Palladium complexes with chiral diphospholane ligands: Comparative catalytic properties and analysis of (η3-allyl)palladium species

FULL PAPER

Palladium Complexes with Chiral Diphospholane Ligands: Comparative

Catalytic Properties and Analysis of (η³-Allyl)palladium Species

[a]

[a][†]

Laurent Barloy,*^[a]

´

Gregory Malaise, Shailesh Ramdeehul, John A. Osborn,*

Nathalie Kyritsakas,^[b]and Roland Graff^[c]

Dedicated to the memory of Jean-Marc Kern

Keywords: Palladium / Allyl ligands / Reaction intermediates / Asymmetric catalysis / Enantioselectivity

The chiral diphospholane ligands Duphos (1) and Dux- and [Pd{(R,R)-2}(η³-allyl)](SbF₆) (13) revealed the presence of

antphos (2), which can be differentiated by their bite angle,

were applied to palladium-catalysed asymmetric reactions,

essentially allylic alkylations with symmetrically (3a−d) or

unsymmetrically (5a,b) substituted allylic acetates as sub-

exchanging isomers in solution. Structural data provided by

these techniques combined with the preferential rotation

model led to a satisfactory interpretation of the catalytic al-

lylic alkylation results. It appears that the bite angle plays a

strates. The most interesting results were found with the first crucial role in the positioning of the proximal methyl groups,

series of substrates in which 2 was more reactive and/or se- which determines the distribution of allyl isomeric interme-

lective than 1. The model complexes of the catalytic interme-

diates [Pd{(S,S)-1}(η³-cyclohexenyl)](BF₄) (10) and exo-

[Pd{(R,R)-2}(η³-cyclohexenyl)](SbF₆) (11) could be charac-

diates and the balance between clockwise and anticlockwise

rotations in the enantioselective step.

terised by X-ray crystallography. ¹H and ³¹P NMR spectro- ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,

scopic analyses of [Pd{(R,R)-2}(η³-diphenylallyl)](SbF₆) (12)

Germany, 2004)

Introduction

been helpful in understanding the intimate steric and/or

electronic grounds for enantioselectivity.^{[1aϪ1c,2a,3]}

The catalytic properties of two chiral xanthene-derived

ligands for use in allylic alkylations have been previously

reported in the laboratory.^[4]These ligands combine the ad-

vantages of having chiral phospholane donor groups, which

have met with great success in catalysis,^[5]and a wide bite

angle. The advantages of large bite angle ligands in homo-

geneous catalysis are well known^[6]and, furthermore, an in-

crease of the PϪPdϪP angle imposed by the ligand is likely

to improve the control of allylic alkylation enantioselectiv-

ity through enhancement of the depth of the chiral pock-

et.^[1b]

Asymmetric palladium-catalysed reactions have attracted

enormous interest over the past years, especially for allylic

substitutions^[1,2a]or the Heck reaction.^[2b]Allylic alkylation

is of considerable use in organic synthesis and proceeds

smoothly under mild conditions. A large amount of work

has been devoted to the design and synthesis of chiral li-

gands, with the aim of increasing the enantioselectivity of

the catalysts. Very good asymmetric inductions have been

achieved, most of them with the benchmark substrate 1-

acetoxy-1,3-diphenylprop-2-ene, but also with sterically less

demanding substrates such as 1-acetoxycyclohex-2-ene.^[1,2a]

Structural analyses of allyl palladium intermediates have

The purpose of this paper is to report a systematic com-

Laboratoire de Chimie des Metaux de Transition et de Catalyse, parison between one of these ligands, 4,5-bis[(2R,5R)-2,5-

[a]

´

UMR 7513 CNRS, Institut Le Bel, Universite Louis Pasteur,

dimethylphospholanyl)]-9,9-dimethylxanthene,

hereafter

4 rue Blaise Pascal, 67070 Strasbourg Cedex, France

Fax: (internat.) ϩ 33-3-90241329

abbreviated (R,R)-Duxantphos [(R,R)-2], with the homolo-

gous bidentate, narrow bite angle ligand (R,R)-Duphos^[5]

[(R,R)-1] or its enantiomer (S,S)-1. This comparison relates

to their application in palladium-catalysed asymmetric re-

actions and to the three-dimensional structural properties

of their respective allyl palladium() complexes in order to

rationalise their catalytic behaviour.

E-mail: barloy@chimie.u-strasbg.fr

[b]

[c]

[†]

´

Service commun des Rayons X, Faculte de Chimie, Institut Le

´

Bel, Universite Louis Pasteur,

4 rue Blaise Pascal, 67070 Strasbourg Cedex, France

´

Service commun de RMN, Faculte de Chimie, Universite

Louis Pasteur,

1 rue Blaise Pascal, 67008 Strasbourg Cedex, France

Deceased; please address further correspondence to L. Barloy

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

DOI: 10.1002/ejic.200400312

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

3987

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

within very short time ranges (5Ϫ10 min, see Entries 10,

12Ϫ14).

With 1-acetoxy-1,3-diphenylprop-2-ene (3a) as substrate,

both ligands give rise to high ee values which is also the

case for a host of chiral ligands reported in the literature.^[1]

Under the same conditions, ligand 1 leads to a 97% ee in

4 h, whereas ligand 2 affords a faster reaction (1 h) but with

a lower enantioselectivity (85% ee) (Entries 1 and 2). Al-

though the configurations of 1 and 2 are the same [viz

(R,R)], the configurations of the major enantiomer of prod-

uct 4a are the opposite. The ee value found for 1 is practi-

cally identical to that reported by Pregosin et al. (98%) for

Results

Homogeneous Catalysis

the same reaction run with (R,R)-1 but with conditions

slightly different from ours (BSA/CH₂Cl₂).^[7]With ligand 2,

the ee could be increased to up to 97% by modifying the

base and solvent and to a higher extent by lowering the

temperature (Entries 3 and 4).

The ligands Duphos (1) and Duxantphos (2) were first

tested in palladium-catalysed allylic alkylation with the ra-

cemic allylic acetates 3Ϫ6 as substrates, the standard di-

methyl malonate anion as the nucleophile and

[Pd(C₃H₅)Cl]₂as the palladium source. The results are

shown in Table 1. All tests were run using ligands with the

The difference between both diphospholane ligands with

(R,R) configuration, except one (Entry 11). The nucleophile regards to enantio-differentiation is far more striking with

was generated from dimethyl malonate either using sodium the sterically less demanding substrates 3bϪd, with a

hydride, or in situ with N,O-bis(trimethylsilyl)acetamide marked superiority of 2 over 1. Low enantiomeric excesses

(BSA) and a catalytic amount of sodium acetate.

were observed using ligand 1 (5Ϫ10%, Entries 5, 8, 11). In

These results show interesting differences between 1 and contrast, ligand 2 affords high ee values with 2-acetoxypent-

2 with regard to reactivity and selectivity. In general, the Pd 3-ene (3b) and 1-acetoxycyclohex-2-ene (3c) (75%, Entry 7

catalyst is less reactive with ligand 1 than with ligand 2. and 83%, Entry 10, respectively). The performance of li-

BSA does not appear to be an appropriate base for 1 (En- gand (R,R)-2 in terms of enantioselectivity is similar to that

tries 6, 9, 11), leading at best to low yields (0Ϫ30%). Quan- previously reported under somewhat different reaction con-

titative yields are obtained only with NaH. On the contrary, ditions.^[4]The asymmetric induction is, however, less ef-

the catalytic alkylation reactions are complete with ligand ficient (50Ϫ54% ee) when one switches to the cyclopentenyl

2 whatever the base, solvent or temperature. Furthermore, substrate 3d (Entries 12Ϫ14) but 2 still remains superior to

with the cyclic substrates 1-acetoxycyclohex-2-ene (3c) or 1- 1. Incidentally, the configurations of the major enantiomers

acetoxycyclopent-2-ene (3d), the reactions are finished of 4c and 4d are the same (S) with ligand 2.

Table 1. Palladium-catalysed asymmetric allylic alkylations of racemic allylic acetates 3aϪd with dimethyl malonate using diphospholane

ligands 1 and 2

Entry

Substrate

Ligand

Base

Solvent

Temperature

Time^[a]

ee [%]^[b]

1

2

3

4

5

6

7

8

3a

3b

3c

3d

(R,R)-1

(R,R)-2

(R,R)-1

(R,R)-2

(R,R)-1

(R,R)-2

(S,S)-1

(R,R)-2

NaH

THF

room temp.

0 °C

room temp.

0 °C

room temp.

0 °C

4 h

1 h

3 h

4 h

24 h

3.5 h

3 h

24 h

10 min

24 h

10 min

5 min

7 min

97 (S)

85 (R)

89 (R)

97 (R)

9 (R)

BSA/KOAc

NaH

BSA/KOAc

NaH

BSA/KOAc

NaH

BSA/KOAc

NaH

CH₂Cl₂

THF

CH₂Cl₂

THF

CH₂Cl₂

THF

CH₂Cl₂

THF

Ϫ^[c]

75 (S)

5 (S)

9

Ϫ^[d]

10

11

12

13

14

83 (S)

10 (R)^[e]

50 (S)

52 (S)

54 (S)

THF

[a]

[b]

The yields determined by GC (using dodecane as internal standard) were quantitative unless otherwise stated.

See Exp. Sect. The

[c]

[d]

[e]

absolute configuration is given in parentheses. No reaction. 1% yield. 30% yield.

3988  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

Table 2. Palladium-catalysed asymmetric allylic alkylations of racemic allylic acetates 5aϪb with dimethyl malonate using diphospholane

ligands 1 and 2

Entry

Substrate

Ligand

Base

Solvent

Time^[a]

Yield (ee) [%] 6

Yield (ee) [%] 7

1

2

3

4

5

6

7

8

5a

5b

(S,S)-1

(R,R)-2

(S,S)-1

(R,R)-2

BSA/KOAc

NaH

CH₂Cl₂

THF

CH₂Cl₂

24 h

2 h

5 min

24 h

1 h

96 (12)^[b]

95 (3)^[c]

98 (16)^[b]

98 (3)^[b]

3 (nd)^[d]

90 (0)

4 (80)

1 (60)

2 (55)

2 (15)

0

BSA/KOAc

24 h

0

20 (0)

^[a]See Exp. Sect.The reactions were performed at room temperature and the yields determined by GC using dodecane as internal standard.

[b]

[c]

[d]

(S) configuration. (R) configuration. nd ϭ not determined.

We were also interested in the regiochemical and stereo- of the simple allyl acetate (2 equiv.) with the prochiral nu-

chemical outcome of the alkylation by dimethylmalonate of cleophile (1 equiv.) derived from 2-methyl-1-tetralone

unsymmetrically substituted allylic acetates, viz 2-acetoxy- (Scheme 1).^[12]However, the selectivity was disappointing.

