Structural control in palladium(II)-catalyzed enantioselective allylic alkylation by new chiral phosphine-phosphite and pyridine-phosphite ligands

Article abstract of DOI:10.1016/S0957-4166(00)00227-5

The ligands 6-[(diphenylphosphanyl)methoxy]-4,8-di-tert-butyl-2,10-dimethoxy-5,7-dioxa-6-phosphadibenzo[a,c]cycloheptene, 1, (S)-4-[(diphenylphosphanyl)methoxy]-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4a']dinaphthalene, (S)-2, and (S)-4-[(diphenylphosphanyl)methoxy]-2,6-bis-trimethylsilanyl-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a']dinaphthalene, (S)-3, (S)-2-(3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a']dinaphthalen-4-yloxymethyl)pyridine, (S)-4, and (S)-2-(3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a']dinaphthalen-4-yloxy)pyridine, (S)-5, have been easily prepared.The cationic complexes [Pd(η³-C₃H₅)(L-L')]CF₃SO₃ (L-L'=1-(S)-5) and [Pd(η³-PhCHCHCHPh)(L-L')]CF₃SO₃ (L-L'=(S)-2-(S)-4) were synthesized by conventional methods starting from the complexes [Pd(η³-C₃H₅)Cl]₂ and [Pd(η³-PhCHCHCHPh)Cl]₂, respectively. The behavior in solution of all the π-allyl- and π-phenylallyl-(L-L')palladium derivatives 6-14 was studied by ¹H, ³¹P{¹H}, ¹³C{¹H} NMR and 2D-NOESY spectroscopy. As concerns the ligands (S)-4 and (S)-5, a satisfactory analysis of the structures in solution was possible only for palladium-allyl complexes [Pd(η³-C₃H₅)((S)-4)]CF₃SO₃, 11, and [Pd(η³-C₃H₅)((S)-5)]CF₃SO₃, 12, since the corresponding species [Pd(η³-PhCHCHCHPh)((S)-4)]CF₃SO₃, 13, and [Pd(η³-PhCHCHCHPh)((S)-5)]CF₃SO₃, 14, revealed low stability in solution for a long time. The new ligands (S)-2-(S)-5 were tested in the palladium-catalyzed enantioselective substitution of (1,3-diphenyl-1,2-propenyl)acetate by dimethylmalonate. The precatalyst [Pd(η³-C₃H₅)((S)-2)]CF₃SO₃ afforded the allyl substituted product in good yield (95%) and acceptable enantioselectivities (71% e.e. in the S form). A similar result was achieved with the precatalyst [Pd(η³-C₃H₅)((S)-3)]CF₃SO₃. The nucleophilic attack of the malonate occurred preferentially at allylic carbon far from the binaphthalene moiety, namely trans to the phosphite group. When the complexes containing ligands (S)-4 and (S)-5 were used as precatalysts, the product was obtained as a racemic mixture in high yield. The number of the configurational isomers of the Pd-allyl intermediates present in solution in the allylic alkylation and the relative concentrations are considered a determining factor for the enantioselectivity of the process. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(00)00227-5

Tetrahedron: Asymmetry 11 (2000) 2765±2779
Structural control in palladium(II)-catalyzed enantioselective
allylic alkylation by new chiral phosphine-phosphite and
pyridine-phosphite ligands
Carmela Grazia Arena, Dario Drommi and Felice Faraone*
Á
Dipartimento di Chimica Inorganica, Chimica Analitica e Chimica Fisica-Universita di Messina, Salita Sperone,
31 Vill. S. Agata-98166 Messina, Italy
Received 15 May 2000; accepted 6 June 2000
Abstract
The ligands 6-[(diphenylphosphanyl)methoxy]-4,8-di-tert-butyl-2,10-dimethoxy-5,7-dioxa-6-phosphadi-
benzo[a,c]cycloheptene, 1, (S)-4-[(diphenylphosphanyl)methoxy]-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4a0]-
dinaphthalene, (S)-2, and (S)-4-[(diphenylphosphanyl)methoxy]-2,6-bis-trimethylsilanyl-3,5-dioxa-4-phospha-
cyclohepta[2,1-a;3,4-a0]dinaphthalene, (S)-3, (S)-2-(3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalen-
4-yloxymethyl)pyridine, (S)-4, and (S)-2-(3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalen-4-yl-
oxy)pyridine, (S)-5, have been easily prepared. 0
The cationic complexes [Pd(Z3-C3H5)(L±L )]CF3SO3 (L±L0=1±(S)-5) and [Pd(Z3-PhCHCHCHPh)-
(L±L0)]CF3SO3 (L±L0=(S)-2±(S)-4) were synthesized by conventional methods starting from the complexes
[Pd(Z3-C3H5)Cl]2 and [Pd(Z3-PhCHCHCHPh)Cl]2, respectively. The behavior in solution of all the p-allyl-
and p-phenylallyl-(L±L0)palladium derivatives 6±14 was studied by H, 31P{1H}, 13C{1H} NMR and 2D-
1
NOESY spectroscopy. As concerns the ligands (S)-4 and (S)-5, a satisfactory analysis of the structures in
solution was possible only for palladium±allyl complexes [Pd(Z3-C3H5)((S)-4)]CF3SO3, 11, and [Pd(Z3-
C3H5)((S)-5)]CF3SO3, 12, since the corresponding species [Pd(Z3-PhCHCHCHPh)((S)-4)]CF3SO3, 13, and
[Pd(Z3-PhCHCHCHPh)((S)-5)]CF3SO3, 14, revealed low stability in solution for a long time. The new
ligands (S)-2±(S)-5 were tested in the palladium-catalyzed enantioselective substitution of (1,3-diphenyl-
1,2-propenyl)acetate by dimethylmalonate. The precatalyst [Pd(Z3-C3H5)((S)-2)]CF3SO3 a€orded the allyl
substituted product in good yield (95%) and acceptable enantioselectivities (71% e.e. in the S form). A
similar result was achieved with the precatalyst [Pd(Z3-C3H5)((S)-3)]CF3SO3. The nucleophilic attack of the
malonate occurred preferentially at allylic carbon far from the binaphthalene moiety, namely trans to the
phosphite group. When the complexes containing ligands (S)-4 and (S)-5 were used as precatalysts, the product
was obtained as a racemic mixture in high yield. The number of the con®gurational isomers of the Pd-allyl
intermediates present in solution in the allylic alkylation and the relative concentrations are considered a
determining factor for the enantioselectivity of the process. # 2000 Elsevier Science Ltd. All rights
reserved.
* Corresponding author.
0957-4166/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(00)00227-5

