Palladium catalytic species containing chiral phosphites: Towards a discrimination between molecular and colloidal catalysts

Article abstract of DOI:10.1002/adsc.200700200

Palladium nanoparticles (Pd1-Pd3) stabilized by chiral diphosphite ligands (1-3), were synthesized and tested as catalysts for the allylic alkylation reaction, using different substrates (rac-I, rac-III and rac-V). Carbohydrate ligands (1 and 2), only differing in the C-3 configuration, led to a remarkable difference in stability of the corresponding nanoparticles: while Pd1 is a robust catalyst, Pd2 decomposes into molecular species. In addition, the high enantioselective systems, Pd1 and Pd3, are only active for a substrate containing phenyl groups. Concerning the catalytic behaviour of the corresponding molecular systems, palladium complexes coordinated to ligands 1 or 3, gave excellent asymmetric inductions, but an analogous catalyst accommodating ligand 2, was not found selective.

Full text of DOI:10.1002/adsc.200700200

FULL PAPERS
DOI: 10.1002/adsc.200700200
Palladium Catalytic Species Containing Chiral Phosphites: Towards
a Discrimination between Molecular and Colloidal Catalysts
Isabelle Favier,a Montserrat Gómez,a,b,* Guillermo Muller,a M. Rosa Axet,c
Sergio Castillón,c Carmen Claver,c,* Susanna Jansat,d Bruno Chaudret,d,*
and Karine Philippotd,*
a
Departament de Química Inorgànica, Universitat de Barcelona, Martí i Franqu›s, 1-11, 08028 Barcelona, Spain
Laboratoire HØtØrochimie Fondamentale et AppliquØe, UMR CNRS 5069, 118, route de Narbonne, 31062 Toulouse
b
cedex 9, France
Fax : (+33)-5-6155-8204; e-mail: gomez@chimie.ups-tlse.fr
Departament de Química Física i Química Inorgànica i Química Analítica i Química Orgànica, Universitat Rovira i
c
Virgili, c/Marcel·lí Domingo, s/n, 43007 Tarragona, Spain
Laboratoire de Chimie de Coordination, UPR CNRS 8241, 205, route de Narbonne, 31077 Toulouse cedex 04, France
d
Received: April 20, 2007; Revised: July 30, 2007
Supporting information for this article is available on the WWW under http://asc.wiley-vch.de/home/.
Abstract: Palladium nanoparticles (Pd1–Pd3) stabi- a substrate containing phenyl groups. Concerning the
lized by chiral diphosphite ligands (1–3), were syn- catalytic behaviour of the corresponding molecular
thesized and tested as catalysts for the allylic alkyla- systems, palladium complexes coordinated to ligands
tion reaction, using different substrates (rac-I, rac-III 1 or 3, gave excellent asymmetric inductions, but an
and rac-V). Carbohydrate ligands (1 and 2), only dif- analogous catalyst accommodating ligand 2, was not
fering in the C-3 configuration, led to a remarkable found selective.
difference in stability of the corresponding nanopar-
ticles: while Pd1 is a robust catalyst, Pd2 decompos-
es into molecular species. In addition, the high enan- Keywords: allylic substitution; catalyst nature; chiral
tioselective systems, Pd1 and Pd3, are only active for phosphites; nanostructures; palladium
Introduction
BINAP-Pd nanoparticles were found to be an effi-
cient catalyst for the asymmetric hydrosilylation of
An ever-increasing interest has been devoted in the styrene under mild conditions, in contrast to the lack
past years to metal nanoparticles because of their of activity showed by the related molecular system.[6]
small size and electronic configuration giving rise to
Palladium particles also recently attracted a high in-
interesting properties in various areas.[1] In this re- terest as catalysts for C C coupling reactions (mainly
spect, the use of metal nanoparticles for catalytic Heck and Suzuki couplings).[7] During the past years,
transformations of organic substrates is a growing some of us have developed the synthesis of metal
area.[2] Metal nanoparticles have been proved to be nanoparticles through an organometallic ap-
efficient and selective catalysts for reactions which proach.[2g,8]
À
are catalyzed by molecular complexes such as olefin
This method leads to nearly monodispersed parti-
À
hydrogenation or C C coupling, for example, but also cles of very small size and displaying interesting sur-
for reactions which are not or only poorly catalyzed face coordination chemistry. This point can, in princi-
by molecular species such as aromatic hydrocarbon ple, be evidenced by spectroscopic methods such as
hydrogenation.[2d,g,b] An unambiguous distinction be- NMR,[9] but also by the reactivity produced by the
tween colloidal and true homogeneous catalysis is presence of ligands, like asymmetric induction in cata-
however often very difficult.[3] Despite impressive lytic processes.
progresses in asymmetric catalysis,[4] only very few
Concerning the chiral ligands, those arising from
colloidal systems have been found to display an inter- the chiral “pool” such as carbohydrate ligands display
esting activity, the best one being Pt(Pd)/cinchonidine several advantages: they are readily available, highly
for the hydrogenation of ethyl pyruvate.[5] Recently, functionalized, and may be systematically modified
Adv. Synth. Catal. 2007, 349, 2459 – 2469
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2459

Isabelle Favier et al.
FULL PAPERS
Scheme 1. Synthesis of palladium nanoparticles Pd1–Pd3.
employing different types of allylic substrates. Di-
phosphite ligand 2, with C1-symmetry differs from di-
phosphite 1 only in the configuration of C-3 of the
sugar backbone, while ligand 3 differs in the flexibility
of its skeleton without the bicyclic backbone. It is
hoped that the various ligand/nanoparticles surface/
substrate environments thus created may help to shed
some light on this novel catalytic system and operat-
ing mechanism. For this purpose, we describe first the
synthesis and characterization of nanoparticles and
Figure 1. Chiral diphosphite ligands 1–3.
and optical resolution is unnecessary. During the past complexes containing chiral diphosphite ligands, to
years some of us have focused on the use of modified follow with the application of both colloids and mo-
furanose derivatives of C1-symmetry, as ligands in lecular systems as catalysts in allylic alkylation of sev-
asymmetric catalysis. Some of these ligands have been eral acetate allylic substrates.