4-phenylbut-3-ene (5a) and 3-acetoxy-2-methylhex-4-ene According to some of the experimental procedures of Trost

(5b), catalysed by palladium associated with the diphospho- and Schroeder,^[13]we used LDA as a base either in THF or

lane ligands (Table 2). With substrate 5a the conversion is DME. Both experiments quantitatively afforded the al-

always complete, but the reaction time strongly depends on lylation product, in 5 min and 1 h, respectively, but with a

the reaction conditions from 5 min (Entry 4) to 24 h (Entry low enantioselectivity (12%). Replacement of LDA with the

2). The nucleophile attacks overwhelmingly on the Me side BSA/KOAc system only led to a 2% conversion in 24 h.

of the allyl intermediate to yield 6a as the major product,

whatever the ligand. This is a general trend with this

substrate,^[8Ϫ10]due to delocalization of the CϭC bond with

the phenyl substituent. The ee values of 6a are poor

(3Ϫ16%), whereas they are far better for the minor re-

gioisomer 7a (up to 80%). This situation mirrors the predic-

˚

tions and experimental results reported by Norrby, Akerm-

Scheme 1. Palladium-catalysed asymmetric allylic alkylation of al-

ark, Helquist et al. for the same substrate.^[9]The differences

lyl acetate with 2-methyl-1-tetralone

between both ligands 1 and 2 are rather minor. Under the

same conditions (BSA/CH₂Cl₂), Duphos (1) performs

slightly better than Duxantphos (2) in terms of enantiose-

lectivity (Entries 1 and 2), but the ee of 6a can be improved

to 16% with 2 using THF as the solvent (Entry 3). Substrate

5b appears less reactive and it is necessary to use NaH as a

base to achieve a good conversion (90%) with ligand 2. Un-

der the same conditions, ligand 1 is almost inactive (3%

conversion in 24 h, Entry 5). Only one regioisomeric prod-

uct, 6b, was formed and no asymmetric induction was ob-

served. Under our conditions, the allyl palladium inter-

mediates derived from (R)-5a,b do not interconvert with

those derived from (S)-5a,b. A powerful memory effect can

therefore be expected which results in poor global ee val-

ues,^[11]taking into account that allylic alkylation proceeds

with overall retention of configuration and that the re-

gioisomers 6a,b are the overwhelmingly major catalysis

products.^[2a]

Finally, we tested the diphospholane ligands in another

Pd-catalysed process, the benchmark asymmetric Heck re-

action between phenyl triflate and 2,3-dihydrofuran

(Scheme 2).^[14]Ligand (R,R)-2 associated with Pd(OAc)₂as

a palladium source led to very low yields and ee values [8:

5% yield, 12% ee (S); 9: 5% yield, 2% ee (R)], whereas no

reaction was observed under the same conditions with li-

gand (R,R)-1. With 2, the Heck reaction with Pd(DBA)₂

instead of Pd(OAc)₂gave racemic products in low yields (8:

3%; 9: 8%).

The good performance of ligand (R,R)-2 with small sub-

strates (vide supra) prompted us to carry out the alkylation Scheme 2. Palladium-catalysed asymmetric Heck reaction

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

www.eurjic.org

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

3989

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

Allylpalladium Complexes

In order to clarify how each of the chiral diphospholane

ligands affects the stereochemical outcome of the catalytic

reactions, we attempted to elucidate the three-dimensional

structures of some cationic allyl palladium catalytic inter-

mediates through solid-state (crystallography) and/or solu-

tion (NMR) studies. For this purpose, a series of cationic

(η³-allyl)palladium complexes (10Ϫ13) was synthesised in

acceptable to very good yields (55Ϫ93%) according to a

classical procedure. Abstraction of Cl^Ϫfrom the chloride-

bridged dimer [Pd(η³-C₆H₉)Cl]₂with AgBF₄or AgSbF₆,

followed by addition of (S,S)-1 or (R,R)-2 afforded the

cyclohexenyl complexes 10 and 11. Complexes 12 and 13

were obtained in a similar way from (R,R)-2, starting from

[Pd(η³-Ph₂C₃H₃)Cl]₂or [Pd(η³-C₃H₅)Cl]₂. The synthesis,

NMR analysis, and X-ray structure of the related com-

pound [Pd{(R,R)-Duphos}(η³-1,3-diphenylallyl)](CF₃SO₃)

have been reported by Drago and Pregosin.^[7]

Figure 2. ORTEP drawing of [Pd{(R,R)-Duxantphos}(C₆H₉)]-

(SbF₆) (11) showing 30% probability thermal ellipsoids and the

Ϫ

atom-numbering scheme; hydrogen atoms and the SbF₆anion

have been omitted for clarity

and 2 and selected bond lengths and angles are given in

Tables 3 and 4. Complexes 10 and 11 crystallise as mono-

cationic, pseudo-square-planar complexes with noncoordi-

nating anions. Each metallic centre is coordinated by the

three allylic carbon atoms and the two phosphorus atoms

of the bidentate diphospholane 1 or 2. Crystallographic dis-

order is present in the structure of 10: the palladium atom

occupies two positions (Pd and PdЈ, 50% occupancy ratio

each). Pd is bound to C3, C4, and C5, whereas PdЈ, which

is located slightly below, is bound to C6, C1, and C2. This

˚

Table 3. Selected bond lengths [A] and angles [°] in [Pd{(S,S)-

Duphos}(C₆H₉)](BF₄) (10)

Crystallographic Studies

Single-crystals of 10 and 11 suitable for X-ray diffraction

analyses were obtained from slow liquid-phase diffusion of

n-heptane and vapour-phase diffusion of pentane, respec-

tively, into a dichloromethane solution of the complex. The

X-ray crystallographic structures are depicted in Figures 1

PdϪP1

PdϪP2

C3ϪC4

C4ϪC5

2.278(3)

2.323(3)

1.43(2)

1.44(3)

PdϪC3

PdϪC4

PdϪC5

2.39(2)

2.23(1)

2.41(3)

P1ϪPdϪP2

P1ϪPdϪC3

P1ϪPdϪC5

PdϪP1ϪC7

PdϪP1ϪC10

PdϪP1ϪC13

C3ϪC4ϪC5

86.40(9)

167.0(4)

105.5(5)

117.5(3)

121.6(3)

107.5(3)

118(1)

C3ϪPdϪC5

P2ϪPdϪC5

P2ϪPdϪC3

PdϪP2ϪC18

PdϪP2ϪC19

PdϪP2ϪC22

61.8(6)

162.8(5)

105.5(5)

106.4(3)

129.6(3)

109.4(3)

˚

Table 4. Selected bond lengths [A] and angles [°] for [Pd{(R,R)-

Duxantphos}(C₆H₉)](SbF₆) (11)

PdϪP1

PdϪP2

C1ϪC2

C2ϪC3

2.321(1)

2.335(1)

1.377(8)

1.403(8)

PdϪC1

PdϪC2

PdϪC3

2.247(5)

2.136(4)

2.206(5)

P1ϪPdϪP2

P1ϪPdϪC1

P1ϪPdϪC3

PdϪP1ϪC7

PdϪP1ϪC10

PdϪP1ϪC19

C1ϪC2ϪC3

105.02(4)

156.4(1)

96.7(1)

122.0(2)

124.2(2)

106.7(2)

118.6(5)

C1ϪPdϪC3

P2ϪPdϪC3

P2ϪPdϪC1

PdϪP2ϪC13

PdϪP2ϪC16

PdϪP2ϪC32

64.9(2)

156.9(1)

92.0(1)

116.3(2)

125.7(2)

107.7(2)

Figure 1. ORTEP drawing of [Pd{(S,S)-Duphos}(C₆H₉)](BF₄) (10)

showing 30% probability thermal ellipsoids and the atom-number-

ing scheme; hydrogen atoms and the BF₄^Ϫanion have been omitted

for clarity

3990

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

disorder results in large thermal ellipsoids for the 6 cyclo- yls of 10 are both syn with regard to the coordinating elec-

hexenyl carbon atoms. For the sake of clarity, only Pd is tron pair of the phosphorus atom, whereas in 11 C17 is syn

shown in Figure 1.

but C12 is anti.

The overall structure of 11 is very close to the reported

In 10 the mean phospholane planes are nearly parallel to

structure of [Pd{(R,R)-Duthixantphospholane}(cyclo- the mean cyclohexenyl plane [5.7(2) and 9.0(1)°] and there-

hexenyl)]^ϩ.^[4a]The conformational characteristics of the fore the shortest spatial distance between a terminal allylic

Duxantphos ligand in complex 11 are similar to those of carbon and a chiral carbon of a phospholane is 4.015(20)

˚

achiral xanthene-based ligands coordinated to pal- A (C3···C22). In contrast, the phospholanes of 11 make an

ladium.^[15,16]The main difference between 10 and 11 lies in angle of 72.94(5) and 89.57(6)° with the cyclohexenyl and

the fact that in 10 the aromatic backbone of the ligand is the C3···C10 and C1···C13 distances are shorter at 3.878(7)

˚

approximately in the coordination plane and keeps a and 3.548(7) A, respectively. This observation mirrors ex-

pseudo-C₂symmetry axis, whereas in 11 the xanthene scaf- actly what Kamer et al. noted about palladium allyl com-

fold is almost perpendicular to the coordination plane. In- plexes coordinated either by 1,2-bis(diphenylphosphanyl)-

deed, the P1ϪPdϪP2 plane subtends an angle of only ethane or xantphos.^[15]The consequence is a stronger pro-

14.8(1)° with the best-fit plane defined by the six aromatic trusion of the proximal methyls into the cyclohexenyl li-

Duphos carbons in 10, while in 11 both xanthene aromatic gand. Their distances from the cyclohexenyl mean plane are

˚

rings (also defined by best-fit six-carbons planes) are bent 2.031(2) A (C12) and 1.962(2) A (C17) in 11, compared

˚

by 77.08(4) and 75.67(4)° with regard to the P1ϪPdϪP2 with 2.657(6) A (C11) and 3.428(6) A (C23) in 10. In con-

plane. In 11, the tricycle is folded along an O···CMe₂axis, nection with the steric repulsion between C12 and the allyl

the oxygenated six-membered ring being in a boat confor- of 11, the P1ϪPdϪC3 angle is slightly but significantly

mation, and the angle between the mean aromatic planes larger than P2ϪPdϪC1 [96.7(1)° vs. 92.0(1)°] and

reaches 45.32(4)°. Another marked difference is the angle PdϪP1ϪC10 is also larger than PdϪP2ϪC13 [124.2(2)°

between the mean phospholane planes (defined by their five vs. 116.3(2)°].

atoms) and the P1ϪPdϪP2 plane: 70.6(1) and 78.3(1)° in

10 compared with only 13.41(4) and 34.52(4)° in 11.

The loss of symmetry of the Duxantphos ligand upon

coordination to the palladium centre induces exo/endo

isomerism in the complex. The exo species can be defined

NMR Investigations

as the conformer where the xanthene backbone is located

on the same side of the coordination plane as the cyclohex-

We have also compared the NMR spectroscopic data of

enyl or, in a more general way, where the central allylic pro- complexes 10 and 11 in solution. Like its crystallographic

ton is on the opposite side with regard to the backbone.^[4a]features, the NMR characteristics of 11 are closely akin to

It is indeed the observed conformer for 11 as well as for those of exo-[Pd{(R,R)-Duthixantphospholane}(cyclo-

[Pd{(R,R)-duthixantphospholane}(cyclohexenyl)]^ϩ.

hexenyl)]^ϩ. Upon coordination of the C₂-symmetric ligands

In both structures, the PdϪP bond lengths (2.278Ϫ2.335 1 and 2 to the Pd centre, the loss of the symmetry axis is

1

A) are within the expected range for complexes with phos- revealed by the ³¹P{¹H}, H, and ¹³C{¹H} NMR spectra.

˚

pholanes coordinated to palladium.^[4a,7,17]As expected, the Many signals are duplicated, e.g. the phosphorus reson-

1

bite angle P1ϪPdϪP2 is much larger in 11 [105.02(4)°] than ances which appear as an AB system, or the four H NMR

in 10 [86.40(9)°]. Because of the strong trans influence of signals of the methyl groups borne by the phospholanes

the phospholanes, the palladiumϪterminal allylic carbon which are doublets of doublets. The J_P,Pvalues measured

bonds (10: 2.39Ϫ2.41 A; 11: 2.206Ϫ2.247 A) are longer from the ³¹P{¹H} spectra are slightly larger in 11 than in

than in other reported PdϪη³-cyclohexenyl complexes.^[17]10 (41 Hz vs. 36 Hz) which may be the consequence of an

In addition, they appear much longer in 10 than in 11 which increased PϪPdϪP angle.

˚

is probably linked to the crystallographic disorder observed

The chemical shifts of the phosphorus nuclei are higher

in 10 (vide supra). For the same reason, the cyclohexenyl in the Duphos complex 10 (δ ϭ 68.2 and 71.9 ppm) than

carbons look, artificially, nearly coplanar in 10, whereas in in the Duxantphos complex 11 (δ ϭ 33.7 and 35.5 ppm).

11 the cyclohexenyl is in a chair conformation.

