Hydrocarboxylation of terminal alkenes in supercritical carbon dioxide

Article abstract of DOI:10.1002/ejic.200800282

The catalytic hydrocarboxylation of linear alkenes to carboxylic acids using supercritical carbon dioxide as a solvent was studied. High selectivities in acids have been obtained. The best results were achieved when adding a perfluorinated surfactant to the reaction mixture (93% conversions and ca. 80% selectivity). Comparative multinuclear high-pressure NMR spectroscopic studies in [D₈]THF and in supercritical CO₂ show the formation of Pd⁰ species. Wiley-VCH Verlag GmbH & Co. KGaA, 2008.

Full text of DOI:10.1002/ejic.200800282

FULL PAPER
DOI: 10.1002/ejic.200800282
Hydrocarboxylation of Terminal Alkenes in Supercritical Carbon Dioxide
[a]
[a]
[a]
[a]
Clara Tortosa Estorach, Arantxa Orejón, Núria Ruiz, Anna M. Masdeu-Bultó,* and
Gábor Laurenczy*[
b]
Keywords: Hydrocarboxylation / Hydroxycarbonylation / Palladium / Supercritical fluids
The catalytic hydrocarboxylation of linear alkenes to carbox-
ylic acids using supercritical carbon dioxide as a solvent was
studied. High selectivities in acids have been obtained. The
best results were achieved when adding a perfluorinated
surfactant to the reaction mixture (93% conversions and ca.
80% selectivity). Comparative multinuclear high-pressure
NMR spectroscopic studies in [D ]THF and in supercritical
show the formation of Pd species.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2008)
8
0
CO
2
Introduction
is a nonpolar medium the best solubility is observed for
nonpolar solutes. Organic fluorocarbons show a high solu-
Carbon dioxide in its liquid or supercritical state (scCO2)
bility in scCO . Therefore, many works report that perfluo-
2
has great potential as an environmentally benign reaction
rochains have been used to solubilize substrates, reagents
medium for chemical synthesis.[ Since the mid-1990s,
1]
[1,4]
or catalysts.
The solubility of polar substances can be
rapidly increasing research efforts have shown that scCO2
substantially increased by adding polar cosolvents or sur-
can replace conventional solvents in a wide range of pro-
[5]
factants to form microemulsions.
[
2]
cesses. Replacing or reducing toxic solvents is also one of
[6]
Palladium-catalyzed hydrocarboxylation of alkenes
[
3]
the postulates of Green Chemistry. There is also increas-
ing evidence that the application of scCO can broaden the
scope of catalytic synthetic methodologies. Carbon dioxide
becomes supercritical at T = 31.1 °C and P = 73.8 bar.
Beyond this point there is no distinction between the liquid
or gas phase and the new supercritical phase presents prop-
erties of both states, this means higher solubility of gases
than with organic solvents.[ Therefore it can improve the
yield of homogeneous catalysis, particularly for those reac-
tions that are first order in the concentration of the gaseous
(Scheme 1) is a straightforward and environmentally
2
friendly method for obtaining carboxylic acids with an
[7]
atom selectivity of 100%. Carboxylic acids have impor-
c
c
tant industrial applications as large volume products and
chemical intermediates such as acrylic or adipic acid deriva-
[6]
tives. Other products of interest are 2-arylproprionic ac-
ids, which are the most important class of nonsteroidal anti-
1]
[8]
inflammatory drugs, and can be obtained by hydrocar-
boxylation of styrene. In this reaction (Scheme 1) two re-
gioisomers can be formed: the branched (b) and the linear
(l) isomers. Byproducts of this reaction are the isomeriza-
tion products (isom) and internal carboxylic acids (int).
reagent. In addition, it is known that the density of CO2
changes dramatically in the near-critical regime, therefore
the solubility of the reagents suffer the same effect. Small
changes in temperature or pressure cause dramatic changes
in the density, viscosity and dielectric constant of supercriti-
cal CO , which makes it a tuneable solvent.
2
The mild critical data and the low cost of the material
itself make carbon dioxide attractive over other solvents
such as hydrocarbons or hydrofluorocarbons. Since scCO2 Scheme 1. Hydrocarboxylation reaction of alkenes.
[
[
a] Departament de Química Física i Inorgànica, Universitat Ro-
vira i Virgili,
The carbonylation of alkenes in organic solvents have
been extensively studied in the last decade. There are many
Marcel.li Domingo s/n, 43007 Tarragona, Spain
Fax: +34-977-559563
papers studying the effects of ligand, solvent, temperature
E-mail: annamaria.masdeu@urv.net
[6,9]
and pressure.
Water has been used as a solvent but when
b] Laboratory of Medicinal and Organometallic Chemistry, Insti-
tut des Sciences et Ingeniérie Chimiques, Ecole Polytéchnique the solubility of the substrates is low the conversion de-
Fédérale de Lausanne,
015 Lausanne, Switzerland
[10]
creases drastically. One efficient way to increase the affin-
ity of water of scCO is to generate water-in-CO (W/C)
1
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
2
2
microemulsions, which can be formed by using ammonium
3524
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3524–3531

Hydrocarboxylation of Terminal Alkenes
perfluoro-polyether-carboxylate surfactants.[5c,11] These dis-
persions have been used to solubilize hydrophilic substances
[11]
into CO , applied to organic reactions
and used to per-
2
form aqueous/scCO2 biphasic metal-catalyzed reactions
with water-soluble catalysts.[ In a preliminary communi-
cation, we have recently reported that the use of a per-
fluoro-polyether surfactant enhances the activity and the
regioselectivity to the linear acid although the selectivity
12]
[
13]
decreases.
In the literature there are different works on
[14,15]
carbonylation reactions that use scCO2,
but none on
Scheme 2. Products obtained in the hydrocarboxylation of different
1-alkenes.
hydrocarboxylation reactions using scCO2.