successfully applied in metal-catalyzed asymmetric re-
actions providing excellent results in rhodium-cata-
lyzed asymmetric hydrogenation, asymmetric hydro- Results and Discussion
formylation or palladium-catalyzed asymmetric allylic
substitution.[10] In particular, the use of furanoside di- Palladium Species Containing Chiral Phosphites
phosphites where the inversion of the configuration in
C-3 (xylo or ribo configuration) influences strongly Palladium Colloids
the catalytic behaviour, becomes especially interest-
ing. For instance, in the rhodium-catalyzed hydrofor- Palladium nanoparticles (colloids Pd1–Pd3) were syn-
mylation of styrene the diphosphite with a xylo con- thesized from [Pd2(dba)3] as metal precursor and the
A
figuration, provided the same ee as the ribo diphos- appropriate chiral diphosphite ligand,[14] 1–3
phite but with the opposite configuration,[11] while in (Scheme 1). As a standard procedure, the decomposi-
the rhodium-catalyzed hydrogenation of dehydroami- tion of [Pd
no acids, only the ribo configuration provides excel- inside a Fischer–Porter bottle, at room temperature
lent ee (up to 99%).[12]
and in the presence of the chosen ligand (1–3; Pd/L=
In a previous preliminary communication, we re- 1/0.2).
G
ported the asymmetric allylic alkylation of rac-3-ace-
The Pd nanoparticles were isolated as black pow-
toxy-1,3-diphenyl-1-propene with dimethyl malo- ders after pentane precipitation. Transmission elec-
nate.[13] using as catalyst Pd nanoparticles stabilized tron microscopy (TEM) revealed the presence of
by the chiral diphosphite ligand 1 (Figure 1). We ob- small spherical, but in some cases agglomerated, par-
served that the nanoparticles display a high selectivity ticles of ca. 4 nm mean size, as seen in Figure 2,
for this asymmetric process. More specifically, the re- Figure 3, and Figure 4. Colloids Pd1 and Pd2 are simi-
action proceeds with one enantiomer of the racemic lar in aspect and mean sizes. These results are not
substrate demonstrating a very high degree of kinetic very surprising since the stabilizing ligands L1 and L2
resolution. In addition, the same reaction catalyzed only differ in the configuration of carbon atom C-3 of
by the molecular species accommodating the same di- their sugar backbones. Size histograms built from the
phosphite ligand, under comparable conditions for displayed micrographs, show narrow size distributions
matching the catalysts rates, proceeded with both en- with a majority of particles having a size correspond-
antiomers. The origin of this selectivity could be due ing to the mean one, most particularly for Pd1.
to a particular combination between the metallic sur-
face, the substrate and the ligand.
Colloid Pd3 contains nanoparticles which are not
well dispersed on the TEM grid in opposition to what
In order to try to understand the origin for this sin- is observed for colloids Pd1 and Pd2. As a conse-
gular effect, we decided to carry out a comparative quence of their agglomeration, the size of Pd3 nano-
study using colloidal and molecular catalytic systems particles could not be precisely determined because
containing three diphosphite ligands (Figure 1), and of their close proximity but is estimated to be ca.
2460
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2459 – 2469

Towards a Discrimination between Molecular and Colloidal Catalysts
FULL PAPERS
Figure 2. Transmission electron micrograph of colloid Pd1 and the corresponding size histogram.
Figure 3. Transmission electron micrograph of colloid Pd2 and the corresponding size histogram.
4 nm. This behaviour can be explained by a lower Palladium Complexes
steric hindrance around the particles, since L3 ligand
shows a more flexible backbone than L1 and L2.
Ionic palladium complexes, 4 and 5, containing the
Wide-angle X-ray scattering analyses evidenced the 1,3-diphenylallyl group were prepared from the corre-
fcc structure of bulk palladium (Figure 5) and a coher- sponding palladium dimer and the appropriate chiral
ence length which in the case of Pd1 and Pd3 matches ligand, 1 and 2 respectively, in the presence of ammo-
the TEM observations. The small coherence length nium hexafluorophosphate (Scheme 2), following the
observed in the case of Pd2 agrees with the presence methodology previously described.[15]
of nanoparticles of irregular shape as observed on
Figure 3.
These compounds were obtained as monometallic
complexes of general formula [Pd(h3-1,3-Ph2C3H3)-
(L)] PF6 (L=1, 2), where L acts as a k2-P,P-bidentate
ligand affording eight-membered palladacycle. Com-
plexes 4 and 5 were fully characterized by the usual
techniques. IR spectra showed strong signals at 1070
Adv. Synth. Catal. 2007, 349, 2459 – 2469
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2461

Isabelle Favier et al.
FULL PAPERS
NMR spectroscopy allowed the structure determi-
nation of these complexes in solution. 31P NMR and
1H NMR spectra of 4 showed the presence of two spe-
cies, with a ratio 2/1, invariable with the temperature
(temperature range studied: 298–223 K). Both iso-
mers are probably due to the relative position of the
central allylic carbon with regard to the ligand back-
bone (see below). For 5, the two isomers could not be
distinguished in the temperature range analyzed
(298–223 K).
We carried out a molecular modelling study of both
isomers of each complex by means of semi-empirical
PM3(tm) calculations (Figure 6). The two isomers
result from the relative position of the central allylic
carbon and the acetal group: exo, if both groups point
to the same direction, endo, if they point to opposite
sides. It is important to notice that the two different
metallacycle arrangements were observed for the two
complexes. For 4, the palladium, two phosphorus and
two oxygen atoms are quasi coplanar, while for 5, the
metallacycle shows an eight-member crown conforma-
tion. These distinct conformations lead to a different
steric hindrance of the biphenyl tert-butyl groups to-
wards the 1,3-diphenylallyl moiety. As a result, the
two terminal allyl carbon atoms are electronically
well differentiated for 4, as observed by the calculated
À
Figure 4. Transmission electron micrograph of colloid Pd3.
distances [D(Pd C)=0.02 and 0.023 Š for endo and
exo isomers, respectively], while for 5, no important
G
À
variations are produced [D
G
0.004 Š for endo and exo isomers, respectively].
These data are in agreement with the remarkable dif-
ferences in the asymmetric induction observed in Pd-
catalyzed allylic alkylation using rac-3-acetoxy-1,3-di-
phenyl-1-propene (rac-I) as substrate (see below).