2766
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
1. Introduction
Palladium-catalyzed allylic substitution provides an e€ective route for enantioselective carbon±
carbon bond formation, under mild conditions.1 Consequently, there is considerable interest in
this research area and a number of groups have synthesized several chiral ligands, which have
achieved good asymmetric inductions.2
Phosphorus±phosphorus, phosphorus±nitrogen and nitrogen±nitrogen chelating ligands have
been successfully used in these reactions. Generally, much lower enantiomeric excess was
observed for ligands with small substituents and low steric bulk. Helmchen et al.2l achieved
enantioselectivities of 99% in the allylic substitution of acyclic substrates using (phosphanyl)-
dihydrooxazoles as ligands. Extensive research by Trost's group has led to many applications of
the allylic alkylation in organic synthesis.3 Pfaltz et al.2e achieved good results (97% e.e., 97%
yield) in the alkylation of racemic 1,3-diphenyl-2-propenyl acetate using substituted bis(oxazoline)
ligands.
However, the origin of the observed enantioselectivity, in the above-mentioned catalytic pro-
cesses, is still under discussion and indications about the features that the chiral coordinated
ligands in the catalyst should have to achieve high enantiomeric excess are ambiguous.
With regard to the mechanism of the Pd-catalyzed enantioselective allylic alkylation, two steps
are very important in order to induce enantiodiscrimination: (i) the activation of the allylic sub-
strate by the Pd0-ligand complex with formation of the Pd0-ole®n intermediate; and (ii) the attack
of the nucleophile on the Pd-(Z3-allyl) cation. Several studies deal with the factors determining
the enantioselectivity of these reaction steps.4 Recently, to understand how the nucleophile can
discriminate between the allylic carbon atoms (C1 and C3), Osborn, van Leeuwen et al.5 intro-
duced the concept of preferential rotation (PR).
Here we report the results of a study on the preparation and applications of a new series of
phosphine-phosphite and pyridine-phosphite ligands in the allylic alkylation reactions. The aim
was to gain an insight into the possible correlation between the intermediate palladium and allyl
isomers containing the chiral bidentate ligand and the enantiodiscrimination of the nucleophilic
attack. As far as we know, this is the ®rst example in which chiral phosphine±phosphite and
pyridine±phosphite mixed ligands are used in Pd-catalyzed enantioselective allylic alkylation.
2. Results and discussion
2.1. Synthesis of ligands
Syntheses of ligands 1±(S)-5 are outlined in Scheme 1. Ligand 1 was easily prepared by reaction
of diphenylhydroxymethylphosphine (Ph2PCH2OH) with 1 equivalent of 4,8-di-tert-butyl-6-
chloro-2,10-dimethoxy-5,7-dioxa-6-phosphadibenzo[a,c]cycloheptene, in the presence of NEt3, in
toluene at 0ꢀC. Following the same synthetic procedure, ligand (S)-2 was obtained using
Ph2PCH2OH and (S)-4-chloro-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene as start-
ing materials.
Ligands 1 and (S)-2 were isolated as white solids, air sensitive in the solid state and in solution.
The ligand 1 is not chiral owing to the low rotational energy barrier of the biphenyl moiety, also
when the two o-phenolic oxygen atoms are bound to phosphorus atom. However, 1 was used
since the NMR characterization of the allylic palladium isomers formed in solution containing

C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
2767
Scheme 1.
this ligand, gave a helpful insight into the NMR study of all the other allylic palladium systems
with coordinated ligands (S)-2±(S)-5.
The enantiopure ortho-bis(trimethylsilyl)substituted dinaphthalene derivative ligand (S)-3 was
obtained similarly to 1 and (S)-2 starting from o-SiMe3-substituted (S)-4-chloro-3,5-dioxa-4-
phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene which was prepared as described by van Leeuwen
et al.6
Ligand (S)-4 was obtained as a white solid by reaction of (S)-4-chloro-3,5-dioxa-4-phospha-
cyclohepta[2,1-a;3,4-a0]dinaphthalene with 2-hydroxypyridine (1:1=equivalent ratio) in THF
solution, as described above for ligand 1. From the crude resulting white powder, ligand (S)-4
was obtained, as a pure sample and in low yield, by extraction with a mixture of toluene:hexane
(1:3). Most probably, the low yield could be ascribed to the presence of a keto±enolic tautomer-
ism7 (Scheme 1) involving the reactant 2-hydroxypyridine.
Ligand (S)-5 was obtained from the 2-hydroxymethylpyridine and (S)-4-chloro-3,5-dioxa-4-
phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene. The ligand was isolated as a white solid, stable in
argon atmosphere.
Selected 31P{1H} and 1H NMR resonances for ligands 1±(S)-5 are reported in Tables 1 and 2.

