X-ray crystallographic and NMR spectroscopical characterization of intermediates in the Pd-catalyzed allylic substitution reaction with 4-substituted phosphinooxazolines. Correlation between intermediate structure and product configuration

Article abstract of DOI:10.1016/S0020-1693(02)00998-2

The novel P,N-ligand 2-[2-(diphenylphosphino)phenyl]-4,5-dihydro-4-methyl-4-phenyloxazole (2a) has been synthesized. The corresponding [Pd^II(η³-diphenylallyl) (2a)]PF₆ (3a) and [Pd^II(η³-1,3-dimethylallyl) (2a)]PF₆ (4a) complexes have been studied by X-ray analysis and NMR spectroscopy. 3a exists as exo-syn-syn isomer in the solid state. In solution, the same isomer predominates. The X-ray structure of 4a reveals that the oxazoline ligand is coordinated in a pseudo-enantiomeric conformation compared with 3a. A syn-anti arrangement of the allyl substituents is favored in the solid state. NMR spectroscopical investigations suggest a formation of six isomers in solution due to endo-exo orientation of the allyl moiety and syn-anti isomerization of the methyl substituents. NMR data of [Pd⁰(η²-dimethylfumarate)(phosphinooxazoline)] complexes give evidence that two isomers exist in solution. The isomeric ratio is strongly dependent on the steric bulk of the oxazoline substituents. The solid state structures of [Pd⁰(η²-dmfu) (2c)] and [Pd⁰(η²-dmfu) (1a)] confirmed the structures of the main isomer found in solution. The asymmetric allylic substitution reaction of 1,3-diphenylallyl acetate with dimethyl malonate catalyzed by 3a proceeds with a selectivity of 97% ee. The ee induced by 2a in catalytic allylic substitution of 1-methylbutenyl acetate is moderate (18%). A comparison of the intermediate 4a and 5a as model of the actual olefinic intermediate suggests that the poor enantioselectivity achieved with ligand 2a is due to the preferred formation of anti-isomers of the allylic intermediate 4a and the conformational instability of the complex.

Full text of DOI:10.1016/S0020-1693(02)00998-2

Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
www.elsevier.com/locate/ica
X-ray crystallographic and NMR spectroscopical characterization of
intermediates in the Pd-catalyzed allylic substitution reaction with 4-
substituted phosphinooxazolines. Correlation between intermediate
structure and product configuration
Margareta Zehnder a, Silvia Schaffner a, Markus Neuburger a, Dietmar A. Plattner b,*
a Institut fur Anorganische Chemie der Universitat Basel, Spitalstrasse 51, CH-4056 Basel, Switzerland
¨
¨
b Laboratorium fur Organische Chemie der Eidgenossischen Technischen Hochschule Zurich, ETH-Honggerberg, HCI, CH-8093 Zurich, Switzerland
¨
¨
¨
¨
¨
Received 11 March 2002; accepted 25 April 2002
Dedicated to Professor Karl Wieghardt on the occasion of his 60th birthday in recognition of his many outstanding contributions to coordination
chemistry.
Abstract
The novel P,N-ligand 2-[2-(diphenylphosphino)phenyl]-4,5-dihydro-4-methyl-4-phenyloxazole (2a) has been synthesized. The
corresponding [PdII(h3-diphenylallyl) (2a)]PF6 (3a) and [PdII(h3-1,3-dimethylallyl) (2a)]PF6 (4a) complexes have been studied by X-
ray analysis and NMR spectroscopy. 3a exists as exo ꢀ
The X-ray structure of 4a reveals that the oxazoline ligand is coordinated in a pseudo-enantiomeric conformation compared with 3a.
A syn ꢀanti arrangement of the allyl substituents is favored in the solid state. NMR spectroscopical investigations suggest a
formation of six isomers in solution due to endo ꢀexo orientation of the allyl moiety and syn ꢀanti isomerization of the methyl
/
syn ꢀ/syn isomer in the solid state. In solution, the same isomer predominates.
/
/
/
substituents. NMR data of [Pd0(h2-dimethylfumarate)(phosphinooxazoline)] complexes give evidence that two isomers exist in
solution. The isomeric ratio is strongly dependent on the steric bulk of the oxazoline substituents. The solid state structures of
[Pd0(h2-dmfu) (2c)] and [Pd0(h2-dmfu) (1a)] confirmed the structures of the main isomer found in solution. The asymmetric allylic
substitution reaction of 1,3-diphenylallyl acetate with dimethyl malonate catalyzed by 3a proceeds with a selectivity of 97% ee. The
ee induced by 2a in catalytic allylic substitution of 1-methylbutenyl acetate is moderate (18%). A comparison of the intermediate 4a
and 5a as model of the actual olefinic intermediate suggests that the poor enantioselectivity achieved with ligand 2a is due to the
preferred formation of anti-isomers of the allylic intermediate 4a and the conformational instability of the complex.
# 2002 Elsevier Science B.V. All rights reserved.
Keywords: Pd(h2-alkene) compounds; Pd(h3-allyl) compounds; Phosphinooxazoline ligands; Enantioselective catalysis; Allylic substitution reaction
1. Introduction
The first highly successful application of these ligands
was in the field of Pd-catalyzed enantioselective sub-
stitution reactions. Up to 99% ee were achieved starting
from racemic 1,3-diphenylpropenyl acetates as sub-
strates. However, selectivities are considerably lower
when alkylallyl substrates are used.
The generally accepted mechanism for the palladium-
catalyzed allylic substitution is shown in Scheme 1 [6]. In
catalytic reactions with phosphinooxazolines the addi-
tion of the nucleophile to the cationic [PdII(h3-allyl)]
intermediate yielding a [Pd0(h2-olefin)] complex is
considered to be the enantiodetermining step of the
reaction [7]. Knowledge of the structures and reactivities
Chiral 4-substituted 2-[2-(diphenylphosphino)phe-
nyl]-4,5-dihydrooxazole ligands (phosphinooxazolines)
(1) have been applied in a number of asymmetric
transition metal-catalyzed reactions including allylic
substitution [1], Heck reaction [2], hydrogenation [3],
hydrosilylation [4] and DielsꢀAlder reaction [5].
/
* Corresponding author. Tel.: ꢁ
1280
E-mail address: plattner@org.chem.ethz.ch (D.A. Plattner).
/
41-1-632 4493; fax: ꢁ41-1-632
/
0020-1693/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 0 - 1 6 9 3 ( 0 2 ) 0 0 9 9 8 - 2