This deshielding could be expected in view of the ring con-

The ’’twist angle’’ τ defined by Helmchen et al.^[18]is use- tribution (∆_R) parameter^[19]which reaches a high value for

ful here for describing the inclination of the allyl in a front a five-membered metallacycle. The chemical shift difference

view. It corresponds to the dihedral angles P1ϪP2ϪC3ϪC5 between the terminal allylic ¹³C signals is a useful tool to

in 10 and P1ϪP2ϪC1ϪC3 in 11 and reaches the values of account for asymmetry in the allyl bonding.^[7]Here, we ob-

Ϫ7.0(6)° and Ϫ8.7(2)°, respectively, which are very close. In served that this difference is only ∆δ ϭ 5.2 ppm in 10, but

other words, in a front view of the molecule where the allyl reaches ∆δ ϭ 12.0 ppm in 11.

is positioned in the foreground, the allyl appears rotated

The case of complex [Pd{(R,R)-2}(η³-1,3-diphenylal-

slightly clockwise so that steric interactions between it and lyl)](SbF₆) (12) appears to be more complicated. In spite of

the proximal phospholane methyls (10: C11 and C23; 11: numerous efforts, we did not manage to grow single crystals

1

C12 and C17) are minimised. It is worth noting that within for X-ray analysis. Its room-temperature ³¹P{¹H} and H

the five-membered phospholane cycles, the proximal meth- NMR spectra exhibit broad signals which sharpen when the

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

www.eurjic.org

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

3991

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

Figure 3. 202.5 MHz ³¹P{¹H} NMR spectra of [Pd{(R,R)-

Duxantphos}(Ph₂C₃H₃)](SbF₆) (12) in CD₂Cl₂in the range

193Ϫ298 K

temperature is decreased (Figure 3). Below 273 K, the ³¹P

NMR spectra exhibit eight narrow doublets between δ ϭ

26 and 42 ppm with 45 Hz Ͻ J_P,PϽ 57 Hz, values which are

all close to those found for complex 11. Measurement of the

PϪP coupling constants associated with a ³¹P{¹H} COSY

spectrum recorded at 193 K allowed us to identify them as

four AB systems, corresponding to two major (12a,b) and

two minor (12c,d) isomeric species where a Pd atom is che-

lated by the diphospholane. If we assume that we can rely

Scheme 3. Representation of relevant NOE contacts in 12a and 12b

from ¹H ROESY; for the sake of clarity, NOEs associated with

scalar couplings are not shown

on the integration of the ³¹P{¹H} NMR signals, their rela- is geminal with the right-side proximal methyl group. In the

tive ratios are 35:25:4:1 at 193 K. 12a and 12b are predomi- background, NOEs are observed between both phosphol-

nant whatever the temperature. Coalescence of the ³¹P and ane distal methylidyne groups (e.g. exo: δ ϭ 2.23 and

¹H signals of the four isomers upon warming to room tem- 2.66 ppm) which is an indication that the xanthene back-

perature indicates that they undergo fast exchange pro- bone is folded as in the X-ray structure of 11.

cesses.

Six allyl signals are spread over a rather large range, i.e.

2D-¹H NMR spectroscopy has proved to be a very useful δ ϭ 4.69Ϫ6.35 ppm. The syn,syn geometry of the allyl li-

tool for the investigation of the 3D structures of allyl Pd gand was established in both isomers from NOE cross-

complexes.^[3]Indeed, although many signals of the 1D-¹H peaks between the lateral allyl signals (δ ϭ 4.69Ϫ5.38 and

NMR spectrum of 12 recorded at 193 K overlap, 2D-NMR 4.92Ϫ5.03 ppm; see Scheme 4). This is also suggested by the

analysis combined with phosphorus-decoupled proton large values of the coupling constants between the central

NMR spectroscopy allowed us to assign most of the signals and lateral allylic protons (10Ϫ14 Hz), although this fact

of both major isomers and identify them as exo,syn,syn alone would not be definite proof.^[3a]We note, however, that

3

(12a) and endo,syn,syn (12b) (Scheme 3). In each isomer, all within each allyl ligand, both values of J_H,Hare not equal

the ¹H signals of the diphospholane ligand (except the which suggests that the allyl group undergoes deformations.

methylene signals) could be assigned with the help of 2D-

The assignment of the allyl configuration (exo or endo)

¹H COSY and ROESY spectra, each proton being connec- is essentially based on the interligand NOEs that the lateral-

ted to another by scalar coupling and/or by spatial proxim- right and central allylic protons cause or do not cause with

ity (NOE effect). The aromatic protons located in the ortho the right-side proximal phospholane methyl group and its

position with regard to the phospholane substituents are gem-methylidyne group (Scheme 3). The endo structure of

3

characterised by a J_H,Pvalue of ca. 7 Hz. We can unam- isomer 12b is supported by the very intense NOE cross-

biguously identify the signals of the key proximal phosphol- peak of the allyl proton signal at δ ϭ 5.03 ppm with the

ane methyl groups that point in the foreground towards the methyl signal at δ ϭ 1.10 ppm (Figure 4). In the exo isomer

allyl moiety. In the 2D-¹H ROESY spectrum, the left-side 12a, the lateral allyl proton (δ ϭ 5.38 ppm) shows a close

proximal methyl group displays a strong NOE with the ad- NOE contact with the proximal methylidyne group (δ ϭ

jacent ortho-xanthene proton (e.g. at δ ϭ 7.35 ppm in the 0.96 ppm) but not with the methyl group (δ ϭ 0.85 ppm).

exo isomer), whereas the other ortho proton (e.g. exo: δ ϭ In addition, an NOE signal between the exo central allyl

6.76 ppm) shows an NOE with the methylidyne group that signal at δ ϭ 6.35 ppm and the proximal methyl signal at

3992

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

normal range (δ ϭ 6.35 ppm). In comparison, the chemical

shift of the central proton of [Pd(Duphos)(1,3-

diphenylallyl)]^ϩis δ ϭ 6.84 ppm, although the signal of the

homologous proton of the endo isomer that lies in the an-

isotropy cone of both xanthene aromatic cycles is shifted

upfield by more than 1 ppm (δ ϭ 5.21 ppm). Conversely,

the signals of several protons belonging to the diphosphol-

ane ligand are shifted upfield, such as (1) on the right side

of the exo isomer, the ortho-xanthene proton (δ ϭ

6.76 ppm), the proximal methylidyne group (δ ϭ 0.96 ppm)

and even the distal methyl group (δ ϭ Ϫ0.08 ppm) as well

as (2) on the endo isomer, both ortho-xanthene protons (δ ϭ

6.46 and 6.74 ppm) and the proximal methylidyne group on

the right (δ ϭ 1.39 ppm). These shifts are probably related

to the anisotropy cone of the phenyl substituents of the

allyl moiety.

1

In each isomer, only the H signals of one phenyl sub-

stituent can be identified, i.e. the right-side phenyl group in

the exo form and the left-side one in the endo form

(Scheme 4). For each of them, five distinct resonances can

be observed. This means that their rotation is frozen at

193 K because of the steric effect of the proximal methyl

groups. However, positive exchange (EXSY) cross-peaks re-

lated to this rotation, between the ortho and meta protons

of each phenyl group, can be found in the ¹H ROESY spec-

trum. The protons oriented downwards resonate at low fre-

quencies (e.g. the endo,ortho proton signal at δ ϭ 5.40 ppm)

because of their location in the anisotropy cone of the xan-

thene backbone (vide supra).

Scheme 4. Representation of relevant NOE contacts (plain arrows)

and chemical exchanges (broken arrows) within the allyl ligand in

12a and 12b from ROESY experiments; for the sake of clarity,

NOEs associated with scalar couplings are not shown

The ¹H NMR signals of both minor isomers 12c and 12d

could hardly be detected in the spectra because of overlap-

ping of the signals of the major isomers 12a and 12b. How-

ever, in view of related (1,3-diphenylallyl)palladium com-

plexes,^[3]it seems highly probable that they are syn,anti iso-

mers (vide infra). Such syn,anti (1,3-diphenylallyl)palladium

complexes are generally less stable than syn,syn isomers be-

cause of unfavorable Ph_antiϪH_antisteric interactions.

1

From the integration of the respective H NMR signals

of 12a and 12b in CD₂Cl₂, we observed that the [12a]/[12b]

ratio in solution increased with the temperature and that a

plot of lnK (K ϭ [12a]/[12b]) vs. 1/T is a straight line (Fig-

ure 5). This allowed us to estimate the thermodynamic data

of the equilibrium, viz. ∆H° ഠ 4 kJ·mol^Ϫ1and ∆S° ഠ 18

J·mol^Ϫ1·K^Ϫ1. Upon warming form 193 to 298 K, the inten-

sity of the AB system at δ ഠ 28 and 39 ppm in the ³¹P

NMR spectrum increased, whereas that of the AB pattern

at δ ഠ 31Ϫ33 and 40Ϫ42 ppm decreased (Figure 3). We

can, therefore, assign the first couple of signals to 12a and

the second to 12b, by comparison with the variable-tem-

Figure 4. Section of the ¹H ROESY spectrum of 12 recorded in

CD₂Cl₂at 193 K; a section of the aliphatic region is represented

horizontally and the three allylic proton signals of 12b vertically

1

perature H NMR spectroscopic data.

1

δ ϭ 0.85 ppm can be observed. However, 12a and 12b can-

The broadening of the H and ³¹P{¹H} NMR signals of

not be differentiated by the left-side lateral allylic proton 12 upon increasing the temperature from 193 K to room

(12a: δ ϭ 4.69 ppm; 12b: δ ϭ 4.92 ppm) because they both temperature is characteristic of fluxional processes. We tried

display NOEs with the proximal, left-side methyl group and to record a high-temperature (323 K) ³¹P{¹H} NMR spec-

methylidyne group.

trum of 12 in CD₃NO₂but unfortunately it decomposed

The low frequency shift of some of the protons is consist- upon heating before any narrowing of the signals could be

ent with their positioning in the anisotropy cone of the aro- observed. We envisioned that the equilibria between the

matic rings. The exo central allylic proton resonates in the isomers of 12 could be visualised by low-temperature 2D

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

www.eurjic.org

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

3993

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

Scheme 5. Hypothetical conformational equilibrium between 12a

and 12b accounting for the phosphorus nuclei exchanges (broken

arrows) observed by ³¹P ROESY

and 12b (Scheme 5), where the xanthene backbone moves

upwards rotating around the P···P axis (flipϪflop). Very

similar NMR observations of a Pd(diphosphane)(η³-allyl)

complex have been reported.^[20]No exchange cross-peaks

were found in the ³¹P ROESY spectrum between the signals

of the major syn,syn and the minor hypothetically syn,anti

isomers, which means that η³Ϫη¹Ϫη³isomerization is too

Figure 5. ln K vs. 1/T curve (K ϭ [12a]/[12b] in CD₂Cl₂)

exchange NMR spectroscopy. No exchange cross-peaks

involving allylic protons were found in the ¹H ROESY spec-

trum, probably because the exchange rates are too low at

193 K. However, we could observe two cross-peaks which

reveal very selective chemical exchanges of phosphorus

nuclei in a ³¹P{¹H} ROESY spectrum recorded at 263 K

(Figure 6). This can be interpreted as the consequence of a

conformational equilibrium between the major isomers 12a

slow to be observed at 263 K. It is plausible that this pro-

cess, which involves PdϪC bond breaking, is slower than a

conformational equilibrium.