2768
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
Table 1
31P{1H} NMR data for compounds 1±14a
2.2. Palladium ꢀ3-allyl complexes
The cationic complexes [Pd(Z3-C3H5)(L±L0)] CF3SO3 [L±L0=1±(S)-5] were synthesized in high
yields by reaction of [Pd(Z3-C3H5)Cl]2 with the corresponding chiral ligand L±L0, in a 1:2 molar
ratio, in the presence of AgCF3SO3 (Pd:Ag=1:1).8 In a similar way, the complexes [Pd(Z3-
PhCHCHCHPh)(L±L0)]CF3SO3 were prepared from [Pd(Z3-PhCHCHCHPh)Cl]2 and the biden-
tate ligands L±L0 [(S)-2±(S)-4].
All the p-allyl- and p-phenylallyl-(L±L0)palladium derivatives 6±14 were obtained as yellow
solids, soluble in chlorinated solvents, acetone and methanol; in acetone solutions they are 1:1
electrolytes.
Tables 1 and 2 show selected 31P{1H} and 1H resonances of complexes 6±14, respectively.
The 31P{1H} NMR spectrum, in CDCl3 solution, of complex [Pd(Z3-C3H5)(1)]CF3SO3, 6,
2
shows two doublets centered at ꢁ 163.5 and 60.4 ppm with coupling constants JP P of 71 Hz,
relative to the phosphite (Pa) and the phosphine (Pb) fragment of the bidentate ligand, respectively.
a
b
1
The H 2D-NOESY spectrum leads to the assignment of the proton signals. The Hc proton
0
shows two NOEs with Ha and Hb (Fig. 1), that are related by further cross-peaks with Ha and
0
0
Hb , respectively. We have assigned Ha as belonging to the allylic site cis to the phosphito group
by the strong cross-peak with the t-butyl protons. On the other hand, the absence of any cross-
peak between Ha and the t-butyl hydrogens led us to conclude that the t-butyl group is lying
below the coordination plane supporting the structure depicted in Fig. 1.

Table 2
1H NMR data for compounds 1±14a

2770
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
1
Figure 1. Allyl section of the phase-sensitive H 2D-NOESY spectrum for the exo-isomer 6
1
The H NMR spectrum shows well-de®ned proton signals belonging only to one species and
the 31P{1H} NMR spectrum presents only one pair of doublet resonances, con®rming the pre-
sence in solution of one species. By 1H 2D-NOESY analysis, this species was assigned as the exo-
conformer and is further supported by the absence of any exchange cross-peak. The presence of
one stable isomer in solution could be ascribed to the high steric hindrance of the t-butyl groups

C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
2771
in the ortho-position in the biphenol fragments. These groups, located below the Pd±allyl
coordination plane, avoid any possible rotation around the Pd±Z1-allyl bond as expected in any
exchange mechanism.
The assignment of all proton signals for the complex 6, where only the exo-isomer is present in
solution, allowed us to elucidate all the species present in solution for complexes 7±10.
As far as compound [Pd(Z3-C3H5)((S)-2)]CF3SO3, 7, is concerned, the 31P{1H} NMR spectrum,
in CDCl3 solution, shows two pairs of doublets relative to the phosphite (ꢁ 170.6 ppm, JP P =74
b
a
Hz and ꢁ 170.5 ppm, JP P =73 Hz) and the phosphine (ꢁ 56.7 ppm, JP P =74 Hz and ꢁ 55.3 ppm,
b
a
b a
JPbPa=73 Hz) moieties of coordinated ligand (S)-2 in the exo- and endo-isomers present in solution
in a ratio of 1:0.8. 2D-NOESY NMR spectroscopy enabled us to assign the allyl hydrogens
resonances for the two isomers. As shown previously for 6, the central allylic proton Hc presents
0
two NOE cross-peaks with Ha and Hb that are related to Ha and Hb , respectively. Phase-sensitive
0
NOEs provides exchange information between the allyl resonances in the Ãendo- and exo-isomers.
Ã
Ã0
0
As shown in Fig. 2, we ®nd exchange cross-peaks between Hc and Hc, Hb and Hb , Hb and Hb ,
Ã0
Ã
0
Ha and Ha , and Ha and Ha.
On the basis of the same conclusions reached by Pregosin et al.,9 this process indicates an Z3-
Z1-Z3 mechanism involving the opening of the Pd±Ca bond cis to the -PPh2 group which is
selective for the allylic carbon atom in the trans position to phosphito moiety since its trans
in¯uence is higher than that for the phosphine donor -PPh2. 13C{1H} assignment of the allylic
carbon atoms was performed by 13C, 1H correlation experiment [13C{1H} ꢁ (CDCl3, 298 K): 123.5
ppm, Cc; 71.4 ppm, Ca; 72.9 ppm, Cb; 124.5 ppm, CÃc; 71.2 ppm, CaÃ; 71.9 ppm, CbÃ]. In both isomers
Cb lies slightly down®eld compared to Ca and this is further evidence that Cb is the labile site in
the allylic exchange as described above.
1H and 31P{1H} NMR spectra of [Pd(Z3-PhCHCHCHPh)((S)-2)]CF3SO3, 8, showed in solu-
tion the resonances relative to only two exo syn±syn and endo syn±anti-isomers con®rming that
the allylic rearrangement came from an Z3-Z1-Z3 mechanism. This feature is already known for
1
compounds of this type.9 Proton assignment of 8 was made possible by H 2D-NOESY (see
Table 2).
1
13C{1H} assignment of the two isomers for complex 8 was performed by 13C, HÃcorrelation
[13C{1H} ꢁ (CDCl3, 298 K): 113.7 ppm, Cc; 90.9 ppm, Ca; 93.5 ppm, Cb; 111.0 ppm, Cc; 82.3 ppm,
CÃa; 98.0 ppm, CÃb]. The two terminal allyl carbons Ca and Cb in the endo syn±anti-isomer have
quite substantially di€erent chemical shifts (CÃa 82.3 ppm versus Cbà 98.0 ppm), while in the exo
syn±syn conformation the resonances have little di€erence, Ca 90.9 ppm versus Cb 93.5 ppm.
Most probably, the high steric hindrance of the two allylic phenyl substituents in the endo syn±
anti conformation led to this marked di€erence in chemical shifts.
The derivatives [Pd(Z3-C3H5)((S)-3)]CF3SO3, 9, and [Pd(Z3-PhCHCHCHPh)((S)-3)] CF3SO3,
10, were studied by NMR spectroscopy.
The behavior in solution of complex 9 is similar to the corresponding complex [Pd(Z3-
C3H5)((S)-2)]CF3SO3, 7, since we found in 31P{1H} NMR spectrum two pair of doublets and the
pattern for the allylic protons in the 1H NMR spectrum indicative of the presence of the exo- and
endo-isomers. As evidenced by 1H and 31P{1H} NMR spectra, complex 10 was present in solution
1
like two di€erent conformers. H 2D-NOESY led us to assign all the proton resonances relative
to the exo- and endo-isomers for complex 9 and to the two exo syn±syn and endo syn±anti-isomers
for complex 10. We gained further structural information from the cross-peaks between the Si±
Me3 group and the ortho-hydrogens in the phenyl substituents of the allyl moiety cis to the
phosphito fragment.

