Effect of trimethylsilyl substitution on the chemical properties of triarylphosphines and their corresponding metal-complexes: Solubilising effect in supercritical carbon dioxide

Article abstract of DOI:10.1039/b406691j

The donor strengths of the following triarylphosphine ligands P(Ar) ₂(Ar′) (Ar = Ar′ = 4-Me₃SiC₆H ₄, 1b; 4-Me₃CC₆H₄, 1d; 4-F ₃CC₆H₄, 1e; Ar = C

Full text of DOI:10.1039/b406691j

View Article Online / Journal Homepage / Table of Contents for this issue
F U L L P A P E R
Effect of trimethylsilyl substitution on the chemical properties
of triarylphosphines and their corresponding metal-complexes:
solubilising effect in supercritical carbon dioxide†
Francisco Montilla,*a Agustín Galindo,a Vitor Rosab and Teresa Avilésb
a
Departamento de Química Inorgánica, Facultad de Química, Universidad de Sevilla,
Aptdo. 553, 41071 Sevilla. E-mail: montilla@us.es
REQUIMTE/CQFB, Departamento de Química, FCT, Universidade Nova de Lisboa,
b
2829-516 Caparica, Portugal
Received 4th May 2004, Accepted 18th June 2004
First published as an Advance Article on the web 21st July 2004
The donor strengths of the following triarylphosphine ligands P(Ar)2(Ar′) (Ar = Ar′ = 4-Me3SiC6H4, 1b; 4-Me3CC6H4,
1d; 4-F3CC6H4, 1e; Ar = C6H5, Ar′ = 4-Me3SiC6H4, 1c) have been evaluated experimentally and theoretically. The
measurements of the J(P–Se) coupling constants of the corresponding synthesised selenides SeP(Ar)2(Ar′), 2b,c and
the DFT calculation of the energies of the phosphine lone-pair (HOMO) reveal insignificant influence on the electronic
properties of the substituted phosphines when the trimethylsilyl group is attached to the aryl ring, in marked contrast
to the strong electronic effect of the trifluoromethyl group. These triarylphosphine ligands P(Ar)2(Ar′) reacted with
(5-C5H5)Co(CO)2, (5-C5H5)Co(CO)I2 or PdCl2 to yield the new compounds (5-C5H5)Co(CO)[P(Ar)2(Ar′)], 3b,d; (5-
C5H5)CoI2[P(Ar)2(Ar′)], 4b–e; and PdCl2[P(Ar)2(Ar′)]2, 5b,c, respectively. These complexes have been characterized
and their spectroscopic properties compared with those reported for the known triphenylphosphine complexes. Again,
the contrast of the 31P NMR and 13C NMR chemical shifts or C–O or M–Cl stretching frequencies, when applied, does
not show an important electronic effect on the metal complex of the trimethylsilyl substituted phosphines with respect to
P(C6H5)3 derivatives. Solubility measurements of complexes 3a and 3b in scCO2 were performed. We conclude that Me3Si
groups on the triarylphosphine improve the solubility of the corresponding metal complex in scCO2.
experimentally and theoretically. Subsequently, we prepared several
new cobalt and palladium complexes and studied the effect of TMS
substitution on the basis of the comparison of their spectroscopic
properties with those of the parent, non-substituted, P(C6H5)3-metal
complexes and, additionally, tert-butyl or trifluoromethyl substituted
triphenylphosphine metal complexes. Finally, we have verified how
theTMS functionalization of the triphenylphosphine ligand increases
the solubility of the metal complexes in the scCO2 media.
Introduction
One of the important areas of development in metal mediated
homogeneous catalysis is concerned with the use of non-
conventional solvents such as water, room-temperature ionic liquids
(RTIL), supercritical carbon dioxide (scCO2), and perfluorocarbons,
as reaction media.1 Triphenylphosphine is a widely used ligand
in homogeneous catalysis,2 but the phosphine itself and their
complexes are not particularly soluble in those solvents. Therefore,
the use of related ligands containing solubilising groups attached
to the aryl phosphine substituents has been a widely applied
strategy in the past few years in order to improve the solubility of
metal complexes. For catalysis in highly polar media (i.e. aqueous
phase or RTIL), this can be done by appending the arylphosphines
with hydrophilic groups such as –SO3Na, –CO2Na and –NR3Cl,3,4
whereas by attaching perfluorinated moieties to the arylphosphines
results in metal-complexes that are fluorous-soluble or scCO2-
soluble.5,6 The latter strategy has been successfully applied to a
wide range of scCO2 metal catalysed homogeneous processes.7–9
A possible drawback to their more generalized use is the strong
electron-withdrawing properties of the fluoroalkyl groups that
affect the reactivity of the catalyst.9 A successful solution to this
problem was found by using an appropriate spacer.10 In any case, the
synthesis of ligands of this sort, as well as of their metal complexes,
is usually difficult and expensive. For these reasons, the design
of new solubilisers, that are easy to prepare but do not alter the
chemical properties of the catalyst, is a highly desirable goal.