A

¹³C{¹H} NMR spectrum of 12 was recorded at 263 K

in CD₃NO₂, a solvent in which the complex is reasonably

soluble (a spectrum recorded in CD₂Cl₂resulted in a poor

signal-to-noise ratio). Although the spectrum is compli-

cated because many species are present in solution, we as-

signed the most intense signals to 12a which is by far the

predominant species at this temperature. One of the

terminal allylic carbon atoms has olefin-like character^[3b]

with a high-frequency chemical shift of δ ϭ 100.2 ppm. As

in the cyclohexenyl series, the ∆δ_{C terminal allyl}value is much

larger with Duxantphos (δ ϭ 22.4 ppm) than with Duphos

(δ ϭ 10.3 ppm^[7]). Indeed, δ ϭ 22.4 ppm is particularly high

for a (1,3-diphenylallyl)Pd complex chelated by two donor

atoms of the same type.^[3b]The terminal allyl carbon signals

of 12a are doublets (as in the cyclohexenyl analogue 11)

and the coupling constants with their respectively trans-po-

sitioned phosphorus atom are quite different (δ

ϭ

77.8 ppm, 30 Hz vs. δ ϭ 100.2 ppm, 20 Hz). All these obser-

vations are consistent with an allyl deformation or tilt

which was suggested by the values of the H-H coupling

constants within the allyl (vide supra).

Like 12, complex [Pd(R,R)-2](η³-allyl)](SbF₆) (13) could

not be analysed by X-ray crystallography. Instead, a de-

tailed NMR spectroscopic analysis was carried out. Its low-

temperature ³¹P{¹H} NMR spectrum revealed the presence

of exo and endo isomers in very different proportions, i.e.

with a ratio Ͼ 10 (Figure 7). At 183 K, the major com-

pound was characterised by an AB system at δ ϭ 33.2 and

34.3 ppm (J_P,Pϭ 40 Hz), whereas only one doublet of the

minor species was visible at δ ϭ 31.3 ppm (J_P,Pϭ 40 Hz).

Upon heating, we observed a coalescence of the signals

which led to a single averaged AB system at room tempera-

ture. The coalescence temperature can be roughly estimated

as 223 K. This suggests that the energy barrier of the

fluxional process is lower for 13 than for 12. A straightfor-

Figure 6. ³¹P ROESY spectrum of 12 recorded in CD₂Cl₂at 263 K;

exchange cross-peaks between 12a and 12b are designated by ar-

rows

3994

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

a methyl group and a methylidyne group, whereas the allylic

anti or central protons show no significant NOE with any

of the Duxantphos protons.

We found two distinct chemical exchange (positive) cross-

peaks in the ROESY spectrum between both ortho-xan-

thene protons and also between two methyl groups

(Scheme 6). Other exchange cross-peaks involving Dux-

antphos protons could not be distinguished because the

chemical shifts of the exchanging nuclei were too close. Be-

sides, the four terminal allyl protons all exchange with each

other. All these exchange cross-peaks can be interpreted as

the result of a flipϪflop of the diphospholane ligand similar

to that shown in Scheme 5 (leading to the minor isomer),

followed by an isomerization back to the major species

through a classical η³Ϫη¹Ϫη³allyl rearrangement mecha-

nism.^[1b]

Figure 7. 121.5 MHz ³¹P{¹H} NMR spectra of [Pd{(R,R)-

Duxantphos}(C₃H₅)](SbF₆) (13) in CD₂Cl₂in the range

183Ϫ298 K

Discussion

The purpose of this discussion is to try to establish a link

between the structural information provided by the analysis

of allyl palladium complexes and some of our most valuable

results of palladium-catalysed allylic alkylations with di-

phospholane ligands, i.e. with substrates 3a and 3c.

ward explanation based on the steric effect of the bulky 1,3-

diphenylallyl moiety can be proposed.

It is highly probable that the major species at low tem-

perature is also largely predominant at room temperature,

i.e. above the coalescence temperature. Therefore, the diag-

As a result of the bite angle of the diphospholane, the

conformation of the scaffold generates exo/endo isomerism

in the Duxantphos complexes 11Ϫ13 and both forms are

in equilibrium. In contrast, Duphos maintains a pseudo-C₂

symmetry when it is coordinated to the metal atom and

there is no such isomerism. We have represented, in

Scheme 7, some simplified front views of the allyl palladium

complexes, in analogy with previously reported represen-

tations.^[4]This allows us to visualise the interactions of the

proximal methyl groups of the (R,R)-diphospholane ligands

with the allyl moiety. The proximal methyl groups [(a) ϭ

down; (b) ϭ up] have dramatically different positions de-

pending on the ligand: with (R,R)-Duphos, (a) is on the

right and (b) on the left, whereas the reverse is true with

(R,R)-Duxantphos. However, this picture does not account

for the distances between the methyl groups and the allyl

plane which are shorter with Duxantphos than with

Duphos (vide supra).

1

nostic cross-peaks observed in the H ROESY NMR spec-

trum recorded at room temperature must be characteristic

of that major isomer. Most of the protons of Duxantphos

in 13 could be identified with the help of that spectrum,

in connection with key Me···H_arand CH···H_arNOE cross-

peaks. Several indications tend to show that it corresponds

to the exo isomer shown in Scheme 6. In an endo isomer, a

strong NOE interaction between H_anti(δ ϭ 2.55 ppm) and

the phospholane methyl group (δ ϭ 1.33 ppm) on the right

side would be expected, which is not the case. Furthermore,

by comparison with reported [Pd(trialkylphosphane)₂-

(allyl)]^ϩcomplexes,^[21]the chemical shifts of the anti pro-

tons are rather low, possibly because they are located in the

anisotropy cone of the backbone aromatic rings in the exo

conformation. Unfortunately, this assignment is not unam-

biguous, since each allylic syn proton displays NOEs with

Scheme 7. Front-view representation of [Pd(diphospholane)-

(allyl)]^ϩcomplexes; black circles ϭ proximal methyl groups; gray

circle ϭ palladium centre

Scheme 6. Representation of relevant NOE contacts (plain arrows)

and chemical exchanges (broken arrows) in 13 from ¹H ROESY;

for the sake of clarity, NOEs associated with scalar couplings are

not shown

The example of the cyclohexenyl complex 11 {or of

[Pd{(R,R)-Duthixantphospholane}(cyclohexenyl)]^{ϩ [4a]}

}

indicates that a methyl group (b) cannot accommodate an

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

www.eurjic.org

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

3995

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

anti substituent on the right in the endo isomer. Therefore,

an endo isomer of 11 could not be detected. On the other

hand, when the allyl group is syn,syn-substituted with large

phenyl groups (complex 12), the exo and endo forms coexist

in comparable proportions. Both undergo some internal

steric congestion, as illustrated by the restricted rotation of

some phenyl groups because of the proximity of phosphol-

ane methyl groups (Scheme 4, vide supra). syn,anti 1,3-Di-

phenylallyl complexes are normally destabilised by

H

_antiϪPh_antisteric repulsion but can be present in signifi-

cant quantity with a large, intrusive ancillary ligand.^[22]In-

deed, minor isomers of 12, which are very probably syn,anti,

coexist with the major syn,syn isomers, whereas

[Pd(Duphos)(η³-1,3-diphenylallyl)^ϩis exclusively syn,syn,^[7]

in connection with the fact that Duphos is less intrusive

than Duxantphos. When Ph is replaced by a smaller Me

on the Duxantphos complex, the syn,anti species become

predominant.^[4b]In brief, it appears that the proportion of

each stereoisomer of a given allyl complex, whether it is

Scheme 8. Interpretation of the selectivity of the nucleophilic attack

of the PdϪcyclohexenyl intermediate with (A) (R,R)-1 or (B)

(R,R)-2 as ligands; Nu ϭ CH(COOMe)₂

related to exo/endo or to syn/anti isomerism, is the result of C1ϪH and the methyl group (a) in the Pd⁰Ϫolefin species,

a subtle balance between various steric effects. whereas no such strain hinders a clockwise rotation of the

A general trend in our catalytic results in the field of allyl group (Scheme 8B). Indeed, compound (S)-4 is formed

allylic alkylation is a higher reactivity of ligand 2 compared with a high ee (83%) in the catalytic process run with ligand

with ligand 1. Here, the reaction rate increases with the bite (R,R)-2.

angle. Such an effect has already been reported in Pd-cata-

It has been reported that in palladium-catalysed asym-

lysed allylic alkylations, but it is not always the rule.^[15,23]It metric allylic alkylation, a twist of the allyl group (observed

can be interpreted as the consequence of a greater stability by X-ray crystallography) toward a product-like state

of the Pd⁰Ϫalkene complex which results from the nucleo- might influence the regioselectivity of the nucleophilic

philic attack on a Pd^IIϪallyl species, assuming that it is the attack.^[16,24c]However, in complex [Pd{(R,R)-Duphos}-

rate-limiting step of the catalytic cycle.^[24a]The origin of this (cyclohexenyl)]^ϩwhich is the mirror image of 10, the allyl

stabilisation can be interpreted as the result of a decrease of group is twisted anticlockwise according to the crystal

the HOMO energy level, as suggested by Walsh dia- structure of 10. Nevertheless, a clockwise rotation is slightly

grams.^[6a]

favoured. Hence, this is another argument against an early

With substrates that generate a symmetrically substituted transition state with these ligands.

η³-allyl ligand (viz. substrates 3aϪd), the enantioselectivity

With the encumbered substrate 3a, chirality transfer was

of the catalytic process arises from the regioselectivity of very efficient in the catalytic reaction. The configurations

the nucleophilic attack where the Pd^IIϪη³-allyl complex is of 4a were found to be opposite with both diphospholane

converted into a Pd⁰Ϫη²-alkene complex. An interpretation ligands, obviously because the positions of the proximal

of the origin of the enantioselectivity is provided by the methyl groups (a) and (b), which undergo repulsive interac-

concept of preferential rotation of the allyl group,^[4,9]also tion with the phenyl substituents, are inverse (Scheme 9). If

referred to as torquoselectivity.^[25]This interpretation is the nucleophile attacks the C3 atom of [Pd{(R,R)-

based on the hypothesis of a medium to late transition Duphos}(cyclohexenyl)]^ϩ, a very severe steric clash occurs

state^[24]in which unhindered, thermodynamically more between barrier (b) and the left-side phenyl group and, to

stable Pd⁰Ϫolefin products would be favoured. In the pre- a lesser extent, between barrier (a) and the right-side phenyl

sent case, enantiodiscrimination is governed by steric inter- group (Scheme 9A). Nucleophilic attack at C1 ac-

actions between the proximal methyl groups and the coordi- companied with anticlockwise rotation of the allyl group

nated alkene product (Schemes 7 and 8).

does not experience a high energy barrier related to steric

An explanation that accounts for the catalytic superiority strain. This is in agreement with the formation of (S)-4a

of 2 over 1 with 3c as the substrate is given in Scheme 8. In with 97% ee.

Scheme 8A, the methyl group (a) of (R,R)-Duphos is lo-

Things are more complicated with Duxantphos, with 4

cated low and far from the allylic carbon atoms and inter- diastereomers of the 1,3-diphenylallyl complex 12 being de-

acts very little with the cyclohexenyl moiety. Therefore, no tected by ³¹P{¹H} NMR spectroscopy (Scheme 9B). It is

major steric clash is expected whether the allyl group rotates reasonable to suppose that the minor isomers 12c and 12d

clockwise or anticlockwise. This correlates well with the are the two syn,anti species shown in Scheme 9 (among four

poor enantioselectivity of the reaction (5% ee). On the con- possible isomers), where no major steric clash exists be-

trary, the situation with (R,R)-Duxantphos is the same as tween the proximal methyl groups and the phenyl substitu-

with (R,R)-Duthixantphos.^[4]That is, an anticlockwise ro- ents. cis-4a was not detected in the catalysed reaction. This

tation is unfavourable because of the steric strain between implies that anticlockwise rotation upon nucleophilic attack

3996

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

aware that the interpretation proposed here is qualitative

since we know nothing of the relative nucleophilic attack

rate on each diastereomer of 12. In particular, it looks prob-

able that this rate is superior for 12a than for 12b: in the

latter, a clockwise rotation may be somewhat hindered by

the methyl group (a).

Conclusion

In this paper we have compared the catalytic and struc-

tural properties of the two chiral diphospholane ligands co-

ordinated to Pd, viz. 1 and 2 which differ in their bite angle.

