Synthesis and characterization of palladium(II) complexes with chiral aminophosphine ligands: Catalytic behaviour in asymmetric hydrovinylation. Crystal structure of cis-[PdCl2(PPh((R)-NHCHCH3Ph)2)2]

Article abstract of DOI:10.1016/j.jorganchem.2007.02.020

Optically active ligands of type Ph₂PNHR (R = (R)-CHCH₃Ph, (a); (R)-CHCH₃Cy, (b); (R)-CHCH₃Naph, (c)) and PhP(NHR)₂ (R = (R)-CHCH₃Ph, (d); (R)-CHCH₃Cy, (e)) with a stereogenic carbon atom in the R substituent were synthesized. Reaction with [PdCl₂(COD)₂] produced [PdCl₂P₂] (1) (P = PhP(NHCHCH₃Ph)₂), whose molecular structure determined by X-ray diffraction showed cis disposition for the ligands. All nitrogen atoms of amino groups adopted S configuration. The new ligands reacted with allylic dimeric palladium compound [Pd(η³-2-methylallyl)Cl]₂ to gave neutral aminophosphine complexes [Pd(η³-2-methylallyl)ClP] (2a-2e) or cationic aminophosphine complexes [Pd(η³-2-methylallyl)P₂]BF₄ (3a-3e) in the presence of the stoichiometric amount of AgBF₄. Cationic complexes [Pd(η⁴3-2-methylallyl)(NCCH₃)P]BF₄ (4a-4e) were prepared in solution to be used as precursors in the catalytic hydrovinylation of styrene. ³¹P NMR spectroscopy showed the existence of an equilibrium between the expected cationic mixed complexes 4, the symmetrical cationic complexes [Pd(η³-2-methylallyl)P₂]BF₄ (3) and [Pd(η³-2-methylallyl)(NCCH₃)₂]BF₄ (5) coming from the symmetrization reaction. The extension of the process was studied with the aminophosphines (a-e) as well as with nonchiral monodentate phosphines (PCy₃ (f), PBn₃ (g), PPh₃ (h), PMe₂Ph (i)) showing a good match between the extension of the symmetrization and the size of the phosphine ligand. We studied the influence of such equilibria in the hydrovinylation of styrene because the behaviour of catalytic precursors can be modified substantially when prepared 'in situ'. While compounds 3 and bisacetonitrile complex 5 were not active as catalysts, the [Pd(η³-2-methylallyl)(η²-styrene)₂]⁺ species formed in the absence of acetonitrile showed some activity in the formation of codimers and dimers. Hydrovinylation reaction between styrene and ethylene was tested using catalytic precursors solutions of [Pd(η³-2-methylallyl)LP]BF₄ ionic species (L = CH₃CN or styrene) showing moderate activity and good selectivity. Better activities but lower selectivities were found when L = styrene. Only in the case of the precursor containing Ph₂PNHCHCH₃Ph (a) ligand was some enantiodiscrimination (10%) found.

Full text of DOI:10.1016/j.jorganchem.2007.02.020

Journal of Organometallic Chemistry 692 (2007) 4005–4019
www.elsevier.com/locate/jorganchem
Synthesis and characterization of palladium(II) complexes with chiral
aminophosphine ligands: Catalytic behaviour in asymmetric
hydrovinylation. Crystal structure of
cis-[PdCl2(PPh((R)-NHCHCH3Ph)2)2]
a
a
a
b
´
Rosa M. Ceder , Carlos Garcıa , Arnald Grabulosa , Fatma Karipcin ,
a
a,
c
c
*
`
`
´
Guillermo Muller , Merce Rocamora , Merce Font-Bardıa , Xavier Solans
a
` `
´
Departament de Quı´mica Inorganica, Universitat de Barcelona, Martı i Franques 1-11, E-08028 Barcelona, Spain
b
Chemistry Department, Su¨leyman Demirel University, Science and Art Faculty, 32260 Isparta, Turkey
`
c
´
`
Departament de CristalÆlografia i Diposits Minerals, Universitat de Barcelona, Martı i Franques, s/n, E-08028 Barcelona, Spain
Received 18 January 2007; received in revised form 19 February 2007; accepted 19 February 2007
Available online 24 February 2007
Abstract
Optically active ligands of type Ph2PNHR (R = (R)-CHCH3Ph, (a); (R)-CHCH3Cy, (b); (R)-CHCH3Naph, (c)) and PhP(NHR)2
(R = (R)-CHCH3Ph, (d); (R)-CHCH3Cy, (e)) with a stereogenic carbon atom in the R substituent were synthesized. Reaction with
[PdCl2(COD)2] produced [PdCl2P2] (1) (P = PhP(NHCHCH3Ph)2), whose molecular structure determined by X-ray diffraction showed
cis disposition for the ligands. All nitrogen atoms of amino groups adopted S configuration. The new ligands reacted with allylic dimeric
palladium compound [Pd(g3-2-methylallyl)Cl]2 to gave neutral aminophosphine complexes [Pd(g3-2-methylallyl)ClP] (2a–2e) or cationic
aminophosphine complexes [Pd(g3-2-methylallyl)P2]BF4 (3a–3e) in the presence of the stoichiometric amount of AgBF4. Cationic com-
plexes [Pd(g3-2-methylallyl)(NCCH3)P]BF4 (4a–4e) were prepared in solution to be used as precursors in the catalytic hydrovinylation of
styrene. 31P NMR spectroscopy showed the existence of an equilibrium between the expected cationic mixed complexes 4, the symmet-
rical cationic complexes [Pd(g3-2-methylallyl)P2]BF4 (3) and [Pd(g3-2-methylallyl)(NCCH3)2]BF4 (5) coming from the symmetrization
reaction. The extension of the process was studied with the aminophosphines (a–e) as well as with nonchiral monodentate phosphines
(PCy3 (f), PBn3 (g), PPh3 (h), PMe2Ph (i)) showing a good match between the extension of the symmetrization and the size of the phos-
phine ligand. We studied the influence of such equilibria in the hydrovinylation of styrene because the behaviour of catalytic precursors
can be modified substantially when prepared ‘in situ’. While compounds 3 and bisacetonitrile complex 5 were not active as catalysts, the
[Pd(g3-2-methylallyl)(g2-styrene)2]+ species formed in the absence of acetonitrile showed some activity in the formation of codimers and
dimers. Hydrovinylation reaction between styrene and ethylene was tested using catalytic precursors solutions of [Pd(g3-2-methylal-
lyl)LP]BF4 ionic species (L = CH3CN or styrene) showing moderate activity and good selectivity. Better activities but lower selectivities
were found when L = styrene. Only in the case of the precursor containing Ph2PNHCHCH3Ph (a) ligand was some enantiodiscrimina-
tion (10%) found.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Chiral aminophosphine ligands; Allylic palladium (II) complexes; Symmetrization reaction; Hydrovinylation reaction
1. Introduction
The catalytic asymmetric hydrovinylation of olefins is
an important stereoselective carbon–carbon bond-form-
*
ing reaction in organic synthesis [1]. The reaction is cat-
Corresponding author. Tel.: +34934039135; fax: +34934907725.
alyzed by a variety of transition metal compounds being
E-mail address: merce.rocamora@qi.ub.es (M. Rocamora).
0022-328X/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2007.02.020

4006
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
nickel and palladium the most frequently used. The
scope of the reaction is limited by the nature of the ole-
fin since excellent stereoselectivity is achieved only with
conjugated dienes, strained olefins or intramolecular
reactions [2]. The catalytic cycle proposed on nickel
and palladium systems shows that the active catalyst is
most likely an unsaturated phosphine-stabilized metal-
hydride. Allylic precursors stabilized with one monoden-
tate phosphine proved to be very useful since cleavage of
the allyl substituent leads to the actual catalyst that ini-
tiates the heterodimerization. The model reaction of the
process involves codimerization between styrene or vinyl-
naphtalene and ethylene (Eq. (1)). When these types of
vinylarenes are used as prochiral olefins, the excellent
regioselectivities obtained originate in the allylic nature
of the intermediate.
may be constructed in large quantities through the use of
relatively simple condensation processes, and from com-
mercial starting materials. Woolins and coworkers have
reported a number of examples of aminophosphines
including these derived from 1,2-diaminobenzene [15],
aminopyridine [16] and several diamines [17]. Burrows
has reported the synthesis of ether functionalized amino-
phosphines [18] and aminophosphines derived from methyl
benzyl amine are also well documented [19]. There are indi-
cations of the potential utility of transition metal com-
plexes with aminophosphines. For example rhodium (I)
[20] and platinum (II) [21] complexes of chiral aminophos-
phines have proved to be efficient catalysts for asymmetric
hydrogenation and hydroformylation reactions respec-
tively, and nickel complexes have been employed in the cyc-
lodimerization of buta-1,3-diene [22]. Some preliminary
H
Me
[cat]
+
+
+
+
+
ð1Þ
(Z) C
2-phenyl-2-butene
(R or S) A
3-phenyl-1-butene
(E) B
cis-1,3-diphenyl-1-butene
1-butene
However the control of the enantioselectivity of the reac-
tion is more difficult. Very good enantioselectivities were
initially obtained by Wilke [3] and more recently by the
groups of Vogt [4], Gibson [5], Rajanbabu [6,7], Leitner
[8] and Zhou [9] using phosphines, planar chromium
phosphines, phosphinites, or phosphoramidites as stabi-
lizing ligands with nickel and palladium systems. We
investigated the model reaction using nickel and palla-
dium precursors. The initial work provided evidence of
the blocking effect of inert bidentate ligands [10]. Further
research with allylic palladium precursors containing P-
stereogenic phosphines enabled the enantioselective ver-
sion of the process to be studied [11,12].
Synthesis of ligands with specific electronic and steric
requirements is an essential part of an asymmetric catalysis
program that relies on ligand tuning to bring about opti-
mum results. Although there now exists a huge body of
work on the chemistry of both phosphines and phosphites,
aminophosphines containing P–N bonds instead of P–C or
P–O bonds have received a lower level of attention. One
potential reason for the underdevelopment of this chemis-
try is that the P–N bond is envisaged as being susceptible
to relatively easily cleavage [13] but several new classes
of P–O and P–N bond containing phosphorous ligands
of high stability have been demonstrated to be capable of
the acceleration and asymmetric catalysis of a number of
synthetic organic reactions [14]. Moreover, these ligands
studies on the use of bidentate aminophosphine ligands
in palladium catalysed allylic alkylation reactions are
reported [23].
We decided to focus the work on the synthesis of
cationic allylic palladium (II) complexes with chiral
monoamino and bisaminoarylphosphines containing one
stereogenic carbon atom in the amino substituent, and
use them as catalytic precursors for hydrovinylation asym-
metric reaction. It is well known that the factors determin-
ing the discrimination ability of the ligands in this process
remain poorly defined. The final efficiency of the reaction is
determined by the delicate competition between the olefins
present as well as all the species contained in the reaction
medium. The fully characterization of the precursors solu-
tions indicate that the cationic allylic palladium (II) com-
plexes undergo symmetrization leading to mixtures of
three different species
2½Pdðg3-2-CH3C3H5ÞPLꢀþ
ð2Þ
ꢀ
½Pdðg3-2-CH3C3H5ÞP2ꢀþ þ ½Pdðg3-2-CH3C3H5ÞL2ꢀþ
The relative amount of each one, using different tertiary
phosphines and aminophosphines, seems related to the ste-
ric parametres of the phosphine ligands. We evaluated the
behaviour of each one in the catalytic hydrovinylation
process.