2772
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
Figure 2. 2D-NOESY of 7 showing the various exchange peaks. The analysis of these data revealed selective exchange
within the two exo- and endo-isomers

C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
2773
1
The 13C and H correlation allowed us to assign allylic carbon atoms for the isomers of com-
plexes 9 and 10. As regards the two isomers for the complex 9, there is a small di€erence between
the Ca and Cb chemical shifts [13C{1H} ꢁ (CDCl3, 298 K): 124.9 ppm, Cc; 73.0 ppm, Ca; 71.7 ppm,
Cb; 125.0 ppm, CÃc; 72.6 ppm, CaÃ; 72.1 ppm, CbÃ], but the Cb resonance is at a higher ®eld than Ca
due to the steric hindrance of the cis SiMe3 on Ca atom. 13C{1H} NMR chemical shifts (ꢁ, in
CDCl3 at 298 K) relative to the two isomers of complex 10 are found as follows: 113.1 ppm, Cc;
92.7 ppm, Ca; 91.7 ppm, Cb; 112.7 ppm, CcÃ; 86.9 ppm, CaÃ; 98.9 ppm, CÃb. As previously described
for compound 8, there is a marked chemical shifts di€erence between the two terminal allyl car-
bons CÃa and CÃb only in the the endo syn±anti-isomer of complex 10.
With regard to ligands (S)-4 and (S)-5, a satisfactory analysis of the structures in solution by
2D 1H-NOESY technique has been possible only for palladium±allyl complexes [Pd(Z3-
C3H5)((S)-4)]CF3SO3, 11, and [Pd(Z3-C3H5)((S)-5)]CF3SO3, 12, since the corresponding species
[Pd(Z3-PhCHCHCHPh)((S)-4)]CF3SO3, 13, and [Pd(Z3-PhCHCHCHPh)((S)-5)]CF3SO3, 14,
revealed a low stability in solution over a long time. For all these complexes, the lower ®eld chemical
shift of the ortho-proton of the pyridine is further evidence for the coordination of the pyridine
nitrogen atom to the palladium center.10 31P{1H} NMR spectra of 11 and 12 species appear in
any case as a broad signal showing that a fast equilibrium in the NMR time scale is taking place.
1
This result is con®rmed by H NMR spectra, where similarly we ®nd again a family of broad
signals. 2D 1H-NOESY has allowed us to identify in solution two di€erent isomers with exo- and
endo-conformation.
The NMR characterization in a solution of 13 and 14 has been possible only by 1H NMR. The
broadness of the signals in the 31P{1H} and 1H NMR in a temperature range of 298±220 K indicated
that exchange processes are present in solutions between all di€erent conformations. The main
reason for the speed of this exchange could be found in the absence of bulky substituents on the
bound pyridine towards the cis allyl side that makes the pyridine unable to stop or even retard the
rotation of the allyl around the Pd±C bond.
2.3. Catalytic reactions
The new ligands (S)-2±(S)-5 have been tested in the palladium-catalyzed enantioselective sub-
stitution of 1,3-diphenyl-1,2-propenyl acetate by dimethylmalonate as shown in Scheme 2.
Scheme 2.
The mixture obtained by reaction of racemic (1,3-diphenyl-1,2-propenyl)acetate with the pallad-
ium catalyst [Pd(Z3-C3H5)((S)-2)]CF3SO3 in dichloromethane solution, followed by dimethylma-
lonate addition, was kept at 25ꢀC for 16 h (Table 3: entry 1) under continuous stirring and
afterwards a€orded the allyl-substituted product in good yield (95%) and acceptable enantio-
selectivities (71% e.e. in the (S) form).

