rev. 1999, 99, 665. (i) Reggelin,
M.; Zur, C. Synthesis 2000, 1.
(9) (a) Trost, B. M.; Lee, C. In Catalytic Asymmetric Synthesis;
Ojima, I., Ed.; Wiley-VCH: New York, 2000; p 593. (b) Trost, B. M.;
Crawley, M. L. Chem. Rev. 2003, 103, 2921.
(10) For KRs with 4 and other nucleophiles, see: (a) Lloyd-J ones,
G. C.; Stephen, S. C. Chem. Commun. 1998, 2321. (b) Trost, B. M.;
Hembre, E. J . Tetrahedron Lett. 1999, 40, 219. (c) Trost, B. M.; Toste,
F. D. J . Am. Chem. Soc. 1999, 121, 3543. (d) Dominguez, B.; Hodnett,
N. S.; Lloyd-J ones, G. C. Angew. Chem. 2001, 113, 4419; Angew. Chem.,
Int. Ed. 2001, 40, 4289. (e) Trost, B. M.; Dudash, J ., J r.; Hembre, E. J .
Chem. Eur. J . 2001, 7, 1619.
4042 J . Org. Chem., Vol. 69, No. 12, 2004

Allylic Alkylation of Thiocarboxylate Ions
TABLE 1. P a lla d iu m -Ca ta lyzed Asym m etr ic Syn th esis
catalyzed enantioselective allylic alkylation of thiocar-
boxylate ions was, however, ambiguous. Although it had
been previously described that the thioacetate ion is
capable to serve as a nucleophile in palladium-catalyzed
allylic alkylation in the presence of achiral ligands,3b its
alkylation with racemic cyclohexenyl carbonate (vide
infra) in the presence of various chiral phosphanes
including 4 proceeded with only medium yields and
enantioselectivities.3a
of Allylic Th ioester s in CH2Cl2/H2O a t 25 °C
en-
convn
[%]a
t
yield
ee
try substrate
salt
[h]
productsb [%] [%]c,d
1
2
3
4
5
6
7
8
rac-1-OMoc KSAc g99 144
(R)-1-SAc 76 84 (92)
(91:9)
(R)-1-SBz 85 87 (92)
(91:9)
(R)-2-SBz 97 90 (49)
(96:4)
(S)-5-SAc 62 73
rac-1-OMoc KSBz
92
96
28e
24
24
rac-2-OMoc KSBz g99
rac-5-OMoc KSAc g99
rac-5-OMoc KSBz g99
rac-6-OMoc KSAc g99
rac-6-OMoc KSBz g99
Beside the development of an asymmetric synthesis of
allylic thioesters, we were also seeking with this study
further information about an interesting phenomenon of
the palladium-catalyzed allylic alkylation, the “memory
effect” (ME) of racemic allylic substrates.15 According to
Scheme 1 both enantiomers of the allylic substrate should
give the same π-allyl-palladium(II) complex or the same
set of equilibrating complexes and thereby, under the
influence of the chiral ligand, the alkylation product with
the same ee value. It has been reported, however, that
the enantiomers of certain allylic esters can give under
certain conditions the substitution product with different
ee values.1e,15-19 Although the ME is mechanistically
highly interesting, it is synthetically undesirable because
it can diminish the efficiency of the palladium-catalyzed
asymmetric synthesis. Most of the information about the
ME has been gathered with ligand 4.1e,16-19 Several
factors have been identified that influence the ME,
including the structure of the carbon skeleton of the
substrate, the nucleofuge, the nucleophile, the catalyst
loading, and the precatalyst.1e,15-19 However, despite
several investigations, the origin of the ME is still a
matter of debate.15b,c Recently a two catalytic cycle model,
based on the formation of monomeric and oligomeric
palladium(0)/4 complexes, has been proposed in order to
account for the ME with this ligand.17 Because 4 is one
of the most successful and widely used ligands in pal-
ladium-catalyzed enantioselective allylic alkylation,7b,9
the collection of further experimental information on the
ME with this ligand is desirable.
(S)-5-OH
(S)-5-SBz 68 59
(S)-5-OH 20 55
5.25 (S)-6-SAc 51 94
(S)-6-OH 30 82
(S)-6-SBz 92 89
(S)-6-OH
(S)-7-SBz 69 86
(S)-7-OH
22 57
24
48
5
f
rac-7-OMoc KSBz
73
g
a
b
Determined by GC. The values in parentheses give the E/Z
ratios. c Determined by GC (thioacetates and alcohols) and HPLC
d
(thiobenzoates) on chiral stationary phase columns. The values
in parentheses give the ee values of the Z isomers. e At reflux
temperature. f Not determined. g Formation of the alcohol was not
detected.
esters as substrates, new and interesting information
about the ME and the dynamic kinetic resolution (DKR).
Resu lts a n d Discu ssion
I. Allylic Alk yla tion of Th ioca r boxyla te Ion s. (i)
Asym m etr ic Syn th esis of Allylic Th ioester s. The
study of the palladium-catalyzed allylic alkylation of
thiocarboxylate ions was carried out with potassium
thioacetate (3a ) (KSAc) and potassium thiobenzoate (3b)
(KSBz) and the racemic cyclic and acyclic carbonates rac-
1-OMoc (Moc ) methoxycarbonyl), rac-2-OMoc, rac-5-
OMoc, rac-6-OMoc, rac-7-OMoc, and rac-8-OMoc, re-
spectively (Scheme 2). The carbonates were chosen
because of their symmetrically substituted carbon skel-
eton, which precludes the formation of regioisomers.
Furthermore, these carbonates had been previously used
in palladium-catalyzed allylic alkylation of other sulfur
nucleophiles in our1b-e and other laboratories.2,3 Thus, a
direct comparison of the reactivity of the vaious sulfur
nucleophiles in palladium-catalyzed allylic alkylation
with ligand 4 would be possible. Carbonates were selected
instead of acetates because of their higher reactivity in
palladium-catalyzed allylic alkylation.7a The bisphos-
phane 4 was used as ligand for the palladium atom
because of the high enantioselectivities it provided in the
allylic alkylation of a number of nucleophiles including
those based on sulfur1,2,6 with a broad range of sym-
metrically substituted cyclic and acyclic allylic esters.7b,9
The carbonates rac-1-OMoc, rac-2-OMoc, rac-5-OMoc,
rac-6-OMoc, and rac-7-OMoc were treated with 1.4
equiv of KSAc or 2.0 equiv of KSBz in the presence of 2
mol % of Pd2(dba)3‚CHCl3 (dba ) dibenzylidene acetone)
and 8 mol % of 4 in CH2Cl2/H2O (9:1).
In this paper we describe the results of a study of the
palladium-catalyzed enantioselective allylic alkylation of
thiocarboxylate ions with racemic allylic esters in the
presence of ligand 4. This investigation not only resulted
in the development of an asymmetric synthesis of allylic
thioesters and widened the scope of the palladium-
catalyzed KR but also gave, by using enantiopure allylic
(14) Although allylic S-thiocarbamates make excellent starting
materials for the synthesis of allylic thiols, their asymmetric synthesis
through a palladium-catalyzed O,S-rearrangement of racemic allylic
O-thiocarbamates is confined to substrates of a sufficient thermal
stablity.6a,b For example, racemic cyclopentenyl O-thiocarbamates
already suffer at room temperature a noncatalyzed rearrangement with
formation of the corresponding racemic cyclopentenyl S-thiocarbam-
ates.
(15) (a) Fiaud, J . C.; Malleron, J . L. Tetrahedron Lett. 1981, 22, 1399.
(b) Lloyd-J ones, G. C.; Stephen, S. C.; Murray, M.; Butts, C. P.;
Vyskocˇil, S.; Kocˇovsky´, P. Chem. Eur. J . 2000, 6, 4348 and references
therein. (c) Fairlamb, I. J . S.; Lloyd-J ones, G. C.; Vyskocˇil, S.; Kocˇovsky´,
P. Chem. Eur. J . 2002, 8, 443 and references therein.
(16) Lloyd-J ones, G. C.; Stephen, S. C. Chem. Eur. J . 1998, 4, 2539.
(17) Lloyd-J ones, G. C.; Stephen, S. C.; Fairlamb, I. J . S.; Martorell,
A.; Dominguez, B.; Tomlin, P. M.; Murray, M.; Fernandez, J . M.;
J effery, J . C.; Riis-J ohannessen, T.; Guerziz, T. Pure Appl. Chem. 2004,
76, 589.
(18) (a) Trost, B. M.; Bunt, R. C. J . Am. Chem. Soc. 1996, 118, 235.