For the analogous complex coordinated to ligand 3,
the corresponding allyl complex, [Pd(h2-1,3-Ph2-C3H3)
(k2-P,P-3)]+, has been also calculated. In this case
Figure 5. WAXS analysis: comparison of the rdf of colloids
Pd1, Pd2 and Pd3 with a theoretical calculated for 4 nm fcc
Pd nanoparticles.
only one allylic structure is possible, due to the steric
hindrance of the ligand (Figure 7 a).
Enantioselective Allylic Alkylation
Asymmetric allylic alkylation of racemic substrates
rac-3-acetoxy-1,3-diphenyl-1-propene (rac-I) and rac-
3-acetoxy-1-cyclohexene (rac-III) with dimethyl malo-
nate under basic Trost conditions,[16] was carried out
using Pd catalytic systems containing ligands 1–3
(Scheme 3).
Scheme 2. Synthesis of palladium complexes 4 and 5.
Two types of catalytic precursors were employed,
molecular systems generated in situ from [Pd
(m-Cl)-
(4) and 1090 (5) cmÀ1 assigned to the P O stretching
ACHTREUNG
À
of the phosphite moiety and at 835 (4) and 801 (5) and colloidal ones, using pre-formed palladium nano-
cmÀ1 assigned to P F stretching of the PF6 anion. particles (see above). These results are summarized in
À
MALDI mass spectra exhibited a peak corresponding Table 1, Table 2 and Table 3 for rac-I and in Table 5
to the [Pd(h3-1,3-Ph2C3H3)(L)]+ fragment. Unfortu- for rac-III.
nately, no monocrystals of 4 or 5 could be obtained.
2462
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2459 – 2469

Towards a Discrimination between Molecular and Colloidal Catalysts
FULL PAPERS
Figure 6. Calculated structures [PM3 (tm)] for cationic species of complexes 4 and 5 (in parenthesis their relative formation
À
enthalpies). Hydrogen atoms are omitted for clarity. Pd C bond distances are in Š.
Figure 7. Modelled structures [PM3 (tm)] for palladium complexes containing ligand 3: cationic palladium allyl complex,
[Pd(h2-1,3-Ph2-C3H3)
A
G
G
A
À
P,P-3)] (b). Hydrogen atoms are omitted for clarity. Pd C bond distances are in Š.
Pd-Catalyzed Allylic Alkylation of rac-3-Acetoxy-1,3- taining the open alkyl chain (entry 3 versus 1 and 2,
diphenyl-1-propene (rac-I)
Table 1). Concerning the catalytic systems containing
carbohydrate ligands 1 and 2, it is important to note
the crucial effect of the configuration at C-3 stereo-
centre: while Pd/1 gave an excellent enantioselectivity
Molecular Catalytic Precursors
The three Pd/L catalytic systems (L=1–3) were active (ee=90% (S)), Pd/2 led to a racemic mixture of II
in the allylic alkylation of rac-I. The highest activity (entries 1 and 2, Table 1).
and enantioselectivity were observed for Pd/3 con-
Adv. Synth. Catal. 2007, 349, 2459 – 2469
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2463

Isabelle Favier et al.
FULL PAPERS
Scheme 3. Allylic alkylation of rac-I and rac-III using Pd/L catalytic systems (L=1–3).
Table 1. Asymmetric allylic alkylation of rac-I catalyzed by
molecular palladium systems containing chiral ligands 1–3.[a]
and 5 (see above). Therefore, the metallacycle formed
in the case of 4 induces a significant electronic differ-
entiation between both terminal allylic carbon atoms,
in contrast to the lack of effect induced by 5 (see rela-
Entry L Time
Conversion
[%][b]
ee I
ee II
[h]
[%][c]
[%][c]
À
tive Pd C bond distances in Figure 6). Comparing Pd/
1[d]
2
3
1
2
3
1.5
1.5
0.15
83
80
100
0
0
-
90 (S)
0
94 (S)
1 and Pd/3 catalytic systems (entries 1 and 3, Table 1),
we observe that the asymmetric induction was excel-
lent in both cases but the activity was significantly su-
perior for Pd/3, which contains a ligand with a flexible
backbone. The calculated structures for intermediates
containing ligand 3 allow to justify the high enantiose-
lectivity observed. Concerning its palladium allyl in-
termediate, only one stable isomer is possible due to
the pocket environment formed upon coordination of
the palladium to ligand 3, leading to a significant elec-
tronic differentiation between both terminal allylic
[a]
[b]
[c]
Results from duplicated experiments; Pd/rac-I/dimethyl
malonate/BSA=1/100/300/300. See Scheme 3.
Conversions based on the substrate I and determined by
1H NMR.
Determined by HPLC on a Chiralcel-OD column. Abso-
lute configuration of II in parentheses, determined by op-
tical rotation: U. Leutenegger, G. Umbricht, C. Fahrni,
P. V. Matt, A. Pfaltz, Tetrahedron 1992, 48, 2143-2156.
See ref.[13]
[d]
À
carbon atoms (Pd C calculated distances are 2.192
and 2.221 Š, Figure 7 a). Moreover, the nucleophile
attacks preferentially the more electrophilic terminal
This fact can be explained by the electronic differ- allylic carbon atom, because the attack on the other
entiation produced at the terminal allylic carbon atom is blocked by the steric hindrance produced in
atoms in the Pd allylic intermediates, complexes 4 the resulting Pd(0) intermediate (Figure 7 b).
Table 2. Asymmetric allylic alkylation of rac-I catalyzed by colloidal palladium systems, Pd1–Pd3.[a]
Conversion [%][b]
ee I [%][c]
ee II [%][c]
k(R)-I/k(S)-I
[d]
Entry
PdL/L
Time [h]
1[e]
2[e]
3[e]
4
5
6
7
8
9
10
11
Pd1/-
Pd1/1
Pd1/1
Pd2/-
Pd2/2
Pd2/2
Pd3/-
Pd3/3
Pd3/3
Pd1/2
Pd2/1
168
10
24
168
24
48
168
3
0
-
-
-
50
56
0
0
100
0
68
88
0
70 (S)
89 (S)
-
-
-
-
94 (S)
>99 (S)
0
-
96 (S)
97 (S)
-
-
11 (S)
-
92 (S)
90 (S)
-
11.6
16.4
-
-
-
-
8.3
4.7
-
24
24
24
100
89 (S)
-
[a]
Results from duplicated experiments; Pd/rac-I/dimethyl malonate/BSA=1/100/300/300. Pd/Lstabilizer/Lfree ligand=
1/0.2/1.05.