2774
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
Table 30
Enantioselective allylic alkylation with [(Z3-C3H5)Pd (L±L )]+ complexes (L±L0=2±5) according to Scheme 3a
The enantioselectivity is not signi®cantly a€ected by any modi®cation of the reaction condi-
tions. Indeed, either the change of solvent from CH2Cl2 to THF, or temperature from 25 to
^20ꢀC, does not signi®cantly modify the e.e. of the product (entry 2 and 3); a decrease of the
temperature gives rise to a decrease of the product yield (entry 2). When the hexa¯uorophosphate
complex [Pd(Z3-C3H5)((S)-2)]PF6, instead of the corresponding CF3SO3 salt, was employed as
catalyst, the enantiomeric excess and product yield were unchanged (entry 5). The absolute con-
®guration of the product was determined from its speci®c rotation according to the literature
data.8 Use of the ortho-bis(trimethylsilyl)substituted binaphthalene derivative (S)-3 does not
signi®cantly a€ect the enantiomeric excess and the yield of the catalytic reaction, even though the
reaction is slower than with ligand (S)-2 (entry 6). As evidenced by NMR spectroscopy, the ortho-
trimethylsilyl substituent on the binaphthalene interacts with the phenyl ring of the allylic moiety
coordinated to palladium(II); this implies a transition state of higher activation energy than for
ligand (S)-2. However, the enantioselectivity of the allylic alkylation was not modi®ed, indicating
that nucleophilic attack of the malonate occurs, as for the ligand (S)-2, preferentially at the allylic
carbon far from the binaphthalene moiety, namely trans to phosphite group in (S)-3, as reported
in the NMR discussion. It has been demonstrated that the nucleophile preferentially attacks the
2f,i,11
allylic carbon atom with the longer Pd±C bond which is consequently more reactive.
Elongation of the Pd±C bond is a consequence either of the trans-in¯uence of the donor-atom in
the trans position or of steric hindrance between the substituent at allylic terminal atoms and the
2f,i,11
coordinated chiral ligand.
The results obtained using the ligand (S)-3 indicate that, for

C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
2775
Scheme 3.
donor-atoms with a high trans in¯uence, the steric factors are slightly important. Thus, with (S)-2
or (S)-3 ligands, the nucleophilic attack of the malonate occurs preferentially at the allylic carbon
atom activated by the higher trans-e€ect of the phosphite than the phosphine group.
By using ligands (S)-4 and (S)-5 in the reaction reported in Scheme 2, product 15 was obtained
as a racemic mixture in high yield (runs 7 and 8). Ligands (S)-4 and (S)-5 are bidentate P,N-ligands
which a€ord mononuclear cationic allylic-palladium(II) chelated complexes by reaction with
[Pd(Z3-C3H5)Cl]2 in the presence of AgCF3SO3 salt. They contain the phosphite framework as
(S)-2 but, di€erently to (S)-2, a pyridine nitrogen atom instead of the PPh2 fragment. Also in (S)-4
and (S)-5 the trans-in¯uence of the phosphite is higher than that of the other donor group. Thus,
the lack of enantiomeric excess using (S)-4 and (S)-5 must be a consequence of the presence of the
pyridine instead of the PPh2; the e€ect is not principally electronic (in this case the enantiomeric
excess should be only modi®ed) but due to smaller steric hindrance of the pyridine with respect to
PPh2. As deduced from NMR spectra, the allylic-intermediates [Pd(PhCHCHCHPh)((S)-4)]+ and
[Pd(PhCHCHCHPh)((S)-5)]+, formed in the reaction course, assume four possible isomeric forms
and undergo fast exchange processes. Under these conditions the steric e€ect in the nucleophilic
attack by the malonate is low and the enantiodiscrimination is mastered. NMR spectra, in solution,
of the complexes [Pd(PhCHCHCHPh)((S)-4)]+ and [Pd(PhCHCHCHPh)((S)-5)]+ support these
assertions. Indeed, they give very broad signals in the temperature range 298±220 K indicating
exchange processes between the four possible di€erent con®gurational isomers (syn±syn exo, syn±
anti exo, syn±syn endo, syn±anti endo). Pregosin et al.12 showed, by a detailed NMR study, that
the palladium±allyl cation, containing the PhCHCHCHMe allyl and the bidentate P,S-ligand (R)-
2-ethylthio-1-(phenylethyl)-(R)-binaphthyl phosphite, forms in solution at least ®ve components