(b) Trost, B. M.; Surivet, J .-P. J . Am. Chem. Soc. 2000, 122, 6291.
(19) Lu¨ssem, B. J .; Gais, H.-J . J . Am. Chem. Soc. 2003, 125, 6066.
The acyclic carbonates rac-1-OMoc and rac-2-OMoc
gave under these conditions (Table 1, entries 1-3) the
thioesters (R)-1-SAc, (R)-1-SBz, and (R)-2-SBz, respec-
tively, with high enantioselectivities in high yields.
Surprisingly, the formation of the (E)-configured allylic
thioesters was accompanied by that of the corresponding
J . Org. Chem, Vol. 69, No. 12, 2004 4043

Lu¨ssem and Gais
Z isomers to a degree of 4-9%. Although the E and Z
isomers were not separated, an unequivocal assignment
of their structures was made possible by 1H and 13C NMR
spectroscopy of the mixture of the isomers. Instrumental
was the observation of similar chemical shifts for the
S-atom bearing C-atoms of the isomers, which allowed
the exclusion of the structures of the isomeric O-allylic
thioesters. Large 13C NMR shift differences are normally
observed for the heteroatom bearing C-atoms of S- and
O-allylic esters, as exemplified by thiobenzoic acid O-
(1-methyl-but-2-enyl) ester (δ ) 72.0 ppm) and the
isomeric thiobenzoic acid S-(1-methyl-but-2-enyl) ester
(δ ) 40.7 ppm).20 We had previously observed formation
of the Z isomer in the palladium-catalyzed reaction of
rac-2-OMoc with 2-pyrimidinethiol and 4 in CH2Cl2 in
the absence of water.1c,e Thus, formation of Z isomers in
palladium-catalyzed allylic alkylation with ligand 4
seems to be more common than previously anticipated.21
This observation may lead to a further refinement of the
mechanistic picture of the palladium-catalyzed allylic
alkylation with 4. Formation of the Z isomers can be
rationalized by proposing the existence of an equilibrium
between the corresponding acyclic (syn,syn)-, (syn,anti)-,
(anti,syn)-, and (anti,anti)-configured π-allyl-palladium-
(II)/4 complexes, which are interconverting by an anion-
promoted π-π-π-isomerization mechanism.22 While the
substitution of the (anti,anti)-configured complex should
give the (Z,S)-configured thioester, that of the (anti,syn)-
configured complex at the (syn)-configured allyl terminus
should deliver the (Z,R)-configured thioester. However,
the absolute configurations of the Z isomers were not
determined, and a prediction as to the reactivity of both
complexes is difficult to make. We note, however, that
with the dimethylallyl substrate rac-1-OMoc the enan-
tioselectivity of the formation of the Z isomer is higher
than with the diethylallyl substrate rac-2-OMoc whose
allyl-palladium complex carries the larger substituents
at the allyl termini (cf. Table 1, entries 1-3).
Treatment of the cyclopentenyl carbonate rac-5-OMoc
with KSAc (entry 4) and KSBz (entry 5) furnished
thioacetate (S)-5-SAc with 73% ee and thiobenzoate (S)-
5-SBz with 59% ee, respectively, in only medium yields.
A similar treatment of the cyclohexenyl carbonate rac-
6-OMoc with KSAc gave thioacetate (S)-6-SAc in an
even lower yield of only 51% but with an higher ee value
of 94% (entry 6). The somewhat low yield of (S)-5-SAc,
(S)-5-SBz, and (S)-6-SAc was due to a competing pal-
ladium-catalyzed “hydrolysis” of the allylic carbonates
with formation of the corresponding allylic alcohols (vide
infra). However, the use of KSBz instead of KSAc in the
reaction of the cyclohexenyl carbonate rac-6-OMoc helped
to alleviate this problem and afforded thioester (S)-6-SBz
with 89% ee in high yield (entry 7). The cycloheptenyl
carbonate rac-7-OMoc showed in the reaction with KSBz,
in comparison to rac-5-OMoc and rac-6-OMoc, a lower
reactivity and gave thioester (S)-7-SBz with 86% ee in
good yield (entry 8).
OMoc, rac-2-OMoc, rac-5-OMoc, rac-6-OMoc, and rac-
7-OMoc in THF/H2O (9:1) (see Supporting Information,
Table S1) led to similar results in regard to yields and
enantioselectivities. The enantioselectivities were, how-
ever, in the single-phase system THF/H2O somewhat
lower than in the two-phase system CH2Cl2/H2O.
In summary, the palladium-catalyzed allylic alkylation
of thiocarboxylate ions with racemic allylic carbonates
allowes an asymmetric synthesis of highly enantioen-
riched acyclic and cyclic allylic thioesters in good to high
yields, the only exemption being the five-membered cyclic
carbonate rac-5-OMoc, which gave thioesters (S)-5-SAc
and (S)-5-SBz only with medium ee values and yields.
However, as will be shown later, thioacetate (S)-5-SAc
can be also obtained with a high ee value in high yield
by using a different racemic cyclopentenyl ester as
substrate.
(ii) Com p etin g “Hyd r olysis” of Allylic Ca r bon -
a tes. Surprisingly, the reactions of the cyclopentenyl and
cyclohexenyl carbonates rac-5-OMoc and rac-6-OMoc,
respectively, with KSAc and KSBz in CH2Cl2/H2O and
THF/H2O were accompanied to various degrees by a
yield-diminishing formation of the allylic alcohols (S)-
5-OH and (S)-6-OH, respectively (cf. Scheme 2, Table 1,
entries 4-7). The alcohols were nonracemic and pos-
sessed the same absolute configuration as the corre-
sponding thioesters. The degree of the alcohol formation
was strongly dependent on the solvent system. Whereas
in the two-phase system CH2Cl2/H2O the alcohols were
formed to a degree of 20-30% (cf. Table 1, entries 4-6),
alcohol formation in the single-phase system THF/H2O
was only negligible (see Supporting Information, Table
S1). Alcohol formation was also dependent on the thio-
carboxylate ion. Although in the reaction of the carbonate
rac-6-OMe with KSBz in CH2Cl2/H2O formation of the
alcohol (S)-6-OH was hardly noticeable, 30% of the
alcohol was formed in the reaction with the less reactive
KSAc. It had been previously found that the treatment
of the carbonates rac-1-OMoc, rac-2-OMoc, rac-5-OMoc,
rac-6-OMoc, rac-7-OMoc, and rac-8-OMoc with Pd2-
(dba)3‚CHCl3 and 4 in the presence of H2O and in the
absence of an external nucleophile leads to a highly
enantioselective and almost quantitative formation of the
corresponding allylic alcohols.19 We have presented evi-
dence showing that the palladium-catalyzed “hydrolysis”
of allylic carbonates is most likely caused by the series
of events depicted in Scheme 3 for rac-5-OMoc and rac-
6-OMoc. Ionization of the racemic allylic carbonates
upon reaction with the palladium(0)/4 catalyst generated
the corresponding π-allyl-palladium(II)/4 complexes and
the methyl carbonate ion. Because of the presence of
water the methyl carbonate ion was hydrolyzed to the
hydrogen carbonate ion and MeOH. The hydrogen car-
bonate ion acted as an oxygen nucleophile and substi-
tuted the π-allyl-palladium(II)/4 complexes with forma-
tion of the hydrogen carbonates (S)-5-OHyc (Hyc )
hydrogen carbonate) and (S)-6-OHyc, the hydrolysis of
which with water gave the allylic alcohols (S)-5-OH and
(S)-6-OH, respectively, and CO2. The formation of the
corresponding allylic alcohols in the reactions of rac-5-
OMoc and rac-6-OMoc with thiocarboxylate ions in the
presence of H2O shows that the hydrogen carbonate ion
can effectively compete with the sulfur nucleophile in the
substitution of the π-allyl-palladium(II)/4 complexes.
A similar study of the palladium-catalyzed allylic
alkylation of KSAc and KSBz with carbonates rac-1-
(20) Lu¨ssem, B. J . Master’s Thesis, RWTH Aachen, 1999.
(21) Trost, B. M.; Toste, F. D. J . Am. Chem. Soc. 1999, 121, 4545.
(22) (a) Hayashi, T.; Yamamoto, A.; Hagihara, T. J . Org. Chem.
1986, 51, 723. (b) Sjogren, M. P. T.; Hansson, S.; A° kermark, B.;
Vitagliano, A. Organometallics 1994, 13, 1963.