Here, we present different aspects of the hydrocarboxyla-
tion of linear alkenes in supercritical carbon dioxide media
using the catalytic precursor formed with [PdCl (NCPh) ]
Hydrocarboxylation in scCO2
2
2
First we examined the solubility of the Pd/1–5 catalyst
and phosphane ligands containing fluorinated groups (1–
, Figure 1). Triphenylphosphane (5, Figure 1) is used as a
precursor systems in scCO under the reaction conditions
4
2
to be used. This was carried out by placing [PdCl2-
reference ligand. Ligand 1 forms the rhodium CO soluble
2
(
NCPh) ] (0.1 mmol) and the corresponding phosphane
complex trans-[Rh(CO)Cl(1) ] (solubility of at least 5.5 m
2
2
(0.4 mmol) in a 100 mL autoclave with sapphire windows,
in scCO at 343 K and 273 atm), which was reported to be
2
[
16]
and the system was charged with liquid CO . Then, the
pressure and temperature were gradually increased over the
ranges used in the catalytic experiments. The conditions of
active in the hydroformylation of 1-octene in scCO2. We
also analysed the ability of perfluorinated surfactants to in-
2
crease the solubility of water in scCO . A study of the spe-
2
solubility for at least 1 m of the [PdCl (NCPh) ]/PR solu-
cies formed under CO pressure by HPNMR is also pre-
sented.
2
2
3
tions (the concentration used in the reactions) are summa-
rized in Table 1. We observed that the palladium system
with the 3,5-CF3 substituted phosphane 2 is soluble at
milder conditions than the systems with the 4-substituted
ones, 1 and 3, regardless of the perfluorinated chain length.
The palladium system with phosphane 4 requires a higher
pressure and temperature to be solubilized and the palla-
dium system with PPh3 was apparently insoluble up to
250 atm and 100 °C.
2 2 3
Table 1. Solubility conditions of [PdCl (NCPh)
]/PR [a].
PR
3
PTotal [atm]
T [°C]
1
2
3
4
5
175
130
160
180
75
90
90
90
Figure 1. Fluorine containing the phosphanes 1–4 and tri-
phenylphosphane (5) used as ligands in the Pd-catalysed hydrocar-
boxylation of 1-alkenes.
insoluble up to 250
insoluble up to 100
] = 1ϫ10–3 , molar ratio Pd/PR
= 1:4, 30 atm
2 3
[a] [PdCl
2
(NCPh)
of CO.
The 31P{ H} NMR spectra, in deuterated chloroform, of
the products obtained after the solubility tests of the sys-
tems Pd/1 or 2 confirmed the in situ formation of the corre-
sponding complexes [PdCl (1 or 2) ] with signals at δ = 23.8
1
Results and Discussion
2
2
The hydrocarboxylation of alkenes 6a–g (Scheme 2) in and 25.5 ppm, respectively.[18,19]
the presence of water was studied using supercritical carbon
Once the initial conditions in which the catalyst precur-
dioxide as the solvent and a catalyst precursor prepared in sors showed solubility in scCO were established, we pro-
2
situ by the addition of the complex [PdCl (NCPh) ], the ceeded with the catalytic studies analysing the influence of
2
2
corresponding ligands 1–5 (Pd/1–5 systems) and an acid. different parameters in the conversion and selectivity of the
The products obtained were the corresponding linear acids hydrocarboxylation reaction.
7a–g and branched acids 8a–f. The internal alkenes 9a–e
Initially, we studied the hydrocarboxylation reaction un-
formed by isomerisation were also detected in the case of der standard conditions (P/Pd = 4, CO pressure 30 atm) in
the 1-alkenes (Scheme 2). For comparative purposes the re- the presence of oxalic acid and added water, and then we
actions were also studied using a standard catalyst precur- studied different pressures and temperatures. With the sys-
sor [PdCl (PPh ) ] in an organic solvent under reaction tem [PdCl (NCPh) ]/1 (Pd/1) we observed that at 120 atm
2
3 2
2
2
conditions selected in previous literature studies.[17]
total pressure the conversion and selectivity in acid was low
Eur. J. Inorg. Chem. 2008, 3524–3531
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3525

A. M. Masdeu-Bultó, G. Laurenczy et al.
FULL PAPER
(Entry 1, Table 2). By increasing the total pressure to version was observed for the catalytic systems Pd/2 (Entries
150 atm, when the precatalytic system is not soluble, the 5 and 6, Table 2) and Pd/3 (Entries 7 and 8, Table 2). The
conversion and selectivity increased to 55% and 90%, palladium catalytic system with phosphane 4 was inactive
respectively (Entry 2, Table 2). By increasing the total pres- under the conditions studied (Entries 9 and 10, Table 2). At
sure to 200 atm the precatalyst becomes soluble but the similar conditions of solubility the conversion decreased in
conversion dropped to 6% (Entry 3, Table 2) and no acids the order Pd/1 ≈ Pd/2 Ͼ Pd/3 Ͼ Pd/4. This is in agreement
were detected at 250 atm (Entry 4, Table 2). Apparently, as with the observation reported in the literature for the hydro-
the solubility in scCO of the catalytic precursor increases, carboxylation of acenaphthylene in organic solvent, where
2
the reactions with a polar reagent, such as water, become the Pd/1 system gave a higher conversion than Pd/4 under
[
20]
less favourable. Some drops of water were always observed similar conditions.
during the reaction period. The same behaviour in the con-
Regarding the regioselectivity, the systems with the 4-
substituted phosphanes 1 and 3 gave the linear product as
the major product, however, with the Pd/2 system the major
product is the branched acid. In all cases the competitive
reaction is the isomerisation to internal alkenes. The iso-
merisation products are formed from the branched alkylic
intermediate (C-b, Scheme 3). Thus, when isomerisation is
important the amount of branched acid decreases. The
change in regioselectivity for the Pd/1 and Pd/2 systems
does not seem to be related to the isomerisation side reac-
tion since the results are similar in both cases (Entries 2
and 5, Table 2). In the styrene hydrocarboxylation it has
Table 2. Hydroformylation of 1-octene using [Rh(acac)(CO)
2
]/L(1–
[
a]
2
3) in scCO .