Palladium-catalysed allylic alkylation tests have confirmed

the success of xanthene-based diphospholane with the

‘‘symmetrical’’ allylic substrates 3aϪd^[4]but this success was

found to be somewhat limited. Attempts to use other kinds

of substrates or nucleophiles, for example, led to modest

selectivities at best. However, a strong positive bite angle

effect was found in allylic alkylation with small, sterically

unhindered substrates such as 3bϪd, where Duxantphos

was much superior to Duphos in terms of reactivity and

especially enantioselectivity.

Investigations of the 3D structures of the allyl complexes

10Ϫ13 were carried out by NMR spectroscopy and, when-

ever possible, by X-ray crystallography. Overall, these

analyses have provided accurate indications about the struc-

tural differences between Duphos and Duxantphos. The in-

crease in the bite angle induces a loss of C₂symmetry of the

ligand upon coordination and renders its proximal methyl

groups more intrusive towards the allyl group. These differ-

ences have major consequences on the nature and distri-

bution of the PdϪallyl intermediates and, furthermore, on

the intimate mechanism of catalytic selectivity which is

Scheme 9. Interpretation of the selectivity of the nucleophilic attack

of the PdϪ1,3-diphenylallyl intermediate with (A) (R,R)-1 or (B)

(R,R)-2 as ligands; Nu ϭ CH(COOMe)₂

of the syn,anti allyl complexes is forbidden because of steric based on steric factors. We applied the preferential rotation

repulsion of one of the phenyl groups with the methyl group concept which proved to be a very suitable model to ac-

(a) (exo) or (b) (endo). This raises the question as to why a count for the catalytic results. With (R,R)-Duxantphos, we

sterically hindered movement of the allyl group is forbidden found that a clockwise rotation in the nucleophilic attack

and not just slowed down. We propose two arguments to step is the preferred, and sometimes the exclusive motion

explain this: (1) from the comparison with the Duxantphos of the allyl group. (R,R)-Duphos, on the other hand, leads

cyclohexenyl complex, it appears that the syn substituent is to anticlockwise rotation but only if the allyl group carries

a bulky phenyl group and not a proton and (2), as stated bulky phenyl substituents. In conclusion, the comparison of

above, the proximal methyl groups of Duxantphos are Duphos (1) with Duxantphos (2) emphasises the advan-

much closer to the allyl plane than those of Duphos. Since tages of large bite angle ligands.

the environment of C1 is the same in the exo,syn,syn and

the exo,syn,anti species represented in Scheme 9, an antic-

lockwise rotation of the exo,syn,syn allyl group must also be

forbidden. The same reasoning is valid for the endo species.

Hence, only clockwise rotations occur in the four isomers

of 12, in analogy with the interpretation reported for the

1,3-dimethylallyl system.^[4b]Both major syn,syn isomers

lead to the (R) alkylated product, whereas the minor syn,-

anti species lead to the (S) product. Our interpretation

based on the hypothesis of CurtinϪHammett con-

ditions^[24a]is in agreement with the high measured ee values

(85Ϫ89%) of (R)-4a. In comparison with the case of

Duphos, the presence of syn,anti catalytic intermediates is

Experimental Section

General Procedures: All experiments were carried out under nitro-

gen or argon using a vacuum line or Vacuum Atmospheres glove-

box equipped with a Dri-Train HE-493 inert gas purifier. Pentane,

benzene, diethyl ether, and THF were distilled from sodium and

benzophenone whereas acetonitrile and dichloromethane were dis-

tilled over calcium hydride under nitrogen immediately before use.

Column chromatography was carried out on Merck silica gel 60,

70Ϫ230 mesh ASTM and thin layer chromatography on

MachereyϪNagel 0.25 mm Polygram^Sil G/UV₂₅₄precoated silica

gel plates. Commercial reagents were used as received: Strem:

detrimental to the enantioselectivity. We are, however, (R,R)- and (S,S)-Duphos, AgBF₄; Aldrich: dimethyl malonate,

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

www.eurjic.org  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 3997

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

N,O-bis(trimethylsilyl)acetamide (BSA), potassium acetate, (rac)-2-

methyl-1-tetralone, allyl acetate, 2,3-dihydrofuran, phenyl triflate, dimethyl (cyclohex-2-en-1-yl)malonate (4c) and dimethyl (cyclo-

N,N-diisopropylethylamine, europium tris[3-(heptafluoropropylhy-

pent-2-en-1-yl)malonate (4d),^[36]dimethyl (E)-(4-phenylbut-3-en-2-

)malonate (4a),^[34]dimethyl (E)-(pent-3-en-2-yl)malonate (4b),^[35]

droxymethylene)-(ϩ)-camphorate] [Eu(hfc)₃], AgSbF₆; Prolabo: n- yl)malonate (6a) and dimethyl (E)-(1-phenylbut-2-en-1-yl)malonate

heptane. NaH (60% in mineral oil; Fluka) was washed three times

(7a),^[8,9]dimethyl (E)-(5-methylhex-3-en-2-yl)malonate (6b) and di-

methyl (E)-(2-methylhex-4-en-3-yl)malonate (7b).^[35]The ee values

with pentane, dried in vacuo, and stored in the glovebox. (R,R)-

Duxantphos was prepared according to ref.^[4]The compounds were measured with the following methods: 4a, chiral HPLC (Chir-

listed hereafter were synthesised according to reported procedures:

(rac)-(E)-1-acetoxy-1,3-diphenylprop-2-ene (3a),^[26](rac)-(E)-2-ace-

toxypent-3-ene (3b),^[27](rac)-1-acetoxycyclohex-2-ene (3c),^[28](rac)-

1-acetoxycyclopent-2-ene (3d),^[29](rac)-(E)-2-acetoxy-4-phenylbut-

3-ene (5a),^[8](rac)-(E)-3-acetoxy-2-methylhex-4-ene (5b),^[30]bis[(η³-

alcel AD, 1 mL·min^Ϫ1; eluent n-hexane/iPrOH, 95:5, v/v); 4b and

4c, chiral GC; 4d, 6a, 7a, 6b, and 7b, ¹H NMR in CDCl₃using

Eu(hfc)₃as a chiral shift reagent.

General Procedure for Palladium-Catalysed Allylic Alkylation with

2-Methyl-1-Tetralone: The procedure was adapted from ref.^[13]To

a solution or suspension of base (1.85 mmol) in THF (2 mL) was

added 2-methyl-1-tetralone (150 mg, 0.94 mmol). A mixture of li-

gand 2 (4.4 mg, 10 µmol) and [Pd(η³-C₃H₅)Cl]₂(1.8 mg, 5 µmol)

in THF (1 mL) was stirred at room temperature for 20 min. To

the catalyst solution was added allyl acetate (187 mg, 1.87 mmol),

followed by the nucleophile solution. The conversion was period-

ically monitored by GC. When the reaction was complete, the

mixture was added to a 0.2  aqueous HCl solution (25 mL) and

extracted with Et₂O (3 ϫ 30 mL). The extract was treated with

saturated NaHCO₃(aq.) and dried with MgSO₄. The solvent was

removed under reduced pressure and the crude 2-allyl-2-methyl-

1-tetralone purified by column chromatography on silica gel with

pentane/EtOAc (v/v, 9:1; R_fϭ 0.52). The ee of the product was

determined by chiral HPLC (Chiralcel OD, 0.7 mL·min^Ϫ1; eluent

n-hexane/iPrOH, 99.9:0.1, v/v) and its absolute stereochemistry was

assigned by comparison with ref.^[13]

allyl)chloropalladium],^[31]

bis[(η³-cyclohexenyl)chloropalladi-

um],^[32]bis[(η³-1,3-diphenylallyl)chloropalladium].^[24f]NMR spec-

tra were recorded at room temperature (unless otherwise indicated)

with Bruker spectrometers. ¹H NMR spectra were recorded at

300.16 MHz (AC-300 instrument) or 500.13 MHz (ARX-500) and

referenced to SiMe₄. The assignments and measurements of coup-

1

ling constants were supported by irradiations. H{³¹P}, H COSY,

and/or ³¹P{¹H} COSY spectra were recorded with the ARX-500

instrument. ¹³C{¹H} NMR spectra (broadband-decoupled) were

recorded at 75.48 MHz (AC-300), 100.62 MHz (AM-400), or

125.76 MHz (ARX-500) and referenced to SiMe₄. The assignments

were supported by DEPT-135°. ³¹P{¹H} spectra (broadband-de-

coupled) were recorded at 121.51 MHz (AC-300) or 202.46 MHz

(ARX-500) and referenced to 85% aqueous H₃PO₄. The phase-sen-

sitive 2D ROESY spectra^[33]were recorded with the ARX-500 spec-

trometer using mixing times of 0.25 s (¹H) or 0.1 s (³¹P{¹H}). FAB

MS spectra were carried out at the corresponding facilities at the

´ ´

´

Federation de Recherche de Chimie, Universite Louis Pasteur,

Strasbourg and elemental analyses by the Service Central d’Ana-

lyses, CNRS, Vernaison. Optical rotations were checked with a

PerkinϪElmer Otopol III polarimeter (λ ϭ 589 nm). GC analyses

were performed with a Hewlett Packard 5890 series II apparatus

equipped with a 30 m ϫ 0.25 mm SGE 25QC2 capillary column

(for conversion measurements) or with a 30 m ϫ 0.25 mm Restek

Rt-βDEXse capillary column (for enantiomeric excess determi-

nations). HPLC analyses were carried out with a Hewlett Packard

series 1100 apparatus equipped with 250 mm ϫ 4.6 mm Daicel Chi-

ralcel^AD or OD analytical columns.

General Procedure for the Heck Reaction: The procedure was ad-

apted from ref.^[14]A mixture of diphospholane ligand (24 µmol)

and the Pd source (12 µmol) in benzene (2 mL) was stirred at room

temperature for 20 min. To the catalyst solution were successively

added a solution of PhOTf (89 mg, 0.394 mmol) in benzene (1 mL),

NiPr₂Et (168 mg, 1.30 mmol), 2,3-dihydrofuran (155 mg,

2.21 mmol) and additional benzene (2 mL). The mixture was

heated to reflux for 48 h. The solvent was removed under reduced

pressure and the crude products were purified by column chroma-

tography on silica gel with pentane/EtOAc (9:1, v/v). The absolute

stereochemistry of the product was determined by comparing the

sign of its optical rotation with data from ref.^[14]The ee values were

measured by chiral GC.