288
M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
Scheme 1. Catalytic cycle of allylic substitution reaction.
of the intermediates involved in this reaction is crucial
for a rationalization of the enantioselectivities observed
in allylic substitution. The chemistry of the [Pd(allyl) (1)]
complexes has been extensively investigated by X-ray
analyses and NMR studies [8] and by computional
studies [9]. Due to their instability the actual [Pd0(al-
kene) (1)] intermediates could not be isolated so far.
Information was derived from NMR spectroscopy [10]
and from [Pd0(alkene)(P,N)] model complexes with the
electron deficient olefins fumarodinitrile [11] and di-
methylfumarate [12].
Plate 1.
2. Results
2.1. X-ray structures of allylic complexes
Single crystals of [PdII(h3-PhÃ
(3a) and [PdII(h3-MeÃ
C3H3ÃMe) (2a)]PF6 (4a) suitable
for X-ray analysis were obtained by crystallization from
1:3:3 CH2Cl2ꢀEtOHꢀhexane. Selected geometrical
/
C3H3Ã
/
Ph) (2a)]PF6
/
/
/
/
[PdII(h3-allyl) (1)]PF6 complexes show dynamic be-
havior in solution. With [Pd(h3-PhC3H3Ph) (1)]PF6
complexes, in each case two diastereoisomers were
parameters are reported in Table 1. For numbering see
Figs. 1 and 2.
2.1.1. Crystal structure of 3a
A close similarity of the coordination geometry and of
found in solution differing in the exo ꢀendo orientation
of the allyl ligand with respect to the oxazoline moiety
[8]. With regard to the orientation of the the allyl
/
the ligand conformation of 3a with the [PdII(h3-PhÃ
/
C3H3Ã
/
Ph) (1)] complexes was found. The coordination
substituents only a syn ꢀ
the central proton was observed. The [Pd(h3-
MeC3H3Me) (1)]PF6 derivatives showed also a syn ꢀ
/syn arrangement with respect to
geometry is pseudo-square-planar. The phosphinooxa-
zoline ligand adopts the characteristic conformation,
which has been found in all coordinated phosphinoox-
azolines of type 1 so far. Due to the non-planarity of the
chelate ring one P-phenyl group adopts a pseudo-axial
and the other a pseudo-equatorial position. The phenyl
substituent of the oxazoline moiety is axially positioned
and lies on the same side of the coordination plane as
the axial P-phenyl ring. The phenyl substituents of the
allyl moiety are in syn position with respect to the
central proton. The steric interaction between the allyl
phenyl group and the oxazoline moiety leads to a
/
anti isomerization of the allylic substituents.
X-ray analyses revealed that the phosphinooxazoline
ligand adopts a rather rigid conformation with the
oxazoline substituent in an axial position with respect
to the coordination plane. As a consequence, the
stereogenic center is remote from the allylic ligand.
In order to bring the chiral information closer to the
allylic terminus we synthesized the novel oxazoline
ligand 2a introducing a methyl substituent in a-position
to the N atom. The corresponding cationic complexes
[PdII(h3-PhC3H3Ph) (2a)]PF6 (3a) and [PdII(h3-
MeC3H3Me) (2a)]PF6 (4a) were prepared and their
structures were investigated in the solid state and in
solution (see Plate 1). A series of [Pd0(h2-dmfu)(oxazo-
Table 1
˚
Selected bond length (A) and angles (8) of 3a and 4a
3a
4a
PdÃN
PdÃP
2.144(3)
2.118(3)
line)] complexes (dmfuꢂdimethylfumarate) was pre-
/
2.279(1)
2.118(3)
2.204(3)
2.365(3)
86.97(8)
65.13(12)
14.87
2.258(1)
2.128(5)
2.133(5)
2.243(5)
88.2(1)
68.0(2)
10.48
pared as model compounds for the elusive olefinic
intermediates. For comparison, the X-ray structures of
complexes [Pd0(h2-dmfu) (2b)] (5b) and [Pd0(h2-dmfu)
(1a)] (6a) were added to this account. The efficiency of
the novel ligand 2a was tested in the standard allylic
substitution reaction [13].
PdÃCcis P
PdÃCcentral
PdÃCtrans P
NÃPdÃP
Ccis PÃPdÃCtrans P
NPdPꢀCPdC
/

M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ
/
298
289
Fig. 1. ORTEP view and labelling scheme of the cation of 3a. Hydrogen atoms have been omitted. Thermal ellipsoids are drawn at 30% probability.
terminus. The allyl substituent cis to the phosphine
moiety is situated in the groove between the P-phenyl
rings. Studies by Pregosin and Albinati confirmed that
this arrangement of the PPh2 moiety and the allyl
substituent is sterically favored [14].
2.2. Crystal structures of olefinic complexes
Crystals of [Pd0(h2-MeO2CÄ
of [Pd0(h2-MeO2CÄ
CO2Me) (1a)] (6a) were obtained by
slow diffusion of pentane into a saturated Et2OꢀCH2Cl2
/
CO2Me) (2b)] (5b) and
/
/
solution. Selected geometrical parameters are reported
in Table 2. For numbering see Figs. 3 and 4.
2.2.1. Crystal structure of 5b
Crystals of 5b contain two crystallographically in-
dependent molecules per asymmetric unit. One molecule
is disordered in the chelate ring portion. The second
molecule shows no signs of disorder and is depicted in
Fig. 3. Except for disorder a close similarity of the two
molecules is observed. The alkene ligand is in a sterically
Fig. 2. ORTEP view and labelling scheme of the cation of 4a. Hydrogen
atoms have been omitted. Thermal ellipsoids are drawn at 30%
probability.
distortion of the square-planar coordination geometry.
N plane forms an angle of 14.98 with the
C(300) plane.
The Pdꢀ
/
Pꢀ
/
PdÃC(100)Ã
/
/
Table 2
˚
Selected bond length (A) and angles (8) for 5b and 6a
2.1.2. Crystal structure of 4a
5b (1)
5b (2)
6a
The coordination geometry of 4a is pseudo-square-
planar. The overall geometry of complex 4a differs
substantially from that of 3a. The allyl group adopts
PdÃN
PdÃP
PdÃCcis P
PdÃCtrans P
CÃC
2.161 (4)
2.162 (7)
2.143 (3)
2.267 (1)
2.067 (3)
2.108 (4)
1.399 (5)
2.268 (1)
2.057 (4)
2.117 (5)
1.414 (8)
85.8 (1)
39.6 (2)
148.6
2.266 (1)
2.061 (4)
2.137 (4)
1.427 (8)
84.2 (4)
39.7 (2)
151.6
exclusively syn ꢀanti configuration, with the allylic
/
methyl substituent cis to the oxazoline moiety in anti
position. The backbone of the phosphinooxazoline
ligand is pseudo-enantiomeric to that of 3a. The phenyl
substituent at the stereogenic center is pseudo-equato-
NÃPdÃP
CÃPdÃC
CÃCÄCÃC
84.9 (1)
39.2 (2)
150.1
rially positioned with respect to the Pdꢀ
/
PꢀN plane. It
/
NPdPꢀ
/
CPdC
6.25
7.62
17.80
lies in the coordination plane close to the adjacent allyl

290
M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
remote from the oxazoline substituent. The steric strain
leads to a rotation of the olefinic bond out of the PdꢀPꢀ
N plane.
The alkene ligand is not planar. The CÃ
CÃH bonds are bent away from the palladium center.
The CÃCÄCÃC torsion angle is about 1508. The length
of the CÄC bond of 1.39 A is significantly greater than
/
/
/CO2Me and
/
/
/
/
˚
/
˚
that for the free olefin (1.36 A). Both the deviation from
planarity and the elongation of the olefin bond provide
evidence of a partial sp3-hybridization of the olefinic
carbons.
2.3. NMR spectroscopy of allylic compounds
Selected NMR data are listed in Table 3.
2.3.1. NMR spectroscopy of 3a
Fig. 3. ORTEP view and labelling scheme of 5b. Hydrogen atoms have
been omitted. Thermal ellipsoids are drawn at 30% probability. For
clarity only the not disorderd molecule is shown.
The NMR spectra of 3a suggest that two isomers are
present in solution. According to the JHH coupling
constants of the allyl protons the phenyl substituents of
the major component (84%) are in the syn position
relative to the central proton. The chemical shift of the
central proton gives evidence that the oxazoline adopts
the same conformation in solution as in the solid state:
The shielding effect of the axially positioned phenyl
substituent at the oxazoline moiety shifts the resonances
of the central allyl proton of the exo isomer to higher
field (5.80 ppm) relative to the corresponding signal of
the endo isomer (ꢀ
HtransP resonance of the minor isomers suggests the
endo ꢀsyn ꢀsyn structure.
/7 ppm). A highfield shift of the
/
/
2.3.2. NMR spectroscopy of 4a
According to the 31P NMR spectrum at least six
isomers exist in CDCl3 solution. These arise from exo ꢀ
/
endo orientation of the allyl group and syn ꢀsyn or syn ꢀ
/
/
Fig. 4. ORTEP view and labelling scheme of 5b. Hydrogen atoms have
been omitted. Thermal ellipsoids are drawn at 30% probability.
anti positions of the allyl substituents. In addition, the
phosphinooxazoline moiety may flip between two con-
formations. The 1H NMR shifts were assigned to
favorable orientation. The CO2Me group cis to the
phosphine moiety lies between P-phenyl rings, the
CO2Me substituent cis to the oxazoline moiety is remote
from the oxazoline substituents.
isomers IꢀIV (Fig. 5). The signals of isomers V and VI
/
are hidden and too small for an unambiguous assign-
ment. The orientation of the allyl substituents was
determined on the basis of the coupling patterns. The
The coordination geometry is trigonal. The CÄ
bond of the alkene ligand is positioned approximately
in the NꢀPdꢀP coordination plane. Bond lengths and
/C
4
coupling constant JHP of a syn methyl group is larger
(ca. 10 Hz) than that of an anti methyl group (ca. 6 Hz).
The conformation of the oxazoline and the orientation
of the allyl group were derived from the high field shifts
of resonances of the allyl ligand due to the anisotropic
ring current effect of the oxazoline phenyl ring. An
axially positioned phenyl ring is situated above the allyl
group and causes an upfield shift of the resonance of the
central proton of the exo isomers. In endo isomers, the
axially positioned phenyl ring affects the resonances of
the adjacent terminal anti proton or methyl substituent,
respectively. An equatorial phenyl group is close to the
allyl terminus and influences both the HtransP and the
/
/
angles of the coordination sphere are in the expected
range. Due to the trans influence of the phosphine
group, the PdÃ
/C distance trans to P is longer than that
trans to N.
2.2.2. Crystal structure of 6a
A close similarity with complex 5b is found. The
oxazoline ligand adopts the characteristic conformation
of coordinated phosphinoxazolines of type 1. The
CO2Me group adjacent to the oxazoline moiety is