[b]
[c]
1
Conversions based on the substrate I and determined by H NMR.
Determined by HPLC on a Chiralcel-OD column. Absolute configuration of II in parentheses, determined by optical ro-
tation: U. Leutenegger, G. Umbricht, C. Fahrni, P. V. Matt, A. Pfaltz, Tetrahedron 1992, 48, 2143-2156.
[d]
[e]
Calculated from kR/kS =ln
[(1ÀC/100)
U
[(1ÀC/100)
(1+ee/100)] (C. H. Chen, Y. Fujimoto, G. Girdaukas, C. J.
Sih, J. Am. Chem. Soc. 1982, 105, 7294-7299).
See ref.[13]
2464
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2459 – 2469

Towards a Discrimination between Molecular and Colloidal Catalysts
FULL PAPERS
Table 3. Asymmetric allylic alkylation of rac-I catalyzed by molecular palladium systems containing chiral ligands 1 and 3, at
different Pd concentrations.[a]
[d]
Entry
L
Pd/I
Time [h]
Conversion [%][b]
ee I [%][c]
ee II [%][c]
k(R)-I/k(S)-I
1[e]
2[e]
3[e]
4
1
1
1
3
3
1/100
1.5
3
24
0.15
0.5
83
43
72
100
88
0
90 (S)
90 (S)
90 (S)
94 (S)
91 (S)
1
1/2,000
1/10,000
1/100
26 (S)
43 (S)
-
2.6
2.0
-
5
1/2,500
>99 (S)
4.7
[a]
Results from duplicated experiments; rac-I/dimethyl malonate/BSA=1/3/3.
Conversions based on the substrate I and determined by H NMR.
Determined by HPLC on a Chiralcel-OD column. Absolute configuration of II in parentheses, determined by optical ro-
tation: U. Leutenegger, G. Umbricht, C. Fahrni, P. V. Matt, A. Pfaltz, Tetrahedron 1992, 48, 2143-2156.
[b]
[c]
1
[d]
[e]
Calculated from kR/kS =ln
A
(1Àee/100)]/ln
U
A
Sih, J. Am. Chem. Soc. 1982, 105, 7294-7299).
See ref.[13]
Table 4. Asymmetric allylic alkylation of rac-V catalyzed by
colloidal and molecular palladium systems containing chiral
ligands 1 and 2.[a]
of Table 2 versus entries 1 and 3 of Table 1). The ki-
netic resolution of the racemic substrate using Pd1
(k(R)-I/k(S)-I ꢀ16; entry 3, Table 2) is better than that
using Pd3 (k(R)-I/k(S)-I ꢀ5; entry 9, Table 2), since for
Pd1 the reaction practically stops after 55% conver-
sion,[13] while for Pd3 the reaction slowly evolves with
the time.
Looking for information about the origin of this ki-
netic resolution, the reaction of rac-I with dimethyl
malonate was carried out with the molecular system
at different Pd concentrations (Table 3). Colloid Pd1
behaves differently than the corresponding molecular
system Pd/1 (entries 2 and 3, Table 2, versus entries 1–
3, Table 3), since for Pd1 the reaction practically
stops after 10 h. In addition, the allylic alkylation of
enriched substrate I [(S)-I/(R)-I=92/8] was studied
using the colloidal system Pd1 and also the very
dilute molecular catalytic system (Pd/1=1/10,000).[13]
After 30 h, only 10% of enriched substrate was con-
verted in the colloidal case, but ca. 70% in the molec-
ular one, giving the alkylated product II in 95% ee in
both cases. This fact shows that for the colloidal cata-
lyst, almost only (R)-I leads to the formation of the
alkylation product.
Entry Catalyst (precur- Time
sor) [h]
Conversion
[%][b]
ee VI
[%][c]
1
2
3
4
Pd1/1 (colloidal) 24
Pd/1 (molecular) 3
Pd2/2 (colloidal) 72
Pd/2 (molecular) 7
0
-
45
82
80
50 (R)
0
0
[a]
Results from duplicated experiments; Pd/rac-V/dimethyl
malonate/BSA=1/100/300/300. For colloidal systems,
Pd/Lstabilizer/Lfree ligand=1/0.2/1.05.
[b]
[c]
Conversions based on the substrate V and determined by
1H NMR.
Determined by optical rotation. Absolute configuration
of VI in parentheses, determined by optical rotation: L.
Xiao, W. Weissensteiner, K. Mereiter, M. Widhalm, J.
Org. Chem. 2002, 67, 2206-2214.
Colloidal Catalytic Precursors
In contrast, for Pd3 and Pd/3 no kinetic differences
were observed between molecular and colloidal sys-
When rac-I reacted with dimethyl malonate using col- tems (entry 5, Table 3, versus entry, 9 Table 2). There-
loids Pd1, Pd2 and Pd3 as catalysts, no reaction was fore for this system, the high activity and selectivity
observed (entries 1, 4, 7, Table 2). However, catalysts showed by both molecular (entries 4 and 5, Table 3)
Pd1 and Pd3 became active after addition of 1 equiv- and colloidal (entries 8 and 9, Table 2) catalysts, do
alent of the corresponding ligand in the catalytic reac- not provide any evidence about the nature of the cat-
tion mixture (entries 2 and 9, Table 2). This fact can alyst.
be related to the presence of free ligand in solution
which prevents nanoparticles agglomeration and/or was only observed at long reaction times (more than
leaching of palladium. 24 h), probably due to some decomposition giving
Colloids Pd1 and Pd3, similarly to the correspond- molecular species.
Concerning Pd2 (entries 5 and 6, Table 2), activity
ing molecular systems, showed excellent enantioselec-
In order to corroborate the robustness of colloidal
tivities (see Table 1). However, in contrast to the mo- Pd1 under catalytic conditions, free ligand 2 (and no
lecular catalysts, Pd1 and Pd3 catalytic systems pro- 1) was added to the catalytic reaction mixture (ratio
vided an excellent kinetic resolution (entries 3 and 8 Pd/1/2=1/0.2/1.05). In this case the catalyst was total-
Adv. Synth. Catal. 2007, 349, 2459 – 2469
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2465

Isabelle Favier et al.