4044 J . Org. Chem., Vol. 69, No. 12, 2004

Allylic Alkylation of Thiocarboxylate Ions
TABLE 2. P a lla d iu m -Ca ta lyzed KR of Allylic Ca r bon a tes a n d Aceta tes w ith Th ioca r boxyla te Ion s a t 25 °C
entry
1
substrate
t [h]
salt
solvent
convn [%]a
productsb
yield [%]
ee [%]c,d
rac-8-OMoc
48c
KSAc
THF/H2O
53
(R)-8-OMoc
(S)-8-SAc
(R)-8-OMoc
(S)-8-SBz
(R)-2-SBz (97:3)
(S)-2-OMoc
(R)-1-SAc (91:9)
(S)-1-OAc
(R)-1-SBz (85:15)
(S)-1-OAc
(S)-5-SAc
(R)-5-OAc
(S)-5-SAc
(S)-6-SAc
(R)-6-OAc
(S)-7-SAc
(R)-7-OAc
48e
42e
28
56
44
47
32
44
53
22
52
40
99
48
43
50
48
72
84
94
73
93 (63)
57
86 (99)
60
83 (95)
99
92
62
58
97
g99
98
g99
2
3
4
5
6
rac-8-OMoc
rac-2-OMoc
rac-1-OAc
rac-1-OAc
rac-5-OAc
48f
3.5g
72
KSBz
KSBz
KSAc
KSBz
KSAc
THF/H2O
67
48
35
56
62
CH2Cl2/H2O
CH2Cl2/H2O
CH2Cl2/H2O
CH2Cl2/H2O
23
1.7
7
8
rac-5-OAc
rac-6-OAc
22
12
KSAc
KSAc
CH2Cl2/H2O
CH2Cl2/H2O
g99
51
9
rac-7-OAc
43
KSAc
CH2Cl2/H2O
51
a
b
Determined by GC. The values in parentheses give the E/Z ratios. c Determined by GC on chiral stationary phase containing column.
The values in parentheses give the ee values of the Z isomers. e Chemical yield. f At 45 °C. At reflux temperature.
d
g
SCHEME 3. Mech a n istic Sch em e for Com p etin g
P a lla d iu m -Ca ta lyzed “Hyd r olysis” of Allylic
Ca r bon a tes in Th eir Rea ction s w ith
ion and the reaction of the allyl-palladium complex with
the sulfur nucleophile dominates.
(iii) Kin etic Resolu tion of Allylic Ester s. In the
reaction of the cyclooctenyl carbonate rac-8-OMoc1e with
KSAc in THF/H2O (cf. Scheme 2) the conversion of the
substrate even at 45 °C did not exceed 53% and a mixture
of thioacetate (S)-8-SAc and carbonate (R)-8-OAc in a
ratio of 53:47 was isolated. Formation of thioacetate (S)-
8-SAc of 84% ee and of carbonate (R)-8-OMoc of 72% ee
in a ratio of approximately 1:1 (Table 2, entry 1) showed
that an efficient KR had occurred (cf. Scheme 1). Similar
results were recorded in the reaction of carbonate rac-
8-OMoc with KSBz (entry 2). The results recorded after
the termination of the reaction of the acyclic carbonate
rac-2-OMoc with KSBz in CH2Cl2/H2O at 48% conversion
also revealed the operation of a KR in this case (entry
3). Because of the strong current interest in both the
catalytic KR23,24 and the palladium-catalyzed KR in
particular,1b-e,10,19,25 the KRs of the racemic allylic ac-
etates rac-1-OAc, rac-5-OAc, rac-6-OAc, and rac-7-OAc
with thiocarboxylate ions and 4 were investigated in more
detail (Scheme 4). The acetates were selected instead of
the corresponding carbonates in order to avoid the
competing formation of the corresponding allylic alcohols
(vide supra). All reactions were carried out in CH2Cl2/
H2O (9:1) by using 2 mol % of Pd2(dba)3‚CHCl3 and 8 mol
% of 4.
Th ioca r boxyla te Ion s in th e P r esen ce of Wa ter
The significantly higher degree of alcohol formation in
CH2Cl2/H2O as compared to THF/H2O can perhaps be
rationalized as follows. In the two-phase system CH2Cl2/
H2O the reactions will predominately take place in the
organic phase and the solubilities of KSAc and KSBz in
the organic phase are most likely only low. The allyl-
palladium complex and its hydrogen carbonate counter-
ion are contained as a contact ion pair in the organic
phase and the thiocarboxylate ion can enter the organic
phase via an anion exchange of the ion pair. Thus, the
concentration of the thiocarboxylate ion in the organic
phase is low and perhaps comparable to that of the
hydrogen carbonate ion. Therefore the latter can ef-
fectively compete in the substitution of the π-allyl-
palladium(II)/4 complex. In the single-phase system THF/
H2O, however, the concentration of the thiocarboxylate
ion is much higher than that of the hydrogen carbonate
Termination of the reaction of the pentenyl acetate rac-
1-OAc with KSAc at 35% conversion showed the opera-
tion of a highly selective KR (entry 4). However, a 50%
conversion of the substrate could be achieved neither at
25 °C nor at reflux temperature. This is in contrast to
the reactivity of carbonate rac-1-OMoc (cf. Table 1, entry
(23) For reviews, see: (a) Ward, R. S. Tetrahedron: Asymmetry
1995, 6, 1475. (b) Hoveyda, A. H.; Didiuk, M. T. Curr. Org. Chem. 1998,
2, 489. (c) Cook, G. R. Curr. Org. Chem. 2000, 4, 869. (d) Keith, J . M.;
Larrow, J . F.; J acobsen, E. N. Adv. Synth. Catal. 2001, 343, 5. (e)
Huerta, F. F.; Mindis, A. B. E.; Ba¨ckvall, J .-E. Chem. Soc. Rev. 2001,
30, 321. (f) Kim, M.-J .; Ahn, Y.; Park, J . Curr. Opin. Biotechnol. 2002,
13, 578. (g) Robinson, D. E. J . E.; Bull, S. D. Tetrahedron: Asymmetry
2003, 14, 1407.
(24) For reviews, see: (a) Gais, H.-J .; Theil, F. In Enzyme Catalysis
in Organic Synthesis; Drauz, K., Waldmann, H., Eds.; Wiley-VCH:
Weinheim, 2002; Vol. II, p 335. (b) Williams, J . M. J .; Parker, R. J .;
Neri, C. In Enzyme Catalysis in Organic Synthesis; Drauz, K.;
Waldmann, H., Eds.; Wiley-VCH: Weinheim, 2002; Vol. I, p 287.
J . Org. Chem, Vol. 69, No. 12, 2004 4045

Lu¨ssem and Gais
SCHEME 4. P a lla d iu m -Ca ta lyzed KR of Allylic
Aceta tes w ith KSAc (3a ) a n d KSBz (3b)
F IGURE 1. Dependencies of the ee values of thioacetate (S)-
6-SAc and acetate (R)-6-OAc on the conversion of the sub-
strate in the reaction of rac-6-OAc with KSAc.
A very high enantiomeric selectivity was exhibited by
the catalyst in the reaction of the cyclohexenyl acetate
rac-6-OAc with KSAc on a 2.5 mmol scale. This led to
the isolation of the thioacetate (S)-6-SAc of 97% ee in
48% yield and the acetate (R)-6-OAc26c of g99% ee in 43%
yield (entry 7). The reaction came practically to a
complete halt after 51% conversion of the substrate. A
further conversion of 2% took place only within a period
of 24 h, and even an increase of the amount of the
precatalyst Pd2(dba)3‚CHCl3 from 2 to 5 mol % and that
of ligand 4 from 8 to 20 mol % saw no conversion of the
remaining acetate (R)-6-OAc. To determine the selectiv-
ity factor S,27 the KR of rac-6-OAc was repeated and the
ee values of the acetate and thioacetate were monitored
over the whole course of the reaction.
1) and perhaps reflects the lower reactivity of allylic
acetates in palladium-catalyzed alkylation.7a This prob-
lem could be overcome, however, by the use of the more
reactive KSBz instead of KSAc. Here, at 56% conversion
of rac-1-OAc the thiobenzoate (R)-1-SBz of 83% ee was
isolated in 53% yield (entry 5) and the acetate (S)-1-
OAc26a of 99% ee was recovered in 22% yield. The low
yield of (S)-1-OAc was due to considerable losses en-
countered during isolation of the acetate because of its
volatility.