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4007
1
2. Results and discussion
erocorrelation NMR experiment allowed us to assign H
NMR spectra unequivocally. H spectra of monoamino-
1
2.1. Synthesis and characterization of the aminophosphines
(a–e)
phosphine ligands (a, b and c) show a broad singlet for the
aminic hydrogen, a multiplet for methinic proton and a dou-
blet for the methyl proton of the amino group. 1H spectra of
the bisaminophosphine ligands (d, e) show two signals for
CH3 groups coupled to the methinic proton, two complex
multiplets for the CH proton and two different broad signals
for the NH proton of amino groups coupled to the phospho-
rus atom as confirmed by 1H–{31P} experiments. It has been
noted that hindered rotation about the P–N bond may exist
as reported in the literature [27]. It is known that monoami-
nophosphines and bisaminophosphines bearing primary
amino groups exist in tautomeric equilibrium with PH-imi-
nophosphines and that bisaminophosphines also are amena-
ble to condensation [27]. NMR spectra allowed us to
observe that freshly prepared samples did not contain signif-
icant amounts of condensation products and that the
position of the prototropic equilibrium is displaced signifi-
cantly towards the NH form.
The monoaminophosphines Ph2P(NHR) were conve-
niently prepared as reported in the literature [19,24]
through the reaction of chlorodiphenylphospine with the
chiral primary amines NH2R (R = (R)-CHCH3Ph (a),
(R)-CHCH3Cy (b), (R)-CHCH3Naph (c)) in the presence
of triethylamine. Bisaminophosphines PhP(NHR)2 (R =
(R)-CHCH3Ph (d), (R)-CHCH3Cy (e)) were prepared by
the same way but starting from the dichlorophenylphos-
phine and the stoichiometric amount of the primary amine
[25] (Scheme 1).
Aminophosphines (a–e) were isolated as spectroscopi-
cally pure colourless oils in very good yields (80–90%)
and can be stored for a long time under inert atmosphere.
Relevant NMR data are shown in Table 1.
31P NMR shows that all compounds exhibit a single sig-
nal and the chemical shift increases from 36 ppm to 59 ppm
depending on the presence of one or two amino groups
bonded to the phosphorus atom but does not change with
the different substituents on the amino group (phenyl, cyclo-
hexyl and naphthyl). The observed chemical shift values
compared with those for PPh3 (ꢁ5 ppm) and for the trisami-
nophosphine P(NHCHCH3Ph)3 (104.8 ppm) depicted by
Kolodiaznyi and Prynada [25] suggest less basicity and big-
ger cone angle [26] when substitution increases. 1H–13C het-
2.2. Synthesis and characterization of the palladium (II)
complexes
2.2.1. Neutral complex cis-[PdCl2(PPh((R)-NHCH-
CH3Ph)2)2] (1)
The reaction of two equivalents of bisaminophosphine
(d) with [PdCl2(COD)] in dichloromethane gave nearly
quantitative yields of [PdCl2P2] (1). The 31P NMR spec-
trum of the palladium complex showed one phosphorus
resonance at 57.1 ppm according to the formation of only
1
one isomer. H NMR spectrum showed two different dou-
i
) toluene, 0 ºC, Ph2PCl
blets for the methyl group, two doublet of doublets for the
aminic protons and two multiplets for CH protons as
shown in Table 1, according to the nonequivalence of the
two aminosubstituents of the same phosphorus atom
observed in the free phosphine. NOESY experiment
allowed a complete assignment of the protons of each
amino group; NOE contacts between protons at 4.8, 3.5
and 1.14 ppm and between protons at 3.26, 4.6 and
1.37 ppm indicate that each set belongs to different amino
group of the same phosphine ligand.
ii) r.t., 90 m
+
H2NR
Et3N
+
Ph2PNHR
Et3N·HCl
R=(
R
)-CHCH3Ph (a)
(
(
R
)-CHCH3Cy (b)
R
)-CHCH3Naph (c)
i
) toluene, 0 ºC, PhPCl2
ii) r.t., 90 m
+
2 H2NR 2 Et3N
+
PhP(NHR)2
2 Et3N·HCl
R=( )-CHCH3Ph (d)
R
(R)-CHCH3Cy (e)
Scheme 1.
Table 1
Selected NMR dataa (d ppm, in CDCl3, 298 K) for aminophosphine ligands (a–e) and [PdCl2(PhP(NHCHCH3Ph)2)2] (1) complex
Compound
d31
P
d1H(NH)
d1H(CH)
d1H(CH3)
Ph2PNHCHCH3Ph (a)
Ph2PNHCHCH3Cy (b)
Ph2PNHCHCH3Naph (c)
PhP(NHCHCH3Ph)2(d)
36.0 (s)
36.4 (s)
36.5 (s)
59.1 (s)
2.18 (bs)
1.70–1.55
2.26 (bs)
2.42 (bd; 6.7)
2.53 (bt; 7.2)
2.05 (bd; 8.2)
1.92 (bd; 6.7)
4.80 (dd; 11.6, 9.6)
3.26 (dd; 9.2, 5.6)
4.22 (m)
2.93 (m)
5.02 (m)
4.18–4.02
1.39 (d; 7.0)
1.06 (d; 6.8)
1.52 (d; 7.0)
1.31 (d; 6.6)
1.48 (d; 6.6)
1.11(pt; 6.0)
PhP(NHCHCH3Cy)2 (e)
60.7 (s)
57.1 (s)
3.05–2.97
[PdCl2P2] (1)
3.50 (m)
4.6 (m)
1.14 (d; 7.2)
1.37 (d; 6.8)
a
Multiplicity: s, singlet; d, doublet; t, triplet; pt, pseudotriplet; m, multiplet; b, broad signal. Coupling constant are given in Hz and are shown in
parentheses after the multiplicity. 1H NMR: 400 Hz. 31P–{1H} NMR: 100.56 Hz.

4008
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
Recrystallization from CH2Cl2/hexane mixture gave
cis isomer exclusively. The values of the angles around the
nitrogen atoms are consistent with significant sp2 character
in good agreement with the partial double bond character of
the P–N bond. The structure of the complex shows the pres-
ence of an intramolecular hydrogen bond between N(2) ami-
orange crystals of [PdCl2P2] (1) suitable for X-ray analysis.
The crystal structure indicate the cis geometry of the
ligands and retention of the R configuration for the stereo-
genic amine carbon atom. All stereogenic nitrogen atoms
show the same S configuration. Selected bond lengths
and angles are given in Table 2, and molecular structure
is shown in Fig. 1.
˚
nic proton and the chloride ligand Cl(2) (2.54(2) A), as well
˚
as between the N(3) aminic proton and Cl(1) (2.24(3) A).
Hydrogen bonds have been observed in a number of other
complexes containing mutually cis aminophosphine and
chloride ligands, and in these previous examples the H ꢂ ꢂ ꢂ Cl
The metallic centre is distorted square-planar, with cis
angles values shown in Table 2. The P–N distances are
shorter than the generally accepted range for P–N single
˚
distances range from 2.1 to 2.5 A [18a,29].
˚
bond (e.g. 1.689–1.727 A in piperidinophosphines [28]), sug-
gesting a degree of double bond character as described [18a].
This situation is generally observed in complexes of
aminophosphines, and a search of the Cambridge Structural
Database revealed the average P–N distance for aminophos-
2.2.2. Neutral allyl complexes [PdCl(g3-2-CH3C3H4)P] (2)
Neutral allyl complexes [PdCl(g3-2-CH3C3H4)P] (2a–2e)
were prepared by reaction at low temperature of the well-
known bridged chloride complex [Pd(g3-2-CH3C3H4)
(l-Cl)]2 with the appropriate ligand as reported in the
literature for similar compounds [12,30,31] (Scheme 2).
Complexes 2 are stable in the solid state and were fully
characterized by the usual techniques. Table 3 collect
NMR data for those complexes.
˚
phines of type Ph2NHR (R = alkyl or aryl) to be 1.67 A
˚
(range 1.63–1.73 A) [25]. A significant p-stacking interac-
tion involves the phenyl rings bonded to the phosphorus
atom; the dihedral angle between the mean planes of the
p-stacked aromatic rings is 10.7(2)ꢁ and the distance
between the centre of the rings is 3.866 A. The existence of
31P NMR spectra for allyl complexes showed two sharp
signals arising from the two isomers formed due to the
presence of a stereogenic carbon atom on the amino group.
The ratio of isomers, depicted in Scheme 2, depends on the
substituents in the stereogenic carbon atom of the amino
groups. Naphthyl substituent in the monoaminophosphine
and the phenyl substituent in bisaminophosphine lead to
some discrimination between the two isomers. The phos-
phorus chemical shifts of complexes containing mono-
aminophosphino ligand (2a, 2b and 2c) lie downfield
˚
this p-stacking interaction may favour the presence of the
Table 2
˚
Selected bond lengths (A) and angles (ꢁ) for cis-[PdCl2P2] (1)
Pd–P(1)
Pd–P(2)
Pd–Cl(1)
Pd–Cl(2)
P(1)–N(3)
N(3)–C(23)
2.2535(17)
2.2605(14)
2.3652(18)
2.3745(16)
1.637(5)
P(2)–N(1)
P(2)–N(2)
N(1)–C(1)
N(2)–C(9)
P(1)–N(4)
N(4)–C(31)
1.636(4)
1.676(5)
1.492(6)
1.501(8)
1.682(9)
1.484(9)
1.414(8)
P(1)–Pd–P(2)
92.42(5)
87.42(7)
129.1(4)
120.7(4)
P(1)–Pd–Cl(1)
P(2)–Pd–Cl(2)
P(1)–N(4)–C(31)
P(2)–N(1)–C(1)
91.13(6)
89.04(6)
125.0(6)
126.3(3)
Cl(1)–Pd–Cl(2)
P(1)–N(3)–C(23)
P(2)–N(2)–C(9)
P
-
+
BF4
Pd
H3C
P
CH2Cl2, 4P
3a- 3i
AgBF4 or TlPF6
P = a, b, c, d, e,
f (PCy3)
g (PBn3)
[Pd(2-CH3C3H4)Cl]2
h (PPh3)
i (PMe2Ph)
CH2Cl2, 2P
P
Pd
H3C
Cl
2a-2i
't
CH3
H
H'c
anti
anti
Htsyn
Hc
syn
H't
H'c
syn
Htanti
Pd
syn
Hc
anti
CH3
Cl Pd
Cl
ratio isomers
P
P
2a 1 / 1.14
2b 1 / 1.09
2c 1 / 1.25
2d 1 / 1.85
2e 1 / 1.13
Fig. 1. Molecular structure of complex 1.
Scheme 2.