General Procedure for Palladium-Catalysed Allylic Alkylation with

Dimethyl Malonate: A mixture of the diphospholane ligand (12

µmol) and [Pd(η³-C₃H₅)Cl]₂(1.8 mg, 5 µmol) in the appropriate

solvent (CH₂Cl₂or THF, 2 mL) was stirred at room temperature

for 20 min. When BSA was used as a base, a suspension of KOAc

(1 mg, 10 µmol) in the solvent (1 mL) was added to the catalyst

solution, followed by a mixture of dimethyl malonate (264 mg,

2 mmol) and BSA (407 mg, 2 mmol) in the solvent (4 mL) and a

solution of the substrate (1 mmol) in the solvent (1 mL). When

NaH was used as a base, a solution of dimethyl sodiomalonate

was prepared by allowing a solution of dimethyl malonate (264 mg,

2 mmol) in THF (1 mL) to react with a suspension of NaH (48 mg,

2 mmol) in THF (4 mL). A solution of the substrate (1 mmol) in

THF (1 mL) was added to the catalyst solution, followed by the

nucleophile solution. In both cases, the conversion was periodically

monitored by GC. When the reaction was complete, water (25 mL)

was added, and the organic phase was extracted with Et₂O (3 ϫ

30 mL). The extract was dried with MgSO₄and the solvent was

removed under reduced pressure. The crude product was purified

by column chromatography on silica gel with pentane/Et₂O (4:1, v/

[Pd{(S,S)-Duphos}(η³-cyclohexenyl)](BF₄) (10): To a solution of

[Pd(η³-C₆H₉)Cl]₂(35 mg, 0.078 mmol) in acetonitrile (4 mL) was

added dropwise in the dark at room temperature a solution of

AgBF₄(35 mg, 0.180 mmol) in acetonitrile (3 mL). After 10 min of

stirring, the suspension was filtered through Celite and a suspen-

sion of (S,S)-Duphos (48 mg, 0,157 mmol) in acetonitrile (1 mL)

was added to the colourless filtrate. A pale orange solution was

obtained after 30 min of stirring. The solvent was evaporated in

vacuo to afford a brown powder which was washed with diethyl

ether. The product was recrystallized from dichloromethane/pen-

tane as colourless crystals. Yield: 50 mg (55%). ³¹P{¹H} NMR

2

(CD₂Cl₂, 121 MHz): δ ϭ 68.2 (d, J_P,Pϭ 36 Hz, 1 P), 71.9 (d,

²J_P,Pϭ 36 Hz, 1 P) ppm. ¹H NMR (CD₃CN, 300 MHz): δ 0.62

3

(dd, J_H,Hϭ 7.0, J_H,Pϭ 15.6 Hz, 3 H, CH₃), 0.85 (dd, J_H,H

ϭ

3

7.0, J_H,Pϭ 15.6 Hz, 3 H, CH₃), 1.12 (dd, J_H,Hϭ 7.0, J_H,P

ϭ

3

20.8 Hz, 3 H, CH₃), 1.35 (dd, J_H,Hϭ 7.1, J_H,Pϭ 20.5 Hz, 3 H,

3

CH₃), 1.60Ϫ3.20 (m, 18 H, CH ϩ CH₂), 5.33 (t, J_H,Hϭ 7.1 Hz,

1 H, H²), 5.90 (m, 2 H, H¹ϩ H³), 7.60Ϫ7.90 (m, 4 H, H_ar) ppm.

v). By comparison with literature data, the alkylation products were ¹³C{¹H} NMR (CD₂Cl₂, 75 MHz): δ ϭ 14.8 (s, 2 CH₃), 18.4 (d,

identified and, if necessary, their absolute stereochemistry was de-

termined by polarimetry: dimethyl (E)-(1,3-diphenylprop-2-en-1-yl-

2

²J_C,Pϭ 12 Hz, CH₃), 18.7 (d, J_C,Pϭ 12 Hz, CH₃), 20.4 (s, CH₂),

2

28.1 (d, J_C,Pϭ 5 Hz, CH₂), 29.4 (d, J_C,Pϭ 5 Hz, CH₂), 36.2 (d,

3998

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

2

3

²J_C,Pϭ 5 Hz, CH₂), 36.5 (d, J_C,Pϭ 5 Hz, CH₂), 37.2 (s, CH₂),

Scheme 3): δ ϭ Ϫ0.08 (dd, J_H,Hϭ 7.0, J_H,Pϭ 14.9 Hz, 3 H,

1

3

37.5 (s, CH₂), 38.6 (d, J_C,Pϭ 24 Hz, CH), 40.7 (d, J_C,Pϭ 22 Hz, CH_{3 phospholane}), 0.85 (dd, J_H,Hϭ 7.4, J_H,Pϭ 22.2 Hz, 3 H,

1

3

CH), 41.0 (d, J_C,Pϭ 26 Hz, CH), 41.3 (d, J_C,Pϭ 24 Hz, CH),

81.9 (dd, J_C,Pϭ 29, J_C,Pϭ 3 Hz, C_{allyl terminal}), 87.1 (dd, J_C,P

CH_{3 phospholane}), 0.87 (dd, J_H,Hϭ 7.1, J_H,Pϭ 14.0 Hz, 3 H,

2

3

ϭ

CH_{3 phospholane}), 1.05 (dd, J_H,Hϭ 7.4, J_H,Pϭ 22.2 Hz, 3 H,

2

29, J_C,Pϭ 3 Hz, C_{allyl terminal}), 113.8 (t, J_C,Pϭ 7 Hz, C_{allyl central}), CH_{3 phospholane}), 1.58 (s,

3 H, CH_{3 xanthene}), 1.65 (s, 3 H,

132.6 (s, 2 CH_ar), 134.1 (d, ²J_C,Pϭ 12 Hz, CH_ar), 134.3 (d, J_C,P

ϭ

CH_{3 xanthene}), 4.69 (t, J_H,H

ϭ

³J_H,Pϭ 10.8 Hz, 1 H, H_allyl), 5.38

2

3

12 Hz, CH_ar), 141.2 (m, 2 C_ar) ppm. MS (FAB^ϩ, NBA matrix): m/

(dd, J_H,Hϭ 13.7, J_H,Pϭ 9.4 Hz, 1 H, H_allyl), 6.35 (ddd, J_H,H

ϭ

3

z (%) ϭ 493.0 (100) [M Ϫ BF₄]^ϩ, 412.9 (37) [M ϩ1 Ϫ BF₄

Ϫ

14.0, J_H,Hϭ 10.0, J_H,Pϭ 2.2 Hz, 1 H, H_allyl) ppm. ¹³C{¹H}

NMR (CD₃NO₂, 125 MHz, 263 K), selected data: δ ϭ 14.6 (s,

CH_{3 phospholane}), 15.2 (s, 2 CH_{3 phospholane}), 17.8 (s, CH_{3 phospholane}),

3

C₆H₉]^ϩ. C₂₄H₃₇BF₄P₂Pd (580.71): calcd. C 49.64, H 6.42, P 10.67;

found C 49.59, H 6.34, P 10.80.

1

21.3 (d, J_C,Pϭ 14 Hz, CH), 22.5 (d, J_C,Pϭ 13 Hz, CH), 25.0 (s,

CH_{3 xanthene}), 29.2 (s, CH_{3 xanthene}), 33.2 (s, CH₂), 34.9 (s, CH₂), 35.1

(s, CH₂), 36.3 (s, CH₂), 36.7 (d, ¹J_C,Pϭ 20 Hz, CH), 38.2 (d, ¹J_C,Pϭ

19 Hz, CH), 38.6 [s, C(CH₃)₂], 77.8 (d, ²J_C,Pϭ 30 Hz, C_{allyl terminal}),

[Pd{(R,R)-Duxantphos}(η³-cyclohexenyl)](SbF₆) (11): To a solution

of [Pd(η³-C₆H₉)Cl]₂(20 mg, 0.045 mmol) in THF (2 mL) was ad-

ded dropwise in the dark at room temperature a solution of

AgSbF₆(32 mg, 0.093 mmol) in THF (2 mL). After 15 min of stir-

ring, the suspension was filtered through Celite and a solution of

(R,R)-Duxantphos (42 mg, 0.096 mmol) in THF (2 mL) was added

to the colourless filtrate. A yellow solution was obtained after

30 min of stirring and the solvent was evaporated in vacuo to afford

a yellow powder which was washed with pentane. The product was

2

100.2 (d, J_C,Pϭ 20 Hz, C_{allyl terminal}), 110.4 (s, C_{allyl central}) ppm.

syn,syn,endo-12b: ³¹P{¹H} NMR (CD₂Cl₂, 202.5 MHz, 193 K):

2

δ ϭ 32.8 (d, J_P,Pϭ 57 Hz, 1 P), 40.6 (d, J_P,Pϭ 57 Hz, 1 P) ppm.

¹H NMR (CD₂Cl₂, 500 MHz, 193 K), selected data (see also

3

Scheme 3): δ ϭ 0.80 (dd, J_H,Hϭ 7.2, J_H,Pϭ 14.5 Hz, 3 H,

3

recrystallized from dichloromethane/diethyl ether. Yield: 72 mg

CH_{3 phospholane}), 0.96 (dd, J_H,Hϭ 7.4, J_H,Pϭ 15.1 Hz, 3 H,

(93%). ³¹P{¹H} NMR (CD₂Cl₂, 121 MHz): δ ϭ 33.7 (d, J_P,P

ϭ

2

3

CH_{3 phospholane}), 1.10 (dd, J_H,Hϭ 7.6, J_H,Pϭ 22.4 Hz, 3 H,

2

1

3

41 Hz, 1 P), 35.5 (d, J_P,Pϭ 41 Hz, 1 P) ppm. H NMR (CD₂Cl₂,

300 MHz): δ ϭ 0.06 (m, 1 H, CH₂), 0.24 (m, 1 H, CH₂), 0.46 (m,

1 H, CH₂), 0.60Ϫ0.80 (m, 2 H, CH₂), 1.13 (dd, ³J_H,Hϭ 7.3, ³J_H,Pϭ

CH_{3 phospholane}), 1.32 (s, 3 H, CH_{3 xanthene}), 1.36 (dd, J_H,Hϭ 7.3,

³J_H,Pϭ 18.8 Hz, 3 H, CH_{3 phospholane}), 1.88 (s, 3 H, CH_{3 xanthene}),

3

4.92 (t, J_H,H

ϭ

³J_H,Pϭ 11.1 Hz, 1 H, H_allyl), 5.03 (dd, J_H,H

ϭ

3

14.7 Hz, 3 H, CH_{3 phospholane}), 1.25 (dd, J_H,Hϭ 7.6, J_H,P

ϭ

13.0, J_H,Pϭ 9.2 Hz, 1 H, H_allyl), 5.21 (t, J_H,Hϭ 12.5 Hz, 1 H,

3

20.5 Hz, 3 H, CH_{3 phospholane}), 1.32 (dd, J_H,Hϭ 7.6, J_H,P

ϭ

H_allyl) ppm.

19.3 Hz, 3 H, CH_{3 phospholane}), 1.20Ϫ1.75 (m, 3 H, CH₂), 1.41 (s, 3

Minor Isomer 12c: ³¹P{¹H} NMR (CD₂Cl₂, 202.5 MHz, 193 K):

δ ϭ 35.3 (d, J_P,Pϭ 45 Hz, 1 P), 37.2 (d, J_P,Pϭ 45 Hz, 1 P) ppm.

3

H, CH_{3 xanthene}), 1.79 (s, 3 H, CH_{3 xanthene}), 1.80 (dd, J_H,Hϭ 7.3,

2

³J_H,Pϭ 14.2 Hz, 3 H, CH_{3 phospholane}), 2.07Ϫ2.40 (m, 5 H, CH₂),

2.63 (m, 1 H, CH₂), 2.80Ϫ3.00 (m, 3 H, CH), 3.09 (m, 1 H, CH),

Minor Isomer 12d: ³¹P{¹H} NMR (CD₂Cl₂, 202.5 MHz, 193 K):

3

5.20 (t, J_H,Hϭ 7.1 Hz, 1 H, H²), 5.60Ϫ5.75 (m, 2 H, H¹ϩ H³),

2

δ ϭ 28.3 (d, J_P,Pϭ 46 Hz, 1 P), 36.0 (d, J_P,Pϭ 46 Hz, 1 P) ppm.