M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ
/
298
291
Table 3
Selected 1H NMR chemical shifts (d ppm) of 3a and 4a with coupling constants (J Hz) in parentheses in comparison with those of 4b [8e]
Isomer
Percentage
Mecis P
Metrans P
Hcis P
Hcentral
Htrans P
31P
a
a
3a
4a
I
84
16
3.96
4.33
5.80 (10.5,13.9)
a
6.33
6.33
4.71
21.33
25.87
a
a
II
(11.7,12.5)
a
a
a
I
11
15
20
47
4
0.75
0.88
3.39
4.89 (ꢀ12)
5.39 (11.0, 13.3)
4.53
4.15
23.85
22.00
24.76
23.73
19.61
18.98
24.32
25.82
20.30
a
a
a
II
III
IV
V
0.73ꢀ/0.78
0.72ꢀ/0.78
3.20 (6.2, 13.3)
a
3.55
3.55 (6.2, 11.9)
a
a
0.91 (6.2, 9.3)
0.97 (6.2, 9.6)
1.29 (6.7, 6.7)
0.70 (6.7, 6.7)
4.35
5.28 (8.4, 11.9)
5.13
4.12
a
a
a
a
a
a
a
a
a
a
a
VI
I
3
a
4b
33
50
17
0.83 (6.3, 11.3)
1.02 (6.3, 9.8)
1.19 (ꢀ7.7)
1.88 (6.0,11.5)
1.35 (6.3,6.9)
3.41 (6.3, 10.6)
3.74 (6.3, 12.1)
3.95 (6.7, 7.7)
5.53 (11.5, 12.4)
5.60 (8.7, 12.1)
4.8
5.95 (ꢀ7)
4.8
III
V
a
a
a
1.90
5.56
a
Chemical shift and/or coupling constants not determined due to signal overlap.
Complexes of type 5 and 6 with chiral ligands exist in
solution as a mixture of two diastereoisomers, which
differ in the orientation of the alkene group with respect
to the oxazoline ligand. Only one isomer is found for 5b.
The signals are assigned according to the JPH coupling
constants, expected to be larger for trans than for cis
coupling. The orientation of the dimethylfumarate
ligand and the conformation of the oxazoline ligand of
5a and 6a were derived from the highfield shift of the
resonances of the alkene proton and the protons of the
methyl group trans to P in analogy to the allylic
complexes.
2.4.1. Spectrum of 5a
According to the highfield shift of the HtransP
resonance (3.04 vs. 3.8 ppm) the alkene ligand has the
same orientation with respect to the phosphinooxazoline
as observed in the solid state structures of 5b and 6a. In
the minor isomer, a significant highfield shift of the
MetransP resonance (2.87 vs. 3.6 ppm) is observed. As the
MetransP resonance (3.55 vs. 3.6 ppm) of the major
diastereoisomer as well as the HtransP resonance (3.57 vs.
3.69 ppm) of the minor isomer are not significantly
affected, it can savely be assumed that the oxazoline
moiety adopts the same conformation as ligands of type
1 with the phenyl ring axially positioned. There is no
evidence of a flipping of the chelate ring.
Fig. 5. Diastereoisomers of 4a in CDCl3 solution. (Exo: the central
CÃH bond of the allyl group points towards the larger substituent R;
endo: the central CÃH bond points away from the larger substituent).
/
/
MetransP signals but not the chemical shift of the Hcentral
resonance, irrespective of the orientation of the ligand.
Following this scheme, the NMR data suggest that
the main isomer (isomer IV) has the same structure as
found in the solid state (highfield shift of the MetransP
and HtransP resonance). In contrast, the highfield shift of
the Hcentral resonance of isomers I and III in comparison
with that of II and IV suggests an axial position of the
oxazoline phenyl ring.
2.4. NMR spectroscopy of olefinic complexes
Selected NMR data are listed in Table 4 (for a
representation of isomers M and m, see Plate 2).
Plate 2.

292
M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
Table 4
Selected 1H NMR chemical shifts (d ppm) of 5a, 5b and 6aꢀ
6d with coupling constants (J Hz) in parentheses
/
Isomer
Percentage
HÃCcis P (JHH, JHP
)
HÃCtrans P (JHHꢀJHP
)
Mecis P
Metrans P (JHP
)
31P(H)
5a
M
m
67
33
dd 3.98 (10.1, 2.1)
dd 4.03
pt 3.04 (10.1)
pt 3.55
s 3.03
s 3.17
s 3.11
s 2.99
s 3.37
s 3.07
s 3.32
s 3.06
s 3.33
s 3.10
s 3.50
s 3.55 (0.8)
s 2.87 (0.8)
d 3.58 (0.9)
d 3.65 (0.7)
d 2.92 (0.7)
d 3.59 (0.5)
d 3.61 (0.5)
s 3.60
20.20
20.09
21.59
20.18
20.23
20.25
21.44
19.89
20.87
20.95
21.62
5b
6a
M
M
M
M
M
M
M
M
M
dd 4.14 (10.0, 2.3)
dd 4.07 (10.3, 1.8)
dd 4.02 (10.1, 2.4)
dd 4.15 (10.1, 0.7)
pt 3.81 (10.2)
dd, 3.39 (10.3, 10.8)
pt 3.69 (10.1)
pt 3.93 (10.2)
pt 3.78 (10.1)
pt 3.91 (10.4)
pt 3.74 (10.4)
pt 3.79 (10.4)
72
28
54
46
62
38
80
20
6b
6c
6d
dd 4.17 (10.1, 0.9)
a
a
s 3.60
d 3.60 (0.8)
d 3.61 (0.8)
a
4.21
4.21
a
a
pt 3.79
s, singlet; d, doublet; pt pseudo-triplet.
Table 5
Asymmetric induction of ligand 2a in comparison with that of ligands
of type 1 [1a]
2.4.2. Spectra of 5c and 6bꢀd
The chemical shifts of the alkene groups in these
complexes display a close similarity. The signal of the
/
proton cis to P (JPH
ppm. The pseudo-triplet of the protons trans to P
(JPH 10) appears at approximately 3.8 ppm. In gen-
eral, the signal of the methyl protons trans to P appears
at 3.6 ppm as a narrow doublet due to a long range
ꢀ
/
2) appears in the range of 4.0ꢀ
/
4.2
L
Substrate
Yield (%)
ee (%)
Product
(S)-2a
(S)-1a
(S)-2a
(S)-1a
(S)-1b
(S)-1c
(S)-1d
Rac.-7
Rac.-7
Rac.-8
Rac.-8
Rac.-8
Rac.-8
Rac.-8
77
99
93
97
95
99
96
97
99
18
50
56
59
71
(S)-9
(S)-9
ꢀ
/
(S)-10
(S)-10
(S)-10
(S)-10
(S)-10
coupling to the phosphorus (JPH
of the methyl protons cis to P of the main isomers
appears at ꢀ3.1 ppm, that of the minor one at ꢀ3.3
B/1 Hz). The resonance
/
/
ppm. This suggests that the major isomers adopt
conformation M in all complexes.
racemic product (S)-9 with a high enantioselectivity
(97%) comparable with that induced by ligands of type
1. For the alkylallyl acetate 8 the enantioselectvity was
dramatically lower with ligand 2a (18%). With ligands of
type 1 the enantiomeric excess is higher than for 2a but
2.4.3. Spectrum of 6a
According to the highfield shift of the HtransP
resonance (3.39 vs. 3.8 ppm) the major isomer has the
same structure as that found in the solid state. In the
minor diastereoisomer, the highfield shift of the MetransP
resonance (2.92 vs. 3.6 ppm) is consistent with a
structure with opposite orientation of the alkene group.
only moderate (50ꢀ71%).
/
For the diphenylallyl acetates 7, the stereochemical
course of the allylic substitution reaction may be
rationalized as proposed in Ref. [6](b). Nucleophilic
attack at the strongly favored isomer I leads to the
favored Pd(olefin) complex (S)-M (Scheme 3). Nucleo-
philic attack at the minor isomer II leads to the
disfavored olefinic diastereoisomer (R)-m. As a conse-
quence, a high enantioselectivitiy results.
An analysis of the equilibria of the (dimethylallyl)pal-
ladium complexes and the relative stabilities of the
(olefin)palladium complexes derived from the isomeric
ratios of the palladium(dimethylfumarate) complexes
gives insight into the origin of enantioselection. It has
been generally accepted that the nuclephilic attack takes
place at the weaker bound allyl terminus trans to P
[6](b). Neglecting nucleophilic attack cis to P and
diastereoisomers leading to Z-isomers, the allylic inter-
2.5. Enantioselective allylic substitution
Asymmetric allylic substitution of the allyl acetates 7
and 8 with dimethyl malonate was carried out with
catalyst 2a following the method with N,O-bis(tri-
methylsilyl)acetamide (BSA) under the same conditions
as with ligands 1 (Scheme 2) [13]. The results of the
catalysis are listed in Tables 5 and 6.
The reaction of allyl acetate rac.-7 with the malonic
ester nucleophile led to the formation of the chiral non
mediates IꢀIV are involved in the catalytic reaction.
/
Nucleophilic attack at the allyl terminus trans to P leads
Scheme 2. Enantioselective allylic substitution reaction.
to four Pd(olefin) diastereoisomers, differing in the