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4009
Table 3
Selected NMR dataa (d ppm, in CDCl3, 298 K) for neutral complexes (2a–2i) and ionic complex (4f)
Complex d31
P
d1H (Htsyn)
d1H (Htanti) d1H (Hcsyn) d1H (Hcanti) d1H (CH3)b d1H (NH)
d1H (CH) d1H (Me)c
2a
Minor
Major
54.6 (s)
55.1 (s)
4.48 (bs)
4.48 (bs)
3.42 (d; 10.4) 3.02 (bs)
3.42 (d; 10.4) 3.02 (bs)
2.37 (s)
2.41 (s)
1.84 (s)
1.84 (s)
4.59 (dd; 19.2,2.4) 4.05 (m)
4.64 (dd; 19.2,4.4) 4.05 (m)
1.18 (d; 6.8)
1.32 (d; 6.8)
2b
Minor
Major
53.63 (s)d 4.37 (m)
53.35 (s)d 4.37 (m)
3.33 (d; 4)
3.30 (d; 4)
2.93 (bs)
2.85 (bs)
2.34 (s)
2.25 (s)
1.81 (s)
1.71 (s)
4.03 (m)
4.03 (m)
2.73 (m)
2.73 (m)
0.85 (d; 6.8)
0.71 (d; 6.8)
2c
Minor
Major
53.9 (s)
54.8 (s)
4.4 (m)
4.4 (m)
3.37 (d; 10)
3.39 (d; 10)
2.92 (bs)
2.99 (bs)
2.27 (s)
2.41 (s)
1.75 (s)
1.82 (s)
4.83 (m)
4.83 (m)
4.83 (m)
4.83 (m)
1.29 (d; 6.4)
1.43 (d; 6.4)
2d
Minor
64.1 (s)
63.3 (s)
4.34–4.24
3.25 (d; 11.2) 2.65 (s)
2.12 (s)
1.40 (s)
1.6–1.5
1.6–1.5
3.7 (bd; 6.8)
4.3 (m)
4.34–4.24
3.52–3.42
4.52–4.4
4.20–4.15
4.76–4.64 1.6–1.5
4.14–4.06
1.6–1.5
Major
4.17 (dd; 7.6, 2.24) 2.83 (d; 11)
2.34 (s)
2e
Isomer I
62.7 (s)
4.35–4.20
4.35–4.28
3.34 (d; 6.8)
3.31 (d; 7.2)
2.75 (s)
2.64 (s)
2.17 (s)
2.33 (s)
0.9–1.4
0.9–1.4
3.78 (bt; 10.8)
3.02 (d; 10)
3.62 (bt; 10)
2.87 (d; 2.4)
3.0–2.9
3.3–3.42
3.2–3.3
3.0–2.9
0.9–1.4
0.9–1.4
Isomer II 63 (s)
2f
2g
2h
2i
4f
a
41.2 (s)
21.6 (s)
22.4 (s)
ꢁ3.9 (s)
42.5 (s)
4.39 (m)
4.41 (s)
4.53 (bs)
4.32 (d; 6.7)
4.95 (m)
3.45 (d; 8.8)
3.28 (m)
3.59 (bs)
3.35 (d; 10.2) 3.05 (s)
3.66 (d; 8.4) 3.26 (bs)
3.15 (bs)
2.55 (s)
2.86 (s)
2.51 (s)
1.68 (s)
2.74 (s)
2.48 (s)
2.61 (bs)
1.94 (s)
1.61 (s)
1.97 (s)
1.83 (s)
1.99 (s)
–
–
–
–
–
–
–
–
–
–
–
–
2.47 (s)e
Multiplicity: s, singlet; d, doublet; t, triplet; m, multiplet; b, broad signal. Coupling constant are given in Hz and are shown in parentheses after the
multiplicity. 1H NMR: 400 Hz. 31P–{1H} NMR: 100.56 Hz. See Scheme 2 for atom labels.
b
Allylic methyl group.
Aminic methyl group.
c
d
273 K.
e
Acetonitrile methyl group.
compared to that of the free ligands, but complexes con-
tive opening of the palladium carbon bond trans to the
taining bisaminophosphine (2d, 2e) ligands show similar
chemical shifts to that in free ligand. H NMR spectra of
phosphorus atom. This has been observed in analogous
compounds [12,32] and suggests the well known g3–g1–
g3 isomerization process. Also cross-peaks that could be
assigned to the whole apparent allyl rotation were
observed. Exchange signals between hydrogens of the
amino group of different isomers in 2e complex are also
detected as expected.
1
these compounds showed two sets of signals for each syn
and anti protons confirming the presence of two isomers.
Proton resonances of the allylic group were assigned by
comparison with literature values [12,31,32]. Monoamino-
phosphine complexes (2a–2c) show two sets of signals for
the aminic proton but bisaminophosphine complexes (2d,
2e) show two pairs of signals because of the presence of
the two chiral aminophosphine substituent in each isomer.
When the isomeric ratio of palladium (II) diastereoisomers
is nearly 1/1 (2e) the 1H NMR signals cannot be associated
with those of 31P NMR since the 31P–13C or 31P–1H corre-
lations have not been carried out. To elucidate the solution
structure of these complexes and their dynamic behaviour,
2D NOESY experiments were carried out for complexes 2a
and 2e. NOE contacts between allylic protons and the
aminophosphine ligands were not detected but interesting
exchange signals between allylic protons were observed.
Exchange between syn/anti allyl protons was observed
selectively with the pair in cis to the phosphorus atom.
The pair that is in trans position exchange only between
syn/syn and anti/anti protons, in accordance with a selec-
Neutral allylic complexes [PdCl(g3-2-CH3C3H4)P] with
monodentate tertiary phosphines (P = PCy3 (2f), PBn3
(2g), PPh3 (2h), PPhMe2 (2i)) were prepared by the same
procedure reported in the literature [12,30,31,33] as shown
in Scheme 2. Complexes 2f and 2g are described, to the best
1
of our knowledge, for the first time. 31P and H NMR sig-
nals are reported in Table 3.
2.2.3. Cationic allyl complexes [Pd(g3-2-CH3C3H4)P2]BF4
(3)
Ionic allylic complexes [Pd(g3-2-CH3C3H4)P2]BF4 (3a–
3e) were obtained by reaction of the dinuclear [Pd(g3-2-
CH3C3H4)(l-Cl)]2 with 2 equiv. of the appropriate
aminophosphine ligand and subsequent treatment with
AgBF4 to remove the chloride ligand (Scheme 2). By the
same procedure, complexes (3f–3i) were prepared

4010
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
(P = PCy3 (3f), PBn3 (3g), PPh3 (3h), PPhMe2 (3i)), some
of them previously reported in the literature [34]. For 3a,
3f, 3g and 3i complexes the best yields were obtained when
reaction was done at 273K and for 3e compound satisfac-
tory elemental analysis was found using TlPF6 salt instead
of AgBF4. The products were stable solids, fully character-
2.2.4. Cationic mixed complexes [Pd(g3-2-CH3C3H4)
(NCCH3)P]BF4 (4)
Cationic complexes [Pd(g3-2-CH3C3H4)(NCCH3)P]BF4
were prepared by removing the chloride by AgBF4 from
neutral complexes 2 in the presence of acetonitrile ligand.
The reaction was followed by 31P NMR spectroscopy
showing more signals than expected, except for 4f complex.
This suggests that [Pd(g3-2-CH3C3H4)(NCCH3)P]BF4 (4a–
4e, 4g–4i) compounds undergo symmetrization, and give a
mixture of the expected mixed complex 4, the bis(acetoni-
trile) complex [Pd(g3-2-methylallyl)(NCCH3)2]BF4 (5)
[37] and the symmetrical bis(phosphine) complex 3
(Scheme 3). [Pd(g3-2-CH3C3H4)(NCCH3)(PCy3)]BF4 (4f)
complex was isolated and fully characterized in solid state.
Characterization data for that compound are coincident
with those reported in the literature for an analogous cat-
ion stabilized by an ether ligand [33].
1
ized in the solid state and in solution by H and 31P NMR
spectroscopy as shown in Table 4.
Complexes with nonchiral phosphines presented only
one signal in the 31P spectra and one broad signal for each
pair of syn and anti allylic protons in the 1H NMR spectra.
31P NMR spectra of complexes with aminophosphines
show two doublets of an AB spin system with coupling
constants similar to those reported for similar complexes
[35]. The two chiral phosphine ligands in cis position are
formally a C2 system but they loose the symmetry in the
presence of the allylic group causing the two phosphorus
1
atoms to be non-equivalent. H NMR spectra show four
Table 5 shows 31P NMR data of the mixed cationic
complexes 4a–4i solutions. Complexes with aminophos-
phine ligands showed two doublets with chemical shift
coincident with the respective symmetric cationic com-
plexes 3a–3e. For 4a, 4c and 4d complexes two broad sig-
nals of different intensity also appeared suggesting the
presence of the two possible isomers of the mixed cationic
complexes [Pd(g3-2-CH3C3H4)(NCCH3)P]BF4. The ratio
between the different isomers was calculated from integra-
tion signals of 31P NMR spectra and was always nearly 1:1
except for 4d. The same discrimination was detected for the
neutral complex 2d. For complexes 4b and 4e only one
broad signal that can be assigned to the mixed complex
appeared, probably because the chemical shift of the iso-
mers are not different enough to be detected separately.
different signals for the anti and syn allylic hydrogens and
a unique methyl allyl signal reflecting the existence of only
one isomer. 2D NOESY NMR experiment for [Pd(g3-2-
CH3C3H4)(Ph2PNHCHCH3Ph)2]BF4 (3a) only showed
intraligand contacts.
In the preparation of [Pd(2-CH3C3H4)(PCy3)2]BF4 (3f)
1
complex, another compound was detected. H NMR spec-
troscopy indicates the presence of g1-allyltricyclohexyl-
phosphonium salt as confirmed by direct comparison
with an authentic sample obtained by reaction of PCy3
and 2-methylallylchloride [36]. The best yields in the
synthesis of compound 3f are achieved starting from
[Pd(g3-2-methylallyl)(NCCH3)(PCy3)]BF4 as reported by
Brookhart [33].
Table 4
Selected NMR dataa (d ppm, in CDCl3, 298 K) for ionic complexes (3a–3i)
Complex
d31
P
d1H (Hsyn)
d1H (Hanti)
d1H (CH3)b
d1H (NH)
d1H (CH)
d1H (Me)c
3a
57.3 (d; 47)
58.4 (d; 47)
3.62 (bs)
3.68 (bs)
3.23 (t; 9)
1.51 (s)
4.09 (dd; 18,10)
4.20 (dd; 18, 10)
3.79 (m)
3.87 (m)
0.95 (d; 7)
1.03 (d;7)
3b
3c
3d
56.8 (s)
3.63 (s)
3.69 (s)
3.24 (bs)
3.27 (bs)
1.55 (s)
1.46 (s)
2.27 (s)
2.93 (m)
3.00 (m)
2.72 (bs)
0.66 (d; 6.5)
0.75 (d; 6.5)
57.7 (d; 47)
58.3 (d; 47)
3.65 (bs)
3.67 (bs)
3.26 (d; 8)
4.70 (m)
4.53 (m)
0.95(d; 6.4)
1.08 (d; 6.4)
69.2 (d; 58)
68.1 (d; 58)
3.54 (t; 3.5)
3.66 (t; 3.5)
2.52 (d; 11.2)
2.87 (d; 11.2)
3.06 (dd; 8.8, 4.2)
3.14 (t; 9.2)
3.70–3.76
3.70–3.76
3.86–3.97
1.11 (d; 6.8)
1.20 (d; 6.8)
1.25 (d; 6.8)
1.41 (d; 6.8)
3e
65.8 (d; 68.8)
66.4 (d; 68.8)
3.86 (s)
3.77 (s)
3.0 (d; 5.6)
3.3 (d; 6.0)
1.6 (s)
3.2–2.6
3.2–2.6
0.86 (d; 6.4)
0.98 (d; 6.8)
1.02 (d; 6.4)
1.12 (d; 6.8)
3f
3g
3h
3i
a
36.0 (s)
9.7 (s)
23.1 (s)
ꢁ8.4 (s)
4.32 (bs)
3.62 (s)
3.72 (bs)
4.06 (s)
3.08 (bs)
2.16 (s)
3.62 (s)
3.20 (bs)
0.95–2.06
0.98 (s)
1.87 (s)
1.82 (s)
–
–
–
–
–
–
–
–
–
–
–
–
Multiplicity: s, singlet; d, doublet; t, triplet; m, multiplet; b, broad signal. Coupling constant are given in Hz and are shown in parentheses after the
multiplicity. 1H NMR: 400 Hz. 31P–{1H} NMR: 100.56 Hz.
b
Allylic methyl group.
Aminic methyl group.
c