7.13Ϫ7.30 (m, 2 H, H_ar), 7.36 (m, 1 H, H_ar), 7.50Ϫ7.62 (m, 3 H,

H_ar) ppm. ¹³C{¹H} NMR (CD₃CN, 75 MHz): δ ϭ 14.5 (s,

[Pd{(R,R)-Duxantphos}(η³-allyl)](SbF₆) (13): To

a

solution of

2

[Pd(η³-C₃H₅)Cl]₂(35 mg, 0.096 mmol) in CH₂Cl₂(4 mL) was ad-

ded dropwise in the dark at room temperature a solution of

AgSbF₆(63 mg, 0.183 mmol) in CH₂Cl₂(1 mL). After 10 min of

stirring, the suspension was filtered through Celite and a solution

of (R,R)-Duxantphos (80 mg, 0.182 mmol) in CH₂Cl₂(2 mL) was

added to the yellow filtrate. A brown solution was obtained after

10 min of stirring. The volume was reduced in vacuo and the prod-

uct was precipitated by slow addition of pentane and collected on

a frit as a brown powder. Yield: 115 mg (77%). ³¹P{¹H} NMR

CH_{3 phospholane}), 17.5 (d, J_C,Pϭ 5 Hz, CH_{3 phospholane}), 19.4 (d,

²J_C,Pϭ 9 Hz, CH_{3 phospholane}), 20.0 (s, CH₂), 20.2 (d, ²J_C,Pϭ 14 Hz,

2

CH_{3 phospholane}), 24.2 (s, CH_{3 xanthene}), 26.9 (d, J_C,Pϭ 5 Hz, CH₂),

2

27.2 (d, J_C,Pϭ 5 Hz, CH₂), 32.1 (s, CH₂), 32.4 (s, CH_{3 xanthene}),

1

32.6 (s, CH₂), 32.9 (d, J_C,Pϭ 22 Hz, CH), 33.6 (s, CH₂), 33.9 (d,

1

¹J_C,Pϭ 20 Hz, CH), 34.0 (d, J_C,Pϭ 14 Hz, CH), 34.7 (s, CH₂),

1

2

36.8 [s, C(CH₃)₂], 37.0 (d, J_C,Pϭ 27 Hz, CH), 81.3 (d, J_C,P

ϭ

2

28 Hz, C_{allyl terminal}), 93.3 (d, J_C,Pϭ 25 Hz, C_{allyl terminal}), 109.1 (t,

²J_C,Pϭ 5 Hz, C_{allyl central}), 119.2 (d, J_C,Pϭ 27 Hz, C_ar), 120.1 (d,

2

²J_C,Pϭ 23 Hz, C_ar), 125.4 (d, ³J_C,Pϭ 4 Hz, CH_ar), 126.2 (d, ³J_C,Pϭ

2

(CD₂Cl₂, 121 MHz) : δ ϭ 33.1 (d, J_P,Pϭ 38 Hz, 1 P), 33.5 (d,

²J_P,Pϭ 38 Hz, 1 P) ppm. ¹H NMR (CD₂Cl₂, 500 MHz), selected

4 Hz, CH_ar), 128.3 (s, CH_ar), 128.8 (s, 2 CH_ar), 130.9 (s, CH_ar),

1

3

134.4 (s, C_ar), 135.2 (s, C_ar), 153.4 (d, J_C,Pϭ 7 Hz, C_ar), 153.8 (d,

data (see also Scheme 6): δ ϭ 1.21 (dd, J_H,Hϭ 7.3, J_H,P

ϭ

¹J_C,Pϭ 7 Hz, C_ar) ppm. MS (FAB^ϩ, NBA matrix): m/z (%) ϭ 625.2

(68) [M Ϫ SbF₆]^ϩ, 545.1 (100) [M ϩ1 Ϫ SbF₆Ϫ C₆H₉]^ϩ.

3

14.6 Hz, 3 H, CH_{3 phospholane}), 1.33 (dd, J_H,Hϭ 7.5, J_H,P

3

20.2 Hz, 3 H, CH_{3 phospholane}), 1.36 (dd, J_H,Hϭ 7.5, J_H,P

3

19.5 Hz, 3 H, CH_{3 phospholane}), 1.52 (dd, J_H,Hϭ 7.2, J_H,P

[Pd{(R,R)-Duxantphos}(η³-1,3-diphenylallyl)](SbF₆) (12): To a sus-

pension of [Pd(η³-Ph₂C₃H₃)Cl]₂(30 mg, 0.045 mmol) in CH₂Cl₂

(3 mL) was added dropwise in the dark at room temperature a solu-

tion of AgSbF₆(31 mg, 0.090 mmol) in CH₂Cl₂(2 mL). After

20 min of stirring, the suspension was filtered through Celite and

a solution of (R,R)-Duxantphos (40 mg, 0.091 mmol) in CH₂Cl₂

(2 mL) was added to the red filtrate. An orange-brown solution was

obtained after 20 min of stirring. The product precipitated upon

slow addition of pentane. It was then separated from the yellow

supernatant, washed with diethyl ether, and dried. Yield:

70 mg (80%). HRMS (FAB^ϩ, NBA matrix): m/z calcd. for

C₄₂H₄₉OP₂¹⁰⁵Pd 736.23094; found 736.23067.

14.5 Hz, 3 H, CH_{3 phospholane}), 1.52 (s, 3 H, CH_{3 xanthene}), 1.68 (s, 3

H, CH_{3 xanthene}), 1.60Ϫ2.10 (m, 4 H, CH₂), 2.20Ϫ2.40 (m, 3 H,

3

CH₂), 2.51 (m, 1 H, CH₂), 2.55 (dd, J_H,Hϭ 13.6, ³J_H,Pϭ 8.8 Hz,

3

1 H, H_anti), 2.73 (dd, J_H,Hϭ 13.1, J_H,Pϭ 9.6 Hz, 1 H, H_anti),

3

4

2.85Ϫ3.10 (m, 4 H, CH), 4.51 (dddd, J_H,Hϭ 7.4, J_H,Hϭ 2.3,

³J_H,Pϭ 2.2, 0.6 Hz, 1 H, H_syn), 4.66 (dddd, J_H,Hϭ 7.2, J_H,H

ϭ

3

4

2.5, ³J_H,Pϭ 2.9, 0.6 Hz, 1 H, H_syn), 5.18 (tt, ³J_H,Hϭ 13.3, ³J_H,Hϭ

3

4

7.3 Hz, 1 H, H_central), 7.28 (td, J_H,Hϭ 7.7, J_H,Pϭ 1.5 Hz, 1 H,

3

4

H_ar), 7.31 (td, J_H,Hϭ 7.6, J_H,Pϭ 1.5 Hz, 1 H, H_ar), 7.37 (ddd,

³J_H,Hϭ 7.7, J_H,Hϭ 1.4, J_H,Pϭ 6.3 Hz, 1 H, H_ar), 7.46 (ddd,

4

3

³J_H,Hϭ 7.7, J_H,Hϭ 1.2, J_H,Pϭ 7.6 Hz, 1 H, H_ar), 7.57 (ddd,

4

3

³J_H,Hϭ 7.7, J_H,Hϭ 1.3, J_H,Pϭ 0.8 Hz, 1 H, H_ar), 7.59 (ddd,

4

5

syn,syn,exo-12a: ³¹P{¹H} NMR (CD₂Cl₂, 202.5 MHz, 193 K): δ ϭ

³J_H,Hϭ 7.8, ⁴J_H,Hϭ 1.4, J_H,Pϭ 0.7 Hz, 1 H, H_ar) ppm. ¹³C{¹H}

5

2

1

2

27.6 (d, J_P,Pϭ 52 Hz, 1 P), 39.2 (d, J_P,Pϭ 52 Hz, 1 P) ppm. H NMR (CD₂Cl₂, 100 MHz):

δ

ϭ

14.9 (d, J_C,P

ϭ

2 Hz,

2

NMR (CD₂Cl₂, 500 MHz, 193 K), selected data (see also

www.eurjic.org

CH_{3 phospholane}), 15.9 (d, J_C,P

ϭ

4 Hz, CH_{3 phospholane}), 19.3

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 3999

´

G. Malaise, S. Ramdeehul, J. A. Osborn, L. Barloy, N. Kyritsakas, R. Graff

FULL PAPER

Table 5. Crystal and refinement data for compounds 10 and 11

10

11

Empirical formula

Formula mass

Colour

C₂₄H₃₇P₂Pd·BF₄

C₃₃H₄₅OP₂Pd·SbF₆

580.72

861.81

colourless

monoclinic

8.2100(2)

18.6150(7)

colourless

orthorhombic

12.041(1)

12.425(2)

Crystal system

˚

a [A]

˚

b [A]

˚

c [A]

8.5920(3)

22.917(4)

β [°]

97.553(2)

1301.7(1)

90

3428(1)

3

˚

V [A ]

Z

2

1.48

4

1.67

D_calcd.[g cm^Ϫ3

]

F000

596

0.71073

0.874

1728

0.71073

1.459

˚

Wavelength [A]

µ [mm^Ϫ1

]

Space group

P 1 2₁1

KappaCCD

0.15 ϫ 0.12 ϫ 0.12

173

P 2₁2₁2₁

MACH3 Nonius

0.40 ϫ 0.25 ϫ 0.20

173

Diffractometer

Cryst dimens [mm]

Temperature [K]

Radiation

Mo-K_αgraphite-monochromated

Scan mode

hkl limits

θ limits [°]

No. of data measured

No. of data with I Ͼ 3σ(I)

Weighting scheme

No. of variables

R

ϕ scans

θ/2θ

Ϫ10/10, Ϫ26/26, Ϫ12/12

0/15, Ϫ15/0, 0/28

2.5/26.29

3917

2.5/30.56

19463

2887

3318

4F_o/[σ²(F_o) ϩ 0.0004 F_o] ϩ 1.0

4F_o/[σ²(F_o) ϩ 0.0016 F_o]

2

4

2

4

297

397

0.067

0.088

0.0(1)

0.738

1.612

0.025

0.031

Ϫ0.02(3)

0.567

1.096

R_w

Flack parameter

Largest peak in final difference [e·A

GOF

Ϫ3

˚

]

2

(d, J_C,P

ϭ 9 Hz, CH_{3 phospholane}), 19.9 (d, J_C,Pϭ 10 Hz,

Acknowledgments

CH_{3 phospholane}), 25.9 (s, CH_{3 xanthene}), 29.4 (s, CH_{3 xanthene}), 32.0 (s,

We thank Denis Sinou for helpful discussions and Rafael Aguado

and Peter Dierkes for a gift of ligand 2. Financial support from the

Centre National de la Recherche Scientifique is gratefully acknowl-

edged.

CH₂), 32.4 (s, CH₂), 32.7 (s, CH₂), 33.3 (s, CH₂), 33.4 (dd, ¹J_C,P

ϭ

3

1

3

23, J_C,Pϭ 2 Hz, CH), 33.7 (dd, J_C,Pϭ 17, J_C,Pϭ 2 Hz, CH),

1

3

1

34.0 (dd, J_C,Pϭ 19, J_C,Pϭ 3 Hz, CH), 35.3 (dd, J_C,Pϭ 22,

³J_C,Pϭ 3 Hz, CH), 36.1 [s, C(CH₃)₂], 66.8 (dd, ²J_C,Pϭ 26, ²J_C,P

ϭ

2

3 Hz, CH_{allyl terminal}), 72.6 (dd, J_C,P

ϭ

25, J_C,P

ϭ

3 Hz,

CH_{allyl terminal}), 118.2 (dd, ²J_C,Pϭ 27, ⁴J_C,Pϭ 2 Hz, C_ar), 118.5 (dd,

²J_C,Pϭ 24, ⁴J_C,Pϭ 2 Hz, C_ar), 118.7 (t, ²J_C,Pϭ 5 Hz, CH_{allyl central}),

[1]

Reviews and selected papers: ^[1a]C. G. Frost, J. Howarth, J. M.

124.8 (d, J_C,Pϭ 6 Hz, CH_ar), 125.3 (d, ²J_C,Pϭ 5 Hz, CH_ar), 127.3

2

[1b]

J. Williams, Tetrahedron: Asymmetry 1992, 3, 1089Ϫ1122.

3

(s, CH_ar), 127.6 (d, J_C,Pϭ 2 Hz, CH_ar), 127.9 (d, J_C,Pϭ 2 Hz,

B. M. Trost, D. L. Van Vranken, Chem. Rev. 1996, 96,

3

CH_ar), 128.6 (s, CH_ar), 134.1 (d, J_C,Pϭ 3 Hz, C_ar), 134.4 (d,

[1c]

395Ϫ422.

203Ϫ214.

G. Helmchen, J. Organomet. Chem. 1999, 576,

K. H. Ahn, C.-W. Cho, J. Park, S. Lee, Tetra-

³J_C,Pϭ 3 Hz, C_ar), 153.2 (d, J_C,Pϭ 6 Hz, C_ar), 153.4 (d, J_C,P

ϭ

1

[1d]

6 Hz, C_ar) ppm. HRMS (FAB^ϩ, NBA matrix): m/z calcd. for

hedron: Asymmetry 1997, 8, 1179Ϫ1185. ^[1e]A. Saitoh, M. Mis-

awa, T. Morimoto, Tetrahedron: Asymmetry 1999, 10,

C₃₀H₄₁OP₂¹⁰⁵Pd 584.1683; found 584.1689.