Surprising results were obtained in the case of the KR
of the cyclic acetates rac-5-OAc, rac-6-OAc, and rac-7-
OAc. The cyclopentenyl acetate exhibited by far the
highest reactivity of all three acetates. Thioacetate (S)-
5-SAc of 92% ee was obtained in 52% yield at 62%
conversion of the substrate (entry 6). Strangely, however,
the ee value of the recovered acetate (R)-5-OAc26b was
only 62% (vide infra). Furthermore, the thioacetate (S)-
5-SAc, which was isolated in 99% yield, had an ee value
of only 58% ee at the full conversion of the substrate
(entry 7) (vide infra).
Figure 128 shows the dependencies of the ee values of
thioacetate (S)-6-SAc and acetate (R)-6-OAc on the
conversion of the substrate. In accordance with the
mechanistic scheme depicted in Scheme 1, the ee value
of the thioacetate (S)-6-SAc was high and remained
practically constant over the whole course of the reaction,
whereas that of the acetate (R)-6-OAc increased steadily.
Nonlinear regression of the selectivity factor S29a for 10
pairs of ee values and conversions gave S ) 72 ( 19.29b
This value corresponds well with S ) 74 ( 7 for the
palladium(0)/4-catalyzed KR of rac-6-OMoc with lithium
tert-butylsulfinate in CH2Cl2/H2O and with S ) 77 ( 11
for that of rac-6-OMoc with 2-pyrimidinethiol in CH2-
Cl2.1e These results demonstrate that the efficiency of the
palladium-catalyzed KR with 4 is independent of the
nucleofuge and the nucleophile. Almost the same results
in terms of selectivities and yields were obtained when
the KR of rac-6-OAc with KSAc was carried out on a 10
mmol scale.
(25) (a) Hayashi, T.; Yamamoto, A.; Yoshihiko, I. J . Chem. Soc.
Chem. Commun. 1986, 1090. (b) Bourghida, M.; Widhalm, M. Tetra-
hedron: Asymmetry 1998, 9, 1073. (c) Brunner, H.; Deml, I.; Dirn-
berger, W.; Ittner, K.-P.; Reisser, W.; Zimmermann, M. Eur. J . Inorg.
Chem. 1999, 51. (d) Ramdeehul, S.; Dierkes, P.; Aguado, R.; Kamer,
P. C. J .; van Leeuwen, P. W. N. M.; Osborn, J . A. Angew. Chem. 1998,
110, 3302; Angew. Chem., Int. Ed. 1998, 37, 3118. (e) Nishimata, T.;
Yamaguchi, K.; Mori, M. Tetrahedron Lett. 1999, 40, 5713. (f) Reetz,
M. T.; Sostmann, S. J . Organomet. Chem. 2000, 603, 105. (g) Longmire,
J . M.; Wang, B.; Zhang, X. Tetrahedron Lett. 2000, 41, 5435. (h)
Okauchi, T.; Fujita, K.; Ohtagura, T.; Ohshima, S.; Minami, T.
Tetrahedron: Asymmetry 2000, 11, 1397. (i) Gilbertson, S. R.; Lan, P.
Org. Lett. 2001, 3, 2237.
(26) (a) McKew, J . C.; Kurth, M. J . J . Org. Chem. 1993, 58, 4589.
(b) Asami, M. Bull. Chem. Soc. J pn. 1990, 63, 721. (c) Hashimoto, T.;
Fukazawa, T.; Shimoji, Y. Tetrahedron: Asymmetry 1996, 6, 1649. (d)
Ito, S.; Kasai, M.; Ziffer, H.; Silverton, J . V. Can. J . Chem. 1987, 65,
574. (e) Wahab, A.; Tavares, D. F.; Rauk, A. Can. J . Chem. 1990, 68,
1559.
A similar efficient KR took place in the reaction of the
cycloheptenyl acetate rac-7-OAc with KSAc. The reaction
came practically to a complete halt at 51% conversion of
(27) Kagan, H. B.; Fiaud, J . C. Top. Stereochem. 1988, 18, 249.
(28) The lines connecting the points of measurement (b, 9, 2) are
only meant to show their togetherness and do not represent further
points of measurement.
(29) (a) Dominguez, B.; Hodnett, N. S.; Lloyd-J ones, G. C. Angew.
Chem. 2001, 113, 4419; Angew. Chem., Int. Ed. 2001, 40, 4289. (b)
The program Origin 6.1 was used.
4046 J . Org. Chem., Vol. 69, No. 12, 2004

Allylic Alkylation of Thiocarboxylate Ions
the substrate. The acetate (R)-7-OAc26d of g99% ee was
obtained in 48% yield and the thioacetate (S)-7-SAc of
98% ee was isolated in 50% yield (entry 9). Even an
increase of the amount of Pd2(dba)3‚CHCl3 from 2 to 4
mol % and that of 4 from 8 to 16 mol % did not lead to a
conversion of the remaining acetate (R)-7-OAc.
The results of the KRs described above show, in
agreement with previous results,1e,10,19 that with ligand
4 the fast-reacting (matched) enantiomer of the acyclic
acetates has the (R)- and that of the slow-reacting
(mismatched) enantiomer the (S)-configuration. In the
case of the cyclic acetates the enantiomer preference is
the opposite.
II. ”Mem or y Effect” a n d Dyn a m ic Kin etic Resolu -
tion of Cyclop en ten yl Aceta te. (i) Rea ction of th e
Ra cem ic Su bstr a te. A comparison of the enantioselec-
tivities of the KRs and substitutions of the acetates rac-
5-OAc, rac-6-OAc, and rac-7-OAc with KSAc (cf. Table
2, entries 6-9; Scheme 1) reveals, beside common
features, interesting differences. While in all three cases
the enantioselectivities of the allylic alkylation of the
nucleophile with the matched acetates are apparently
high, the selectivities of the KRs of the three racemates
differ significantly. The selectivity of the KR is high for
the six- and seven-membered cyclic acetates but is
apparently low for the five-membered cyclic acetate.
Furthermore, the enantioselectivity of the allylic alky-
lation of KSAc with the five-membered acetate rac-5-OAc
under complete turnover conditions is much lower than
under incomplete turnover conditions, for example at 62%
conversion. A similar low enantioselectivity was observed
in the case of the alkylation of thiocarboxylate ions with
the carbonate rac-5-OMoc under complete turnover
conditions (cf. Table 1, entry 5). This stands in contrast
to the reactivity of the six- and seven-membered carbon-
ates rac-6-OMoc and rac-7-OMoc, respectively, which
delivered under high turnover conditions the correspond-
ing thioesters with high ee values (cf. Table 1, entries 7
and 8). Although the six- and seven-membered acetates
rac-6-OAc and rac-7-OAc were not studied under high
turnover conditions, they are expected to show a reactiv-
ity similar to that of the carbonates. These differences
in reactivity of rac-5-OAc, rac-6-OAc, and rac-7-OAc
point to (1) the operation of a strong Me in the case of
rac-5-OAc,1e,16-19 the alkylation with the matched acetate
(S)-5-OAc proceeding with a high enantioselectivity and
that with the mismatched acetate (R)-5-OAc with a low
enantioselectivity and (2) a racemization of the mis-
matched acetate (S)-5-OAc during the course of the
substitution reaction. To verify these assumptions, the
reaction of the racemic acetate rac-5-OAc with KSAc was
repeated and the ee values of (S)-5-SAc and (R)-5-OAc
were monitored over almost the whole course of the
reaction (Figure 2 and Supporting Information, Table S2).
Figure 228 shows the dependencies of the ee values of (S)-
5-SAc and (R)-5-OAc on the conversion of the substrate.
The ee value of (S)-5-SAc remained high at 96% until
approximately 50% conversion. An increase of the con-
version from 50% up to g99% led to a strong decrease of
the ee value of the thioacetate. These results support the
assumption of the operation of a strong ME. A control
experiment demonstrated that a possible partial racemi-
zation of (S)-5-SAc under high turnover conditions can-
F IGURE 2. Dependencies of the ee values of the thioacetate
(S)-5-SAc and the mismatched acetate (R)-5-OAc on the
conversion of the substrate in the reaction of rac-5-OAc with
KSAc.
SCHEME 5. Con figu r a tion a l Sta bility of
Th ioa ceta te (S)-5-SAc d u r in g P a lla d iu m -Ca ta lyzed
Allylic Alk yla tion
not be held responsible for the decrease of the ee value
of the thioacetate.