FULL PAPERS
Table 5. Asymmetric allylic alkylation of rac-III catalyzed by molecular and colloidal palladium systems containing chiral li-
gands 1 and 3.[a]
[b]
Entry
L
Pd/III
Time [h]
Conversion [%][a]
ee III (%)[a]
ee IV [%][a]
k(S)-III/k(R)-III
Molecular systems[c]
1
2
3
4
5
1
1
3
3
3
1/100
1/5,000
1/100
1/100
1/5,000
0.5
24
0.25
0.5
24
100
19
48
73
-
22 (S)
27 (S)
88 (S)
73 (S)
nd[d]
-
28 (R)
>98 (R)
>98 (R)
0
14.0
156.5
8.3
5
-
Colloidal systems[e]
6
7
1
3
1/100
1/100
24
24
0
7
0
0
-
-
-
nd[d]
[a]
Conversions based on the substrate III. Both conversions and enantioselectivities were determined by GC. Absolute con-
figuration of IV in parentheses, determined by optical rotation: L. Xiao, W. Weissensteiner, K. Mereiter, M. Widhalm, J.
Org. Chem. 2002, 67, 2206-2214.
[b]
Calculated from kR/kS =ln
Sih, J. Am. Chem. Soc. 1982, 105, 7294-7299).
A
G
[(1ÀC/100)(1+ee/100)] (C. H. Chen, Y. Fujimoto, G. Girdaukas, C. J.
[c]
[d]
[e]
Results from duplicated experiments; rac-III/dimethyl malonate/BSA=1/3/3. See Scheme 3.
Not determined.
Results from duplicated experiments; rac-III/dimethyl malonate/BSA=1/3/3. Pd/Lstabilizer/Lfree ligand=1/0.2/1.05. See
Scheme 3.
ly inactive (entry 10, Table 2), analogously to the Pd-Catalyzed Allylic Alkylation of rac-3-Acetoxy-1-
system without free ligand added (entry 1). Probably, cyclohexene (rac-III)
ligand 2 does not interact with Pd1 nanoparticles.
And what is more important, Pd1 does not leach mo- The asymmetric allylic alkylation of the cyclic sub-
lecular species. On the contrary, the “cross-system” strate rac-III with colloids Pd1 and Pd3 was carried
formed by Pd2 nanoparticles and free ligand 1, exhib- out (Scheme 3, Table 5). Concerning the molecular
its a high activity and enantioselectivity (entry 11, systems, Pd/3 was more selective than Pd/1, giving IV
Table 2), similar to that observed for molecular with ee up to 88% (entry 1 versus entry 3, Table 5).
system Pd/1 (entry 1, Table 1), pointing to the decom- The high enantioselectivity observed for Pd/3 is prob-
position of Pd2 nanoparticles to give molecular spe- ably due to the more important steric hindrance of
cies, mainly molecular Pd/1 species due to the relative the ligand 3 towards the allyl moiety than that of
ratio of both ligands (1/2=1.05/0.2). The same trend ligand 1 (see structures for allyl complexes containing
between palladium systems containing 1 and 2 was 1,3-diphenyl allyl, Figure 6 and 7), due to the pocket
also found using rac-3-acetoxy-1,3-dimethyl-1-propene environment around the “Pd-allyl” moiety.
(rac-V) as substrate (Table 4). The colloidal Pd1 cata-
But what is more remarkable is the high kinetic res-
lyst was found inactive, in contrast to the molecular olution resulting with the Pd/3 catalytic system. The
system (entry 1 versus 2, Table 4). However a similar substrate (S)-III was resolved more than 150 times
reactivity was observed for both molecular and colloi- faster than its enantiomer at ca. 50% conversion
dal catalytic systems involving ligand 2. Again, the ac- (entry 3, Table 5).
tivity and selectivity observed for Pd/2 molecular
When the palladium content was low (Pd/sub-
system was analogous to that observed for colloidal strate=1/5,000), both Pd/1 and Pd/3 systems were not
catalyst (entries 3 and 4, Table 4), demonstrating de- very active, giving 19 and 5% conversion after 24 h,
finitively the formation of molecular species from col- respectively (entries 2 and 5, Table 5).
loid Pd2.
With regard to colloidal systems, both Pd1 and Pd3
In order to shed more light concerning the interac- exhibited a negligible activity at long reaction time
tion between the metallic surface and the substrate, (entries 6 and 7, Table 5), showing both their stability
the cyclic racemic substrate, 3-acetoxy-1-cyclohexene, under the catalytic conditions used and the low affini-
has been considered.
ty for alkyl substrates, in contrast to the behaviour ob-
served for the allylic alkylation of rac-I (entries 3 and
9, Table 2). In addition, these colloidal systems do not
give palladium leaching, independently of the sub-
strate used and confirming their robustness.
2466
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2459 – 2469

Towards a Discrimination between Molecular and Colloidal Catalysts
FULL PAPERS
Conclusions
agostic interactions. Therefore, aryl-substituted allyl
substrates could be envisaged as substrate models to
We report in this paper, novel asymmetric catalytic distinguish molecular from colloidal catalysts.
systems based on palladium nanoparticles. These sys- In summary this study demonstrates that : i) cataly-
tems where found to be dependent both upon the li- sis by nanoparticles may be very effective in terms of
gands and the substrates in a much more dramatic enantioselectivity; ii) the system is much more sensi-
way than corresponding molecular systems. We find tive to the adjustment between the metal, the ligand
that small modifications in carbohydrate phosphite and the substrate than the corresponding molecular
derivatives lead to important differences in selectivity. catalysts, in a way which may resemble enzymatic sys-
Therefore, chiral diphosphites, 1 and 2, which only tems. This may explain the very limited number of
differ at C-3 configuration of the carbohydrate back- enantioselective catalytic systems which involved
bone, exhibit dramatic differences in their catalytic nanoparticles. Further studies in this field will have to
behaviour. While Pd/1 molecular system is highly integrate the necessity of this tuning if one wants to
enantioselective (ee=90% for the allylic alkylation of exploit the tremendous potential of these new catalyt-
rac-I), the analogous Pd/2 system does not induce any ic systems.
selectivity (ee=0% for the same reaction). This fact
is in agreement with the modelling of palladium allyl-
ic intermediates (compounds 4 and 5), showing that
upon coordination to palladium, ligand 1 causes an ef-
Experimental Section
ficient electronic discrimination between both termi-
nal allylic carbon atoms, in contrast to the symmetri-
cal environment produced by 2. In addition, palladi-
um nanoparticles stabilized with these ligands (Pd1
and Pd2) show a remarkable difference of stability.