M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ
/
298
293
Table 6
Isomeric ratio of [Pd(dimethylallyl)(oxazoline)] and the [Pd(dmfu)(oxazoline)] complexes vs. asymmetric induction
[Pd(dimethylallyl)(oxazoline)]
[Pd(dmfu)(oxazoline)
Catalysis
Ratio of isomers involved in catalytic reaction (%) I:II:III:IV (deS )
2a 11:16:22:51 (24)
Ratio of isomers (%) M:m
ee (%)
18
67:33
72:28
54:46
62:38
80:20
a
b
1a 50:31:19 (0)
1b 75:16:9 (50)
50
56
a
b
a
b
1c 74:18:8 (48)
1d 72:15:13 (44)
59
71
a
b
(deS ) diastereomeric excess of isomers leading to S-10.
a
obtained from Ref. [8e].
obtained from Ref. [1a].
b
isomer leads to the disfavored olefinic product (S)-m. A
decrease of the asymmetric induction with respect to the
diastereoisomer ratio is observed.
For ligands of type 1 isomers IꢀIII are involved in the
/
formation of product 10. Isomer IV was not detectable
in solution. For the phenyl derivative 1a, an equal
abundance of isomers leading to (S)-10 or (R)-10,
respectively, was detected by NMR spectroscopy. With
respect to the olefin complexes, however, the isomer (S)-
M derived from isomer I leading to (S)-10 is clearly
favored, thus a respectable ee of 50% results.
For ligands 1bꢀd, an almost equal abundance of the
/
major Pd(h3-allyl) isomer I was found. A comparison of
the diastereomeric ratios of complexes 6 reveals that the
stability of the major Pd(h2-olefin) isomer M relative to
that of the minor isomer m increases with the bulk of the
oxazoline substituent. The methyl derivative 1b induces
a poor discrimination between the major and minor
isomer. As a consequence, the enantiomeric excess
induced by 1b is only marginally higher (56%) than the
diastereomeric excess of the allylic isomer I (50%). In
contrast, the major isomer of the t-butyl derivative is
clearly favored (M:mꢂ80:20). In spite of a lower
/
diastereomeric excess of the [Pd(h3-allyl)] isomer I
(44%) ligand 1d induces an enantioselectivity of 71%.
The results give evidence that the enantiomeric excess is
strongly correlated with the relative stabilities of the
diastereomeric olefin complexes. The influence of the
relative concentrations of the allylic intermediates on the
enantioselectivity of the catalytic reaction is much lower,
but is observable.
Scheme 3. Intermediates in allylic substitution reaction.
3. Conclusion
orientation of the olefin moiety with respect to the
oxazoline ligand and in the configuration of the stereo-
genic center, respectively.
In the catalytic reaction with 1-methylbut-2-enyl
acetate 8 and ligand 2a isomers I and IV lead to the
(S)-configured product, the major enantiomer in the
catalytic reaction. The main isomer is diastereoisomer
IV. Nucleophilic attack at the allyl terminus of this
The enantioselectivity induced by ligand 2a in the
enantioselective allylic substitution reaction is in the
same range as that obtained with the analogous ligands
of type 1 using 1,3-diphenylprop-2-enyl acetate as
substrate. With 1-methylbut-2-enyl acetate the enantio-
meric excess obtained with ligand 2a is drastically lower
in comparison with that of ligands of type 1. Investiga-