Treatment of thioacetate (S)-5-SAc of 92% ee with
KSAc in CH2Cl2/H2O (9:1) in the presence of 2 mol % of
Pd2(dba)3‚CHCl3 and 8 mol % of 4 for 2 d and a
subsequent workup led to an almost quantitative recov-
ery of (S)-5-SAc that still had an ee value of 87% (Scheme
5). Thus, only a minor racemization of the thioacetate
occurred within this period of time. Interestingly, the
omission of the nucleophile KSAc (condition of no turn-
over) in the above experiment did also not lead to a
significant racemization of (S)-5-SAc. Treatment of (S)-
5-SAc of 59% ee with 2 mol % of Pd2(dba)3‚CHCl3 and 8
mol % of 4 in CH2Cl2 for 1 d (3 d) and a subsequent
workup afforded (S)-5-SAc with 56% ee (52% ee).
(ii) Rea ction s of th e En a n tiop u r e Su bstr a tes. To
gain more insight into the ME and the origin of the
apparently low selectivity of the KR of rac-5-OAc, a
determination of the stereochemical course of the reac-
tions of both the matched acetate (S)-5-OAc and the
mismatched acetate (R)-5-OAc, in separate experiments,
by using the enantiopure acetates was desirable. Al-
though the ME of the cyclopentenyl acetate with 4 had
been previously studied with NaCH(CO2R)2 by employing
(S)-5-OAc of 55% ee,17a the low enantiomeric enrichment
J . Org. Chem, Vol. 69, No. 12, 2004 4047

Lu¨ssem and Gais
SCHEME 6. Syn th esis of En a n tiop u r e
Cyclop en ten yl Aceta tes (R)-5-OAc a n d (S)-5-OAc
F IGURE 3. Dependencies of the ee values of the thioacetate
(S)-5-SAc and the matched acetate (S)-5-OAc on the conver-
sion of the substrate in the reaction of (S)-5-OAc with KSAc.
SCHEME 7. ME, Ra cem iza tion , a n d DKR of
Aceta tes (S)-5-OAc a n d (R)-5-OAc in Th eir
P a lla d iu m -Ca ta lyzed Rea ction s w ith KSAc
of the acetate precluded the observation of more subtle
effects including a possible racemization of the remaining
substrate.30
Preparative HPLC of the racemic cyclopentenyl naph-
thoate rac-5-ONa p h on a Chiralcel-OD column on a 6-g
scale readily afforded the naphthoates (S)-5-ONa p h 31
and (R)-5-ONa p h each of g99% ee and in 46% yield
(Scheme 6). Hydrolysis of (R)-5-ONa p h and (S)-5-
ONa p h furnished the enantiopure alcohols (R)-5-OH26e
and (S)-5-OH, respectively. Their acylation gave the
corresponding enantiopure acetates (R)-5-OAc26b and (S)-
5-OAc.
Both acetates (S)-5-OAc and (R)-5-OAc were sepa-
rately submitted to a palladium-catalyzed reaction with
1.4 equiv of KSAc in the presence of 2 mol % of Pd2(dba)3‚
CHCl3 and 8 mol % of 4 in CH2Cl2/H2O (9:1) (Scheme 7).
The ee values of the thioacetate and the remaining
acetate were monitored over the whole range of the
reactions. The reaction of the matched acetate (S)-5-OAc
was faster than that of the mismatched acetate (R)-5-
OAc.
was in excess and its ee value increased with increasing
conversion of the substrate to reach g99% at 88%
conversion. The ee value of the thioacetate (S)-5-SAc
remained high up to approximately 75% conversion but
decreased at higher conversion. These results show that
at the beginning of the reaction of (S)-5-OAc with KSAc
a partial racemization of the matched acetate occurred.
The matched acetate (S)-5-OAc reacted preferentially
and delivered the thioacetate (S)-5-SAc with high enan-
tioselectivity up to 80% conversion. At the same time the
concentration of the mismatched acetate (R)-5-OAc
gradually increased. After the consumption of most of the
matched acetate (S)-5-OAc, the ee value of the thioac-
etate (S)-5-SAc decreased presumably because of the
onset of a substitution of the mismatched acetate (R)-5-
OAc, which proceeded with a lower enantioselectivity.
Whether at high conversion, where GC analysis only
showed the presence of the mismatched acetate (R)-5-
OAc, also a partial racemization of (R)-5-OAc occurred
is difficult to decide because of the possibility of a fast
consumption of the matched acetate (S)-5-OAc.
Figure 328 shows the dependencies of the ee values of
the matched acetate (S)-5-OAc and the thioacetate (S)-
5-SAc on the conversion of the substrate (Supporting
Information, Table S3).
The ee value of acetate (S)-5-OAc, which was g98%
at the beginning, decreased with increasing conversion;
in other words the mismatched acetate (R)-5-OAc was
formed. At approximately 50% conversion only the pres-
ence of the racemic acetate rac-5-OAc was detected.
From that point on the mismatched acetate (R)-5-OAc
Having recorded these unexpected results in the reac-
tion of the matched acetate (S)-5-OAc with KSAc, it was
of interest to see whether the mismatched acetate (R)-
5-OAc would exhibit a behavior similar to that of the
(30) For an in-depth study of the ME of cyclopentenyl esters by using
isotopically labeled racemic substrates, see ref 16.
(31) Kawasaki, K.; Katsuki, T. Tetrahedron 1997, 53, 6337.
4048 J . Org. Chem., Vol. 69, No. 12, 2004

Allylic Alkylation of Thiocarboxylate Ions
SCHEME 8. P a lla d iu m -Ca ta lyzed Ra cem iza tion of
Aceta te (S)-5-OAc
It was now of interest to see whether a palladium-
catalyzed racemization of the cyclopentenyl acetate would
also occur in the absence of the nucleophile, the condition
of no turnover.10a Thus, the enantiopure acetate (S)-5-
OAc was treated with 2 mol % of Pd2(dba)3‚CHCl3 and 8
mol % of 4 in CH2Cl2/H2O under the same conditions used
above but by omitting KSAc (Scheme 8). After a short
reaction time the acetate (S)-5-OAc had been completely
racemized and rac-5-OAc was isolated in almost quan-
titative yield. The racemization of (S)-5-OAc and (R)-5-
OAc without nucleophile perhaps proceeds through a
Pd(0)/4-catalyzed ionization depicted in Scheme 1 fol-
lowed by an addition of the acetate ion to the palladium
atom and a subsequent reductive elimination.10a,33 The
racemization of the acetates in the presence of KSAc, the
condition of complete turnover, is perhaps more complex,
as hinted by the dependencies of the ee values of the
substrates on their conversion (cf. Figures 3 and 4).
In summary, the results of the reactions of the matched
acetate (S)-5-OAc and the mismatched acetate (R)-5-OAc
with KSAc confirm the operation of a ME. The matched
acetate reacts with an enantioselectivity significantly
higher than that of the mismatched acetate. This effect
is, however, superimposed by a temporary concomitant
racemization and DRK of the acetates. As a consequence
of these effects both acetates deliver the thioacetate (S)-
5-SAc up to a relatively high conversion with high
enantioselectivities. However, the DKR of the mis-
matched acetate (R)-5-OAc is only a temporary one, and
thus the ee value of the thioacetate decreases strongly
at a high conversion of the substrate because of the
preferential reaction of the mismatched acetate. In the
case of the matched acetate (S)-5-OAc this effect also
operates, although to a much lesser degree. A rational-
ization of the various aspects of the palladium(0)/4-
catalyzed reactions of the cyclopentenyl esters including
the ME, which have been described previously1e,16-19 and
in this Article, is difficult at present. A major obstacle to
the development of a mechanistic scheme is the formation
of not one but several monomeric, oligomeric, and dia-
stereomeric Pd(0)/4 as well as π-allyl-Pd(II)/4 complexes
upon reaction of the racemic allylic ester with the
precatalyst and the bisphosphane 4.17,34 These complexes,
which are in rapid equilibrium, are most likely endowed
with not only a different reactivity but also selectivity.
Recently a two catalytic cycle working model has been
proposed in order to rationalize the ME. It features a
F IGURE 4. Dependencies of the ee values of the thioacetate
(S)-5-SAc and the mismatched acetate (R)-5-OAc on the
conversion of the substrate in the reaction of (R)-5-OAc with
KSAc.
matched acetate or, according to Figure 2, show a ME
and react with low enantioselectivity.
Figure 428 shows for the reaction of the mismatched
acetate (R)-5-OAc with KSAc complex similar depend-
encies of the ee values of the components on the conver-
sion of the substrate as in the case of the matched
acetate. From the onset of the reaction the ee value of
the mismatched acetate (R)-5-OAc, which was g99%,
decreased; in other words, the matched acetate (S)-5-OAc
was formed. The ee value of (R)-5-OAc remained almost
constant between approximately 18% and 45% conversion
and increased again strongly at higher conversion.