With respect to this, we have recently reported a new methodology
that increases the solubility of cyclopentadienyl metal complexes
in scCO2 through the functionalisation of the Cp ligand with one
trimethylsilyl group (TMS).11 We are currently investigating the
effect of introducing a TMS on several ligands and here we report
the extension of our studies to triphenylphosphine. Initially, we
examined the -donor ability of the free phosphine ligands, both
Experimental
All preparations and other operations were carried out under dry
oxygen-free nitrogen or argon atmosphere following conventional
Schlenk techniques. Solvents were dried and degassed before
use. Tris(4-trimethylsilylphenyl)phosphine, P(4-Me3SiC6H4)3
(1b)12 and diphenyl(4-trimethylsilylphenyl)phosphine, P(4-
Me3SiC6H4)(C6H5)2 (1c)13 were prepared, according to previously
described procedures, through the reaction of the appropriate aryl
magnesium bromide with phosphorus trichloride or (C6H5)2PCl,
respectively. Phosphines P(4-Me3CC6H4)3 (1d)14 and P(4-
16
F3CC6H4)3 (1e)15 and compounds (5-C5H5)Co(CO)2 and (5-
17
C5H5)Co(CO)I2 were also prepared by the methods reported in
the literature. CO2, employed for supercritical measurements, was
purchased from Carburos Metálicos (99.9999% purity). Infrared
spectra were recorded on Perkin-Elmer Model 883 and ATI
1
Mattson Genesis FTIR spectrophotometer. H, 13C and 31P NMR
spectra were run on Bruker AMX-300, Bruker AMX-500 or Bruker
ARX-400 spectrometer. 1H NMR spectra were recorded using
SiMe4 as internal reference, and 13C NMR spectra were referenced
using the 13C resonance of the solvent as internal standard. 31P shifts
were measured with respect to external 85% H3PO4. Microanalyses
(C, H, N) were carried out by the Microanalytical Service of the
University of Sevilla (Spain) or by the Microanalytical Laboratory
of the Universidade Técnica de Lisboa (Portugal).
† Electronic supplementary information (ESI) available: selected computed
data of the arylphosphines and correlations between their computed HOMO
energies and spectroscopic parameters of the corresponding selenides. See
http://www.rsc.org/suppdata/dt/b4/b406691j/
Preparation of phosphine selenides 2
The known derivatives 2a18 and 2e19 were prepared as previously
reported by heating a solution in toluene of the corresponding
2 5 8 8
D a l t o n T r a n s . , 2 0 0 4 , 2 5 8 8 – 2 5 9 2
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 4

View Article Online
1
phosphine in contact with elemental selenium. The new phosphine
selenide compounds 2b and 2c were similarly obtained, as described
below for 2b.
56%). H NMR (CDCl3):  8.10–8.05 (m, 6, Ar), 7.40 (s, H, Ar),
4.75 (s, 5 H, Cp), 0.21 (s, 27, Si(CH3)3). 13C{1H} NMR (CDCl3): 
144.61, 134.69, 134.05 (s, Ar), 87.48 (s, Cp), −0.66 (Si(CH3)3). 31
P
NMR (CDCl3):  31.01 (s). Elemental analysis (found: C, 44.91; H,
5.01: calc. for C32H44PSi3CoI2: C, 44.87; H, 5.18%).
SeP(4-Me3SiC6H4)3, 2b.
A solution of 1b (0.026 g,
0.054 mmol) in toluene (1 ml) was heated at 110 °C in the presence
of elemental selenium (0.02 g, 0.26 mmol, excess) for three
days. The resulting yellow solution was filtered and the volatiles
were removed under vacuum. Compound 2b was obtained as an
air-sensitive yellow solid. Analysis by 31P NMR indicated that
(5-C5H5)CoI2[P(4-Me3SiC6H4)(C6H5)2], 4c. A mixture of (5-
C5H5)Co(CO)I2 (0.65 g, 1.6 mmol) and P(4-Me3SiC6H4)(C6H4)2
(0.19 g, 1.6 mmol) in 60 ml of CH2Cl2 was treated as described
above. Compound 4c was obtained as dark green crystals (0.21 g,
18%). 1H NMR (C6D6):  8.02–7.98 (m, 7,Ar), 7.41–7.11 (m, 7,Ar),
4,71 (s, 5, Cp), 0.15 (s, 9, Si(CH3)3). 13C NMR (100 MHz, C6D6): 
134.74, 134.67 (d, Ar) 134.04, 133.96 (d, Ar), 130.91 (s, Ar), 86.89
(s, Cp), −1.26 (s, CH3). 31P NMR (C6D6):  30.80. Elemental analy-
sis (found: C, 44.03; H, 3.88: calc. for C26H28PSiCoI2: C, 43.84; H,
3.96%).