[d]
Entry
L
P
Total [atm] % C[b]
% S7,8[c]
7/8 [%]
S
9
1
2
3
4
5
6
7
8
9
1
1
1
1
2
2
3
3
4
4
1
1
5
5
120
150
200
250
120
150
150
200
120
150
150
200
150
–
10
55
6
0
57
2
11
1
0
61
90
75
–
85
77
55
–
55:45
75:25
60:40
–
41:59
34:66
73:27
–
39
10
25
–
15
–
45
100
–
–
–
–
–
been reported that Pd systems with PPh mainly give the
3
10
11
12
13
14
0
–
[
[
e]
e]
branched isomer, while systems with cis-chelated diphos-
phanes favour the formation of the linear isomers. This has
been related to the formation of the trans species with the
monophosphane ligands favouring the branched product
16
26
29
89
18
83
33
58:42
33:67
55:45
51:49
82
17
67
8
[
f]
[g]
86
] = 1ϫ10–3 , and the palladium with cis-chelated diphosphanes favour-
[
a] Reaction conditions: time = 12 h, [PdCl
molar ratio Pd/L/H ·2H O/1-octene/H
CO = 30 atm, T = 90 °C. [b] Total conversion. [c] Selectivity in
acids. [d] Selectivity in internal alkenes. [e] T = 120 °C. [f]
2
(NCPh)
2
C
2
O
4
2
2
O = 1:4:62.5:62.5:500,
[17b]
2
ing the linear species.
It has been reported that the for-
P
mation of cis- or trans-palladium phosphane species (spe-
cies A–D, Scheme 3) depends on the steric requirements of
[
PdCl
2 3 2 3
(PPh ) ]/4PPh as catalyst in dimethoxyethane as solvent.
[
21]
[g] 6% of oligomers also formed.
the phosphane. Thus, phosphanes with Tolman angles of
Scheme 3. General hydride mechanism for hydrocarboxylation.[9b]
526
3
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3524–3531

Hydrocarboxylation of Terminal Alkenes
θ Ͼ 140° favour the trans species while phosphanes with Table 3. Effect of the acid used in the hydrocarboxylation of 1-
octene with Pd/1.[
a]
smaller θ values also give the cis species. So, phosphane 2
[
19]
(θ = 160°) would favour the formation of the trans inter-
% C[b]
% S7,8[c]
Entry Acid
7/8 [%]
mediate species in the catalytic cycle and therefore favours
the formation of the branched acid.
The effect of the temperature has been studied with the
Pd/1 catalytic system. In Table 2 it can be observed that by
1
2
3
4
–
0
–
–
H
2
C
2
O
4
·2H
2
O
55
18
3
90
76
35
75:25
73:27
75:25
[
d]
HCl (aq.)
CF COOH
3
increasing the temperature at 150 atm of total pressure the [a] Reaction conditions: PCO = 30 atm, P = 150 atm, T = 90 °C,
T
–
3
time = 12 h, [PdCl
2 2
(PhCN) ] = 1ϫ10 , molar ratio Pd/L/acid/
conversion and the selectivity decrease drastically (Entry 11
versus Entry 2, Table 2). At 200 atm, the conversion is bet-
ter but still very low and the selectivity in the linear isomer
decreases (Entry 12, Table 2). An increase in the branched
1
-octene/H O = 1:4:62.5:62.5:500. [b] Total conversion. [c] Selectiv-
2
ity in acids. [d] 35%.
products were internal octenes. Other byproducts such as
aldehydes and oligomers were also detected (Entry 1,
Table 4). By increasing the molar ratio of H O/Pd the selec-
tivity improved significantly but the conversion under these
conditions remained low. For a Pd /PPh system reported
in the literature, increasing the concentration of water re-
sulted in the conversion reaching a maximum, but a higher
addition of water led to competition for coordination va-
cancies and decreased the conversion.
low affinity for water in the scCO phase might reduce this
isomer with an increase in temperature was also reported
II
for the hydrocarboxylation of alkenes using Pd /PPh as
3
the catalytic system.[22] The increase of isomerisation prod-
2
ucts observed in Entry 11 (Table 2) could be related to an
II
3
increase of β-elimination with an increase in the tempera-
[
9b]
ture.
Regarding the effect of partial pressure, at higher pres-
sures of carbon monoxide (50 atm) the Pd/1 system is inac-
tive, probably because competition for the coordination to
the palladium centre promotes the formation of the carbon-
[
22]
In our case, the
2
0
[23]
competition, and no decrease in the conversion was ob-
served up to a molar ratio of H O/Pd = 825.
ylic Pd inactive species.
At lower pressures of carbon
2
monoxide (15 atm) no acids were detected. A decrease in
the reaction conversion was also reported when the CO
pressure decreased for the hydrocarboxylation of 1-octene
Table 4. Effect of the amount of water used in hydrocarboxylation
of 1-octene with Pd/1.
[a]
[22]
using PdCl /PPh as the catalytic system.
2
3
% C[b]
% S7,8[c]
Entry
H
2
O/Pd
7/8 [%]
These initial experiments lead us to conclude that the
best conditions of pressure and temperature for hydrocar-
boxylation of 1-octene are 30 atm of carbon monoxide pres-
sure, 90 °C and 150 atm of total pressure for system Pd/1
and 120 atm of total pressure for catalytic system Pd/2. In
both cases the results obtained were better than the results
1
2
3
125
500
825
8
6
26
8[d]
75
96
62:38
60:40
67:33
[a] Reaction conditions: PCO = 30 atm, P
T
–3
= 200 atm, T = 90 °C,
2 2
time = 12 h, [PdCl (PhCN) ] = 1ϫ10 , molar ratio Pd/L/
H C O ·2H O/1-octene = 1:4:62.5:62.5. [b] Total conversion. [c]
2 2 4 2
obtained with the palladium catalytic system with PPh (5) Selectivity in acids. [d] 50% isomerisation and 42% aldehydes/
3
oligomers.
in scCO (Entry 13, Table 2) but the conversion was lower
2
than the one obtained in an organic solvent (dimethoxy-
Once the reaction conditions were optimised, we con-
ethane) under similar conditions with the [PdCl (PPh ) ]/
2
3 2
cluded that the low solubility of water in scCO was the
2
4
PPh catalytic system (Entry 14, Table 2).
The acid used also had an important effect on the results.
3
main problem to be overcome in order to achieve a higher
conversion. Therefore we considered the addition of a sur-
factant as a mass transfer agent.
The first evidence of this was with water only, i.e. no added
acid, where the hydrocarboxylation of 1-octene using the
catalyst precursor Pd/1 did not take place in scCO (Entry
2
1
, Table 3). Presumably the palladium hydride species (spe-
Addition of a Surfactant
cies A, Scheme 3) is not formed in the absence of a stronger
acid. Oxalic acid provided the best conversion while hydro-
chloric acid or a more CO -soluble acid such as trifluoro-
acetic acid gave a lower conversion and selectivity (Entries
We used an ammonium perfluoro-polyether-carboxylate
®
[24]
(PFPE), Krytox
(11, Figure 2), as the surfactant. The
2
syntheses were performed according to literature pro-
[
25]
cedures. The commercially available ammonium pentade-
cafluorooctanoate (12, Figure 2) was also employed for
comparative purposes.