Under catalytic conditions using rac-I or rac-V as sub-
strates, Pd2 decomposes into molecular species,
whereas Pd1 does not.
When the carbohydrate skeleton is substituted by
the flexible alkyl 2,4-dimethylpentyl backbone (ligand
3) high activity and excellent selectivity are observed
when molecular catalytic precursor is used (for allylic
alkylation of both rac-I and rac-III), mainly for the
less hindered substrate, cyclohexenyl derivative rac-
III, reaching an enantiomeric excess up to 88%. The
excellent enantioselectivity observed in both alkylated
products, II and IV, can be explained by the pocket
General Remarks
All compounds were prepared under a purified nitrogen at-
mosphere using standard Schlenk and vacuum-line tech-
niques. The solvents were purified by standard procedures
and distilled under nitrogen. [Pd
G
G
[15a]
1,3-Ph2-C3H3)
A
and ligands 1–3[14] were prepared as
previously described. NMR spectra were recorded on
Bruker DRX 500 (1H, standard SiMe4), Varian Gemini (13C,
50 MHz, standard SiMe4) and Bruker DRX 250 (31P,
101.2 MHz, standard H3PO4) spectrometers in CDCl3 unless
otherwise cited. Chemical shifts were reported downfield
from standards. IR spectra were recorded on an FT-IR
Nicolet Impact 400 spectrometer. MALDI mass chromato-
grams were obtained on a Fisons V6-Quattro instrument
using dithranol as a matrix. Enantiomeric excess were deter-
mined by HPLC on a Hewlettt–Packard 1050 Series chro-
matograph (Chiralcel-OD chiral column) with a UV detec-
tor, and by GC on a Hewlett–Packard 5890 Series II gas
effect of the ligand upon coordination to the metal, as chromatograph (25 m FS-cyclodex-b-I/P column: hepta-
kis(2,3,6-tri-O-methyl)-b-cyclodextrin/polysiloxane) with an
FID detector. Optical rotations were measured in a Perkin–
Elmer 241MC polarimeter. Elemental analyses were carried
out by the “Serveis Cientifico-T›cnics” of the Universitat
Rovira i Virgili in an Eager 1108 microanalyzer, by the
“Service d’Analyses du LCC” in Toulouse and by the “Serv-
ice Central d’Analyses du CNRS” in Lyon for phosphorus
and palladium determinations. Samples for TEM analysis
were prepared in a glove-box by slow evaporation of a drop
of each crude colloidal solution deposited onto a holey
stated by modelling studies of palladium intermedi-
ates containing both 1,3-diphenyl and cyclohexenyl
allyl moieties. Concerning the palladium nanoparticles
stabilized by 3, Pd3 is a highly active and selective
system for the allylic alkylation of rac-I in a similar
way as the molecular catalytic system. But, in contrast
to the molecular catalyst, it is almost inactive for the
alkylation of rac-III.
It is also noteworthy the excellent kinetic resolution
showed by Pd1 in the allylic alkylation of rac-I, which carbon-covered copper grid. The TEM experiments were
performed at the “Service Commun de Microscopie Electro-
nique de l’UniversitØ Paul Sabatier” (TEMSCAN) on a
JEOL 200 CX-T electron microscope operating at 200 kV
with a point resolution of 4.5 Š and at the “Centre de mi-
croscopie Ølectronique appliquØe à la Biologie de l’Univer-
sitØ Paul Sabatier” on a Philips CM12 electron microscope
operating at 120 kV with a resolution point of 5 Š. The ap-
proximation of the particle mean size was made through a
manual analysis of enlarged micrographs by measuring a
number of particles on a given grid. For wide-angle X-ray
has not been observed using Pd2. This fact points to a
key-lock matching between substrate and catalyst.
From the point of view of the substrate using stable
colloidal catalysts (Pd1 and Pd3), alkyl-substituted
allyl acetates (rac-III and rac-V) are not alkylated,
while the phenylallyl acetate (rac-I) is. This fact is in
agreement with the strong preference shown by met-
allic surfaces for aryl groups, probably due to p-coor-
dination between the metallic surface and the aromat-
ic group whereas alkyl groups can only lead to weak scattering samples (analysis performed at the CEMES-
Adv. Synth. Catal. 2007, 349, 2459 – 2469
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2467

Isabelle Favier et al.
FULL PAPERS
CNRS, Toulouse), samples were sealed in 1-mm diameter
Lindemann glass capillaries. Modelling studies have been
carried out using the following software: MACSPARTAN
PRO Version 1.0.4. WaveFunction, Inc. 18401 Von Karman
Avenue, suite 370. Irvine, CA 92612 U.S.A.
62.6 mg of diphosphite 3 (0.070 mmol). TEM: mean size
around 4 nm; elemental analysis: 71.75% Pd, 0.46% P,
11.78% C, 0.13% H.
General Procedure for Palladium-Catalyzed Allylic
Alkylation
Synthesis of Complexes 4 and 5
Allylic alkylation of rac-3-acetoxy-1,3-diphenyl-1-propene
In a purged Schlenk tube, 20 mg (0.03 mmol) of [Pd(h3--1,3-
Ph2-C3H3)Cl]2 and 63 mg (0.06 mmol) of ligand (5 or 6) were
dissolved in 10 cm3 of freshly distilled dichloromethane. The
yellow solution was stirred for 30 min at room temperature
and 28 mg of NH4PF6 (0.23 mmol) dissolved in 1 cm3 of di-
chloromethane was then added. After 1 hour of stirring,
washings with degassed water were carried out (4”5 cm3).
The organic phase was dried over anhydrous Na2SO4, fil-
tered off and the solvent removed under reduced pressure.