Surprisingly, the ee value of the thioacetate (S)-5-SAc
was much higher than anticipated and even increased
with increasing conversion. Finally, the ee value of the
thioacetate decreased at higher conversion where only
the presence of (R)-5-OAc was detected by GC analysis.
These results suggest that at the beginning of the
reaction a partial racemization of the mismatched acetate
(R)-5-OAc occurred with formation of the matched ac-
etate (S)-5-OAc, which reacted with high enantioselec-
tivity. The observation of almost constant ee values of
the acetate (R)-5-OAc and the thioacetate (S)-5-SAc
between 18% and 45% conversion of the substrate
strongly indicates that at this point of the reaction a
dynamic kinetic resolution (DKR)23e-g was established,
which involved a racemization of the mismatched acetate
and a preferentially substitution of the matched acetate,
both in the presence of the chiral catalyst. Surprisingly,
however, the DKR apparently ceased at a conversion
higher than 45%. From that point on the ee value of (R)-
5-OAc increased, and as a consequence of the lower
enantioselectivity of the reaction of the mismatched
acetate (R)-5-OAc, the ee value of the thioacetate (S)-5-
OAc strongly decreased. Examples of this type of DKR
that involves a racemization of the substrate under full
turnover condition are scarce in a palladium-catalyzed
allylic alkylation.7 The DKR is different from the dynamic
kinetic asymmetric transformation of an racemic allylic
ester in palladium-catalyzed allylic alkylation, which
features an equilibration of diastereomeric π-allyl-pal-
ladium(II) complexes.32
(32) Trost, B. M.; Toste, F. D. J . Am. Chem. Soc. 2003, 125, 3090.
(33) Grennber, H.; Langer, V.; Ba¨ckvall, J .-E. J . Chem. Soc., Chem.
Commun. 1991, 1190.
(34) (a) Butts, C. P.; Crosby, J .; Lloyd-J ones, G. C.; Stephen, S. C.
Chem. Commun. 1999, 1707. (b) Fairlamb, I. J . S.; Lloyd-J ones, G. C.
Chem. Commun. 2000, 2447.
J . Org. Chem, Vol. 69, No. 12, 2004 4049

Lu¨ssem and Gais
SCHEME 10. Assign m en ts of Absolu te
SCHEME 9. Asym m etr ic Syn th esis of
Cyclop en ten yl Th ioa ceta te (S)-5-SAc
Con figu r a tion s of Th ioester s (R)-1-SAc, (S)-6-SAc,
a n d (S)-6-SBz
highly selective monomeric cyclic for the matched enan-
tiomer and a less enantioselective oligomeric cycle for the
mismatched enantiomer.17 In accordance with this model
would be the previous observation that the ME increases
with increasing catalyst loading.34b,18b This effect is
particularly expressed in the case of cyclopentenyl
esters.18b For example, although at a catalyst loading of
2 mol % of Pd2(dba)3‚CHCl3 and 8 mol % of 4 the six-
and seven-membered cyclic allylic esters showed no ME
in the reactions with thiocarboxylate ions, the five-
membered cyclic esters exhibited a strong ME (cf. Tables
1 and 2). We have now also observed a strong ME in the
reaction of the cyclohexenyl carbonate rac-6-OMoc with
KSAc in THF/H2O. An increase of the catalyst loading
from 2 to 5 mol % of Pd2(dba)3‚CHCl3 and that of ligand
4 from 8 to 20 mol % saw under the condition of complete
turnover a decrease of the ee value of the thioacetate (S)-
6-SAc from 73% to 33% (see Supporting Information,
Table S1, entries 6 and 7).
(iii) Asym m etr ic Syn th esis of Cyclopen ten yl Th io-
a ceta te. An efficient asymmetric synthesis of the cyclo-
pentenyl thioacetate (S)-5-SAc from acetate rac-5-OAc
or carbonate rac-5-OMoc was precluded because of the
operation of a strong ME and other effects. Therefore the
palladium-catalyzed reaction of the naphthoate rac-5-
ONa p h with 1.4 equiv of KSAc in the presence of 2 mol
% of Pd2(dba)3‚CHCl3 and 8 mol % of 4 in CH2Cl2/H2O
was studied. It was hoped that the ME would be sup-
pressed in this case because of the other nucleofuge.16,19
Indeed, by using the naphthoate rac-5-ONa p h as
starting material the thioacetate (S)-5-SAc was isolated
with 92% ee in high yield (Scheme 9). Although this
result is synthetically welcome, it also gives testimony
to the complexity of the palladium-catalyzed allylic
alkylation with ligand 4 since a similar reaction of rac-
5-ONa p h with 2-pyrimidinethiol as nucleophile in CH2-
Cl2 still showed a strong ME.1e
chemical correlation with sulfide (S)-6- SC4H3N2.1e Al-
though the eight-membered thiobenzoate (S)-8-SBz ex-
hibits the opposite sign of optical rotation as the seven-
membered thiobenzoate (S)-7-SBz, it was also assigned
the S configuration. We had previously noted a similar
change of the sign of optical rotation in the case of the
corresponding seven- and eight-membered cyclic tert-
butyl sulfones, which both had the same absolute
configuration.1e The absolute configurations of acetates
(S)-2-OMoc,1e (S)-1-OAc,26a (R)-6-OAc,26c and (R)-7-
OAc,26d carbonate (R)-8-OMoc,1e and alcohols (S)-5-
OH26e and (S)-6-OH26b were assigned on the basis of their
chiroptical data.
Con clu sion
The palladium-catalyzed allylic alkylation of thiocar-
boxylate ions with racemic allylic esters in the presence
of ligand 4 allows the asymmetric synthesis of cyclic and
acyclic allylic thioesters. The necessary use of water in
the allylic alkylations of the thiocarboxylate ions led with
some allylic carbonates to a competing palladium-
catalyzed “hydrolysis” with formation of the correspond-
ing enantioenriched allylic alcohols. Formation of allylic
alcohols in palladium-catalyzed allylic alkylation of
external nucleophiles with allylic carbonates in the
presence of water has not been previously reported. It
may limit in certain cases the application of allylic
carbonates in palladium-catalyzed allylic alkylation in
water, a topic that has found much interest recently.7d
In the present cases the unwanted side reaction could
be avoided by choosing the more reactive thiobenzoate
ion as nucleophile.
The reaction of the racemic allylic acetates with
thiocarboxylate ions was accompanied by a KR of the
substrates. The KRs were highly efficient for the six- and
seven-membered cyclic acetates in terms of both the
selectivities of enantiomer differentiation and of the
allylic alkylation. The similar S values of the KRs of
cyclohexenyl acetate with KSAc, lithium tert-butylsulfi-
nate, and 2-pyrimidinethiol show that the selectivity is
nearly independent of the nucleophile and the solvent.
III. Con figu r a tion a l Assign m en ts
The absolute configuration of thioacetate (R)-1-SAc
was assigned through chemical correlation with sulfide
1e
(R)-1-SC4H3N2 as outlined in Scheme 10.
The acyclic thioesters (R)-1-SAc, (R)-1-SBz, and (R)-
2-SBz were also assigned the R configuration since all
of them possess the same sign of optical rotation. The
absolute configuration of the cyclic thioacetate (S)-6-SAc
was assigned through chemical correlation with thiol (S)-
6-SH.10d The cyclic thioesters (S)-5-SAc, (S)-6-SAc, (S)-
7-SAc, (S)-5-SBz, (S)-6-SBz, and (S)-7-SBz were also
assigned the S configuration because of the same sign of
optical rotation. In addition the absolute configuration
of the cyclic thiobenzoate (S)-6-SBz was assigned by
4050 J . Org. Chem., Vol. 69, No. 12, 2004

Allylic Alkylation of Thiocarboxylate Ions
temperature, a complete conversion of the carbonate was
observed. For workup the mixture was diluted with hexane/
EtOAc (10:1) (200 mL) and filtered through a layer of flash
silica gel (3 cm × 2 cm). The filtrate was dried (MgSO4) and
concentrated in vacuo. Purification of the residue by chroma-
tography (hexane/EtOAc, 20:1) afforded thioacetate (S)-6-SAc
(199 mg, 51%) of 94% ee [GC, Lipodex γ, tR ((S)-6-SAc) ) 16.95
min, tR ((R)-6-SAc) ) 16.79 min] and alcohol (S)-6-OH (73 mg,
30%) of 82% ee [GC, CP-Chirasil-Dex-CB, tR ((S)-6-OH ) 27.05
min, tR ((R)-6-OH) ) 28.70 min)]; [R]D -92.9 (c 0.98, CH2Cl2).