3
and 4, Table 3). Stronger acids may favour the proton-
ation of the phosphane, and thus decrease the formation of
the actives species. Furthermore, the coordinating ability of
the counteranions of these acids may interfere with the co-
ordination of the substrate and decrease the activity of the
systems.
Finally, we studied the influence on the conversion of the
amount of added water to the Pd/1 system at 200 atm total
pressure, 30 atm of CO pressure and 90 °C (Table 4) when
Figure 2. Perfluorinated surfactants used to form the water-scCO
microemulsion; Krytox® (11), ammonium pentadecafluorooctano-
of H O/Pd = 125 the selectivity was very low and the main ate (12).
2
the catalytic system was soluble in scCO . At a molar ratio
2
2
Eur. J. Inorg. Chem. 2008, 3524–3531
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3527

A. M. Masdeu-Bultó, G. Laurenczy et al.
Table 5. Hydrocarboxylation of 1-octene using surfactants (surf.) 11 and 12 and system Pd/1 as the catalyst precursor.[a]
FULL PAPER
Entry
T [°C]
PTot [atm]
H
2
O/Pd (molar ratio)
H
2
O/Surf (molar ratio)
% C[b]
% S[c]
7/8 [%]
1
2
3
4
5
6
7
8
9
90
90
120
90
90
90
90
90
90
150
150
150
150
150
200
150
150
150
500
250
250
250
500
500
500
500
500
14
14
14
28
28
14
28
14
14
93
83
11
67
80
26
0
77
78
10
78
80
42
–
82:18
83:17
66:44
82:18
80:20
76:24
–
[
[
[
d]
e]
f]
0
37
–
45
–
80:20
[
1
1
a] Reaction conditions: PCO = 30 atm, time = 12 h, [PdCl
2
(PhCN)
:4:62.5:62.5, surfactant 11. [b] Total conversion. [c] Selectivity in acids. [d] No PR
2.
2
] = 1ϫ10–3 , molar ratio Pd/L/H
2
C
2
O
4
·2H
2
O/1-octene =
3
addition. [e] No oxalic acid addition. [f] Surfactant
The results obtained using the catalyst precursor Pd/1 solvent are high even for the long chain alkenes 1-dodecene
are shown in Table 5. The addition of 11 in a H O/11 molar and 1-hexadecene (up to ca. 90%). Although the reason is
2
ratio of 14 promoted an increase in the conversion of up to unclear, an unexpected low conversion was also reported in
9
3% (Entry 1, Table 5) when working at a molar ratio of the hydroformylation of 1-decene in water micellar sys-
[
26]
H O/Pd = 500. The selectivity in acids was ca. 10% lower tems.
The selectivities in acids are around 80%, the rest
2
than when no surfactant was used (Entry 1 in Table 5 versus being isomerisation products. The regioselectivities in the
Entry 2 in Table 2). Furthermore, when the reactor was linear acid for substrates 6b and 6d are around 80%, in the
vented at pressures and temperatures above the critical case of 6a and 6e they are lower and internal alkenes are
point of pure CO2 the cold trap contained detectable formed (Entry 1 and 5, Table 6). When no isomerisation is
amounts of the acids. This indicates that, at optimised con- possible, as in the case of styrene, no other byproducts were
ditions, extraction of the acids from the reaction mixture is detected (Entry 6, Table 6). In the case of styrene, the pre-
possible. At a lower H O/Pd molar ratio (250) the conver- ferred product is the branched acid, but since the addition
2
sion was also high (83%). A general increase in the regiose- of 11 favours the formation of the linear product it results
lectivity of the linear acid of up to ca. 80% is observed in in a low regioselectivity in the branched isomer (56%). For
all cases at 90 °C. This increase of linear acid could be re- the hindered substrate cyclohexene only 9% of conversion
lated to the higher amount of isomerisation products that was achieved (Entry 7, Table 6).
are formed from the branched alkylic intermediate. Also the
Table 6. Hydrocarboxylation with Pd/1/11 of different substrates in
coordination of the anion could increase the steric effect
[a]
2
scCO .
around the metal and favour the formation of the linear
isomer. The addition of surfactant 11 did not improve the
results obtained at 120 °C (Entry 3, Table 5). At lower sur-
factant concentrations the conversion decreased to 67%
[d]
Entry
Substrate
% C[b]
% S7,8[c]
7/8
% S
9
1
2
3
4
5
6
7
6a
6b
6c
6d
6e
6f
60
93
16
88
95
99
9
76
77
75
86
83
Ͼ99
–
64:36
82:18
63:37
80:20
64:25
44:56
–
24
23
25
14
17
–
(Entry 4 versus 2 and Entry 5 versus 1, Table 5). It was
observed that, like without surfactant, increasing the
amount of water increased the conversion (Entry 5 versus
Entry 4, Table 5). Under the conditions in which the cata-
[e]
6g
–
lyst precursor is soluble in scCO (200 atm, 90 °C) the ad-
dition of surfactant promoted a higher conversion but the = 150 atm, [PdCl
2
[a] Reaction conditions: PCO = 30 atm, time = 12 h, total pressure
–
3
2
(NCPh)
2
] = 1ϫ10 , molar ratio Pd/L/
O/11 = 14. [b]
H
2
C
2
O
4
·2H
2
O/H
2
O/substrate = 1:4:62.5:500:62.5, H
2
value was still low (Entry 6, Table 5). In the presence of the
surfactant, the palladium complex [PdCl (NCPh) ] without
Total conversion. [c] Selectivity in acids. [d] Selectivity in internal
alkenes. [e] 11% of internal acids also formed.
2
2
ligand 1 was an inactive catalyst (Entry 7, Table 5). Without
the acid the system Pd/1 with the surfactant was also inac-
tive (Entry 8, Table 5). Employment of the surfactant 12
with a shorter perfluorinated chain did not improve the re-
sults (Entry 9, Table 5).