2776
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
which undergo selective exchange; they considered that the presence of much components in
solution is the reason for the modest selectivities observed in the allylic alkylation by this P,S-
ligand.
Di€erently from the complexes 13 and 14, the intermediates [Pd(PhCHCHCHPh)((S)-2)]+ and
[Pd(PhCHCHCHPh)((S)-3)]+, containing ligands (S)-2 and (S)-3, which a€ord the highest
enantioselectivity, exist in solution only as a couple of diastereomeric form (syn±syn exo; syn±anti
endo) in the ratio of 1:0.8. As shown in Scheme 3, the nucleophilic attack occurs at the allyl
carbon Cb trans to the phosphito group. The fact that the (S)-product is obtained as a major
enantiomer indicates higher reactivity of the endo syn±anti-isomer A than the exo syn±syn-isomer
B. It is important to consider that the observed e.e. for the allylation is di€erent from the isomeric
distribution of the intermediate allyl-complexes, indicating di€erent rates for the nucleophilic
attack at the carbon atom of diastereomeric forms. In conclusion, the number and the relative
concentrations of the con®gurational isomers of the Pd±allyl complexes present in solution as
intermediate in the allylic alkylation is a factor determining the enantioselectivity of the process.
This aspect was previously also considered by Togni, Pregosin et al.12,13
Studies that consider ligands such as (S)-4 and (S)-5 containing bulky substituents on the pyridine
ring are in progress, with the aim to further ascertain that the enantioselectivity of the process
increases with a decrease in the number of Pd±allyl con®gurational isomers formed as a con-
sequence of the steric requirements of the pyridine moiety.
3. Experimental
Published methods were used to prepare the compounds Ph2PCH2OH,14 4,8-di-tert-butyl-
6-chloro-2,10-dimethoxy-5,7-dioxa-6-phosphadibenzo[a,c]cycloheptene,15 (S)-4-chloro-3,5-dioxa-
4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene,15 (S)-4-chloro-2,6-bis-trimethylsilanyl-3,5-dioxa-
4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene,6 [Pd(Z3-PhCHCHCHPh)(m-Cl)]2.2f All other
reagents were purchased from Sigma±Aldrich and were used as supplied. Solvents were dried by
standard procedures. All experiments were performed under puri®ed argon. For column chroma-
tography, silica gel 60 (220±440 mesh) purchased from Fluka was used. 1D and 2D NMR
1
experiments were carried out using a Bruker AMX R300 spectrometer. H NMR spectra were
referenced to internal tetramethylsilane and 31P{1H} spectra to external 85% H3PO4. Standard
1
1
1
pulse sequences were employed for H±2D-NOESY,16 13C, H ,17 31P, H correlation studies.18
The phase sensitive NOESY experiments used mixing times of 0.8 s. Elemental analyses were
performed by Redox s.n.c., Cologno Monzese, Milano.
3.1. 6-[(Diphenylphosphanyl)methoxy]-4,8-di-tert-butyl-2,10-dimethoxy-5,7-dioxa-6-phospha-
dibenzo[a,c]cycloheptene 1
A solution of Ph2PCH2OH (0.71 g, 3.28 mmol) and Et3N (0.653 g, 6.57 mmol) in toluene (5 ml)
was added dropwise to a solution of 4,8-di-tert-butyl-6-chloro-2,10-dimethoxy-5,7-dioxa-6-phos-
phadibenzo[a,c]cycloheptene (1.390 g, 3.28 mmol) in the same solvent (15 ml) at 0ꢀC. The reac-
.
tion mixture was stirred overnight at room temperature and then the precipitate of Et3N HCl
formed was removed by ®ltration. The toluene was evaporated from the ®ltrate. The residue was
washed with hexane (5 ml) and dried. A white solid was obtained in 90% yield (1.781 g, 2.95
mmol). Anal. calcd for C35H40O5P2: C, 69.76; H, 6.69. Found: C, 68.56; H, 6.63.

C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
2777
3.2. (S)-4-[(Diphenylphosphanyl)methoxy]-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4a0]dinaph-
thalene 2
This compound was obtained by an analogous procedure to 1, as a white solid, by reaction of
(S)-4-chloro-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene (0.612 g, 1.75 mmol) and
Ph2PCH2OH (0.377 g, 1.75 mmol). Yield: 85% (0.789 g, 1.49 mmol). Anal. calcd for C33H24O3P2:
C, 74.71; H, 4.56. Found: C, 73.52; H, 4.61.
3.3. (S)-4-[(Diphenylphosphanyl)methoxy]-2,6-bis-trimethylsilanyl-3,5-dioxa-4-phosphacyclo-
hepta[2,1-a;3,4-a0]dinaphthalene 3
This compound was prepared by an analogous procedure to 1, as a white solid, by reaction of
(S)-4-chloro-2,6-bis-trimethylsilanyl-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalene
(0.124 g, 2.50 mmol) and Ph2PCH2OH (0.540 g, 2.50 mmol). Yield: 90% (1.518 g, 2.25 mmol).
Anal. calcd for C39H40O3P2Si2: C, 69.41; H, 5.97. Found: C, 64.27; H, 5.89.
3.4. (S)-2-(3,5-Dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalen-4-yloxymethyl)pyridine 4
A solution of PyOH (0.166 g, 1.77 mmol) and Et3N (0.358 g, 3.54 mmol) in THF (8 ml) was
added dropwise to a solution of (S)-4-chloro-3,5-dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]di-
naphthalene (0.620 g, 1.77 mmol) in the same solvent (15 mL) at 0ꢀC. The resulting solution was
stirred overnight at room temperature and then the solvent was vacuum evaporated. The pure ligand
4 was extracted by toluene:hexane (1:3) from the resulting white powder. Yield 35% (0.253 g, 0.619
mmol). Anal. calcd for C25H16NO3P: C, 73.35; H, 3.94; N, 3.42. Found: C, 73.61; H, 3.99; N, 3.46.
3.5. (S)-2-(3,5-Dioxa-4-phosphacyclohepta[2,1-a;3,4-a0]dinaphthalen-4-yloxy)pyridine 5
This compound was obtained by an analogous procedure to 1, as a white solid, by reaction of
2-(hydroxymethyl)pyridine (0.225 g, 2.06 mmol) and (S)-4-chloro-3,5-dioxa-4-phosphacyclo-
hepta[2,1-a;3,4-a0]dinaphthalene (0.723 g, 2.06 mmol). Yield: 90% (0.785 g, 1.85 mmol). Anal.
calcd for C26H18NO3P: C, 73.75; H, 4.29; N, 3.31. Found: C, 73.25; H, 4.23; N, 3.30.
3.6. [Pd(ꢀ3-C3H5)(1)]CF3SO3 6
The following procedure for the preparation of 6 is typical and it was used for the synthesis of all
[Pd(Z3-C3H5)(L±L0)]CF3SO3 complexes (L±L0=1±5). A solution of 1 (0.457 g, 0.76 mmol) in
CH2Cl2 (5 ml) was added to a stirred solution of [Pd(Z3-C3H5)(m-Cl]2 (0.111 g, 0.30 mmol) in the
same solvent (15 ml). After ca. 30 min, AgCF3SO3 (0.156 g, 0.60 mmol) was added, and the pre-
cipitated AgCl ®ltered on Celite. The ®ltrate was vacuum reduced to ca. 3 ml and addition of hex-
ane (20 ml) gave a yellow solid. Yield 80% (0.432 g, 0.48 mmol). Anal. calcd for
C39H45F3O8P2PdS: C, 52.09; H, 5.04. Found: C, 51.10; H, 5.07.
3.7. [Pd(ꢀ3-C3H5)((S)-2)]CF3SO3 7
Yellow powder. Yield 86%. Anal. calcd for C37H29F3O6P2PdS: C, 53.73; H, 3.53. Found: C,
53.80; H, 3.55.