Data for (S)-6-SAc: 1H NMR (300 MHz, CDCl3) δ 1.62-1.81
(m, 3 H), 1.96-2.10 (m, 3 H), 2.31 (s, 3 H), 4.16-4.20 (m, 1
H), 5.58-5.66 (m, 1 H), 5.83 (dtd, J ) 10.0, J ) 3.7, J ) 1.7
Hz, 1 H); 13C NMR (75 MHz, CDCl3) δ 12.0 (u), 24.7 (u), 29.5
(u), 30.5 (d), 39.8 (d), 126.2 (d), 130.8 (d), 195.7 (u); IR (neat)
ν˜ 3028 (m), 2925 (s), 2860 (m), 2836 (m), 1690 (s), 1444 (m),
1431 (m), 1353 (m), 1256 (w), 1212 (w), 1136 (s), 1110 (s), 1039
(m), 1000 (w), 988 (w), 955 (m), 918 (w) cm-1; MS (EI) m/z 156
(8) [M+], 113 (10), 81 (100), 80 (46), 79 (48), 77 (12). Anal. Calcd
for C8H12OS (156.25): C 61.50, H 7.74. Found: C 61.25; H 7.76.
Data for (S)-6-OH: 1H NMR (400 MHz, CDCl3) δ 1.54-1.77
(m, 4 H), 1.83-2.10 (m, 3 H), 4.17-4.23 (m, 1 H), 5.72-5.78
(m, 1 H), 5.81-5.87 (m, 1 H); 13C NMR (100 MHz, CDCl3) δ
18.9 (u), 25.0 (u), 32.0 (u), 65.4 (d), 129.6 (d), 130.4 (d).
Rea ction of Ca r bon a te r a c-6-OMoc w ith KSBz. Th io-
ben zoic Acid (S)-S-Cycloh ex-2-en yl Ester ((S)-6-SBz).
Following GP1, to a solution of Pd2dba3‚CHCl3 (52 mg, 0.05
mmol) and ligand 4 (138 mg, 0.20 mmol) in CH2Cl2 (18 mL)
were successively added a solution of KSBz (881 mg, 5 mmol)
in degassed water (2 mL) and carbonate rac-6-OMoc (390 mg,
2.5 mmol). After the mixture stirred for 24 h at room temper-
ature, a complete conversion of the carbonate was observed
and a mixture of (S)-6-SBz and (S)-6-OH in a ratio of 88:12
(GC) was formed. Workup and chromatography (hexane/
EtOAc, 20:1) afforded thiobenzoate (S)-6-SBz (504 mg, 92%)
as a colorless oil: 89% ee [HPLC, Chiralcel-OD-H, heptane/
2-propanol, 100:1, 1.00 mL/min tR ((S)-6-SBz) ) 17.09 min, tR
((R)-6-SBz) ) 18.92 min]; [R]D -243.9 (c 1.02, CH2Cl2). A
similar reaction of rac-6-OMoc with KSBz in THF/water
(9:1) for 15 h at room temperature furnished after workup
thiobenzoate (S)-6-SBz (487 mg, 89%) of 72% ee: 1H NMR (400
MHz, CDCl3) δ 1.70-1.92 (m, 3 H), 2.04-2.15 (m, 3 H), 4.38-
4.46 (m, 1 H), 5.70-5.76 (m, 1 H), 5.86-5.92 (m, 1 H), 7.40-
7.45 (m, 2 H), 7.52-7.57 (m, 1 H), 7.94-7.98 (m, 2 H); 13C NMR
(100 MHz, CDCl3) δ 20.38 (u), 25.1 (u), 23.0 (u), 40.2 (d), 126.4
(d), 127.4 (d), 128.8 (d), 131.3 (d), 133.5 (d), 137.3 (u), 191.9
(u); IR (neat) ν˜ 3028 (w), 2926 (m), 2859 (w), 2834 (w), 1660
(s), 1597 (w), 1581 (w), 1448 (m), 1430 (w), 1201 (s), 1175 (s),
1000 (w), 910 (s) cm-1; MS (EI) m/z 218 (37) [M+], 139 (12),
105 (100), 81 (57), 80 (31), 79 (18), 77 (33). Anal. Calcd for
The racemic cyclopentenyl acetates showed a strong
ME in the palladium-catalyzed reaction with the thio-
carboxylate ions. The matched acetate reacted with an
enantioselectivity higher than that of the mismatched
acetate. Most interestingly, this effect was superimposed
by a concomitant racemization of both acetates. This led
to the establishment of a DKR in the case of the
mismatched acetate, and as a result both acetates deliver
the thioacetate with high enantiomeric excess up to
approximately 70% conversion. However, because of yet
unknown reasons, a lasting DKR was not established
over the whole range of the reaction, and thus the
enantioselectivity of the allylic alkylation decreased
substantially at higher conversion. Rationalization of the
ME with ligand 4 is hampered because of the most likely
involvement of not only one but several monomeric,
oligomeric, and diastereomeric Pd(0)/4 and π-allyl-Pd-
(II)/4 complexes in several catalytic cycles. The recently
proposed two catalytic cycle model for the equilibrating
monomeric and oligomeric complexes and their diaster-
eomers17 may perhaps represent a good starting point
for the development of a mechanistic scheme for the
various aspects of the palladium-catalyzed allylic alky-
lation with this ligand, including the influence of the
catalyst loading on the ME.
Exp er im en ta l Section
Gen er a l P r oced u r e for P a lla d iu m -ca ta lyzed Rea c-
tion s of Ca r bon a tes w ith KSAc a n d KSBz (GP 1). A
Schlenk flask was successively charged with Pd2dba3‚CHCl3,
ligand 4, and CH2Cl2 or THF, and the resulting orange solution
was stirred at room temperature for 15 min. In the meantime
a second Schlenk flask was charged with KSAc or KSBz and
degassed water. The content of the first flask was added via a
syringe to that of the second flask followed by the addition of
the neat racemic allylic carbonate. Subsequently, the reaction
mixture was stirred at room temperature if not stated other-
wise, and the progress of the reaction was monitored by GC.
For workup the reaction mixture was diluted with hexane/
EtOAc, 20:1 (200 mL) and filtered through a layer (3 cm × 2
cm) of flash silica gel. The filtrate was dried (MgSO4) and
concentrated in vacuo. Purification of the residue by chroma-
tography and/or Kugelrohr distillation gave the allylic thioester.
Gen er a l P r oced u r e for P a lla d iu m -Ca ta lyzed KR of
Allylic Aceta tes w ith KSAc a n d KSBz (GP 2). A Schlenk
flask was successively charged with Pd2dba3‚CHCl3, ligand 4,
and CH2Cl2, and the resulting orange solution was stirred at
room temperature for 15 min. In the meantime a second
Schlenk flask was charged with KSAc or KSBz and degassed
water. The content of the first flask was added via a syringe
to that of the second flask followed by the addition of the neat
racemic allylic acetate. Subsequently, the reaction mixture was
stirred at room temperature if not stated otherwise, and the
progress of the reaction was monitored by GC. After the stated
conversion of the substrate, the reaction mixture was diluted
with hexane/EtOAc (20:1) (200 mL) and filtered through a
layer of flash silica gel (3 cm × 2 cm). The filtrate was dried
(MgSO4) and concentrated in vacuo. Purification of the residue
by chromatography and/or kugelrohr distillation gave the
allylic thioester and the allylic acetate.
C
13H14OS (218.32): C 71.52, H 6.46. Found: C 71.54, H 6.54.
Resolu tion of Aceta te r a c-6-OAc w ith KSAc. Following
GP2, to a solution of Pd2dba3‚CHCl3 (52 mg, 0.05 mmol) and
ligand 4 (138 mg, 0.20 mmol) in CH2Cl2 (18 mL) were
successively added a solution of KSAc (400 mg, 3.5 mmol) in
degassed water (2 mL) and acetate rac-6-OAc (350 mg, 2.5
mmol). After the mixture stirred for 12 h at room temperature,
a 51% conversion of the acetate was observed. Workup and
chromatography (hexane/EtOAc, 20:1) afforded thioacetate
(S)-6-SAc (187 mg, 48%) of 97% ee [GC, Lipodex γ, tR ((S)-6-
SAc) ) 16.95 min, tR ((R)-6-SAc) ) 16.79 min]; [R]D -265.9
(c 1.00, CH2Cl2) and acetate (R)-6-OAc (150 mg, 43%) of g
99% ee [GC, Lipodex γ, tR ((R)-6-OAc ) 10.35 min, tR ((S)-6-
OAc) ) 10.73 min (co-injection)]; [R]D +216.9 (c 0.99, CH2Cl2)
as colorless oils. Data for (R)-6-OAc: 1H NMR (400 MHz) δ
1.53-2.06 (m, 9 H), 5.15-5.21 (m, 1 H), 5.60-5.66 (m, 1 H),
5.85-5.91 (m, 1 H); 13C NMR (100 MHz, CDCl3) δ 19.2 (u),
21.7 (d), 25.2 (u), 28.6 (u), 68.3 (d), 125.9 (d), 132.8 (d), 170.8
(u).