The yellow solid obtained was washed with diethyl ether
(3”5 cm3) and dried under reduced pressure. Yield of 4:
81.6 mg (90%); yield of 5: 77.1 mg (85%). Anal. calcd. for
C79H105F6P3O9Pd: C 62.76, H 6.95%; found for 4: C 62.05, H
(I): For molecular catalytic systems containing ligands 1–3,
the catalytic precursor was generated in situ from [Pd
(h3-
A
C3H5)Cl]2 and the appropriate ligand (0.02 mmol of Pd and
0.025 mmol of chiral ligand) dissolved in 2 cm3 of CH2Cl2
and stirred for 30 min before adding the substrate. For col-
loidal catalytic systems, 2.5 mg of preformed nanoparticles,
Pd1–0Pd3, with 0.005 mmol of the appropriate ligand (for 1
and 2: 5.2 mg; for 3: 4.7 mg) were dissolved in 4 cm3 of
CH2Cl2 for 30 min before adding the substrate. rac-3-ace-
toxy-1,3-diphenyl-1-propene (252 mg, 1 mmol), dissolved in
CH2Cl2 (2 cm3), was then added to the catalyst solution fol-
lowed by dimethyl malonate (396 mg, 3 mmols), BSA
(610 mg, 3 mmols), and a catalytic amount of KOAc. The
mixture was stirred at room temperature until total conver-
sion of substrate (monitored by TLC). Then, the solution
was diluted with diethyl ether, filtered over celite, and
washed with saturated ammonium chloride solution (4”
10 cm3) and water (4”10 cm3). The organic phase was dried
over anhydrous Na2SO4, filtered off, and solvent removed
under reduced pressure. Conversions were determined by
1H NMR and enantiomeric excesses determined by HPLC
on a Chiralcel OD column, using hexane/2-propanol, 99/1,
as eluent, in a flow of 0.3 cm3minÀ1.
Allylic alkylation of rac-3-acetoxy-1-cyclohexene (III):
The procedure was analogous to the one described for the
rac-3-acetoxy-1,3-diphenyl-1-propene. Conversions were de-
termined by 1H NMR and enantiomeric excesses were deter-
mined by GC on a FS-cyclodex-b-I/P column.
Allylic alkylation of rac-3-acetoxy-1,3-dimethyl-1-pro-
pene (V): The procedure was analogous to the one described
for the rac-3-acetoxy-1,3-diphenyl-1-propene. Conversions
were determined by 1H NMR and enantiomeric excesses
were determined by polarimetry.
7.10%; found for 5: C 62.45, H 6.80%. IR (KBr): for 4: n=
À1
À
À
À
1070 (st, P O), 835 (st, P F) cm ; for 5: n=1090 (st, P O),
À1
À
801 (st, P F) cm . MS (MALDI positive): m/z=1365.6
([MÀPF6]+) for both complexes. 31P NMR (CDCl3,
101.3 MHz, 298 K) for 4: major species d=129.6 (d,
190.7 Hz) and 128.2 (d, 190.7 Hz); minor species d=128.7
(d, 190.7 Hz) and 126.9 (d, 199.2 Hz); for 5: d=125.5 (d,
196.9 Hz) and 121.8 (d, 196.9 Hz).
General Procedure for the Synthesis of Pd Colloids
In a standard procedure, 160 mg of [Pd2(dba)3] (0.175 mmol)
E
were dissolved under argon at À1108C (ethanol/N2 bath)
and under vigorous magnetic stirring in a solution of
160 mL of THF containing 0.2 equivs./Pd of the chosen
ligand in a Fischer–Porter bottle. The mixture was then pres-
surized under dihydrogen (3 bars) at room temperature. The
colour of the solution turned from purple to black in a few
minutes. The vigorous stirring and the dihydrogen pressure
were maintained for 18 h at room temperature, leading to
black and homogeneous colloidal solutions. After depres-
surization, a drop of each colloidal solution was deposited
under argon on a holey carbon-covered copper grid for
TEM analysis. The colloidal solution was then concentrated
to ca. 5 mL. Addition of cold pentane allowed the precipita-
tion of the particles as black solids which were washed with
pentane (2”40 mL) and dried under vacuum. The colloids
were characterized by TEM analysis, elemental analysis and
WAXS. Size histograms could not be built for colloid Pd3
because of the close proximity of the particles.
Acknowledgements
The authors thank Ministerio de Educación y Ciencia
(CTQ2004–01546/BQU,
CTQ2004–04412/BQU
and
CTQ2005–03124, Consolider Ingenio 2010, CSD2006–0003),
Generalitat de Catalunya (2005SGR007777), CNRS (PICS
N82428) and RØgion Midi-PyrØnØes (CTP N803007519) for
financial support, and Pierre Lecante (CEMES-CNRS) for
WAXS analysis. M.R.A. thanks Fons Social Europeu for a
predoctoral grant.
Colloid Pd1 was synthesized according to the standard
procedure from 160 mg of [Pd2(dba)3] (0.175 mmol) and
R
74.7 mg of diphosphite 1 (0.070 mmol). TEM: mean size
4.2 nm; eemental analysis: 81.72% Pd, 0.17% P, 14.68% C,
0.94% H.
Colloid Pd2 was synthesizd according to the standard
procedure from 160 mg of [Pd2
A
74.7 mg of diphosphite 2 (0.070 mmol). TEM: mean size
4.2 nm; elemental analysis: 72.97% Pd, 0.37% P, 13.99% C,
0.38% H.
Colloid Pd3 was synthesized according to the standard
procedure from 160 mg of [Pd2(dba)3] (0.175 mmol) and
A
2468
asc.wiley-vch.de
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2007, 349, 2459 – 2469

Towards a Discrimination between Molecular and Colloidal Catalysts
FULL PAPERS
Dekker, New York, 2002; c) Nanoparticles. From
theory to application, (Ed.: G. Schmid), Wiley-VCH,
Weinheim, 2004.
Cassol, A. P. Umpierre, G. Machado, S. I. Wolke, J.
Dupont, J. Am. Chem. Soc. 2005, 127, 3298–3299;
f) J. G. de Vries, Dalton Trans. 2006, 421–429.
[8] a) C. Pan, K. Pelzer, K. Philippot, B. Chaudret, F. Das-
senoy, P. Lecante, M.-J. Casanove, J. Am. Chem. Soc.