1025Ϫ1028. ^[1f]W. Zhang, T. Shimanuki, T. Kida, Y. Nakatsuji,

X-ray Crystallographic Studies: Details of data collection param-

eters and refinement results are listed in Table 5. The structures

were solved by direct methods. Hydrogen atoms were introduced

[1g]

I. Ikeda, J. Org. Chem. 1999, 64, 6247Ϫ6251.

K. R. Campos, J. D. Tedrow, F. E. Michael, M. R. Gagne, J.

D. A. Evans,

´

[1h]

Am. Chem. Soc. 2000, 122, 7905Ϫ7920.

K. Ito, R. Kashi-

˚

as fixed contributors in calculated positions [CϪH ϭ 0.95 A,

wagi, S. Hayashi, T. Uchida, T. Katsuki, Synlett 2001,

[1i]

B(H) ϭ 1.3 B_equiv]. Final difference maps revealed no significant

maxima. All calculations were performed using the Nonius Open-

MoleN package.^[37]Neutral atom scattering factor coefficients and

anomalous dispersion coefficients were taken from a standard

source.^[38]CCDC-236309 (10) and -236310 (11) contain the sup-

plementary crystallographic data for this paper. These data can be

obtained free of charge at www.ccdc.cam.ac.uk/conts/

retrieving.html [or from the Cambridge Crystallography Data Cen-

tre, 12 Union Road, Cambridge CB2 1EZ, UK [Fax: ϩ 44-1223/

336-033; E-mail: deposit@ccdc.cam.ac.uk].

284Ϫ286.

S. J. Greenfield, A. Agarkov, S. R. Gilbertson,

Org. Lett. 2003, 5, 3069Ϫ3072.

[2] [2a]

A. Pfaltz, M. Lautens, in: Comprehensive Asymmetric Ca-

talysis (Eds.: E. Jacobsen, A. Pfaltz, H. Yamamoto), Springer,

Berlin, 1999, vol. II, chapter 24, pp. 833Ϫ884.

[2b]

M. Shiba-

saki, E. M. Vogl, in: Comprehensive Asymmetric Catalysis

(Eds.: E. Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin,

1999, vol. I, chapter 14, pp. 457Ϫ487.

[3] [3a]

P. S. Pregosin, G. Trabesinger, J. Chem. Soc., Dalton Trans.

1998, 727Ϫ734. ^[3b]P. S. Pregosin, R. Salzmann, Coord. Chem.

Rev. 1996, 155, 35Ϫ68.

4000

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eurjic.org

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

Palladium Complexes with Chiral Diphospholane Ligands

FULL PAPER

Strijdonck, J. N. H. Reek, P. C. J. Kamer, P. W. N. M. van

^[4a]P. Dierkes, S. Ramdeehul, L. Barloy, A. De Cian, J. Fischer,

[4]

P. C. J. Kamer, P. W. N. M. van Leeuwen, J. A. Osborn, Angew.

Leeuwen, Eur. J. Inorg. Chem. 1999, 1237Ϫ1241.

[4b]

[24]

Chem. Int. Ed. 1998, 37, 3116Ϫ3118.

S. Ramdeehul, P. Di-

The question of an early or late transition state for this step

[24a]

erkes, R. Aguado, P. C. J. Kamer, P. W. N. M. van Leeuwen, J.

A. Osborn, Angew. Chem. Int. Ed. 1998, 37, 3118Ϫ3121.

M. J. Burk, Acc. Chem. Res. 2000, 33, 363Ϫ372.

has been a matter of debate:

P. B. Mackenzie, J. Wheland,

[24b]

B. Bosnich, J. Am. Chem. Soc. 1985, 107, 2046Ϫ2054.

J.

[5]

[6]

Sprinz, M. Kiefer, G. Helmchen, M. Reggelin, G. Huttner, O.

^[6a]P. Dierkes, P. W. N. M. van Leeuwen, J. Chem. Soc., Dalton

Trans. 1999, 1519Ϫ1529. ^[6b]P. W. N. M. van Leeuwen, P. C. J.

Walter, L. Zsolnai, Tetrahedron Lett. 1994, 35, 1523Ϫ1526. ^[24c]

A. Togni, U. Burckhardt, V. Gramlich, P. S. Pregosin, R. Salz-

[24d]

Kamer, J. N. H. Reek, P. Dierkes, Chem. Rev. 2000, 100,

mann, J. Am. Chem. Soc. 1996, 118, 1031Ϫ1037.

B. M.

[6c]

2741Ϫ2769.

Z. Freixa, P. W. N. M. van Leeuwen, Dalton

Trost, L. Weber, P. E. Strege, T. J. Fullerton, T. J. Dietsche, J.

Trans. 2003, 1890Ϫ1901.

Am. Chem. Soc. 1978, 100, 3416Ϫ3426. ^[24e]J. M. Brown, D. I.

[7]

[8]

[9]

[24f]

D. Drago, P. S. Pregosin, J. Chem. Soc., Dalton Trans. 2000,

Hulmes, P. J. Guiry, Tetrahedron 1994, 50, 4493Ϫ4506.

von Matt, G. C. Lloyd-Jones, A. B. E. Minidis, A. Pfaltz, L.

P.

3191Ϫ3196.

T. Hayashi, A. Yamamoto, T. Hagihara, J. Org. Chem. 1986,

51, 723Ϫ727.

Macko, M. Neuburger, M. Zehnder, H. Rüegger, P. S. Pregosin,

[24g]

Helv. Chim. Acta 1995, 78, 265Ϫ284.

P. E. Blöchl, A.

[24h]

E. Pen˜a-Cabrera, P.-O. Norrby, M. Sjögren, A. Vitagliano, V.

Togni, Organometallics 1996, 15, 4125Ϫ4132.

H. Stein-

˚

De Felice, J. Oslob, S. Ishii, D. OЈNeill, B. Akermark, P. Helqu-

hagen, M. Reggelin, G. Helmchen, Angew. Chem. Int. Ed. Engl.

[24i]

˚

ist, J. Am. Chem. Soc. 1996, 118, 4299Ϫ4313.

1997, 36, 2108Ϫ2110.

H. Hagelin, B. Akermark, P.-O.

[10]

K. Selvakumar, M. Valentini, M. Wörle, P. S. Pregosin, A. Al-

binati, Organometallics 1999, 18, 1207Ϫ1215.

Norrby, Chem. Eur. J. 1999, 5, 902Ϫ909.

G. C. Lloyd-Jones, S. C. Stephen, M. Murray, C. P. Butts, S.

[25]

[26]

ˇ

[11] [11a]

ˇ

´

I. J. S. Fairlamb, G. C. LloydϪJones, S. Vyskocil, P. Kocov-

Vyskocil, P. Kocovsky, Chem. Eur. J. 2000, 6, 4348Ϫ4357.

P. R. Auburn, P. B. Mackenzie, B. Bosnich, J. Am. Chem. Soc.

1985, 107, 2033Ϫ2046.

[11b]

´

sky, Chem. Eur. J. 2002, 8, 4343Ϫ4353.

L. Gouriou, G. C.

ˇ

´

LloydϪJones, S. Vyskocil, P. Kocovsky, J. Organomet. Chem.

2003, 687, 525Ϫ537.

J. C. Fiaud, A. Hibon de Gournay, M. Larcheveque, H. B.

[27]

[28]

J. C. McKew, M. J. Kurth, J. Org. Chem. 1993, 58, 4589Ϫ4595.

A. Heumann, B. Akermark, S. Hansson, T. Rein, Org. Synth.

[12]

[13]

[14]

˚

ˆ

Kagan, J. Organomet. Chem. 1978, 154, 175Ϫ185.

B. M. Trost, G. M. Schroeder, J. Am. Chem. Soc. 1999, 121,

6759Ϫ6760.

F. Ozawa, A. Kubo, Y. Matsumoto, T. Hayashi, E. Nishioka,

K. Yanagi, K. Moriguchi, Organometallics 1993, 12,

4188Ϫ4196.

M. Kranenburg, P. C. J. Kamer, P. W. N. M. van Leeuwen, Eur.

J. Inorg. Chem. 1998, 25Ϫ27.

R. J. van Haaren, K. Goubitz, J. Fraanje, G. P. F. van

Strijdonck, H. Oevering, B. Coussens, J. N. H. Reek, P. C. J.

Kamer, P. W. N. M. van Leeuwen, Inorg. Chem. 2001, 40,

3363Ϫ3372.

Coll. Vol. 1993, 8, 137Ϫ140.

[29] [29a]

W. H. Bunnelle, T. A. Isbell, J. Org. Chem. 1992, 57,

[29b]

729Ϫ740.

B. M. Trost, M. G. Organ, G. A. O’Doherty, J.

Am. Chem. Soc. 1995, 117, 9662Ϫ9670.

I. Fleming, D. Higgins, N. J. Lawrence, A. P. Thomas, J. Chem.

Soc., Perkin Trans. 1 1992, 3331Ϫ3349.

F. R. Hartley, S. R. Jones, J. Organomet. Chem. 1974, 66,

465Ϫ473.

B. M. Trost, P. E. Strege, L. Weber, T. J. Fullerton, T. J. Diets-

che, J. Am. Chem. Soc. 1978, 100, 3407Ϫ3415.

D. G. Davis, A. Bax, J. Mag. Res. 1985, 64, 533.

P. von Matt, A. Pfaltz, Angew. Chem. Int. Ed. Engl. 1993, 32,

566Ϫ568.

P. A. Evans, J. D. Nelson, J. Am. Chem. Soc. 1998, 120,

5581Ϫ5582.

P. Sennhenn, B. Gabler, G. Helmchen, Tetrahedron Lett. 1994,

35, 8595Ϫ8598.

[30]

[31]

[32]

[15]

[16]

[33]

[34]

[17]

[18]

´

A. C. Albeniz, P. Espinet, B. Martın-Ruiz, Chem. Eur. J. 2001,

[35]

[36]

[37]

[38]

7, 2481Ϫ2489.

M. Kollmar, B. Goldfuss, M. Reggelin, F. Rominger, G.

Helmchen, Chem. Eur. J. 2001, 7, 4913Ϫ4927.

P. E. Garrou, Chem. Rev. 1981, 81, 229Ϫ266.

I. J. S. Fairlamb, G. C. Lloyd-Jones, Chem. Commun. 2000,

2447Ϫ2448.

[19]

[20]

OpenMoleN, Interactive Structure Solution, Nonius B. V., Delft,

The Netherlands, 1997.

^{[21] [21a]}R. S. Paonessa, A. L. Prignano, W. C. Trogler, Organomet-

D. T. Cromer, J. T. Waber, International Tables for X-ray Crys-

[21b]

allics 1985, 4, 647Ϫ657.

F. Ozawa, T.-i. Son, S. Ebina, K.

tallography, The Kynoch Press, Birmingham, 1974, vol. IV, ^[38a]

[38b]

Osakada, A. Yamamoto, Organometallics 1992, 11, 171Ϫ176.

P. Barbaro, P. S. Pregosin, R. Salzmann, A. Albinati, R. W.

Kunz, Organometallics 1995, 14, 5160Ϫ5170.

Table 2.2b,

Table 2.3.1.

[22]

[23]

Received April 19, 2004

Early View Article

Published Online September 10, 2004

R. J. van Haaren, H. Oevering, B. B. Coussens, G. P. F. van

Eur. J. Inorg. Chem. 2004, 3987Ϫ4001

www.eurjic.org

 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

4001

Article Doi

Palladium complexes with chiral diphospholane ligands: Comparative catalytic properties and analysis of (η3-allyl)palladium species

DOI: 10.1002/ejic.200400312

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejic.200400312

Full text of DOI:10.1002/ejic.200400312

Products guided by the article

R&D Labs maybe for 90383-10-1

Relevant to this article

Hot Product