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4011
P
4 / 3
ratio
Pd
H3C
4a / 3a 1 / 0.45
4b / 3b 1 / 0.38
4c / 3c 1 / 0.32
4d / 3d 1 / 0.22
4e / 3e 1 / 0.27
4f / 3f 1 / 0
Cl
2a-2i
, r. t.,
CH2Cl2
1 AgBF4,10 CH3CN
4g / 3g 1 / 0.03
4h / 3h 1 / 0.13
4i / 3i 1 / 0.68
P
P
P
NCCH3
NCCH3
Pd
H3C
Pd
3a- 3i
H3C
+
Pd
H3C
NCCH3
4a-4i
5
Scheme 3.
Table 5
A solution of complexes 4a–4i in CH2Cl2 was reacted
with the stoichiometric amount of AgBF4 in the presence
of a ten fold excess of acetonitrile. According to the results
of the 31P NMR spectra, the crude mixtures contained the
mixed (phosphine/acetonitrile) ionic complex (4) and the
bis(phosphine)ionic complex (3) in a relation shown in
Scheme 3, the bisacetonitrile complex 5 must be also present.
The results show few variation of the symmetrization reac-
tion within complexes (4a–4c) and bisaminophosphine com-
plexes (4d, 4e). The different extension of the symmetrization
reaction observed for mixed cationic complexes with
monodentate tertiary phosphines seems to be related to
the steric over the electronic parameter [26]. The compound
containing the phosphine with the highest value of the cone
angle (PCy3) does not suffer the symmetrization reaction
and, in fact, is the only case where the cationic mixed com-
plex [Pd(g3-2-CH3C3H4)(NCCH3)(PCy3)]BF4 was fully
characterized in the solid state (selected NMR data are
shown in Table 3). The same observation can be described
for complexes 4a–4e, those containing monoaminophos-
phines show a greater tendency to the symmetrization than
those with bisaminophosphines. Moreover, reaction of the
stoichiometric amount of 4c and 5 complexes in CH2Cl2
leads to an equilibrium mixture of 3c, 4c and 5. Symmetriza-
tion reaction has been reported in the literature [31] for com-
plexes [Pd(g3-allyl)L1L2]+ (L1 = PPh3, L2 = NCCH3, py).
The synthesis and characterization of the mixed complex
[Pd(g3-2-CH3C3H4)(NCCH3)(PBnCyPh)]BF4 reported in
the literature [11a] agrees with the statement that symmetri-
zation reaction is nearly unobservable with bulky
phosphines.
Selected 31P NMR dataa (d ppm, 298 K) for CH2Cl2 solution of complexes
[Pd(g3-2-CH3C3H4)(NCCH3)(P)]BF4 (4a–4i)
P
[Pd(g3-2-CH3C3H4)(NCCH3)(P)]+b
[Pd(g3-2-CH3C3H4)(P)2]+
a
52.6 (bs) [1.2]
57.0 (bs) [1.0]
57.4 (d; 47)
58.6 (d; 47)
b
c
57.3 (bs)
58.3 (s)
56.2 (bs) [1.0]
57.0 (bs) [1.0]
57.9 (d; 47)
58.5 (d; 47)
d
e
65.9 (bs) [1.5]
67.9 (bs) [1.0]
69.2 (d; 58)
68.1 (d; 58)
67.1 (bs)
65.8 (d; 68.8)
66.4 (d; 68.8)
f
42.5 (s)
19.8 (bs)
24.9 (bs)
ꢁ4.5 (bs)
–
g
h
i
9.7 (s)
23.3 (s)
ꢁ8.4 (s)
a
Multiplicity: s, singlet; d, doublet; b, broad signal. Coupling constant
are given in Hz and are shown in parentheses after the multiplicity. 31P-
{1H} NMR: 100.56 Hz.
b
Isomer ratio is shown in brackets.
Presumably broad signals are due to the lability of the ace-
tonitrile ligand. Solutions of complexes 4g–4i showed two
broad signals, one of them coincident with the respective
symmetrical complex and the other assigned to the mixed
acetonitrile cationic complex 4g–4i. Solutions of complex
4f only showed one signal corresponding to the mixed com-
plex and no signals of the products coming from the sym-
metrization reaction were observed.
2.3. Study of the symmetrization reaction of complexes
[Pd(g3-2-CH3C3H4)(NCCH3)P]BF4 (4)
3. Hydrovinylation reaction
Solutions of cationic mixed complexes type 4 obtained
‘in situ’, are used as catalytic precursors in several reac-
tions, so the study of the symmetrization equilibrium is
important to evaluate the consequences in the catalytic
processes. The reaction was studied with the chiral phos-
phines a, b, c, d and e, and also with nonchiral phosphines
f, g, h and i in order to relate the steric and/or electronic
effects of phosphine ligand with the extension of the sym-
metrization process.
The solution of mixed allylic cationic complexes (4a–4h)
were tested as catalytic precursors in the hydrovinylation
reaction of styrene with ethylene and the results are shown
in Table 6.
The hydrovinylation reaction was always carried out
using a [Pd]/styrene ratio of 1/1000, CH2Cl2 as solvent,
15 bar of pressure and at 25 ꢁC temperature. The products
of the catalytic process were analyzed after 60 or 360 min

4012
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
Table 6
Hydrovinylation of styrenea
Entry
Phosphine
Time (min)
Conversionb (%)
Codimerc (%A + B + C)
Selectivityd (%A)
TOFe (total Pd)
TOFf(active Pd)
23
1
2
3
4
5
6
7
8
9
ag
60
60
360
60
60
360
60
60
360
60
360
60
360
60
60
60
60
60
120
240
1.6
10.4
76.4
1.6
22.6
97.3
3.2
8.6
72.7
7.8
1.6
9.7
99.8
97.8
69.5
99.5
93.9
34.7
99.6
98.8
72.9
98.4
82.5
99.3
95.2
99.6
94.2
89.8
96.4
86.1
78.5
60.0
16
104
127
16
226
162
32
86
121
78
49
55
33
143
624
777
222
19
ah
ah
74.1
74.1
20.9
93.1
0.7
8.2
71.0
6.6
bg
22
43
bh
bh
cg
ch
ch
10
11
12
13
14
15
16
17
18
19
20
a
dh
dh
29.3
5.5
25.9
4.5
eh
eh
20.1
14.3
62.4
77.7
22.2
1.9
17.8
12.6
61.5
75.9
21.8
1.9
fi
143
643
gg
gh
hg
252
styrenej
styrenej
styrenej
5.5
14.4
5.5
14.4
28
36
Reaction carried out at 25 ꢁC, and 15 bar of initial pressure of ethylene in 10 mL of CH2Cl2; ratio styrene/Pd 1000/1.
Conversion of starting styrene.
Codimer: total amount of codimers: see Eq. (1).
b
c
d
e
f
Selectivity: % of 3-phenyl-1-butene with respect to the codimers.
TOF = (mol codimer) (mol neutral complex)ꢁ1 (h)ꢁ1
.
TOF = (mol codimer) (mol cationic mixed complex)ꢁ1 (h)ꢁ1
.
g
h
i
Precursor obtained by reaction of 2 with AgBF4 and CH3CN (method A).
Precursor obtained by reaction of 2 with AgBF4 (method B).
[Pd(g3-2-CH3C3H4)(NCCH3)(PCy3)]BF4.
j
Precursor obtained by reaction of 1 equiv. [PdCl(g3-2-CH3C3H4)]2 with 2 equiv. of AgBF4.
reaction. Actually, only in the case of 4f complex was pos-
sible to introduce a well-defined precursor (entry 14). After
1 h reaction low conversion (12% codimers) but good selec-
tivity (99 % of 3-phenyl-1-butene) was found. As described
above, mixed allylic cationic complexes (4a–4e, 4g and 4h)
undergo symmetrization in some degree, leading to solu-
tions were symmetric bisphosphine complexes type 3,
ture [38] for olefin isomerization and oligomerization
reaction. In consequence, when solutions of the mixed cat-
ionic compounds 4 were used as catalytic precursors, only
[Pd(g3-2-CH3C3H4)(NCCH3)P]BF4 complexes were able
to generate the active species. Moreover, 31P NMR spectra
of the solutions after one or 3 h of reaction showed the
presence of the symmetrical bisphosphine complexes
[Pd(g3-2-CH3C3H4)P2]BF4 (3) suggesting that the equilib-
rium was not shifted towards the formation of active mixed
complex in the reaction conditions. Therefore, in order to
properly compare the activity of the different precursors
it is necessary to use TOF values calculated from the
amount of the mixed complex in the equilibrium and not
from the amount of the starting neutral complex used in
the preparation of the precursor solution. Both TOF values
are depicted in Table 6 when data are available.
The best catalytic properties for the formation of 3-phe-
nyl-1-butene is observed with the mixed complex [Pd(g3-2-
CH3C3H4)(NCCH3)PBn3]BF4 (4g) (entries 15 vs. 1, 4, 7,
14, 17). This fact suggests that the steric bulk and electronic
properties in PBn3 ligand allow a good stabilization of
hydride active species and a fast reaction, as described
for nickel complexes [39]. The data also reflect that substi-
tution of one or two P–C bonds by P–N bonds in phos-
phine ligands cause a great decrease in the activity of the
precursor (entries 17 vs 2, 5, 8, 10, 12). It seems that these
aminophosphine ligands are less suitable to stabilize the
mixed complexes type
4
and [Pd(g3-2-methylal-
lyl)(NCCH3)2]BF4 (5) complex are in equilibrium. To ana-
lyze the results from the hydrovinylation reaction under
comparable experimental conditions to those used in the
study of the symmetrization equilibrium, we prepared the
solutions of the precursors by reaction of the stoichiome-
tric amounts of the neutral complex (2a–2e, 2g and 2h)
and AgBF4 in Cl2CH2 to which a 10-fold excess of acetoni-
trile was added and the AgCl filtered off. The activity of
complexes 3 and 5 was studied. All the cationic complexes
[Pd(g3-2-CH3C3H4)P2]BF4 (3a–3i) were tested under the
same catalytic conditions showing no conversion and
recovered unaltered. Although the literature states that
[PdR(NCCH3)P2]BF4 complexes are readily activated in
the presence of ethylene, it seems that ethylene is not able
to perform by itself the substitution of certain phosphines
or induce the g3–g1 shift on the allyl moiety [12]. Complex
[Pd(g3-2-CH3C3H4)(NCCH3)2]BF4 (5) did not present any
catalytic activity for hydrovinylation under the experimen-
tal conditions. The same result was reported in the litera-