1
the reaction was complete. H NMR (C6D6):  7.91 (m, 6, Ar),
7.26 (m, 6, Ar), 0.12 (s, 27, Si(CH3)3). 13C{1H} NMR (C6D6): 
144.51–131.34 (Ar), −1.72 (s, Si(CH3)3). 31P NMR (C6D6):  34.62
(1JPSe = 758 Hz).
1
SeP(4-Me3SiC6H4)(C6H4)2, 2c. H NMR (C6D6):  7.85–6.94
(m, 14, Ar), 0.11 (s, 9, Si(CH3)3). 13C{1H} NMR (C6D6): 
144.22–130.98 (Ar), −1.74 (s, Si(CH3)3). 31P NMR (C6D6):  34.65
(1JPSe = 758 Hz).
(5-C5H5)CoI2[P(4-Me3CC6H4)3], 4d. A solution of (5-
C5H5)Co(CO)I2 (0.18 g, 0.4 mmol) and P(4-Me3CC6H4)3 (0.19 g,
0.4 mmol) in 60 ml of CH2Cl2 was treated as mentioned above.
Compound 4d was obtained as dark green crystals (0.16 g, 56%).
1H NMR (CDCl3):  8.30 (s, 12, Ar), 5.04 (s, 5, Cp), 1.18 (s, 27,
C(CH3)3). 13C{1H} NMR (CDCl3):  155.33, 133.25, 132.12 (s, Ar),
85.08 (s, Cp), 32.35 (C(CH3)3), 29.28 (C(CH3)3). 31P NMR (CDCl3):
 27.71 (s). Elemental analysis (found: C 52.05, H 5.61: calc. for
C35H44CoI2P: C 52.00 H 5.49%).
Synthesis of (5-C5H5)Co(CO)(phosphine) complexes 3
Carbonyl-cobalt complexes were prepared by reacting (5-
C5H5)Co(CO)2 with the appropriate phosphine as described below
for 3b. The preparation of (5-C5H5)Co(CO)[(P(C6H5)3] 3a has
already been reported.20
(5-C5H5)Co(CO)[P(4-Me3SiC6H4)3], 3b. A mixture of (5-
C5H5)Co(CO)2 (0.353 g, 1.96 mmol) and 1b (0.94 g, 1.96 mmol)
in 60 ml of petroleum ether (boiling range 100–140 °C) was kept
under reflux for 24 h. The resulting red–black reaction mixture was
cooled to room temperature, a red crystalline solid precipitating out
of the solution. After several hours the mixture was filtered and the
resulting crystals were washed with portions of petroleum ether
(boiling range 40–60 °C) and air-sensitive red crystals of compound
3b were obtained (0.9 g, 72%). 1H NMR (C6D6):  7.93 (m, 6, Ar),
7.46 (m, 6,Ar), 4.81 (s, 5, Cp), 0.25 (s, 27, Si(CH3)3). 13C{1H} NMR
(C6D6):  208.16 (br, CO), 142.24–133.22 (Ar), 83.11 (s, Cp), −1.28
(s, Si(CH3)3). 31P NMR (C6D6):  69.32 (s). IR (Nujol): /cm−1 1924
(M–CO). Elemental analysis (found: C, 61.90; H, 7.63: calc. for
C33H44OPSi3Co: C, 62.83; H, 7.03%).
(5-C5H5)CoI2[P(4-F3CC6H4)3], 4e. A solution of (5-C5H5)Co-
(CO)I2 (0.52 g, 1.3 mmol) and P(4-F3CC6H4)3 (0.5 g, 1.3 mmol)
in 60 ml of CH2Cl2 was treated as stated above. Compound 4e was
1
obtained as dark green crystals (0.74 g, 68%). H NMR (CDCl3):
 8.01–7.79 (m, 12, Ar), 5.15 (s, 5, Cp). 13C{1H} NMR (CDCl3):
 134.65, 132.74, 125.98 (s, Ar), 87.10 (s, Cp). 31P NMR (C6D6):
 32.1 (s). Elemental analysis (found: C, 38.10; H, 2.04: calc. for
C26H17PF9CoI2: C, 37.00; H, 2.03%).