2778
C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
3.8. [Pd(ꢀ3-C3H5)((S)-3)]CF3SO3 9
Yellow powder. Yield 83%. Anal. calcd for C43H45F3O6P2PdSSi2: C, 53.17; H, 4.67. Found: C,
53.20; H, 4.70.
3.9. [Pd(ꢀ3-C3H5)((S)-4)]CF3SO3 11
White powder. Yield 80%. Anal. calcd for C29H21F3NO6PPdS: C, 49.34; H, 3.00; N, 1.98.
Found: C, 49.20; H, 2.97; N, 1.95.
3.10. [Pd(ꢀ3-C3H5)((S)-5)]CF3SO3 12
White powder. Yield 82%. Anal. calcd for C30H23F3NO6PPdS: C, 50.05; H, 3.22; N, 1.95.
Found: C, 50.12; H, 3.26; N, 1.98.
3.11. [Pd(ꢀ3-PhCHCHCHPh)((S)-3)]CF3SO3 10
In a similar manner to the preparation of 6, [Pd(Z3-PhCHCHCHPh)(L±L0)]CF3SO3 (L±L0=1±5)
complexes were prepared. Compound 10 was synthesized by reaction of [Pd(Z3-PhCHCHCHPh)-
(m-Cl)]2 (0.131 g, 0.20 mmol) and (S)-3 (0.500 g, 0.49 mmol). Yield: 85% (0.382 g, 0.34 mmol).
Anal. calcd for C55H53F3O6P2PdSSi2: C, 58.79; H, 4.75. Found: C, 58.84; H, 4.77.
3.12. [Pd(ꢀ3-PhCHCHCHPh)((S)-2)]CF3SO3 8
Orange powder. Yield 82%. Anal. calcd for C49H37F3O6P2PdS: C, 60.10; H, 3.81. Found: C,
60.17; H, 3.83.
3.13. [Pd(ꢀ3-PhCHCHCHPh)((S)-4)]CF3SO3 13
Orange powder. Yield 81%. Anal. calcd for C41H29F3NO6PPdS: C, 57.39; H, 3.41; N, 1.63.
Found: C, 57.22; H, 3.48; N, 1.65.
3.14. [Pd(ꢀ3-PhCHCHCHPh)((S)-5)]CF3SO3 14
Orange powder. Yield 83%. Anal. calcd for C42H31F3NO6PPdS: C, 57.84; H, 3.58; N, 1.61.
Found: C, 57.92; H, 3.53; N, 1.67.
3.15. Allylic alkylation
The procedure applied for allylic alkylation was the same as described by Pfaltz et al.2e In
many cases [Pd(Z3-C3H5)(L±L0)]CF3SO3 complexes [L±L0=(S)-2±(S)-5] were used as catalyst
precursors. For yields and enantiomeric excesses see Table 1.
Acknowledgements
We thank MURST for ®nancial support.