2001, 123, 7584–7593; b) K. Philippot, B. Chaudret, C.
R. Chimie 2003, 6, 1019–1034; c) M. Gómez, K. Philip-
pot, V. Colli›re, P. Lecante, G. Muller, B. Chaudret,
New J. Chem. 2003, 27, 114–120; d) E. Ramirez, S.
Jansat, K. Philippot, P. Lecante, M. Gómez, A.
Masdeu, B. Chaudret, J. Organomet. Chem. 2004, 689,
4601–4610.
[9] T. Pery, K. Pelzer, G. Buntkowsky, K. Philippot, H.-H.
Limbach, B. Chaudret, ChemPhysChem. 2005, 6, 605–
607.
[10] a) M. DiØguez, O. Pàmies, C. Claver, Chem. Rev. 2004,
104, 3189–3215; b) M. DiØguez, O. Pàmies, A. Ruiz, Y.
Diaz, S. Castillón, C. Claver, Coord. Chem. Rev. 2004,
248, 2165–2192; c) S. Castillón, C. Claver, Y. Diaz,
Chem. Soc. Rev. 2005, 34, 702–713.
[2] a) J. D. Aiken III, R. G. Finke, J. Mol. Catal. A: Chemi-
cal A 1999, 145, 1–44; b) M. A. El-Sayed, Acc. Chem.
Res. 2001, 34, 257–264; c) H. Bçnnemann, R. M. Ri-
chards, Eur. J. Inorg. Chem. 2001, 2455–2480; d) A.
Roucoux, J. Schulz, H. Patin, Chem. Rev. 2002, 102,
3757–3778; e) J. Dupont, G. S. Fonseca, A. P. Um-
pierre, P. F. P. Fichtner, S. R. Teixeira, J. Am. Chem.
Soc. 2002, 124, 4228–4229; f) A. Roucoux, K. Philip-
pot, Hydrogenation with noble metal nanoparticles, in:
Handbook of Homogenous Hydrogenations, (Ed.: G.
de Vries), Wiley-VCH, Weinheim, 2006; g) S. Jansat, D.
Picurelli, K. Pelzer, K. Philippot, M. Gómez, G. Muller,
P. Lecante, B. Chaudret, New J. Chem. 2006, 30, 115–
122; h) D. Astruc, F. Lu, J. Ruiz Aranzaes, Angew.
Chem. Int. Ed. 2005, 44, 7852–7872; i) B. Karimi, S.
Abedi, J. H. Clark, V. Budarin, Angew. Chem. Int. Ed.
2006, 45, 4776–4779.
[3] a) I. W. Davies, L. Matty, D. L. Hughes, P. J. Reider, J.
Am. Chem. Soc. 2001, 123, 10139–10140; b) J. A. Wi-
degren, R. G. Finke, J. Mol. Catal. A: Chemical 2003,
191, 187–207.
[11] M. DiØguez, O. Pàmies, A. Ruiz, S. Castillón, C.
Claver, Chem. Eur. J. 2001, 7, 3086–3094.
[12] M. DiØguez, A. Ruiz, C. Claver, J. Org. Chem. 2002,
67, 3796–3801.
[4] a) Comprehensive Asymmetric Catalysis, Vols. I–III,
(Eds.: E. N. Jacobsen, A. Pfaltz, H. Yamamoto),
Springer, Berlin, 1999; b) Catalytic Asymmetric Synthe-
sis, (Ed.: I. Ojima), 2nd edn., Wiley, New York, 2000.
[5] a) M. Studer, H.-U. Blaser, C. Exner, Adv. Synth. Catal.
2003, 345, 45–65; b) H. Bçnnemann, G. A. Braun,
Angew. Chem. Int. Ed. Engl. 1996, 35, 1992–1995;
c) H. Bçnnemann, G. A. Braun, Chem. Eur. J. 1997, 3,
1200–1202; d) X. Zuo, H. Liu, D. Guo, X. Yang, Tetra-
hedron 1999, 55, 7787–7804; e) J. U. Kçhler, J. S. Brad-
ley, Catal. Lett. 1997, 45, 203–208; f) J. U. Kçhler, J. S.
Bradley, Langmuir 1998, 14, 2730–2735.
[13] S. Jansat, M. Gómez, K. Philippot, G. Muller, E. Guiu,
C. Claver, S. Castillón, B. Chaudret, J. Am. Chem. Soc.
2004, 126, 1592–1593.
[14] Ligand 1: a) G. J. H. Buisman, M. E. Martin, E. J. Vos,
A. Klootwijk, P. C. J. Kamer, P. W. N. M. van Leeuwen,
Tetrahedron: Asymmetry 1995, 6, 719–738; ligand 2:
b) O. Pàmies, G. Net, A. Ruiz, C. Claver, Tetrahedron:
Asymmetry 1999, 10, 2007–2014; ligand 3: c) J. E.
Babin, G. T. Whiteker, World Patent WO93/03839,
1993.
[15] a) P. von Matt, G. C. Lloyd-Jones, A. B. E. Minidis, A.
Pfaltz, L. Macko, M. Neuburger, M. Zehnder, H. Rꢁeg-
ger, P. S. Pregosin, Helv. Chim. Acta 1995, 78, 265–284;
b) M. Gómez, S. Jansat, G. Muller, M. A. Maestro, J.
Mahía, Organometallics 2002, 21, 1077–1087.
[16] B. M. Trost, D. J. Murphy, Organometallics 1985, 4,
1143–1145.
[6] M. Tamura, H. Fujihara, J. Am. Chem. Soc. 2003, 125,
15742–15743.
[7] a) M. T. Reetz, R. Breinbauer, K. Wanninger, Tetrahe-
dron Lett. 1996, 37, 4499–4502; b) M. T. Reetz, E.
Westermann, Angew. Chem. Int. Ed. 2000, 39, 165–
168; c) M. Moreno-MaÇas, R. Pleixats, S. Villarroya,
Organometallics 2001, 20, 4524–4528 d) C. Rocaboy,
J. A. Gladysz, Org. Lett. 2002, 4, 1993–1996; e) C. C.
[17] Y. Tatsuno, T. Yoshida, S. Otsuka, Inorg. Synth. 1990,
28, 342–346.
Adv. Synth. Catal. 2007, 349, 2459 – 2469
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
2469