294
M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
tion of the (dimethylallyl)palladium intermediates and
the corresponding olefinic (dmfu)palladium complexes
suggests that the moderate selectivity is due to the
conformational lability of ligand 2a leading to prefer-
ential formation of (allyl)palladium species other than
4.2.2. 2-[2?-(Diphenylphosphino)phenyl]-4,5-dihydro-4-
methyl-4-phenyloxazole (2a)
a-Phenylglycinol (0.18 g, 1.2 mmol), cyanophenyl(di-
phenyl)phosphine (0.32 g, 1.1 mmol) and ZnCl2 (0.17 g,
1.3 mmol) under N2 were heated to reflux for 94 h. The
the exo ꢀ
/
syn ꢀ
/
syn diastereoisomer. These either lead to
mixture was submitted to CC (4ꢄ
/
5 cm, silica-gel). The
the formation of the product of opposite chirality (II
and III) or are less reactive due to an unfavorable
transition state (IV).
unreacted benzonitrile was removed with 1:2 CH2Cl2ꢀ
/
hexane (50 ml). A mixture of 2a and [Zn (2a)Cl2] was
eluted with AcOEt and isolated as a colorless solid (0.42
g). The mixture was dissolved in CHCl3 (3 ml) and
treated with 2,2?-bipyridine (0.14 g, 0.9 mmol) in CHCl3
(2 ml). The resulting suspension was stirred for 30 min
A comparison of the isomeric ratios of the [Pd(di-
methylallyl) (1)] and the [Pd(dmfu) (1)] complexes versus
the degree of asymmetric induction demonstrates a
strong correlation between the relative stabilities of the
olefinic intermediates and the asymmetric induction.
The relationship between the relative concentrations of
the allylic intermediates and the enantiomeric excess is
less pronounced but clearly observable.
and submitted to CC (2ꢄ
ml)). A colorless solid resulted (0.32 g, 68.4%). 1H NMR
(CDCl3): 1.35 (s, 3H, CH3), 4.15, 4.20 (2 d, 2H, Jꢂ8.0,
CH2), 6.87ꢀ6.92 (m, 1H, arom), 7.14ꢀ7.40 (m, 19 H,
arom), 7.95ꢀ7.80 (m, 1H, arom).
/3 cm, silica-gel, (CHCl3 100
/
/
/
/
4.3. Preparation of allylic compounds
4. Experimental
A mixture of 1 equiv. of the [(allyl)PdCl]2 dimer and
of 2.1 equiv. of phopshinooxazoline was stirred in EtOH
for 15 min. The resulting solution was treated with 3
equiv. of NH4PF6. After 10 min the complexes were
4.1. General
Pd2(dba)3×CHCl3 [15], racemic a-methylphenylglycine
[16] and enantiomerically pure a-methylphenylglycine
[17] were prepared according to literature procedures.
/
filtered off and recrystallized from CH2Cl2ꢀ
/
EtOHꢀ
/
hexane.
Preparation of ligands 1aꢀd and 2b is described in [18].
/
The solvents for chromatography were distilled before
use. The reactions were carried out under N2 using dried
glassware. Column chromatography (CC): SiO2, C 560,
4.3.1. (h3-1,3-diphenylallyl){2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydro-4-methyl-4-
phenyloxazole}palladium(II) hexafluorophosphate (3a)
0.035ꢀ
When not stated otherwise, NMR spectra were recorded
Gemini 300
spectrometer with CDCl3 as solvent; 1H NMR: 300
MHz, chemical shifts in ppm relative to CHCl3 as
internal reference (7.26 ppm), coupling constants J in
Hz; 31P{1H} NMR: 121 MHz, triphenyl phosphate as
/
0.070 mm, F 254, Chemische Fabrik, Uetikon.
Yield: 76%. Recrystallization from CH2Cl2ꢀ
/
EtOHꢀ
/
hexane 1:3:3 afforded air-stable crystals suitable for X-
ray analysis.
1H NMR (CDCl3, r.t.): major isomer (84%): d 0.92 (s,
at room temperature (r.t.) on a Varianꢀ
/
3H, Me), 3.96 (d, 1H, Jꢂ
4.37 (d, 1H, Jꢂ8.9, CH2), 5.80 (dd, 1H, Jꢂ
Hcentral), 6.27ꢀ6.37 (m, 3H, arom, HtransP), 7.35ꢀ
(m, 21H, arom), 8.04ꢀ
8.09 (m, 1H, arom). 31P{H}
NMR: 21.33.
Minor isomer (16%): d 1.31 (s, 3H, Me), 4.03, 4.11 (
2d, 2H, Jꢂ8.6, CH2), 4.33 (d, 1H, HcisP), 4.71 (dd, 1H,
Jꢂ10.5, 13.0, HtransP), 6.27ꢀ6.37 (m, 3H, arom, Hcen-
8.00 (m, 1H,
/
8.9, CH2), 3.96 (d, 1H, HcisP),
11.7, 12.5,
7.73
/
/
/
/
external reference (ꢃ
/
18.0 ppm).
/
4.2. Preparation of ligand 2a
/
4.2.1. a-Methylgycinol
/
/
a-Methylphenylglycine (0.3 g, 1.8 mmol) was added
to a stirred suspension of NaBH4 (0.3 g, 8.1 mmol) in
THF (10 ml). The reaction mixture was immersed in an
ice bath and a solution of concentrated H2SO4 (0.18 ml)
in Et2O (0.42 ml) was added dropwise. The stirring was
continued at r.t. overnight. MeOH (1 ml) and 5 N
NaOH (10 ml) were successively added. After removing
the organic solvents, the aqueous solution was refluxed
for 2 h and, after cooling, extracted with CH2Cl2.
Evaporation of the solvent yielded a-methylgycinol
tral), 7.35ꢀ
/
7.73 (m, 21H, arom), 7.96ꢀ
/
arom). 31P{H} NMR: 25.87.
4.3.2. (h3-1,3-dimethylallyl){2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydro-4-methyl-4-
phenyloxazole}palladium(II) hexafluorophosphate (4a)
Yield: (82%). Recrystallizing from CH2Cl2ꢀ
/
EtOHꢀ
/
hexane 1:3:3 afforded crystals suitable for X-ray analysis
containig 1 equiv. of CH2Cl2.
1H NMR (400 MHz, CDCl3, r.t.): major isomer
1
(0.18 g, 81%) as a colorless solid. H NMR (CDCl3):
(isomer IV): d 0.70 (pt, 3H, JHH
0.97 (dd, 3H, JHH 6.2, JHP 9.6, MecisP), 1.51 (s, 1H,
Me), 3.55 (dq, 1H, Jꢂ6.2, 11.9, HcisP), 4.12 (ddq, 1H,
transP), 4.30, 4.59 (2 d, 2H, Jꢂ9.0, CH2), 5.28 (dd, 1H,
ꢀ
/
JHP
ꢂ6.7, MetransP),
/
d 1.43 (s, 3H, CH3), 2.21 (s (br), 3H, OH, NH), 3.54 (d,
ꢂ
/
ꢂ
/
1H, Jꢂ
/
10.8, CH2) 3.62 (d, 1H, Jꢂ
/
10.8, CH2), 7.24ꢀ
/
/
7.44 (m, 5H, arom).
H
/