Multinuclear High-Pressure NMR Experiments
In order to gain an insight into the species formed during
the reaction we have performed an in situ high-pressure
NMR (HP-NMR) spectroscopic study of the hydrocarbox-
ylation of 1-octene with palladium precursors for both sys-
Other Substrates
The system Pd/1 with the surfactant 11, under the best tems in organic solvents and in supercritical carbon dioxide.
conditions obtained for the hydrocarboxylation of 1-octene, In this study the HP-NMR spectra were carried out using
was studied as a catalyst for substrates 6a–6g. The results a palladium precursor [PdCl (NCPh) ] modified with the
2
2
obtained are shown in Table 6. With the exception of 1- fluorinated phosphanes 1–4. Oxalic acid was used as the
decene, the total conversions obtained using scCO as the acid under CO pressure in [D ]THF or supercritical carbon
2
8
3528
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3524–3531

Hydrocarboxylation of Terminal Alkenes
dioxide as a solvent. Because of the detection limits in the were added the 31P{ H} NMR spectrum of the solution at
1
1
3
C experiments of the HP-NMR spectroscopy, the concen- room temperature shows a signal at δ = 19.7 corresponding
tration used to record these spectra was higher than the to the palladium(0) species, Pd(2) , and other minor signals
3
ones used in the catalytic experiments. Selected HP-NMR at 12.6, 7.2 and –4.2 ppm. When 1-octene was added to
spectra are presented in the Supporting Information.
this solution no substantial changes were observed for both
systems.
HP-NMR Spectroscopy in Organic Solvents
For all the catalytic systems studied, Pd/1–3, the 13C{1H}
The 31P{ H} NMR spectrum of the palladium precursor NMR spectra show the corresponding singlets of the car-
1
–
2
[
PdCl (NCPh) ] (1ϫ10 ) and the phosphane 1 (molar boxylic acids, CH –CH(CH )(CH ) –COOH and CH –
2 2 3 3 2 5 3
ratio P/Pd = 4:1) in [D ]THF at room temperature shows a (CH ) –COOH, at δ = 178.5 and 174.4 ppm, respectively,
8
2 7
singlet signal at δ = 23.5 ppm, which corresponds to the confirming that the hydrocarboxylation took place.
complex trans-[PdCl (1) ], synthesized and characterized For the palladium system [PdCl (NCPh) ]/3 no signals
2
2
2
2
[
18]
0
previously,
and the signal corresponding to free phos- attributable to the Pd species were detected, which may
phane (δ = –5.6 ppm). Oxalic acid (ratio H C O ·2H O/Pd account for the low activity of the Pd/3 system in hydrocar-
2
2
4
2
=
62.5), deuterium oxide (0.1 mL) and 30 atm of CO were boxylation catalytic experiments.
added to this solution and heated at 90 °C for 1.5 h. The
3
1
1
P{ H} NMR spectrum at room temperature shows a
broad signal centred at δ = 18 ppm, a singlet at δ =
HP-NMR Spectroscopy in scCO2
II
2
4.4 ppm attributed to the Pd trans-[PdCl (1) ] and a sing-
Once the reference HP-NMR spectroscopic experiments
2
2
0
let corresponding to the Pd species Pd(1) appears at δ = in [D ]THF were completed the same experiments were car-
3
8
19.7 ppm. The palladium(0) species were formed by oxi- ried out using supercritical carbon dioxide as a solvent. For
dation of the phosphane to the corresponding oxide [step this, 0.083 mmol of the complex was placed in the sapphire
a) in Scheme 3], which appears at δ = 25.6 ppm. The Pd0 tube (8.2 mL) in order to obtain a 1ϫ10  concentration.
–2
(
species with PPh reported in the literature present signals The different components for each experiment were also
3
[27]
at δ = 22.6 ppm at 70 °C
and δ = 21.48 ppm at room placed in the tube. Then, liquid CO was introduced to 1/4
2
[
28]
temperature.
of the total volume and the system was heated until the
Upon cooling to –50 °C the broad signal splits into one total pressure achieved 100 bar. The total pressure was mea-
at δ = –6.4 ppm corresponding to the free ligand 1 and sig- sured by a digital manometer connected directly to the
nals at δ = 22.7, 25.1 and 77.6, which may correspond to NMR tube. Because of CO solidification no experiments
2
0
Pd carbonyl species in equilibrium, Pd(CO)(1) (x = 2, 3). were performed at low temperature. For this reason very
x
[
29]
A similar behaviour was reported for the PPh system.
The Pd/1 systems generate the expected Pd species under
broad signals were observed.
3
0
31
1
The P{ H} NMR spectrum of the palladium precursor
the conditions studied.
[PdCl (NCPh) ] with phosphane 1 (molar ratio P/Pd = 4:1)
2
2
Finally, the substrate 1-octene was added (molar ratio 1- showed the signal of the free phosphane at δ = –5.8 ppm
octene/Pd = 62.5) and pressurized again at 30 atm of CO. and that of the trans-[PdCl (1) ] at δ = 23.2 ppm. After ad-
2
2
3
1
1
The P{ H} NMR spectrum of this solution at –70 °C dition of oxalic acid and D O, and pressurizing the system
2
shows the same signals as it did without the substrate, ex- with CO, the formation of the Pd(1) and phosphane oxide
3
cept that two more peaks at δ = 19.6 and 19.1 ppm were took place as indicated by the signals at δ = 19.9 ppm and
1
3
1
31
1
observed with a ratio of 62:38. In the C{ H} NMR spec- 26.6 ppm, respectively, in the P{ H} NMR spectra. When
tra a signal appears at δ = 233.7 ppm. From literature data 1-octene was added no substantial changes were observed
these new signals are attributed to the acyl isomeric species in the 31P NMR spectra although the C NMR spectra
13
[
Pd{C(O)C H }Cl(1) ]
and
[Pd{C(O)(CH )C H }- showed the formation of the acids. The ratio of the linear/
8
17
2
3 7 15
[
9,30]
Cl(1)2],
which form by migratory insertion of CO into branched acid was ca. 65:35, which is close to the catalytic
the alkylic species C [step (e) in Scheme 3] After the reac- results (Entry 2, Table 2). In the 13C NMR spectra the sig-
tion was completed the peaks of the acyl species disap- nal corresponding to free CO can be observed at δ =
peared and only the acid signals remained.