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4013
proposed hydride species in contrast with some results
reported in the literature for nickel complexes with chiral
phosphoramidite ligands [8].
Disappointing values of enantioselectivity were obtained
with precursors containing chiral monoamino and bisami-
nophosphine ligands (only for b ligand a 11% ee was
obtained, (S)-3-phenyl-1-butene isomer [11b]) showing that
the chirality over the carbon backbone of the amino group
has hardly any discriminating effect.
It is known that the presence of acetonitrile in the reac-
tion medium can compete with styrene and ethylene by
the open coordination position. This kind of competence
alters the precursor activity, so we prepared mixed cationic
compounds in the absence of acetonitrile and we analyzed
the species existing in these solutions. To a dichlorometane
solution of the neutral complex [Pd(g3-2-CH3C3H4)-
(Cl)(Ph2PNHCHCH3Naph)] (2c), the stoichiometric
amount of AgBF4 was added in the presence of 1000-fold
styrene excess. The AgCl was filtered off and the solution
monitored by 31P NMR spectroscopy. A quantitative anal-
ysis of the species in equilibrium was not possible because
very broad signals appeared probably due to their lability.
The spectrum showed two broad doublets at 57.6 ppm
and 56.8 ppm (J = 47 Hz), assigned to the [Pd(g3-2-
CH3C3H4)P2]BF4 (3c) complex and a very broad signal at
58.2 ppm tentatively assigned to the mixed species [Pd(g3-
2-CH3C3H4)(styrene)(Ph2PNHCHCH3Naph)]BF4. Conse-
quently, [Pd(g3-2-CH3C3H4)(styrene)2]BF4 complex must
be also present in the solution although it is not detected.
We prepared the symmetrical bisstyrene palladium(II) com-
plex by reaction of the dinuclear [Pd(g3-2-CH3C3H4)Cl]2
with AgBF4 in the presence of styrene and the filtered
solution was introduced in the reactor in the catalytic con-
ditions. Table 6 shows that [Pd(g3-2-CH3C3H4)(sty-
rene)2]BF4 is moderately active (entries 18, 19, 20),
showing high selectivity to 3-phenyl-1-butene at low con-
versions but a considerable amount of styrene dimerization
(1,3-diphenyl-1-butene) is obtained. Isomerization com-
pounds (B, C) were obtained at higher conversions (entries
18, 19, 20).
Precursors obtained in the absence of acetonitrile leads
to higher conversions and lower selectivites (entries 1, 4,
7, 15 vs. 2, 5, 8, 16). The same result is obtained when
reaction time is increased to 6 h (entries 2, 5, 8, 10, 12
vs. 3, 6, 9, 11, 13). This fact indicates that 3-phenyl-1-
butene formed in the reaction may coordinate to the
palladium centre without the competence of CH3CN.
Consequently 3-phenyl-1-butene so formed isomerizes to
2-phenyl-2-butene. Moreover the presence of [Pd(g3-2-
CH3C3H4)(styrene)2]BF4 complex also increases the for-
mation of the isomerization products. When reaction time
changes from 1 to 6 h TOF values (calculated with total
amount of Pd(II) introduced) increase for complexes with
aminophosphine ligand a and c (entries 2, 3, 8 and 9),
while decrease for d and e (entries 10, 11, 12 and 13) sug-
gesting that monoaminophosphine ligands stabilize better
the active hydride species than does bisaminophosphine.
The increase in the TOF values for a and c ligands can
be argued taking account of the existence of an initial
activation time for the precursor. For ligand b only a
small decrease of TOF is observed, probably because
after 6 h of reaction styrene was exhausted (entries 5
and 6).
4. Conclusions
The susceptibiliy of aminophosphines and bis(amino-
phosphine) ligands to air, moisture and protic solvents
was not a limitation for the preparation of stable allylic
neutral and cationic palladium complexes [Pd(g3-2-methyl-
allyl)ClP] and [Pd(g3-2-methylallyl)P2]BF4. The results
described here in relation to the symmetrization equilib-
rium experienced by cationic complexes [Pd(g3-2-methylal-
lyl)(NCCH3)P]BF4 may be considered as a warning in the
study of a catalytic reaction where the catalyst is prepared
in situ. Such a procedure is correct and may provide reli-
able results only when the quantitative formation of the
catalytically active precursor is proved to occur within
the residence time preceding the addition of the other
reagents involved in the process. The species formed
because of symmetrization of catalytic precursors are not
innocent although in our case the results of the hydroviny-
lation reaction are not substantially modified. The presence
of [Pd(g3-2-CH3C3H4)(styrene)2]BF4 complex increases the
amount of styrene dimerization and isomerization com-
pounds (B, C) and contribute to a decrease of the enanti-
oselectivity of the process at high conversions. The
cationic complexes [Pd(g3-2-CH3C3H4)P2]BF4 are respon-
sible of a decrease in the conversion of the process.
Although chiral monodentate aminophosphine ligands
are easily prepared, those presented in the present paper
show moderate activity, low enantioselectivity but good
selectivity towards 3-phenyl-1-butene in the asymmetric
hydrovinylation catalytic process. Probably aminophos-
phine (and particularly bisaminophosphine) ligands do
not stabilize the hydride active species and, furthermore,
the chiral centre is too far from the metal to induce good
enantiodiscrimination.
5. Experimental
5.1. General methods
All compounds were prepared under a purified nitrogen
atmosphere using standard Schlenk and vacuum-line tech-
niques. The solvents were purified by standard procedures
and distilled under nitrogen. (R)-a-methylbenzylamine,
(R)-1-(1-Naphthyl)ethylamine and (R)-1-cyclohexylethyl-
amine (Aldrich) and PPh3, PMe2Ph, PCy3 (Strem) were
used as supplied. PBn3 [40], [Pd(g3-2-CH3C3H4)(l-Cl)]2
[41], [PdCl(g3-2-CH3C3H4)(PPh3)] [31], [PdCl(g3-2-
CH3C3H4)(PMe2Ph)] [30], [Pd(g3-2-CH3C3H4)(PPh3)2]BF4
[34] and [Pd(g3-2-CH3C3H4)(NCCH3)2]BF4 [31] were pre-
pared as previously described. The 1H, 13C, and 31P

4014
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
NMR spectra were recorded on either a Varian XL-500
and Mercury-400 MHz (1H, standard SiMe4), Varian
Gemini (13C, 50 MHz, standard SiMe4) and Bruker DRX
250 (31P, 101 MHz, standard H3PO4) spectrometers in
CDCl3 unless otherwise cited. Chemical shifts in ppm were
reported downfield from standards. The two-dimensional
experiments were carried out with a Bruker DMX500 or
a Varian XL-500 instruments, at 298 K with mixing time
of 500 ms. IR spectra were recorded on the following spec-
trometers: FT-IR Nicolet 520, FT-IR Nicolet Impact 400,
FT-IR Avatar 330 and FTIR Nicolet 5700. FAB mass
chromatograms were obtained on a Fisons V6-Quattro
instrument. The routine GC analyses were performed on
a Hewlett–Packard 5890 Series II gas chromatograph (50-
m Ultra 2 capillary column 5% phenylmethylsilicone and
95% dimethylsilicone) with a FID detector. Enantiomeric
excess was determined by GC on a Hewlett–Packard
5890 Series II gas chromatograph (30-m Chiraldex DM
column) with a FID detector. Elemental analyses were car-
5.2.3. Synthesis of diphenyl[1-(R)-(1-naphthyl)ethylamino]
phosphine (c)
Prepared analogously to (a) from (R)-1-(1-naphthyl)eth-
ylamine. The colorless oil obtained was used without puri-
fication. Yield: 85%. 1H NMR (CDCl3, 200 MHz): [d/
ppm] 1.52 (d, JHH = 7, 3H), 2.26 (bs, 1H), 5.02 (m, 1H),
7.13–8.03 (17H). 13C NMR (CDCl3, 50 MHz): [d/ppm]
24.12 (d, JCP = 6.4, CH3), 51.19 (d, JCP = 23.2, CH),
31
120.83–141.37 (Ar).
P NMR (toluene, 101 MHz): [d/
ppm] 36.5. MS-CI (NBA, m/z): 355 [M]+. [a]298
(c = 0.858, CH2Cl2) = ꢁ61.4.
5.2.4. Synthesis of phenylbis[1-(R)-phenylethylamino]-
phosphine (d)
Following a slight modification of the method described
in the literature for PhP[NH(S)-CH(CH3)Ph]2 [25] (R)-(+)-
a-methylbenzylamine (2.69 mL, 20 mmol) and NEt3
(4.18 mL, 30 mmol) were dissolved in 20 mL of toluene
and the solution was cooled at 0ꢁC. A solution of PPhCl2
(1.40 mL, 10 mmol) in toluene (15 mL) was added drop-
wise. After the addition was finished the mixture was stir-
red for 1.5 h at room temperature and the precipitate was
filtered off. The solvent and excess of triethylamine were
removed in vacuo. A colourless oil was obtained and used
without purification. Yield: 80%. 1H NMR (CDCl3,
250 MHz): [d/ppm] 1.31 (d, JHH = 6.6, 3H), 1.48 (d,
`
ried out by the Serveis Cientificotecnics of the Universitat
Rovira i Virgili in an Eager 1108 microanalyzer. Optical
rotations were measured on a Perkin Elmer 241MC spec-
tropolarimeter at 23 ꢁC.
5.2. Synthesis of phosphines
5.2.1. Synthesis of diphenyl [(R)-(1)-phenylethylamino]-
phosphine (a)
JHH = 6.6, 3H), 2.42 (d, JHH = 6.7, 1H), 2.53 (t, JHH–
JPH = 7.2, 1H), 4.18–4.02 (2H), 7.18–7.65 (15H). 13C
NMR (CDCl3, 50 MHz): [d/ppm] 26.02 (d, JCP = 8.3,
CH3), 26.76 (d, JCP = 3.9, CH3), 54.85 (d, JCP = 19.0,
CH), 53.53 (d, JCP = 12.2, CH),125.83–143.68 (Ar). 31P
Following the same method described in the literature
for Ph2PNH(S)-CHCH3Ph [24]. (R)-a-methylbenzylamine
(1.30 ml, 10 mmol) and NEt3 (2.10 ml, 15 mmol) were dis-
solved in 15 mL of toluene and the solution was cooled to
0 ꢁC. A solution of PPh2Cl (1.80 mL, 10 mmol) in toluene
(15 mL) was added dropwise. The mixture was then stirred
for 1.5 h at room temperature and the colourless precipitate
of triethylamine hydrochloride was filtered off. The solvent
and excess of triethylamine were removed in vacuum. A col-
ourless oil was obtained and used without purification.
Yield: 80%. 1H NMR (CDCl3, 200 MHz): [d/ppm] 1.39
(d, JHH = 7.0, 3H), 2.18 (bs, 1H), 4.22 (m, 1H), 7.19–7.85
(m, 15H). 13C NMR (CDCl3, 50 MHz): [d/ppm] 24.82 (d,
JCP = 7.3, CH3), 55.36 (d, JCP = 22.7, CH), 124.79–131.37
(Ar). 31P NMR (toluene, 101 MHz): [d/ppm] 36.0. MS-CI
(NBA, m/z): 305 [M]+. [a]298 (c = 0.813, CH2Cl2) = +6.5.
NMR (CDCl3, 101 MHz): [d/ppm] 59.1. MS-CI (NBA,
298
m/z): 348 [M]+. [a]
(c = 2.67, CH2Cl2) = +47.94.
5.2.5. Synthesis of phenylbis[1-(R)-cyclohexylethylamino]-
phosphine (e)
The procedure used was the same as for (d) starting from
the amine (R)-1-cyclohexylethylamine. A viscous oil was
obtained and used as such for the next step. Yield: 75%.
1H NMR (CDCl3, 250 MHz): [d/ppm] 0.9–1.9 (28H), 1.92
(dd, JHH = 6.7, 1H), 2.05 (d, JHH = 8.2, 1H), 2.9–3.1
(2H), 7.18–7.84 (5H). 13C NMR (CDCl3, 50 MHz): [d/
ppm] 20.87 (d, JCP = 6.1, CH3), 21.32 (d, JCP = 3.8, CH3),
26.46–29.95 (CH2), 46.10–45.90 (CH), 55.49 (d, JCP = 18,
CH), 55.84 (d, JCP = 19, CH), 127.84–145.69 (Ar). 31P
NMR (CDCl3, 101 MHz): [d/ppm] 60.7. MS-CI (NBA,
m/z): 360 [M]+. [a]298 (9ic = 0.860, CH2Cl2) = ꢁ19.19.
5.2.2. Synthesis of diphenyl [(R)-1-cyclohexylethylamino]-
phosphine (b)
Prepared analogously to (a) from (R)-1-cyclohexylethyl-
amine. The colourless oil obtained was used without puri-
fication. Yield: 84%. 1H NMR (CDCl3, 200 MHz): [d/ppm]
0.99 (d, JHH = 6.8, 3H), 0.86–1.74 (m, 12H), 2.93 (m, 1H),
7.22–7.90 (10H). 13C NMR (CDCl3, 50 MHz): [d/ppm]
19.74 (d, JCP = 5.9, CH3), 25.34, 25.43, 25.66, 27.51,
5.3. Synthesis of dichlorobis[phenylbis((R)-1-
phenylethylamino)phosphine]palladium(II) (1)
0.714 g (2.5 mmol) of PdCl2(COD) and 1.74 g (5 mmol)
of phosphine d were dissolved in 20 mL of CH2Cl2 at room
temperature. After 2 h stirring, the resulting red solution
was concentrated under vacuum and 10 mL of ether were
added. The orange precipitate was filtered and dried. Yield:
1.0 g (49% ). 1H NMR (250 MHz, CDCl3): [d/ppm] 1.14 (d;
28.35 (s, CH2), 44.43 (d, JCP = 6.4, CH), 57.10 (d,
31
JCP = 24.1, CH), 126.9–131.22 (Ar).
P NMR (toluene,
101 MHz): [d/ppm] 36.4. MS-CI (NBA, m/z): 311 [M]+.
[a]298 (c = 0.908, CH2Cl2) = ꢁ18.