M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ
/
298
295
Jꢂ
/
8.4, 11.9, Hcentr.), 7.21ꢀ
/
7.78 (m, 13H, arom), 8.20ꢀ
/
CH2), 4.14 (dd, JHH
ꢂ/10.0, JHP
ꢂ
/
2.3, 1H, HcisP), 6.89ꢀ
/
8.60 (m, 1H, arom).
Isomer I: d 0.75 (3H, MecisP), 0.88 (3H, MetransP),
1.85 (3H, Me), 3.39 (1H, HcisP), 4.53 (1H, HtransP), 4.89
6.95 (m, 1H, arom), 7.20ꢀ
/
7.52 (m, 12H, arom), 7.89ꢀ
/
7.93 (m, 1H, arom). 31P{H} NMR: d 21.59.
4.4.3. (h2-dimethylfumarate){2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydro-4-
phenyloxazole]} palladium(0) (6a)
(pt, 1H, J ꢀ
8.00ꢀ8.05 (m, 1H, arom).
Isomer II: d 0.75 (2dd, 6H, MecisP, MetransP), 1.71 (s,
1H, Me), 3.20 (dq, 1H, Jꢂ6.2, 13.3, HcisP), 4.15 (ddq,
1H, HtransP), 5.39 (dd, 1H, Jꢂ
7.78 (m, 13 H, arom), 8.00ꢀ8.05 (m, 1H, arom).
Isomer III: d 0.91 (dd, 3H, JHH 6.2, JHP
MecisP), 1.29 (pt, JHH JHP
3H, Me), 3.55 (dq, 1H, HcisP), 4.21, 4.73 (2 d, 2H, Jꢂ
9.1, CH2), 4.35 (dd, 1H, Hcentr), 5.13 (ddq, 1H, HtransP),
6.97ꢀ7.00 (m, 2H, arom), 7.21ꢀ7.78 (m, 11H, arom),
8.00ꢀ8.05 (m, 1H, arom).
/
12, Hcentral), 7.21ꢀ7.78 (m, 13 H, arom),
/
/
/
Yield: 64%. Diffusion of pentane into a saturated
/
11.0, 13.3, Hcentral), 7.21ꢀ
/
Et2OꢀpentaneꢀCH2Cl2 solution yielded crystals suita-
/
/
/
ble for X-ray analysis.
1H NMR: major isomer: d 2.99 (s, 3H, MecisP), 3.39
ꢂ
/
ꢂ/9.3,
ꢀ
/
ꢂ6.7, 3H, MetransP), 1.96 (s,
/
(dd, 1H, JHH
Jꢂ0.71, MetransP), 4.07 (dd, 1H, JHH
cisP), 4.24 (dd, 1H, Jꢂ7.9, 8.7, CH2), 4.76 (dd, 1H,
Jꢂ8.7, 10.5, CH2), 5.43 (dd, 1H, Jꢂ7.9, 10.5, CH),
6.92ꢀ7.53 (m, 18H, arom), 8.10ꢀ8.14 (m, 1H, arom).
31P{H} NMR: d 20.19.
Minor isomer: d 2.92 (d, 3H, Jꢂ
(s, 3H, MecisP), 3.69 (pt, 1H, JHH JHP
4.02 (dd, 1H, JHH 10.1, JHP 2.8, HcisP), 4.24 (dd,
1H, Jꢂ5.0, 8.4, CH2), 4.71 (dd, 1H, Jꢂ8.4, 10.1, CH2),
5.70 (dd, 1H, Jꢂ5.0, 10.1, CH), 6.81ꢀ7.53 (m, 18H,
arom), 8.15ꢀ
8.19 (m, 1H, arom). 31P{H} NMR: d
20.23.
ꢂ
/
10.3, JHP
ꢂ
/
10.8, HtransP), 3.65 (d, 3H,
/
/
ꢂ
/
10.3, JHP 1.8,
ꢂ
/
H
/
/
/
/
/
/
/
/
The synthesis and characterization of the [Pd(di-
methylallyl)(1)] complexes is described elsewhere [8e].
/0.7, MetransP), 3.37
ꢂ
/
ꢂ/10.1, HtransP),
4.4. Preparation of [Pd(h2-
dimethylfumarate)(oxazoline)] complexes
ꢂ
/
ꢂ
/
/
/
/
/
A solution of 1 equiv. of Pd2(dba)3×/CHCl3 and 2.1
/
equiv. of the oxazoline ligand in CH2Cl2 was stirred for
30 min at r.t. until the dark purple solution turned dark
brown. Subsequently 2.5 equiv. of dimethylfumarate
was added. The mixture was stirred for 2 h finally
turning yellow. The solution was concentrated and
transferred onto a SiO2 column and eluted with 2:1
4.4.4. (h2-dimethylfumarate){2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydro-4-methyl-
oxazole]}palladium(0) (6b)
Yield: 62%.
hexaneꢀAcOEt. Evaporation of the solvent yielded the
/
1H NMR: major isomer: d 1.33, (d, 3H, Jꢂ
/
6.5, Me),
0.5, MetransP),
10.2, 1H, HtransP), 3.95ꢀ4.01 (m,
10.1, JHP 0.7, 1H, HcisP), 4.33ꢀ
7.11 (m, 1H, arom), 7.20ꢀ7.52 (m,
8.06 (m, 1H, arom). 31P{H} NMR: d
dimethylfumarato complexes as yellow solids.
3.07 (s, 3H, MecisP), 3.59 (d, 3H, JHP
3.93 (pt, JHH JHP
1H), 4.15 (dd, JHH
4.48 (m, 2H), 7.06ꢀ
12H, arom), 8.00ꢀ
20.25.
Minor isomer: d 1.04, (d, 3H, Jꢂ
3H, MecisP), 3.61 (d, 3H, JHP 0.8, MetransP), 3.76 (pt,
JHH JHP 10.1, 1H, HtransP), 3.95ꢀ4.01 (m, 1H), 4.17
(dd, JHH 10.1, JHP 0.9, 1H, HcisP), 4.33ꢀ4.48 (m,
2H), 6.94ꢀ6.99 (m, 1H, arom), 7.20ꢀ7.52 (m, 12H,
arom), 7.97ꢀ
8.00 (m, 1H, arom). 31P{H} NMR: d
21.44.
ꢂ
/
4.4.1. (h2-dimethylfumarate){2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydro-4-methyl-4-
phenyloxazole]}palladium(0) (5a)
Yield: 40%.
ꢀ
/
ꢂ
ꢂ
/
/
/
ꢂ
/
/
/
/
/
1H NMR: major isomer: d 1.78 (s, 3H, Me), 3.03 (s,
/6.4, Me), 3.32 (s,
3H, MecisP), 3.04 (pt, JHH
3.55 (d, 3H, JHP 0.85, MetransP), 3.98 (dd, JHH
JHP 2.1, 1H, HcisP), 4.30, 4.41 (2 d, Jꢂ8.6, 2H, CH2),
6.86ꢀ7.71 (m, 18H, arom), 8.02ꢀ8.07 (m, 1H, arom).
31P{H} NMR: d 20.20.
Minor isomer: d 1.94 (s, 3H, Me), 2.87 (d, 3H, JHP
0.8, MetransP), 3.17 (s, 3H, MecisP), 3.55 (pt, 1H,
transP), 6.86ꢀ7.71 (m, 18H, arom), 8.15ꢀ8.20 (m, 1H,
arom). 31P{H} NMR: d 20.09.
ꢀ
/
JHP
ꢂ/10.1, 1H, HtransP),
ꢂ
/
ꢂ
/
ꢂ/10.1,
ꢀ
/
ꢂ
/
/
ꢂ
/
/
ꢂ
/
ꢂ
/
/
/
/
/
/
/
ꢂ
/
H
/
/
4.4.5. (h2-dimethylfumarate){2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydro-4-
isopropyloxazole]}palladium(0) (6c)
Yield: 51%.
4.4.2. (h2-dimethylfumarate){4,4-dimethyl-2-[2?-
(diphenylphosphino)phenyl]-4,5-dihydrooxazole]}
palladium(0) (5b)
1H NMR: major isomer: d 0.32 (d, 3H, Jꢂ
/
6.9, Me),
2.54 (m, 1H, HCMe2), 3.06
(s, 3H, MecisP), 3.60 (d, 3H, JHP 0.9, MetransP), 3.91
(pt, JHH JHP 10.4, 1H, HtransP), 4.12ꢀ4.52 (m, 4H, 3
aliph, HcisP), 7.03ꢀ7.09 (m, 1H, arom), 7.15ꢀ7.50 (m,
12H, arom), 8.05ꢀ
8.12 (m, 1H, arom). 31P{H} NMR: d
19.90.
Yield: 40%. Diffusion of pentane into a saturated
0.81ꢀ
/
0.90 (d, 3H, Me), 2.42ꢀ
/
Et2Oꢀ
/
pentaneꢀ/CH2Cl2 solution yielded crystals suit-
ꢂ
/
able for X-ray analysis.
1H NMR: d 1,20, 1.30 (2 s, 6H, Me), 3.11 (s, 3H,
ꢀ
/
ꢂ
/
/
/
/
MecisP), 3.58 (d, 3H, JHP
ꢂ/0.9, MetransP), 3.81 (pt, JHH
ꢀ
/
/
JHP 10.2, 1H, HtransP), 3.98, 4.08 (d, Jꢂ
ꢂ
/
/8.3, 2H,