The same studies were carried out with the catalytic sys- the presence of species with coordinated CO, which could
182 ppm and a small signal at δ = 181 ppm could indicate
3
1
1
31
tems [PdCl (NCPh) ]/2 and 3. The P{ H} NMR spectra not be differentiated in the P NMR spectra because of the
2
2
of the solution of the palladium precursor [PdCl (NCPh) ] broad signals.
2
2
When the 13C{ H} NMR spectrum of [PdCl (NCPh) ]/
1
with the phosphanes 2 or 3 in [D ]THF at room tempera-
8
2 2
ture exhibit a major narrow singlet signal at δ = 26.2, which 1/H C O ·2H O/D O/CO/1-octene is obtained in the pres-
2
2
4
2
2
[19]
corresponds to the complex trans-[PdCl (2) ]
and a ence of the surfactant 11 the signals of the linear and
2
2
broad signal at δ = 23 ppm corresponding to trans- branched acids at δ = 177.6 and 181 ppm, respectively, in-
[
31]
[
PdCl (3) ]. For the system [PdCl (NCPh) ]/2, two minor crease and a ratio of 90:10 is observed. This is a similar
2 2 2 2
signals appear at δ = 20.2 and 35.5 ppm that are attributed ratio to the one observed in the catalytic reaction (Entry 1,
to cationic species resulting from the chloride exchange Table 5). Unfortunately, the 31P{ H} NMR spectrum at
1
+
with the solvent, [PdCl(2) (Solvent)]
or [Pd(2)2- 90 °C showed broad signals and no conclusions could be
2
2
+ [32]
(
Solvent) ] .
When the acid D O and 30 atm of CO inferred about the species formed.
2
2
Eur. J. Inorg. Chem. 2008, 3524–3531
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3529

A. M. Masdeu-Bultó, G. Laurenczy et al.
FULL PAPER
For the [PdCl (NCPh) ]/2 and 3 systems in scCO the Solubility and Catalytic Studies: The solubility studies in supercriti-
2
2
2
3
1
1
cal carbon dioxide were performed in a stainless steel Thar reactor
(volume 100 mL) containing sapphire windows. The palladium pre-
P{ H} NMR spectrum shows the signals of the trans-
0
[
PdCl (1–2) ] species and signals corresponding to the Pd
2 2
cursor [PdCl
phane (0.4 mmol) were placed in the autoclave, the system was
purged with nitrogen/vacuum and it was charged with liquid CO
Then, the pressure and temperature were gradually increased and
the solubility was examined by ocular inspection through the sap-
phire windows.
2 2
(NCPh) ] (0.1 mmol) and the corresponding phos-
species in the range 20–21 ppm. The signals of the free li-
gand and the phosphane oxide are also observed. After 1-
1
3
1
2
.
octene is added the C{ H} NMR spectrum confirms the
3
1
1
formation of the acids but the P{ H} NMR spectrum
shows no significant changes.
3
1
1
The P{ H} NMR spectrum of the system [PdCl -
2
For catalytic experiments, the palladium precursor [PdCl
(NCPh) ] (0.025 mmol) with the acid (1.562 mmol), the phosphane
0.1 mmol), and the corresponding amount of surfactant, when
2
-
(
NCPh) ]/4 shows a broad small signal at δ = –31.2 ppm
2
[21]
2
corresponding to trans-[PdCl (4) ] together with the sig-
2
2
(
nal of the free phosphane. After the acid addition no sig-
used, were loaded into a stainless steel reactor vessel (25 mL). The
system was purged with nitrogen/vacuum. The corresponding
amount of degassed water, substrate (1.562 mmol) and undecane
nificant changes were observed and no acid formation was
1
3
1
detected in the C{ H} NMR spectrum. The fact that the
0
formation of the Pd species was not detected could explain
(
97.5 mg), as GC internal standard, were mixed and charged in
the low catalytic activity of this system.
vacuo. Then the CO gas was charged, the reactor pressurized to
In summary, for all the systems, the major species formed the desired pressure and the liquid carbon dioxide introduced at
by the reaction of [PdCl (NCPh) ] with 1–4 was the trans- room temperature via a syringe pump (Thar) until ca. 60 atm, and
2
2
the reactor was heated to the desired temperature. The compressed
carbon dioxide was introduced to attain the desired reaction pres-
sure, adjusted by software controlling the syringe pump, and mag-
netically stirred (750 rpm). After the reaction was completed, the
vessel was cooled with ice water to 0 °C and slowly depressurized
to atmospheric pressure through a cold trap. The reaction mixture
was extracted with diethyl ether and analyzed by gas chromatog-
raphy using a Hewlett–Packard 5890A apparatus in an HP-5 (5%
diphenyl silicone/95% dimethyl silicone) column (25 mϫ0.2 mm).
GC conditions: 45 °C (3.5 min), 10 °C/min, 150°°C (3 min) 8a (t =
[
PdCl (1–4) ] species. After the reagents and CO were
2 2
added palladium(0) species were formed except with ligand
4
these Pd species ([PdP ], Scheme 3) will react with the acid
to form a Pd–H species (A, Scheme 3), which has not been
detected. At the end of the experiment the formation of the
acids was confirmed by C NMR spectroscopy.
[
9]
. According to the proposed mechanism for this reaction
0
n
1
3
Conclusions
9.8 min), 7a (10.5 min); 32 °C (7 min), 10 °C/min, 150°°C (5 min)
In conclusion, hydrocarboxylation reactions can be per- 8b (t = 19.7 min), 7b (20.8 min); 80 °C (3.5 min), 10 °C/min, 150°°C
formed in supercritical carbon dioxide as a solvent using (3 min) 8c (t = 12.5 min), 7c (13.5 min); 35 °C (3 min), 10 °C/min,
2
2
00°°C (10 min) 8d (t = 18.1 min), 7d (18.7 min); 80 °C (3 min),
0 °C/min, 200°°C (10 min) 8e (t = 30.5 min), 7e (31.8 min); 150 °C
palladium catalysts with phosphanes containing –CF3
groups. When using [PdCl (NCPh) ] and phosphanes as
2
2
(
4
3.5 min), 10 °C/min, 200°°C (10 min) 8f (t = 5.1 min), 7f (5.6 min);
5 °C (3.5 min), 10 °C/min, 150°°C (3 min) 7g (t = 9.8 min).
catalyst precursors the conversion in the carbonylation of
-octene is better when the catalyst is not soluble in scCO2.