C. G. Arena et al. / Tetrahedron: Asymmetry 11 (2000) 2765±2779
2779
References
1. (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395. (b) Consiglio, G.; Waymouth, M. Chem. Rev.
1989, 89, 257. (c) Tsuji, J. Palladium Reagents and Catalysts; John Wiley & Sons: Chichester, 1995. (d) Hayashi, T.
Catalytic Asymmetric Synthesis; Ojima, I., Ed.;VCH: New York, 1993; p. 325.
2. (a) Hayashi, T.; Kawatsura, M.; Uozumi, Y. J. Am. Chem. Soc. 1998, 120, 1681. (b) Trost, B. M.; Rasdinov, R.;
Grenzer, E. M. J. Am. Chem. Soc. 1997, 119, 7879. (c) Trost, B. M.; Ariza, X. Angew. Chem., Int. Ed. Engl. 1997,
36, 2635. (d) Trost, B. M. J. Am. Chem. Soc. 1998, 120, 70. (e) Pretot, R.; Pfaltz, A. Angew Chem., Int. Ed. Engl.
1998, 37, 323. (f) von Matt, P.; Lloyd-Jones, G. C.; Minidis, A. B. E.; Pfaltz, A.; Macko, L.; Neuburger, M.;
Zehnder, M.; Ruegger, H.; Pregosin, P. S. Helv. Chim. Acta 1995, 78, 265. (g) Boog-Wick, K.; Pregosin, P. S.;
Trabesinger, G. Organometallics 1998, 17, 3254. (h) Brown, J. M.; Hulmes, D. I.; Guiry, P. J. Tetrahedron 1994,
50, 4493. (i) Sprinz, J.; Kiefer, M.; Helmchen, G.; Reggelin, M.; Huttner, G.; Walter, O.; Zsolnai, L. Tetrahedron
Lett. 1994, 35, 1523. (l) Kudis, S.; Helmchen, G. Angew. Chem., Int. Ed. Engl. 1998, 37, 3047. (m) von Matt, P.;
Loiseleur, O.; Koch, G.; Pfaltz, A.; Lefeber, C.; Feucht, T.; Helmchen, G. Tetrahedron Asymmetry 1994, 5, 573.
3. (a) Trost, B. M.; Strege, P. E. J. Am. Chem. Soc. 1977, 99, 1649. (b) Trost, B. M.; Murphy, D. J. Organometallics
1985, 4, 1143. (c) Trost, B. M.; Van Vranken, D. L. Angew. Chem., Int. Ed. Engl. 1992, 31, 228. (d) Trost, B. M.;
Van Vranken, D. L.; Bingel, C. J. Am. Chem. Soc. 1992, 114, 9327. (e) Trost, B. M.; Bunt, R. C. J. Am. Chem.
Soc. 1994, 116, 4089. (f) Trost, B. M.; Verhoeven, T. R. Comprehensive Organometallic Chemistry Wilkinson, G.,
Ed.; Pergamon: Oxford, UK, 1982; Vol. 8, 799.
4. See, for example: (a) Kranenburg, M.; Kamer, P. C. J.; van Leeuwen, P. W. N. M. Eur. J. Inorg. Chem. 1998, 25.
(b) Burckhardt, U.; Baumann, M.; Trabesinger, G.; Gramlich, V.; Togni, A. Organometallics 1997, 16, 5252. (c)
Trabesinger, G.; Albinati, A.; Feiken, N.; Kunz, R. W.; Pregosin, P. S.; Tschoerner, M. J. Am. Chem. Soc. 1997,
119, 6315, and references cited therein. (d) Aranyos, A.; Szabo K. J.; Castano, A. M.; Backvall, Jan-E.
Organometallics 1997, 16, 1058.
5. (a) Dierkes, P.; Ramdeehul, S.; Barloy, L.; Decian, A.; Fischer, J.; Kamer, P. C. J.; van Leeuwen, P. W. N. M.;
Osborn, J. A. Angew. Chem., Int. Ed. Engl. 1998, 37, 3116. (a) Ramdeehul, S.; Dierkes, P.; Aguado, R.; Kamer,
P. C. J.; van Leeuwen, P. W. N. M.; Osborn, J. Angew. Chem., Int. Ed. Engl. 1998, 37, 3118.
6. Buisman, G. J. M.; van der Veen, L. A.; Klootwijk, A.; de Lauge, W. G. J.; Kamer, P. C. J.; van Leeuwen,
P. W. N. M.; Vogt, D. Organometallics 1997, 16, 2929.
7. Beak, P.; Fry, F. S.; Lee, J.; Steele, F. J. Am. Chem. Soc. 1976, 98, 171.
8. Auburn, P. R.; Mackenzie, P. B.; Bosnich, B. J. Am. Chem. Soc. 1985, 107, 2033.
9. Pregosin, P. S.; Salzmann, R. Coord. Chem. Rev. 1996, 155, 35.
10. Arena, C. G.; Drommi, D.; Faraone, F.; Lanfranchi, M.; Nicolo, F.; Tiripicchio, A. Organometallics 1996, 15,
3170.
11. (a) Allen, J. V.; Coote, S. J.; Dawson, G. J.; Frost, G. C.; Martin, C. J.; Williams, J. M. J. Chem. Soc., Perkin
Trans. 1994, 2065. (b) Chesney, A.; Bryce, M. R.; Chubb, R. W. J.; Batsanov, A. S.; Howard, J. A. K.
Tetrahedron: Asymmetry 1997, 8, 2337. (c) Dawson, G. J.; Williams, J. M. J. Tetrahedron: Asymmetry 1995, 6,
2535.
12. Selvakumar, K.; Valentini, M.; Pregosin, P. S.; Albinati, A. Organometallics 1999, 18, 4591.
13. Togni, A.; Burckhardt, U.; Gramlich, V.; Pregosin, P. S.; Salzmann, R. J. Am. Chem. Soc. 1996, 118, 1031.
14. Helmann, H.; Badu, J.; Birkner, H.; Schumacher, O. Liebigs Ann. Chem. 1962, 659, 49.
15. Francio, G.; Arena, C. G.; Faraone, F.; Grai€, C.; Lanfranchi, M.; Tiripicchio, A. Eur. J. Inorg. Chem. 1999,
1219.
16. Jenner, J.; Meier, B. H.; Bachmann, P.; Ernst, R. R. J. Chem. Phys. 1979, 71, 4546-4553.
17. Summers, M. F.; Marzilli, L. G.; Bax, A.; Ernst, R. R. J. Am. Chem. Soc. 1986, 108, 4285-4294.
18. Sklenar, V.; Miyashiro, H.; Zon, G.; Miles, H. T.; Bax, A. FEBS Lett. 1986, 208, 94.