296
M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
Minor isomer: d 0.11 (d, 3H, Jꢂ
(d, 3H, Me), 2.00ꢀ2.10 (m, 1H, HCMe2), 3.33 (s, 3H,
MecisP), 3.60 (d, 3H, JHP 0.9, MetransP), 3.74 (pt, JHH
JHP 10.4, 1H, HtransP), 4.12ꢀ4.52 (m, 4H, 3 aliph,
cisP), 6.94ꢀ6.99 (m, 1H, arom), 7.15ꢀ7.50 (m, 12H,
arom), 8.00ꢀ
8.05 (m, 1H, arom). 31P{H} NMR: d
20.88.
/
6.7, Me), 0.81ꢀ
/
0.90
4.5.2. Palladium-catalyzed allylic alkylation of 1-
methylbut-2-enyl acetate
/
ꢂ
/
ꢀ
/
A mixture of [Pd(C3H5)Cl]2 (2.32 mg, 6.36 mmol) and
of 2a (6.7 mg, 15.9 mmol) in CH2Cl2 (0.6 ml) was stirred
for 30 min at r.t. To the resulting pale yellow mixture, a
solution of rac.-1-methylbut-2-enyl acetate (84.5 mg,
0.635 mmol) in CH2Cl2 (3 ml) was added. Then dimethyl
malonate (252 mg, 1.9 mmol), BSA (388 mg, 1.90 mmol)
and KOAc (1.1 mg) were added. The mixture was stirred
at r.t. for 1 h. The solution was diluted with CH2Cl2 (25
ml) and washed twice with ice cold sat. aq. NH4Cl
solution. The organic phase was dried (Na2SO4) and
ꢂ
/
/
H
/
/
/
4.4.6. (h2-dimethylfumarate){4-t-butyl-2-[2?-
(diphenylphosphino)phenyl]-4,5-
dihydroxazole]}palladium(0) (6d)
Yield: 48%.
evaporated. CC (SiO2, hexaneꢀ
mg (93.1%) of (E)-2-methoxycarbonyl-3-methylhex-4-
ene acid methyl ester. 1H NMR: d 1.06 (d, Jꢂ
6.8, Me),
1.63 (dd, Jꢂ6.4, 1.3, Me) 2.83ꢀ2.96 (m, HÃC(3)), 3.27
(d, Jꢂ9.1, HÃC(2)), 3.69, 3.73 (2 s, Me), 5.34 (ddq, Jꢂ
15.2, 8.1, 1.3, HÃC(4)), 5.51 (ddq, Jꢂ15.2, 6.4, 0.8, HÃ
/EtOAc 4:1) afforded 123
1H NMR: major isomer: d 0.71, (s, 9H, 3 Me), 3.10 (s,
3H, MecisP), 3.60 (d, 3H, JHP
JHH JHP 10.4, 1H, HtransP), 4.14ꢀ
aliph, HcisP), 6.96ꢀ7.01 (m, 1H, arom), 7.17ꢀ
12H, arom), 8.10ꢀ
8.14 (m, 1H, arom). 31P{H} NMR: d
20.95.
Minor isomer: d 0.53, (s, 9H, 3 Me), 3.50 (s, 3H,
MecisP), 3.61 (d, 3H, JHP 0.8, MetransP), 3.79 (pt, 1H,
transP), 4.14ꢀ4.35 (m, 3H, aliph, HcisP), 4.50ꢀ4.53 (m,
1H, aliph), 6.79ꢀ6.84 (m, 1H, arom), 7.17ꢀ7.50 (m,
12H, arom), 7.16ꢀ
8.20 (m, 1H, arom). 31P{H} NMR: d
21.62.
ꢂ
/
0.8, MetransP), 3.79 (pt,
4.35 (m, 4H, 3
7.50 (m,
/
ꢀ
/
ꢂ
/
/
/
/
/
/
/
/
/
/
/
/
/
/
C(5)). An enantiomeric excess of 18% was determined
by 1H NMR spectroscopy, using the splitting of the
signal of the Me group at the stereogenic carbon center
by the chiral shift reagent Eu(hfc)3 (CDCl3, 300 MHz,
0.5 equiv.).
ꢂ
/
H
/
/
/
/
/
4.6. Structure determinations
4.6.1. X-ray structure analysis of 3a
4.5. Palladium catalyzed allylic alkylation using BSA
procedure
Crystal data and parameters of the data collection are
compiled in Table 7. Unit-cell parameters were deter-
mined by accurate centering of 25 strong reflections
(11.655
/
u 5
/
20.81). Reflection intensities were collected
Nonius CAD-4),
0.71069 A, graphite monochro-
2u scan. Three standard reflections
4.5.1. Palladium-catalyzed allylic alkylation of 1,3-
diphenylprop-2-enyl acetate
on a four-circle diffractometer (Enrafꢀ
Mo Ka radiation (lꢂ
mator) using v ꢀ
/
˚
/
A mixture of [Pd(C3H5)Cl]2 (2.3 mg, 6.36 mmol) and
of 2a (7 mg, 15.9 mmol, 1.25 equiv./Pd) in CH2Cl2 (0.6
ml) was stirred for 15 min at r.t. To the resulting pale
yellow reaction mixture a solution of rac.-1,3-diphenyl-
prop-2-enyl acetate (160.1 mg, 0.68 mmol) in CH2Cl2 (2
ml) was added. Then dimethylmalonate (252 mg, 1.9
mmol), N,O-bis(trimethylsilylacetamide) (388 mg, 1.90
mmol) and KOAc (1.1 mg) were added. The mixture
was stirred at r.t. for 22 h. Only traces of diphenylpro-
/
were monitored every hour during data collection and
showed no decrease during data collection. The usual
corrections were applied. The absorption correction was
determined using the c-scans of seven reflections [19].
The structure was solved by direct methods [20].
Anisotropic least-squares refinement against F was
carried out on all non-H-atoms using the program
CRYSTALS [21]. The allylic protons were refined iso-
tropically, the positions of the remaining H-atoms were
calculated. Scattering factors were taken from the
International Tables of Crystallography, Vol. IV. The
anion was found to be disordered. The disordered atoms
were refined in two positions. Geometric restraints were
applied to the disordered parts of the structure during
the structure refinement.
penyl acetate were detected by TLC (hexaneꢀAcOEt
/
3:1). The solution was diluted with Et2O and washed
twice with ice cold saturated aqueous (aq.) NH4Cl
solution. The organic phase was dried (Na2SO4) and
concentrated. CC (SiO2, hexaneꢀEtOAc 3:1) afforded
/
158 mg (76.7%) of colorless (E)-2-methoxycarbonyl-3,5-
1
diphenylpent-4-enacid-methylester. H NMR (CDCl3):
d 3.51 (s, 3H, Me), 3.70 (s, 3H, Me), 3.95 (d, 1H, Jꢂ
10.8, H-C(2)), 4.27 (dd, 1H, Jꢂ8.4, 10.7, HÃC(3)), 6.49
(d, 1H, Jꢂ15.6, HÃC(5)), 6.33 (dd, 1H, Jꢂ8.4, 15.8,
HÃC(4)), 7.19ꢀ7.32 (m, 10H, arom). The ee value (97%)
was determined on a Daicel Chiralcel OD column at
lꢂ
254 nm, flow rate 0.3 ml minꢃ1, eluent: hexaneꢀ
isopropanol 99:1.
/
/
/
4.6.2. X-ray structure analyses of 4a, 5b and 6a
Data were collected at ambient temperature with a
/
/
/
/
/
KappaCCD diffractometer, Mo Ka radiation (lꢂ
/
˚
0.71069 A, graphite monochromator). The data were
/
/
indexed and scaled by using the programs DENZO and
SCALEPACK [22] and corrected for Lorenz and polariza-

M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ
/
298
297
Table 7
Data for the X-ray analyses
3a
4a
5b
6a
Chemical formula
Molecular weight
Crystal size (mm)
C
866.11
43H37F6NOP2Pd
C33H33F6NOP2Pd.CH2Cl2
826.90
C29H30NO5PPd
609.94
C33H30NO5PPd
657.98
0.19ꢄ0.23ꢄ0.44
10.833 (3)
11.373 (3)
16.377 (3)
105.92 (2)
91.07 (2)
96.91 (2)
1923.6
0.20ꢄ0.39ꢄ0.40
10.2268 (5)
10.9691 (4)
15.8743 (9)
90
0.05ꢄ0.12ꢄ0.13
9.5217 (2)
20.2317 (5)
29.1562 (6)
90
0.11ꢄ0.21ꢄ0.23
11.3108 (4)
12.2633 (4)
12.3985 (3)
62.025 (2)
76.071 (2)
81.362 (2)
657.98
˚
a (A)
˚
b (A)
˚
c (A)
a (8)
b (8)
g (8)
95.611 (7)
90
1772.2
90
90
3
˚
V (A )
5616.7
orthorhombic
Pbc21
Crystal system
Space group
Z
triclinic
¯
monoclinic
P21
triclinic
¯
P/1/
P/1/
2
878
2
834
8
2488
2
670
F (000)
rcalcd (g cmꢃ3
)
1.50
0.630
1.55
0.825
1.44
0.756
1.48
0.727
Absorption coefficient (mmꢃ1
)
u max (8)
30.37
12169
11635
30.06
25196
9831
27.50
36209
11749
28.96
14214
10768
Number of measured reflections
Number of independent reflections
hmin hmax
ꢃ155h 515
ꢃ165k 516
ꢃ155l 516
6559 I ]3s(I)
554
ꢃ115h 514
ꢃ155k 515
ꢃ225l 522
7727 I ]2s(I)
515
ꢃ125h 512
ꢃ245k 526
ꢃ375l 537
8764 I ]2s(I)
800
ꢃ145h 515
ꢃ165k 516
ꢃ155l 516
7400 I ]2s(I)
378
kmin kmax
lmin lmax
Reflections used in reference
Number of parameters
Final R
0.047
0.047
0.045
0.045
0.042
0.040
0.049
0.047
Final Rw
Goodness-of-fit
1.051
ꢃ0.50/2.11
Chebychev polynominal [23]
1.013
ꢃ0.78/1.05
0.989
ꢃ0.86/1.44
1.047
ꢃ0.82/1.11
ꢃ3
˚
Drmax/min (e A )
Weighting scheme
tion effects. The structures were solved by direct
methods [20]. Anisotropic least-squares refinement
against F was carried out on all non-hydrogen-atoms
using the program CRYSTALS [21]. The hydrogen atoms
of the allylic and the alkene moieties, respectively, were
located by Fourier difference synthesis and refined
isotropically. All other hydrogen atoms were included
in calculated positions. Scattering factors were taken
from the International Tables of Crystallography, Vol.
IV. The anion and the CH2Cl2 of 4a were found to be
disordered. One of the two crystallographically inde-
pendent molecules per asymmetric unit of 5b was
partially disorderd. The disordered atoms were refined
in two positions with the occupancy of 0.5 for each
position. Geometric restraints were applied to the
disordered parts of the structure during the structure
refinement.
pies of the data can be obtained, free of charge, on
application to the CDCC, 12 Union Road, Cambridge
CB2 1EZ, UK (fax: ꢁ44 (1223-336-033; e-mail: depos-
/
it@ccdc.cam.ac.uk or www; http://www.ccdc.cam.a-
c.uk).
Acknowledgements
The support of this project by the Swiss National
Foundation
Science
(project
Number
2000-
05?8855.99.1) is gratefully acknowledged.
References
[1] (a) P. von Matt, A. Pfaltz, Angew. Chem., Int. Ed. Engl. 32 (1993)
566;
(b) J. Sprinz, G. Helmchen, Tetrahedron Lett. 34 (1993) 1769;
(c) G.J. Dawson, C.G. Frost, J.M.J. Williams, Tetrahedron Lett.
34 (1993) 3149.
5. Supplementary material
[2] O. Loiseleur, M. Hayashi, N. Schmees, A. Pfaltz, Synthesis (1997)
1338.
Crystallographic data (excluding structure factors) for
the structures reported in this paper have been deposited
with the Cambridge Crystallographic Date Centre as
[3] A. Lightfood, P. Schnider, A. Pfaltz, Angew. Chem., Int. Ed.
Engl. 37 (1998) 2897.
[4] (a) L.M. Newman, J.M. Williams, R. McCague, G.A. Potter,
Tetrahedron Asymmetry 7 (1996) 1597;
deposition Number CCDC-179970ꢀCCDC-179973. Co-
/