Rea ction of Ca r bon a te r a c-6-OMoc w ith KSAc. Th io-
a cetic Acid (S)-S-Cycloh ex-2-en yl Ester ((S)-6-SAc). Fol-
lowing GP1, to a solution of Pd2dba3‚CHCl3 (52 mg, 0.05 mmol)
and ligand 4 (138 mg, 0.20 mmol) in CH2Cl2 (18 mL) were
successively added a solution of KSAc (400 mg, 3.5 mmol) in
degassed water (2 mL) and carbonate rac-6-OMoc (390 mg,
2.5 mmol). After the mixture stirred for 5.25 h at room
A similar resolution of rac-6-OAc with KSAc in CH2Cl2/
water (9:1) was carried out on a 10 mmol scale. Following GP2,
to a solution of Pd2dba3‚CHCl3 (204 mg, 0.20 mmol) and ligand
J . Org. Chem, Vol. 69, No. 12, 2004 4051

Lu¨ssem and Gais
4 (552 mg, 0.80 mmol) in CH2Cl2 (72 mL) were successively
added a solution of KSAc (1.60 g, 14 mmol) in degassed water
(8 mL) and acetate rac-6-OAc (1.4 g, 10 mmol). After the
mixture stirred for 12 h at room temperature, a 51% conversion
was observed. Workup furnished thioacetate (S)-6-SAc (750
mg, 48%) of 97% ee and acetate (R)-6-OAc (616 mg, 44%) of
g99% ee.
Rea ction of Aceta te r a c-5-OAc w ith KSAc. To a solution
of Pd2dba3‚CHCl3 (52 mg, 0.05 mmol) and ligand 4 (138 mg,
0.20 mmol) in CH2Cl2 (18 mL) were successively added a
solution of KSAc (400 mg, 3.5 mmol) in degassed water (2 mL)
and acetate rac-5-OAc (320 mg, 2.5 mmol). After the mixture
stirred at room temperature for 22 h, a complete conversion
of the acetate was observed. Workup and chromatography
(hexane/EtOAc, 20:1) afforded thioacetate (S)-5-SAc (352 mg,
99%) of 58% ee [GC, Lipodex γ, tR ((S)-5-SAc) ) 16.36 min, tR
((R)-5-SAc) ) 16.07 min] as a colorless oil.
Rea ction of Aceta tes (S)-5-OAc a n d (R)-5-OAc w ith
KSAc. Following GP1, to a solution of Pd2dba3‚CHCl3 (52 mg,
0.05 mmol) and ligand 4 (138 mg, 0.20 mmol) in CH2Cl2 (18
mL) were successively added a solution of KSAc (400 mg, 3.5
mmol) in degassed water (2 mL) and acetate (S)-5-OAc (315
mg, 2.5 mmol) of g99% ee. During the reaction samples were
withdrawn, and the conversion and the ee values of the acetate
(R)-5-OAc and the thioacetate (S)-5-SAc were determined by
GC (Lipodex γ). The reaction of enantiomerically pure acetate
(R)-5-OAc of g99% ee with KSAc was carried out in a similar
way.
Ra cem iza tion of (S)-5-OAc. To a solution of Pd2dba3‚
CHCl3 (52 mg, 0.05 mmol) and ligand 4 (138 mg, 0.20 mmol)
in CH2Cl2 (18 mL) was added degassed water (2 mL) and
acetate (S)-5-OAc (315 mg, 2.5 mmol, g99% ee). GC analysis
after 45 min showed the acetate to be racemic. Workup
afforded rac-5-OAc (305 mg, 96%).
Resolu tion of Na p h th oa te r a c-5-ONa p h by HP LC on
Ch ir a l Sta tion a r y P h a se Con ta in in g Colu m n . Na p h th oic
Acid (R)- a n d (S)-Cyclop en t-2-en yl Ester (R)-5-ONa p h
a n d (S)-5-ONa p h . Resolution of rac-5-ONa p h (6.00 g) on a
Chiralcel-OD column (250 mm × 50 mm) (hexane/ethanol,
99.5:0.5, 40 flow/min, 16 bar, 50 mg racemate/1 mL solvent,
(tR ((S)-5-ONa p h ) ) 49.0 min, tR ((R)-5-ONa p h ) ) 61 min, 75
mg per injection) gave (S)-5-ONa p h (2.76 g, 46%) and (R)-5-
ONa p h (2.76 g, 46%). Data for (R)-5-ONa p h : g99% ee
[HPLC, Chiralcel-OD-H, heptane/2-propanol, 99:1, 0.6 mL/min,
Rea ction of Na p h th oa te r a c-5-ONa p h w ith KSAc. To
a solution of Pd2dba3‚CHCl3 (52 mg, 0.05 mmol) and ligand 4
(138 mg, 0.20 mmol) in CH2Cl2 (18 mL) were successively
added a solution of KSAc (400 mg, 3.5 mmol) in degassed water
(2 mL) and naphthoate rac-5-ONa p h (596 mg, 2.5 mmol).
After the mixture stirred at room temperature for 3 h, a
complete conversion of the naphthoate was observed. Workup
and chromatography (hexane/EtOAc, 20:1) afforded thioacetate
(S)-5-SAc (317 mg, 89%) as a colorless oil of 92% ee [GC,
Lipodex γ, tR ((S)-5-SAc) ) 16.36 min, tR ((R)-5-SAc) ) 16.07
min]; [R]D -202.3 (c 1.00, CH2Cl2).
t
R ((S)-5-ONa p h ) ) 13.73 min, tR ((R)-5-ONa p h ) ) 14.57 min];
[R]D +190.0 (c 0.99, CH2Cl2). Data for (S)-5-ONa p h : g99%
ee; [R]D -189.5 (c 1.00, CH2Cl2); 1H NMR (400 MHz, CDCl3) δ
2.01-2.09 (m, 1 H), 2.32-2.50 (m, 2 H), 2.54-2.66 (m, 1 H),
6.00-6.08 (m, 2 H), 6.16-6.19 (m, 1 H), 7.42-7.52 (m, 2 H),
7.56-7.62 (m, 1 H), 7.82-7.86 (m, 1 H), 7.94-7.98 (m, 1 H),
8.13-8.16 (m, 1 H), 8.88-8.94 (m, 1 H); 13C NMR (100 MHz,
CDCl3) δ 29.9 (u), 31.1 (u), 81.1 (d), 124.3 (d), 125.6 (d), 125.9
(d), 127.4 (u), 127.5 (d), 128.3 (d), 129.2 (d), 129.8 (d), 131.1
(u), 132.9 (d), 133.6 (u), 137.6 (d), 167.3 (u). IR (neat) ν˜ 3054
(w), 2971 (w), 2943 (w), 2853 (w), 1709 (s), 1593 (w), 1576 (w),
1510 (m), 1453 (w), 1364 (w), 1337 (m), 1277 (t), 1244 (t), 1197
(t), 1135 (t), 1073 (w), 1030 (s), 1009 (m), 947 (w) cm-1; MS
(EI) m/z 238 (23) [M+], 173 (10), 172 (10), 155 (30), 127 (27),
67 (19). Anal. Calcd for C16H14O2: C 80.65, H 5.92. Found: C
80.62, H 5.92.
Ack n ow led gm en t. Support of this work by the
Deutsche Forschungsgemeinschaft (SFB 380 “Asym-
metric Synthesis With Chemical and Biological Meth-
ods”) is gratefully acknowledged. We thank Dr. Guy C.
Lloyd-J ones for providing us with a preprint (ref 17) and
Cornelia Vermeeren for the HPLC separation.
Su p p or tin g In for m a tion Ava ila ble: General experimen-
tal procedures, spectral, analytical and mass spectrometric
data for all new compounds not described in the Experimental
Section, and Tables S1-S4. This material is available free of
charge via the Internet at http://pubs.acs.org.
J O049756X
4052 J . Org. Chem., Vol. 69, No. 12, 2004