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4015
7.2, 3H), 1.37 (d; 6.8, 3H), 3.26 (dd; 9.2, 5.6, 1H) 3.5 (m,
5.4.4. Chloro(g3-2-methyallyl){phenylbis[1-(R)-
phenylethylamino]phosphine} palladium(II) (2d)
1H), 4.6 (m, 1H), 4.8 (dd; 11.6, 9.6, 1H), 6.4–7.31 (15H).
31P (101 MHz, CH2Cl2): [d/ppm] 57,1 (s). Anal. Calc. for
C44H50Cl2N4P2Pd: C, 58.16; N, 6.78; H, 6.10. Found: C,
56.63; N, 5.92; H, 5.97%. MS/ESI (+): m/z 839.4 [MꢁCl]+.
To [Pd(g3-2-methylallyl)(l-Cl)]2 (0.98 g, 2.5 mmol) in
CH2Cl2 (5 mL) a solution of phenylbis(1-(R)-phenylethyla-
mino)phosphine (d) (5 mmol) in 5 mL of CH2Cl2 was
added. The mixture was stirred at room temperature for
1 h. The solvent was evaporated under reduced pressure.
Addition of 5 mL of diethylether lead to a yellow solid.
Yield: 2.109 g (78%). 1H NMR (400 MHz, CDCl3,
298 K): [d/ppm] 1.5–1.6 (9H); 1.40, 2.12 (s, 1H); 2.34,
2.65 (s, 1H); 2.83, 3.25 (d, JPH = 11, 1H); 3.7–4.8 (5H);
7.16–7.57 (15H). 31P NMR (101 MHz, CDCl3, 298 K):
[d/ppm] 64.1(s), 63.3 (s). Anal. Calc. for C26H32ClN2PPd:
C, 57.26; H, 5.91; N, 5.14. Found: C, 57.82; H, 6.83; N,
5.17%. MS-FAB (NBA, m/z): 509 [M+]ꢁCl.
5.4. Synthesis of complexes [PdCl(g3-2-methylallyl)P] (2)
The complexes were obtained by similar procedures
described in the literature for complexes (2h) [31] and (2i)
[30].
5.4.1. Chloro(g3-2-methylallyl){diphenyl[(R)-(1)-
phenylethylamino]phosphine}palladium(II) (2a)
0.98 g (2.5 mmol) of [Pd(g3-2-methylallyl)(l-Cl)]2 were
dissolved in 5 mL of dichloromethane and 1.53 g (5 mmol)
of phosphine (a) dissolved in 10 mL of CH2Cl2 were added.
The mixture was stirred at ꢁ20 ꢁC for 1 h and the solvent
5.4.5. Chloro(g3-2-methyallyl){phenylbis[1-(R)-
cyclohexylethylamino]phosphine}palladium(II) (2e)
[Pd(g3-2-methylallyl)(l-Cl)]2 (0.98 g, 2.5 mmol) was dis-
solved in 5 mL of CH2Cl2 and the solution was cooled to
0 ꢁC. A solution of phenylbis (1-(R)-cyclohexylethyla-
mino)phosphine (e) (5 mmol in 5 mL of CH2Cl2) was
added. The mixture was stirred at 0 ꢁC for 1 h and then
warmed to room temperature. After stirring for 1 h, the
solvent was evaporated under reduced pressure. On addi-
tion of ether (5 mL) a yellow powder is formed. Yield:
2.358 g (63%). 1H NMR (400 MHz, CDCl3, 298 K): [d/
ppm] 0.9–1.8 (31H); 2.17, 2.33 (s, 1H); 2.64, 2.75 (s, 1H);
2.9–3.8 (5H); 4.3–4.4 (1H); 7.4–7.7 (5H). 31P NMR
(101 MHz, CDCl3, 298 K): [d/ppm] 63.0(s), 62.7 (s). Anal.
Calc. for C26H44ClN2PPd: C, 55.37; H, 8.92; N, 4.71.
Found: C, 56.02; H, 7.96; N, 4.71%. MS-FAB (NBA,
m/z): 521 [M]+ꢁCl.
removed under reduced pressure. Addition of ether
1
(5 mL) lead to an orange solid. Yield: 0.51 g (20%).
H
NMR (400 MHz, CDCl3, 298 K): [d/ppm] 1.18, 1.32 (d,
JHH = 6.8, 3H); 1.84 (s, 3H); 2.37, 2.41 (s, 1H); 3.02 (bs,
1H); 3.42 (s, JPH = 10.4, 1H); 4.05 (m, 1H); 4.48 (bs,
1H); 4.64, 4.59 (dd, JPH = 19.2, JHH = 4.4, 1H); 7.00–
7.74 (15H). 31P NMR (101 MHz, CH2Cl2, 298 K): [d/
ppm] 54.6 (s), 55.1 (s). Anal. Calc. for C24H27ClNPPd: C,
57.39; H, 5.42; N, 2.29. Found: C, 56.74; H, 6.0; N, 2.82%.
5.4.2. Chloro(g3-2-methylallyl){diphenyl[(R)-1-
cyclohexylethylamino]phosphine}palladium(II) (2b)
Complex (2b) was synthesized in a similar way to that
used for (2a). Starting materials for 2b: 0.942 g (2.39 mmol)
of palladium [Pd(g3-2-methylallyl)(l-Cl)]2 and 1.5 g
(4.82 mmol) of phosphine (b). A yellow solid was obtained.
1
Yield: 0.48 g (24%). H NMR (400 MHz, CDCl3, 298 K):
5.4.6. Chloro(g3-2-methylallyl)(tricyclohexylphosphine)-
palladium(II) (2f)
[d/ppm] 0.71, 0.85 (d, JHH = 6.8, 3H); 0.73–1.61 (22H),
1.72, 1.81 (s, 3H); 2.25, 2.34 (s, 2H); 2.73 (m, 2H); 2.85,
2.93 (bs, 2H); 3.30, 3.33 (d, JPH = 4.0, 2H); 4.03 (m, 2H);
4.37 (m, 2H); 7.32–7.69 (20H). 31P NMR (101 MHz,
CDCl3, 273 K): [d/ppm] 53.63 (s), 53.35 (s). Anal. Calc.
for C24H33ClNPPd: C, 56.70; H, 6.54; N, 2.76. Found: C,
56.21; H, 6.99; N, 2.68%.
Compound (2f) is synthesized in a similar way to that
used for the preparation of (2a). Starting materials:
4.27 mmol of PCy3 (7.56 ml of a 20% toluene solution)
and 2.13 mmol (0.841 g) of [Pd(g3-2-methylallyl)(l-Cl)]2.
The solvent was evaporated under reduced pressure and
ether/hexane was added. After 2 h at 0 ꢁC a pale yellow
solid was formed. Yield: 1.134 g (56%). 1H NMR
(200 MHz, CDCl3, 298 K): [d/ppm] 1.94 (s, 3H), 1.26–
2.22 (33H), 2.51 (s, 1H), 3.15 (bs, 1H), 3.45 (d,
JPH = 8.8, 1H), 4.39 (m, 1H). 31P NMR (101 MHz,
CH2Cl2, 298 K): [d/ppm] 41.2 (s). Anal. Calc. for
C22H40ClPPd: C, 55.35; H, 8.45. Found: C, 55.62; H,
9.42%.
5.4.3. Chloro(g3-2-methylallyl){diphenyl[1-(R)-(1-
naphtyl)ethylamino]phosphine]palladium(II) (2c)
Complex (2c) was synthesized in a similar way to that
used for preparation of (2a). Starting materials for (2c):
0.729 g (1.85 mmol) of [Pd(g3-2-methylallyl)(l-Cl)]2 and
1.33 g (375 mmol) of phosphine (c). A pale yellow solid
1
was obtained. Yield: 1.35 g (66%). H NMR (400 MHz,
CDCl3, 298 K): [d/ppm] 1.29, 1.43 (d, JHH = 6.4, 3H);
1.75, 1.82 (s, 3H); 2.27, 2.41 (s, 2H); 2.92, 2.99 (bs, 2H);
3.37, 3.39 (d, JPH = 10.0, 2H); 4.44 (m, 2H); 4.83 (m,
4H); 6.65–7.69 (17H). 31P NMR (101 MHz, CH2Cl2,
298 K): [d/ppm] 54.8 (s), 53.9 (s). Anal. Calc. for
C28H29ClNPPd: C, 60.88; H, 5.29; N, 2.54. Found: C,
60.54; H, 5.43; N, 2.57%.
5.4.7. Chloro(g3-2-methylallyl)(tribenzylphosphine)-
palladium(II) (2g)
The tribenzylphosphine (g) (0.77 g, 1.28 mmol) and
[Pd(g3-2-methylallyl)(l-Cl)]2 (0.5. g, 2.54 mmol) were dis-
solved in 40 ml of dichloromethane. The solution was stir-
red for 1 h at room temperature. The solvent was
removed under reduced pressure and hexane addition