1
With these systems the best conversions were up to 60% 1,2-Dimethoxyethane (DME) was used as the organic solvent in
with selectivities in the acid of up to 90%. The conversion the comparative experiments. These were conducted with a 100 mL
in the hydrocarboxylation of 1-octene improved to 93% by autoclave following the preceding procedure without the CO
adding a perfluorinated surfactant with a selectivity in acids
2
charge.
of 77%. The regioselectivity in the linear acid increased
with the use of the surfactant. Additionally this catalytic
system was satisfactorily applied to the hydrocarboxylation
HP-NMR Experiments in [D
8
]THF: The high-pressure nuclear
magnetic resonance spectra were performed in a 10 mm sapphire
[34]
tube.
The tube enables experiments to be performed with gas,
of other long chain alkenes. The multinuclear high-pressure liquid or supercritical fluid samples up to 100 bar pressure. The
NMR spectroscopy allowed us to follow the formation of palladium precursor [PdCl (NCPh) ] (7.7 mg, 0.02 mmol) and
the Pd species under CO pressure in [D ]THF and in phosphane (0.08 mmol) were introduced into the tube under an
2
2
0
8
8
inert atmosphere. Degassed [D ]THF (2 mL) was then introduced,
scCO2.
1
3
and the tube was charged with CO (1–5 atm) and with CO up
to 30 atm pressure. The other reagents were introduced stepwise
following the same procedure. Measurements were performed with
Bruker DRX 400 MHz and Varian 300 MHz NMR spectrometers.
Experimental Section
General: Compounds 1–5 are commercially available and were used
without purification. The syntheses of the palladium complex
HP-NMR Experiments in scCO
out these experiments. The corresponding amount of palladium
precursor [PdCl (PhCN) ] (31.7 mg, 0.08 mmol) and phosphane
0.33 mmol) were introduced into the tube under an inert atmo-
sphere. The tube was then charged with CO (1–5 atm) and with
CO up to 30 atm pressure. The tube was cooled with a dry ice/
2
: The same tube was used to carry
[
2 2
PdCl (NCPh) ] was performed according to literature pro-
[
33]
cedures. All olefins used as substrates were filtered through alu-
mina before use. All these material were from Aldrich and Fluo-
rochem. Carbon dioxide (SCF Grade, 99.999%) was supplied by
Air Products and Linde and carbon monoxide (99.99%) was sup-
plied by Air Liquide.
2
2
(
13
2
acetone bath and liquid CO was introduced up to 1/4 of the total
Safety Warning: Experiments involving pressurized gases can be
hazardous and must only be conducted with suitable equipment
and following appropriate safety conditions.
volume. The tube was heated to achieve 100 atm of pressure. The
total pressure was measured by a digital manometer connected di-
rectly to the NMR tube. The other reagents were introduced step-
3530
www.eurjic.org
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Inorg. Chem. 2008, 3524–3531

Hydrocarboxylation of Terminal Alkenes
wise following the same procedure. Measurements were carried out
with a Bruker DRX 400 MHz NMR spectrometer.
[10] a) S. Tilloy, E. Monflier, F. Bertoux, Y. Castanet, A. Mortreux,
New J. Chem. 1997, 21, 529–531; b) G. Verspui, J. Feiken, G.
Papadogianakis, R. A. Sheldon, J. Mol. Catal. A: Chem. 1999,
146, 299–307.
Supporting Information (see also the footnote on the first page of
this article): Five figures showing 31P{ H} NMR spectra.
1
[11] G. B. Jacobson, C. T. Lee, K. P. Johnston, J. Org. Chem. 1999,
64, 1201–1206.
[
[
[
12] G. B. Jacobson, C. T. Lee, K. P. Johnston, W. Tumas, J. Am.
Chem. Soc. 1999, 121, 11902–11903.
13] C. Tortosa, N. Ruiz, A. M. Masdeu-Bultó, Chem. Commun.
Acknowledgments
2006, 2789–2791.
We gratefully acknowledge the Generalitat de Catalunya (DURSI
and Fons Social Europeu) for a Fellowship (to F. I.) to C. T. and
14] a) K. Nozaki, F. Shibahara, S. Elzner, T. Hiyama, Can. J.
Chem. 2001, 79, 593–597; b) M. C. Chen, H. F. Jiang, J. H. Li,
Chin. J. Chem. 2001, 19, 1153–1156.
2005SGR00777, the Ministerio de Educacion y Ciencia (MEC)
(CTQ2004-03831/PPQ, CSD2006-0003), Swiss State Secretariat for
[
15] a) L. Jia, H. Jiang, J. Li, Green Chem. 1999, 1, 91–93; b) C.
Education and Research (Grant SER C03.0020) and COST D30
Action (European Science Foundation) for financial support. We
also thank Brugarolas, S.A. for supplying the Krytox® FS-L 57,
and Prof. J. Iggo for helpful discussions.
Tortosa, A. M. Masdeu-Bultó, Catal. Lett. 2008, 122, 76–79.
[16] a) D. R. Palo, C. Erkey, Ind. Eng. Chem. Res. 1998, 37, 4203–
4206; b) D. R. Palo, C. Erkey, Ind. Eng. Chem. Res. 1999, 38,
3
786–3792; c) A. C. J. Koeken, M. C. A. van Vliet, L. J. P.
van de Broeke, B. J. Deelman, J. T. F. Keurentjes, Adv. Synth.
Catal. 2006, 348, 1553–1559.
[
[
1] P. G. Jessop, W. Leitner, Chemical Synthesis Using Supercritical
Fluids, Wiley-VCH, Weinheim, New York, 1999.
[17] a) B. El Ali, H. Alper, J. Mol. Catal. A: Chem. 1993, 80, 377–
381; b) I. del Río, N. Ruiz, C. Claver, L. A. van der Veen,
P. W. N. M. van Leeuwen, J. Mol. Catal. A: Chem. 2000, 161,
39–48.
2] a) P. G. Jessop, T. Ikariya, R. Noyori, Chem. Rev. 1999, 99,
4
75–493; b) W. Leitner, Acc. Chem. Res. 2002, 35, 746–756; c)
D. J. Cole-Hamilton, Science 2003, 299, 1702–1706; d) P. G.
Jessop, J. Supercrit. Fluids 2006, 38, 211–231.