298
M. Zehnder et al. / Inorganica Chimica Acta 337 (2002) 287ꢀ298
/
(b) T. Langer, J. Janssen, G. Helmchen, Tetrahedron Asymmetry
7 (1996) 1599.
[11] (a) B. Jedlicka, R.E. Rulke, W. Weissensteiner, R. Ferna´ndez-
¨
Gala´n, F.A. Jalo´n, B.R. Manzano, J. Rodr´ıguez-de la Fuente, N.
Veldman, H. Kooijman, A.L. Spek, J. Organomet. Chem. 516
(1996) 97;
[5] I. Sagasser, G. Helmchen, Tetrahedron Lett. 39 (1998) 261.
[6] For a review see: (a) B.M. Trost, D.L. Van Vranken, Chem. Rev.
96 (1996) 395;
(b) M. Zehnder, M. Neuburger, S. Schaffner, M. Jufer, D.A.
Plattner, Eur. J. Inorg. Chem. (2002) 1511.
(b) M.A. Lautens, A. Pfaltz, in: E.N. Jacobsen, A. Pfaltz, H.
Yamamoto (Eds.), Allylic Substitution Reactions in Comprehen-
sive Asymmetric Catalysis, vol. 2, Springer, Heidelberg, 1999, pp.
[12] R. Ferna´ndez-Gala´n, F.A. Jalo´n, B.R. Manzano, J. Rodr´ıguez-de
la Fuente, M. Vrahami, B. Jedlicka, W. Weissensteiner, G. Jogl,
Organometallics 16 (1997) 3758.
833ꢀ884.
/
[7] (a) A. Pfaltz, Acta Chem. Scand. 50 (1996) 189;
(b) G. Helmchen, J. Organomet. Chem. 576 (1999) 203;
(c) G. Helmchen, A. Pfaltz, Acc. Chem. Res. 33 (2000) 336.
[8] (a) J. Sprinz, M. Kiefer, G. Helmchen, M. Reggelin, G. Huttner,
O. Walter, L. Zsolnai, Tetrahedron Lett. 95 (1994) 1523;
(b) N. Baltzer, L. Macko, S. Schaffner, M. Zehnder, Helv. Chim.
Acta 79 (1996) 803;
[13] U. Leutenegger, G. Umbricht, C. Fahrni, P. von. Matt, A. Pfaltz,
Tetrahedron 48 (1992) 2143.
[14] P. Barbaro, P.S. Pregosin, R. Salzmann, A. Albinati, R.W. Kunz,
Organometallics 14 (1995) 5160.
[15] T. Ukai, H. Kawazura, Y. Ishii, J.J. Bonnet, J.A. Ibers, J.
Organomet. Chem. 65 (1974) 253.
[16] D. Obrecht, U. Bohdal, R. Ruffieux, K. Muller, Helv. Chim. Acta
¨
(c) G. Helmchen, S. Kudis, P. Sennhenn, H. Steinhagen, Pure
Appl. Chem. 69 (1997) 513;
77 (1994) 1423.
[17] D. Obrecht, U. Bohdal, C. Broger, D. Bur, C. Lehmann, R.
(d) S. Schaffner, L. Macko, M. Neuburger, M. Zehnder, Helv.
Chim. Acta 80 (1997) 463;
(e) S. Liu, J.F.K. Muller, M. Neuburger, S. Schaffner, M.
¨
Ruffieux, P. Schonholzer, C. Spiegler, K. Muller, Helv. Chim.
¨
¨
Acta 78 (1995) 563.
[18] G. Koch, G.C. Lloyd-Jones, O. Loiseleur, A. Pfaltz, R. Pre´toˆt, S.
Schaffner, P. Schnider, P. von Matt, Recl. Trav. Chim. Pays-Bas.
114 (1995) 206.
Zehnder, J. Organomet. Chem. 549 (1997) 283;
(f) S. Schaffner, J.F.K. Muller, M. Neuburger, M. Zehnder, Helv.
¨
Chim. Acta 81 (1998) 1223;
(g) K. Selvakumar, M. Valentini, M. Worle, P.S. Pregosin, A.
¨
[19] A.C.T. North, D.C. Phillips, F.S. Mathews, Acta Crystallogr. 24
(1968) 351.
Albinati, Organometallics 18 (1999) 1207;
(h) G.H. Bernardinelli, E.P. Kundig, P. Meier, A. Pfaltz, K.
¨
[20] A. Altomare, M.C. Burla, M. Camalli, G. Cascarano, C.
Giaccovazzo, A. Gualgliardi, G. Polidori, J. Appl. Crystallogr.
27 (1994) 435.
Radkowski, N. Zimmermann, M. Neuburger-Zehnder, Helv.
Chim. Acta 84 (2001) 3233.
[9] (a) P.E. Blochel, A. Togni, Organometallics 15 (1996) 4125;
¨
[21] D. Watkin, Crystals (issue 9), Chemical Crystallography Labora-
tory, Oxford, 1990.
˚
(b) H. Hagelin, B. .Akermark, P. Norrby, Organometallics 18
[22] J.R. Carruthers, D.J. Watkin, Acta Crystallogr., A 35 (1979) 698.
[23] Z. Otwinowsky, W. Minor, Processing of X-ray diffraction data
in oscillation mode in methods in enzymology, in: C.W. Carter,
R.M. Sweet (Eds.), Macromolecular Crystallography, vol. 276
(1999) 2884.
[10] (a) H. Steinhagen, M. Reggelin, G. Helmchen, Angew. Chem.,
Int. Ed. Engl. 36 (1997) 2108;
(b) J. Junker, B. Reif, H. Steinhagen, B. Junker, I.C. Felli, M.
Reggelin, C. Griesinger, Chem. Eur. J. 6 (2000) 3281.
(Part A), Academic Press, London, 1997, pp. 307ꢀ326.
/