4016
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
(10 mL) caused the precipitation of a white solid. Yield:
0.97 g (76%). 1H NMR (200 MHz, CDCl3, 298 K): [d/
ppm] 1.61 (s, 3H), 1.68 (s, 1H), 2.55 (s, 1H), 3.28 (7H),
4.41 (bs, 1H), 7.27 (m, 15H). 31P NMR (101 MHz,
CH2Cl2, 298 K): [d/ppm] 21.6 (s). Anal. Calc. for
C25H28ClPPd: C, 59.90; H, 5.63. Found: C, 59.68; H,
5.61%.
Cl)]2 and 8.6 mL (1.89 mmol) of a 0.21 M solution of
AgBF4 in THF. The product was obtained as a brownish
solid. Yield: 1.49 g (83%). H NMR (400 MHz, CDCl3):
[d/ppm] 0.95 (d, JHH = 6.4, 3H), 1.08 (d, JHH = 6.4,
3H), 1.46 (s, 3H), 3.27 (d, JPH = 8.0, 2H), 3.65 (bs, 1H),
3.67 (bs, 1H), 4.53 (m, 2H), 4.70 (m, 2H), 6.74–7.83
(34H). 31P NMR (101 MHz, CDCl3): [d/ppm] 57.7 (d,
JPP = 47.4), 58.2 (d, JPP = 47.4). Anal. Calc. for
C52H51BF4N2P2Pd: C, 65.12, H, 5.36; N, 2.92. Found:
C, 65.54; H, 6.07; N, 2.95%.
1
5.5. Synthesis of [Pd(g3-2-methylallyl)(P)2]BF4 complexes
(3)
5.5.4. Synthesis of (g3-2-methylallyl)bis[phenylbis((R)-1-
phenylethylamino)phosphine] palladium(II)
5.5.1. Synthesis of (g3-2-methylallyl)bis[diphenyl[(R)-(1)-
(phenylethylamino)]phosphine palladium(II)
tetrafluoroborate (3d)
tetrafluoroborate (3a)
To a solution of [Pd(g3-2-CH3C3H4)(l-Cl)]2 (0.49 g,
1.25 mmol) in toluene (7 mL) a solution of phenylbis(1-
phenylethylamino)phosphine (d) (5 mmol) in toluene
(7 mL) was added. The mixture was cooled to 0 ꢁC and
2.5 mmol of a solution 0.21 M of AgBF4 in THF was
added. The mixture was stirred for 1 h at room tempera-
ture. The AgCl formed was filtered and the solvent was
removed. The pasty solid obtained was washed and stir-
red several times with ether until a yellow solid was
obtained. Yield: 0.75 g (32%). 1H NMR (400 MHz,
CDCl3): [d/ppm] 1.11 (d, J = 6.8, 3H), 1.20 (d, J = 6.8,
3H), 1.25 (d, J = 6.8, 3H), 1.27 (s, 3H), 1.41 (d, J = 6.8,
3H), 2.52 (d, JPH = 11, 1H), 2.87 (d, JPH = 11, 1H),
3.06 (dd, J = 8.8, J = 4, 1H), 3.14 (t, J = 9.2, 1H), 3.54
(m, 1H), 3.66 (m, 1H), 3.70–3.76 (2H), 3.86–3.97 (4H),
6.8–7.6 (30H).31P NMR (250 MHz, CDCl3): [d/ppm]
69.2 (d, J = 58), 68.1 (d, J = 58 ). Anal. Calc. for
C48H57BF4N4P2Pd Æ CH2Cl2: C, 57.13; H, 5.77; N, 5.44.
Found: C, 56.01; H, 6.08; N, 5.93%. MS/FAB (+): m/z
857 [M]+.
To 0.24 g (0.61 mmol) of [Pd(g3-2-CH3C3H4)(l-Cl)]2
and 0.76 g (2.49 mmol) of the phosphine (a) dissolved
in 20 mL of dichloromethane at 0 ꢁC, 1.24 mmol of a
0.21 M solution of AgBF4 in THF were added. The mix-
ture was stirred for 1 h. After filtering through a Celite
pad, the solvent was partially removed. Addition of hex-
ane caused the precipitation of the complex (3a) as a
brownish solid. This solid was filtrated and washed with
ether and pentane. Yield: 0.62 g (58%). 1H NMR
(CDCl3, 500 MHz): [d/ppm] 0.95 (d, JHH = 7.0, 3 H),
1.03 (d, JHH = 7.0, 3H), 1.50 (s, 3H), 3.23 (t, JPH = 9,
2H), 3.62 (bs, 1H), 3.68 (bs, 1H), 3.79 (m, 2H), 3.87
(m, 2H), 4.09 (dd, JPH = 18, JHH = 10, 1H), 4.20 (dd,
JPH = 18, JHH = 10, 1H), 6.95–7.85 (30H). 31P NMR
(101 MHz, CH2Cl2, 298 K): [d/ppm] 57.3 (d, JPP = 47),
58.4 (d, JPP = 47). Anal. Calc. for C44H47BF4N2P2Pd:
C, 61.52; H, 5.51; N, 3.26. Found: C, 61.46; H, 6.58;
N, 3.25%.
5.5.2. Synthesis of (g3-2-methylallyl)bis[diphenyl[(R)-(1-
cyclohexylethylamino)phosphine]palladium(II)
tetrafluoroborate (3b)
5.5.5. Synthesis of (g3-2-methylallyl)bis[phenylbis((R)-1-
cyclohexylethylamino)]palladium(II) hexafluorophosphate
(3e)
This complex was prepared with the same method used
for (3a). Starting from 1.21 g (3.96 mmol) of the phosphine
(b), 0.38 g (0.97 mmol) of [Pd(g3-2-CH3C3H4)(l-Cl)]2 and
8.8 mL (1.94 mmol) of a 0.21 M solution of AgBF4 in
THF. The product was obtained as a pale orange solid.
Yield: 1.37 g (81%). 1H NMR (500 MHz, CDCl3): [d/
ppm] 0.66 (d, JHH = 6.5, 3H), 0.75 (d, JHH = 6.5, 3H),
1.55 (s, 3H), 0.78–1.75 (22H), 2.72 (bs, 2H), 2.93 (m,
1H), 3.00 (m, 1H), 3.24 (bs, 1H), 3.27 (bs, 1H), 3.63 (s,
1H), 3.69 (s, 1H), 7.09–7.87 (20H). 31P NMR (101 MHz,
To a solution of [Pd(g3-2-CH3C3H4)(l-Cl)]2 (0.49 g,
1.25 mmol) in Cl2CH2 (7 mL) a solution of phenylbis(1-
cyclohexylethylamino)phosphine (e) (5 mmol) in CH2Cl2
(7 mL) was added. The mixture was cooled to ꢁ20 ꢁC
and 2.5 mmol of TlPF6 was added. The mixture was stir-
red at ꢁ20 ꢁC for 1 h and then warmed to room temper-
ature and stirred for 1 h. After filtering the TlCl the
solvent was removed and the residue was dissolved again
in ether. After washing and stirring several times with
ether a light yellow powder was obtained. Yield: 1.47 g
(86%). 1H NMR (400 MHz), CDCl3: [d/ppm] 0.86 (d,
J = 6.4, 3H), 0.97 (d, J = 6.8, 3H), 1.02 (d, J = 6.4,
3H), 1.12 (d, J = 6.8, 3H), 1.8–0.8 (44 H), 2.98–2.64
(8H), 3.00 (d, J = 5.6, 1H), 3.03 (d, J = 6, 1H), 3.25 (bs,
1H), 3.77 (bs, 1H), 7.6–7.5 (10H). 31P NMR (250 MHz,
CDCl3): [d/ppm] 66.6 (d, J = 57), 67.2 (d, J = 57). Anal.
Calc. for C48H81BF6N4P3Pd: C, 56.11; H, 7.95; N, 5.45.
Found: C, 56.84; H, 7.95; N, 5.45%. MS/FAB (+): m/z
881 [M]+.
CDCl3):
[d/ppm]
56.8
(s).
Anal.
Calc.
for
C44H59BF4N2P2Pd: C, 60.67; H, 6.83; N, 3.22. Found: C,
61.69; H, 8.0; N, 3.21%.
5.5.3. Synthesis of (g3-2-methylallyl)bis[diphenyl(1-(R)-1-
naphtylethylamino)phosphine]palladium(II)
Tetrafluoroborate (3c)
This complex was prepared with the same method used
with (3a). Starting from 0.37 g (0.94 mmol) of the phos-
phine (c), 1.35 g (3.79 mmol) of [Pd(g3-2-CH3C3H4)(l-

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4017
5.5.6. Synthesis of (g3-2-methylallyl)bis(tricyclohexyl-
phosphine)palladium(II) tetrafluoroborate (3f)
5.7. Synthesis of acetonitrile (g3-2-methylallyl)
(tricyclohexylphosphine)palladium(II) tetrafluoroborate
(4f)
The complex was prepared following the same proce-
dure described for (3a) by using phosphine (f). A white
solid was obtained after removing the solvent under vac-
To a solution of 0.40 g (0.84 mmol) of [PdCl(g3-2-
CH3C3H4)(PCy3)] (2f) in 30 mL of CH2Cl2, 0.22 mL
(4.19 mmol) of acetonitrile and 0.84 mmol of AgBF4 in
4 mL of THF are were added. The mixture was stirred
for 1 h at ꢁ20 ꢁC. The AgCl precipitate formed was filtered
off and the solution evaporated to dryness. The oil formed
was washed with n-pentane several times and 0.23 g (57%
yield) of a white precipitate was obtained. 1H NMR
(250 MHz, CDCl3): [d/ppm] 1.25–1.89 (m, 33H); 1.99 (s,
3H); 2.47 (s, 3H); 2.61 (bs, 1H), 3.26 (bs, 1H), 3.66 (d,
J = 8.4, 1H); 4.95 (m, 1H). 31 P NMR (101 MHz, CH2Cl2):
[d/ppm] 42.52 (s).
1
uum and adding hexane. Yield: 0.08 g (19%). H NMR
(200 MHz, CDCl3): [d/ppm] 0.95–2.06 (69H), 3.08 (bs,
2H), 4.32 (bs, 2H). 31P (101 MHz, CH2Cl2), [d/ppm]:
[36.0 (s). Anal. Calc. for BC40F4H73P2Pd: C, 59.37; H,
9.09. Found: C, 59.49; H, 10.21%.
5.5.7. Synthesis of (g3-2-methylallyl)bis(tribenzyl-
phosphine)palladium(II) tetrafluoroborate (3g)
This complex was prepared with the same method
used with (3a). Starting from 0.72 g (2.36 mmol) of the
phosphine (g), 0.23 g (0.58 mmol) of [Pd(g3-2-
CH3C3H4)(l-Cl)]2 and 5.4 mL (1.18 mmol) of a 0.21 M
solution of AgBF4 in THF. The product was obtained
as a white solid after addition of hexane. Yield: 0.16 g
(16%). 1H NMR (400 MHz, CDCl3): [d/ppm] 0.98 (s,
3H), 2.16 (m, 2H), 3.17–3.35 (12H), 3.62 (s, 2H), 7.16–
7.34 (30H). 31P NMR (101 MHz, CH2Cl2): [d/ppm] 9.7
(s).
5.8. Structure determination
Orange crystals of cis-[PdCl2(PPh((R)-NHCHCH3-
Ph)2)2] (1) were obtained by slow diffusion of hexane over
dichloromethane solution of the complex at room
temperature.
A prismatic crystal (0.1 · 0.1 · 0.2 mm) was selected and
mounted on a MAR345 diffractometer with a image plate
detector. Unit-cell parameters were determined from auto-
matic centering of 482 reflections (3 < h < 31ꢁ) and refined
by least-squares method. Intensities were collected with
graphite monochromatized Mo Ka radiation. 10285 reflec-
tions were measured in the range 3.50 < h < 31.62. 9677
reflections were assumed as observed applying the condi-
tion I > 2r(I). Lorentz-polarization but no absorption cor-
rections were made. The structure was solved by Patterson
synthesis, using the SHELXS computer program [42] and
refined by the full-matrix least-squares method with SHELX
97 computer program [43], using 10285 reflections, (very
5.5.8. Synthesis of (g3-2-methylallyl)bis(dimethylphenyl-
phosphine)palladium(II) Tetrafluoroborate (3i)
This complex was prepared with the same method
used with (3a). Starting from 0.4 mL (2.81 mmol) of
the phosphine (i), 0.28 g (0.7 mmol) of [Pd(g3-2-
CH3C3H4)(l-Cl)]2 and 8.6 mL (1.9 mmol) of a 0.21 M
solution of AgBF4 in THF. The product was obtained
1
as a very pale brown oil. H NMR (250 MHz, CDCl3):
[d/ppm] 1.62 (bs, 12 H), 1.82 (s, 3H), 3.20 (bs, 2H),
31
4.06 (s, 2H), 7.32 (m, 10H);
P NMR (101 MHz,
CH2Cl2): [d/ppm] ꢁ8.4 (s).
5.6. 31P NMR study of the reactions of [PdCl(g3-2-
methylallyl)(P)] (2) with AgBF4 in CH2Cl2 acetonitrile
negative intensities were not assumed). The function mini-
P
2
2
2
mized was
wiFoj ꢁ jFcj j , where w = [r2(I) +
(0.0920P)2+4.0438P]ꢁ1, and P = (jFoj + 2 jFcj )/3, values
of f, f0 and f00 were taken from [44]. The chirality of struc-
ture was define from the Flack coefficient, which it is equal
to 0.00(5) for the given results [45]. All H atoms were com-
puted and refined, using a riding model, with an isotropic
temperature factor equal to 1.2 time the equivalent temper-
ature factor of the atom which are linked. The final R (on
2
2
To a solution of the appropriate neutral palladium
complex
2
(0.44 mmol) and acetonitrile (0.23 mL,
4.4 mmol) in CH2Cl2 (20 mL) cooled to 0 ꢁC, a solution
of AgBF4 (85 mg, 0.44 mmol) in THF (2.1 mL) was
added. After stirring for an hour at room temperature
the silver chloride was filtered trough a celite pad and
31
2
the crude product was analysed by 31P NMR.
P
F) factor was 0.054, wR (on jFj ) = 0.152 and goodness-of-
NMR (101 MHz, CH2Cl2); chemical shift (ppm) and iso-
mer ratio (I/II) for the mixed cationic complexes (4):
(4a): [d/ppm] 55.6 (s), 57.0 (s), 1.16/1; (4b): [d/ppm]
57.1 (bs); (4c): [d/ppm] 56.2 (s), 57.0 (s); 1/1; (4d): [d/
ppm] 67.9 (s), 65.9 (s); 1/1.5; (4e): [d/ppm] 67.1 (bs);
(4f): [d/ppm] 42.5 (s); (4g): [d/ppm] 19.8 (s); (4h): [d/
ppm] 24.9; (4i): [d/ppm] ꢁ4.5; ratio between the mixed
cationic complex (4) and the bis(phosphine) cationic
complex (3): 4a/3a, 1/0.45; 4b/3b, 1/0.38; 4c/3c, 1/0.32;
4d/3d, 1/0.22; 4e/3e, 1/0.27; 4f/3f, 1/0; 4g/3g, 1/0.03;
4h/3h, 1/0.13; 4i / 3i,1/0.68.
fit = 1.047 for all observed reflections. Number of refined
parameters was 408. Max. shift/esd = 0.0, Mean shift/
esd = 0.0. Max. and min. peaks in final difference synthesis
were 0.699 and ꢁ0.424 e Aꢁ3, respectively.
˚
5.9. Hydrovinylation reaction
5.9.1. General procedure
Hydrovinylation reactions were performed in a stain-
less-steel autoclave fitted with an external jacket connected
to an isobutanol bath and the temperature controlled using