[18] D. Hâncu, J. Green, E. Beckman, Ind. Eng. Chem. Res. 2002,
41, 4466–4474.
[
3] a) P. T. Anastas, J. C. Warner, Green Chemistry: Theory and
Practice, Oxford University Press, Oxford, 1998; b) P. T. An-
astas, M. M. Kirchhoff, Acc. Chem. Res. 2002, 35, 686–694.
4] a) A. V. Yazdi, E. J. Beckman, J. Mater. Res. 1995, 10, 530–
[19] M. L. Clarke, D. Ellis, K. L. Mason, A. G. Orpen, P. G. Prin-
gle, R. L. Wingad, D. A. Zaher, R. T. Baker, Dalton Trans.
2005, 1294–1300.
[20] J. Real, E. Prat, S. González-Cabello, M. Pagès, A. Polo, Orga-
nometallics 2000, 19, 4715–4719.
[
5
37; b) A. V. Yazdi, E. J. Beckman, J. Ind. Eng. Chem. 1997,
136, 2368–2374; c) S. Kainz, D. Koch, W. Baumann, W. Le- [21] A. M. Trzeciak, H. Bartosz-Bechowski, Z. Ciunik, K. Niesyty,
itner, Angew. Chem. Int. Ed. Engl. 1997, 36, 1628–1630; d) M.
Giménez Pedrós, A. M. Masdeu-Bultó, J. Bayardon, D. Sinou,
Catal. Lett. 2006, 107, 205–208; e) M. Giménez Pedrós, C. Tor-
tosa-Estorach, A. Bastero, A. M. Masdeu-Bultó, M. Solinas,
W. Leitner, Green Chem. 2006, 8, 875–877.
J. J. Ziółkowski, Can. J. Chem. 2001, 79, 752–759.
[22] D. M. Fenton, J. Org. Chem. 1973, 38, 3192–3198.
[23] a) T. E. Kron, M. I. Terekhova, Y. G. Noskov, E. S. Petrov, Ki-
net. Katal. 2001, 42, 182–188; b) M. C. Bonnet, A. L. Mon-
teiro, I. Tkatchenko, J. Mol. Catal. A: Chem. 1999, 143, 131–
136.
[
5] a) K. P. Johnston, K. L. Harrison, M. J. Clarke, S. M. Howdle,
M. P. Heitz, F. V. Bright, C. Carlier, T. W. Randolph, Science
[24] Krytox® is a trademark of DuPont.
1
996, 271, 624–626; b) M. J. Clarke, L. Kristi, L. Harrison,
P. K. Johnston, S. M. Howdle, J. Am. Chem. Soc. 1997, 119,
399–6406; c) C. T. Lee, P. A. Psathas, K. P. Johnston, J.
[25] H. Kondo, Y. Hisamichi, T. Kamei, J. Magnet. Magn. Mater.
1996, 155, 332–338.
[26] M. Giménez-Pedrós, A. Aghmiz, C. Claver, A. M. Masdeu-
Bultó, D. Sinou, J. Mol. Catal. A: Chem. 2003, 200, 157–163.
[27] B. E. Mann, A. Musco, J. Chem. Soc., Dalton Trans. 1975,
1673–1677.
[28] C. Amatore, A. Jutand, F. Lemaître, J. L. Ricard, S. Kozuch,
S. Shaik, J. Organomet. Chem. 2004, 689, 3728–3734.
[29] F. Morandini, G. Consiglio, F. Wenzinger, Helv. Chim. Acta
1979, 62, 59–61.
[30] A. Seayad, S. Jayasree, K. Damodaran, L. Toniolo, R. V.
Chaudhari, J. Organomet. Chem. 2000, 601, 100–107.
[31] J. Fawcett, E. G. Hope, R. D. W. Kemmitt, D. R. Paige, D. R.
Rusell, A. M. Stuart, J. Chem. Soc., Dalton Trans. 1998, 3751–
3763.
6
de Grazia, T. W. Randolph, Langmuir 1999, 15, 6781–6791.
6] a) A. Mullen, in New synthesis with carbon monoxide (Ed.: J.
Falbe), Springer-Verlag, Berlin, 1980, pp. 243–258; b) B.
Cornils, W. A. Hermann, Applied Homogeneous Catalysis with
Organometallic Compounds, VCH, Weinheim, 1996; c) H. M.
Colquhom, D. J. Thompson, M. V. Twigg, Carbonylation, Ple-
num Press, New York, 1991, 102; d) G. Kiss, Chem. Rev. 2001,
[
1
01, 3435–3456.
[
[
7] R. A. Sheldon, J. Mol. Catal. A: Chem. 1996, 107, 75–83.
8] B. Hinz, C. P. Dorn Jr, T. Y. Shen, K. Brune, in Ullmanns Ency-
clopedia of Industrial Chemistry (Ed.: W. Gerhartz), VCH,
Weinheim, Germany, 2002.
[
9] a) E. Drent, P. H. M. Budzelaar, Chem. Rev. 1996, 96, 663–681;
b) I. Del Río, C. Claver, P. W. N. M. van Leeuwen, Eur. J. Inorg.
Chem. 2001, 2719–2738; c) D. Kruis, N. Ruiz, M. D. Janssen,
[32] V. N. Zudin, V. D. Chinakov, V. M. Nekipelov, V. A. Rogov,
V. A. Likholobov, Y. I. Yermakov, J. Mol. Catal. A: Chem.
1989, 52, 27–48.
J. Boersma, C. Claver, G. van Koten, Inorg. Chem. Commun. [33] F. R. Hartley, Org. Chem. Rev. A 1970, 119–137.
1
998, 1, 295–298; d) N. Ruiz, I. Del Río, J. L. Jiménez, C. Cla-
[34] A. Cusanelli, U. Frey, D. Richens, A. E. Merbach, J. Am.
Chem. Soc. 1996, 118, 5265–5271.
ver, J. Forniés-Cámer, C. J. Cardin, S. Gladiali, J. Mol. Catal.
A: Chem. 1999, 143, 171–180; e) A. Ionescu, G. Laurenczy, O.
Wendt, Dalton Trans. 2006, 32, 3934–3940.
Received: March 18, 2008
Published Online: July 4, 2008
Eur. J. Inorg. Chem. 2008, 3524–3531
© 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
3531