4018
R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
Applied Homogeneous Catalysis with Organometallics Compounds,
vol. 3, VCH, New York, 2002, p. 1164;
(c) L.J. Goosen, Angew. Chem., Int. Ed. 41 (2002) 3775–3778.
[2] (a) U. Boothe, H.C. Rudbeck, D. Tanner, M. Johansen, J. Chem.
Soc., Perkin Trans. 1 (2001) 3305–3311;
a thermostat to 0.5 ꢁC. Internal temperature was con-
trolled with a termopair and pressure was controlled with
a transductor, pressure was registered as a function of time
with a Linseis L-200 recorder. The catalyst precursor solu-
tion was placed in the autoclave, which had previously
been purged by successive vacuum/nitrogen cycles and
thermostatted at 15 ꢁC. Ethylene was admitted until a pres-
sure of 15 bar was reached. After the time indicated in
Table 6 for each reaction, the autoclave was slowly depres-
surized and NH4Cl 10% solution (10 mL) was added. The
mixture was stirred 30 min in order to quench the catalyst.
The CH2Cl2 layer was decanted off and dried with Na2SO4.
The quantitative distribution of products and their enan-
tiomeric excess was determined by GC analysis.
(b) G. Dieno Buono, C. Siv, G. Peiffer, C. TriantaPhylides, P. Denis,
A. Mortreux, F. Petit, J. Org. Chem. 50 (1985) 1781–1782;
(c) G. Hilt, F.X. du Mesnil, S. Luers, Angew. Chem., Int. Ed. 40
¨
(2001) 387–389;
(d) R. Kumareswaran, M. Nandi, T.V. Rajanbabu, Org. Lett. 5
(2003) 4345–4348;
(e) G. Muller, J.I. Ordinas, J. Mol. Catal. 125 (1997) 97–108.
[3] (a) G. Wilke, Angew. Chem., Int. Ed. Engl. 27 (1988) 185–206;
(b) K. Angerlunnd, A. Eckerle, F. Lutz, Z. Naturfosch 50b (1995)
488–501.
[4] R. Bayersdo¨rfer, B. Ganter, U. Englert, W. Keim, D. Vogt, J.
Organomet. Chem. 552 (1998) 187–194.
[5] S.E. Gibson, H. Ibrahim, Chem. Commun. (2002) 2465–2473.
[6] H. Park, T.V. Rajanbabu, J. Am. Chem. Soc. 124 (2002)
734–735.
[7] A. Zhang, T.V. Rajanbabu, Org. Lett. 6 (2004) 1515–1517;
A. Zhang, T.V. Rajanbabu, J. Am. Chem. Soc. 128 (2006) 5620–5621.
[8] G. Francio, F. Faraone, W. Leitner, J. Am. Chem. Soc. 124 (2002)
736–737.
[9] W.J. Shi, J.H. Xie, Q.L. Zhou, Tetrahedron: Asymmetry 16 (2005)
705–710.
[10] R.M. Ceder, G. Muller, J.I. Ordinas, J. Mol. Catal. 92 (1994) 127–
139.
5.9.2. Preparation of catalyst precursor solutions
Method A: To a mixture of 0.03 mmol of the appropri-
ate neutral palladium complex 2 solved in 10 mL of
CH2Cl2, 12.3 mg (0.3 mmol) of acetonitrile and 0.03 mmol
of a 0.22 M solution of AgBF4 in THF were added. After
stirring in the dark at room temperature for 30 min and fil-
tering off the AgCl formed, 3.45 mL (30 mmol) of styrene
were added. The resulting solution was introduced in the
reactor previously purged.
[11] (a) J. Albert, J.M. Cadena, J. Granell, G. Muller, J.I. Ordinas, D.
Method B: To 0.03 mmol of the appropiate neutral pal-
ladium complex 2 solved in 10 mL of CH2Cl2, 3.45 mL
(30 mmol) of styrene and 5.85 mg (0.03 mmol) of AgBF4
were added. After stirring in the dark at room temperature
for 30 min and filtering off the AgCl formed, the resulting
solution was introduced in the reactor.
Method C: The complex 4f (0.03 mmol) and 3.45 mL
(30 mmol) of styrene were solved in 10 mL of CH2Cl2
and introduced in the reactor.
Panyella, C. Puerta, C. Sanudo, P. Valerga, Organometallics 18
˜
(1999) 3511–3518;
(b) J. Albert, R. Bosque, J.M. Cadena, S. Delgado, J. Granell, G.
Muller, J.I. Ordinas, M. Font-Bardia, X. Solans, Chem. Eur. J. 8
(2002) 2279–2287.
[12] A. Grabulosa, G. Muller, J.I. Ordinas, A. Mezzetti, M.A. Maestro,
M. Font-Bardia, X. Solans, Organometallics 24 (2005) 4961–4973.
[13] L.A. Hamilton, P.S. Landis, in: G.M. Kosolapoff, L. Maier (Eds.),
Organic Phosphorous Compounds, vol. 4, Wiley-Interscience, New
York, 1972, p. 504.
[14] J. Ansell, M. Wills, Chem. Soc. Rev. 31 (2002) 259–268.
[15] T.Q. Ly, A.M.Z. Slawin, J.D. Woolins, J. Chem. Soc., Dalton Trans.
(1997) 1611–1616.
Acknowledgements
[16] S.M. Aucott, A.M.Z. Slawin, J.D. Woolins, J. Chem. Soc., Dalton
Trans. (2000) 2559–2575.
[17] M.R.I. Zubiri, H.L. Milton, D.J. Cole-Hamilton, A.M.Z. Slawin,
J.D. Woolins, Polyhedron 23 (2004) 693–699;
This work was supported by the Spanish Ministerio de
Ciencia y Tecnologı´a (CTQ2004-01546). Financial support
from MEC (AP2000-2866) is gratefully acknowledged by
A.G.
M.R.I. Zubiri, H.L. Milton, A.M.Z. Slawin, J.D. Woolins, Inorg.
Chim. Acta 357 (2004) 1243–1246.
[18] (a) A.D. Burrows, M.F. Mahon, M.T. Palmer, J. Chem. Soc., Dalton
Trans. (2000) 1669–1677;
(b) A.D. Burrows, M.F. Mahon, M.T. Palmer, J. Chem. Soc., Dalton
Trans. (2000) 3615–3619.
[19] R.P. Kamalesh Babu, S.S. Krishnaurthy, M. Nethaji, Tetrahedron:
Asymmetry 6 (1995) 427–438.
Appendix A. Supplementary material
CCDC 633742 contains the supplementary crystallo-
graphic data for 1. These data can be obtained free of
[20] K. Osakada, T. Ikariya, M. Saburi, S. Yoshikawa, Chem. Lett.
(1981) 1691–1694;
charge
via
http://www.ccdc.cam.ac.uk/conts/retriev-
ing.html, or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax:
(+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.jorganchem.2007.02.020.
A. Roucoux, I. Suisse, M. Devocelle, J.F. Carapentier, F. Agbossou,
A. Mortreux, Tetrahedron: Asymmetry 7 (1996) 379–382;
R. Guo, X. Li, J. Wu, W.H. Kwok, J. Chen, M.C.K. Choi, A.S.
Chan, Tetrahedron Lett. 43 (2002) 6803–6806.
[21] S. Na¨ıli, J.F. Carapentier, F. Agbossou, A. Mortreux, Organomet-
allics 14 (1995) 401–406.
[22] I. Suisse, H. Bricout, A. Mortreux, Tetrahedron Lett. 35 (1994) 413–
416.
[23] S.K. Mandal, G.A.N. Gowda, S.S. Krishnamurthy, C. Zheng, S. Li,
N.S. Hosmane, J. Organomet. Chem. 676 (2003) 22–37;
I.C.F. Vasconcelos, G.K. Anderson, N.P. Rath, C.D. Spilling,
Tetrahedron: Asymmetry 9 (1998) 927–935;
References
[1] (a) T.V. Rajanbabu, Chem. Rev. 103 (2003) 2845–2860;
(b) P.W. Jolly, G. Wilke, in: B. Cornils, W.A. Herrmann (Eds.),

R.M. Ceder et al. / Journal of Organometallic Chemistry 692 (2007) 4005–4019
4019
X. Chen, R. Guo, Y. Li, G. Chen, C.H. Yeung, A.S.C. Chan,
Tetrahedron : Asymmetry 15 (2004) 213–217.
[24] H. Brunner, J. Doppelberger, Chem. Ber. 111 (1978) 673–691.
[25] O.I. Kolodiazhnyi, N. Prynada, Tetrahedron Lett. 41 (2000) 7979–
8000.
[33] G.M. DiRenzo, P.S. White, M. Brookhart, J. Am. Chem. Soc. 118
(1996) 6225–6234.
[34] J. Powell, B.L. Shaw, J. Chem. Soc. (A) (1968) 774–777.
[35] P. Dotta, P.G.A. Kumar, P.S. Pregosin, Helv. Chim. Acta 87 (2004)
272–278;
[26] C.A. Tolman, Chem. Rev. 77 (1977) 313–348.
[27] B. Eichhorn, H. No¨th, T. Seifert, Eur. J. Inorg. Chem. (1999) 2355–
2368.
[28] R.A. Burrow, D.H.C.H. Honeyman, Acta Crystallogr., Sect. C 50
(1994) 681.
S. Filipuzzi, P.S. Pregosin, A. Albinati, S. Rizzato, Organometallics
25 (2006) 5955–5964.
[36] C. Amatore, A. Jutand, M.A.M. Barki, G. Meyer, L. Mottier, Eur.
J. Inorg. Chem. (2001) 873–880.
[37] (a) D.J. Mabbott, B.E. Mann, P.M. Maitlis, J. Chem. Soc., Dalton
Trans. (1977) 294–299;
[29] (a) A.M.Z. Slawin, M.B. Smith, D.J. Woolins, J. Chem. Soc., Dalton
Trans. (1996) 1283–1293;
(b) G. Carturan, M. Biasolo, S. Daniele, G.A. Mazzocchin, P. Ugo,
Inorg. Chim. Acta 119 (1986) 19–24.
(b) A.M.Z. Slawin, M.B. Smith, D.J. Woolins, J. Chem. Soc., Dalton
Trans. (1996) 4567–4573;
(c) A. Badia, L.R. Falvello, R. Navarro, E.P. Urriolabeitia, J.
Organomet. Chem. 547 (1997) 121–128;
(d) R.P. Kamlesh Babu, S.S. Krishnamurthy, M. Nethaji, Polyhedron
15 (1996) 2689–2699.
[38] A. Sen, T. Lai, Organometallics 2 (1983) 1059–1060.
[39] R. Ceder, G. Muller, J.I. Ordinas, J. Mol. Catal. 92 (1994) 127–139.
[40] R.C. Hinton, F.G. Mann, J. Chem. Soc. (1959) 2835.
[41] Y. Tatsuno, T. Yoshida, S. Otsuka, Inorg. Synth. 28 (1990) 342–343.
[42] G.M. Sheldrick, SHELXL 97. A Computer Program for Automatic
Solution of Crystal Structures, University of Go¨ettingen, 1997.
[30] J. Powell, B.L. Shaw, J. Chem. Soc. (A) (1967) 1839–1850.
[31] B. Akermark, B. Krakrenberger, S. Hansson, A. Vitagliano, Organo-
metallics 6 (1987) 620–628.
[43] G.M. Sheldrick, SHELXL 97.
A Program for Crystal Structure
Refinement, University of Go¨ttingen, 1997.
[32] M.D.K. Boele, P.C.J. Kamer, M. Lutz, A.L. Spek, J.G. de Vries,
P.W.N.M. van Leeuwen, G.P.F. van Strijdonck, Chem. Eur. J. 10
(2004) 6232–6246.
[44] International Tables for X-Ray Crystallography. Kynoch Press,
Birmingham, 1974, vol. IV, pp. 99–100 and 149.
[45] H.D. Flack, Acta Crystallogr., Sect. A 39 (1983) 876.