Synthesis of PdCl2(phosphine)2 complexes 5
Palladium complexes 5a21 and 5e22 were prepared as reported in the
literature by reacting a solution of the corresponding phosphine in
ethanol with PdCl2. The new phosphine-palladium compounds 5b
and 5c were prepared by the same method.
(5-C5H5)Co(CO)[P(4-Me3CC6H4)3], 3d. A mixture of (5-
C5H5)Co(CO)2 (0.3 g, 1.6 mmol) and 1d (0.7 g, 1.6 mmol) in
60 ml of petroleum ether (boiling range 100–140 °C) was treated
as before in 3b. Compound 3d was obtained as air-sensitive red
PdCl2[P(4-Me3SiC6H4)3]2, 5b. A mixture of 1b (0.22 g,
0.47 mmol) and PdCl2 (0.042 g, 0.24 mmol) were mixed in ethanol
(25 ml) at 40 °C for 72 h. A long reaction time was required due
to the low solubility of reagents in the solvent. The resulting
suspension was filtered off and the yellow solid washed with
petroleum ether in order to eliminate the unreacted phosphine.
1
crystals (0.56 g, 60%). H NMR (C6D6):  7.85 (s, 4, Ar), 4.77 (s,
5, Cp), 1.20 (s, 27, C(CH3)3). 13C{1H} NMR (C6D6):  209.67 (br,
CO), 153.37, 136.01–135.56, 134.64, 125.77 (Ar), 83.47 (Cp),
34.77 (s, C(CH3)3), 31.32 (s, C(CH3)3). 31P NMR (C6D6):  66.09
(s). IR (Nujol): /cm−1 = 1924 (M–CO). Elemental analysis (found:
C, 74.02; H, 7.70: calc. for C36H44OPCo: C, 74.21; H, 7.61%).
1
Compound 5b was obtained as a yellow solid (0.18 g, 66%). H
NMR (CDCl3):  7.67, 7.52 (m, 6, Ar), 0.26 (s, 27, Si(CH3)3).
13C{1H} NMR (CDCl3):  143.54, 134.27, 133.01, 130.12 (s,
Ar), −1.06 (s, Si(CH3)3). 31P NMR (CDCl3):  22.25 (s). IR (Nujol):
/cm−1 360 (M–Cl). Elemental analysis (found: C, 56.35; H, 6.75:
calc. for C54H78Cl2Si6P2Pd: C, 57.14; H, 6.93%).
Synthesis of (5-C5H5)CoI2(phosphine) complexes 4
Diiodo-cobalt complexes were prepared by reacting (5-C5H5)CoI2-
(CO) with the appropriate phosphine as described below for 4b.
The synthesis of complex (5-C5H5)Co[(P(C6H5)3]I2 4a has already
been described.17
PdCl2[P(4-Me3SiC6H4)(C6H5)2]2, 5c. This compound was
prepared as described above for 5b but using 1c (0.22 g,
0.66 mmol). The higher solubility of 1c in ethanol allowed the
completion of the reaction in 2 h. Compound 5c was obtained as
1
a yellow solid (0.22 g, 79%). H NMR (CDCl3):  7.70, 7.67 (m,
(5-C5H5)CoI2[P(4-Me3SiC6H4)3], 4b. A solution of (5-
C5H5)Co(CO)I2 (0.85 g, 2.09 mmol) in 40 ml of dichloromethane
was added slowly at r.t. to a stirred solution of P(4-Me3SiC6H4)3
(1 g, 2.09 mmol) in 30 ml of dichloromethane. Rapid gas evolution
occurred and the original dark purple colour of (5-C5H5)Co(CO)I2
became dark green. After stirring overnight at room temperature,
the reaction mixture was filtered and 5 ml of petroleum ether was
added to the filtrate. The solvent was gradually removed under
vacuum until precipitation of the dark green crystals appeared to be
complete. The product was then filtered, washed with diethyl ether
and petroleum ether, and dried to give dark green crystals of 4b (1 g,
1
3H, Ar), 7.52 (d, JHH = 7.8 Hz, 2H, Ar), 7.42 (m, 3H, Ar), 7.38
1
(t, JHH = 7.1 Hz, 3H, Ar), 0.25 (s, 9, Si(CH3)3). 13C{1H} NMR
(CDCl3):  143.57, 135.053, 134.09, 132.87, 130.44, 128.02 (s,
Ar), −1.27 (s, Si(CH3)3). 31P NMR (C6D6):  22.57 (s). IR (Nujol):
/cm−1 = 356 (M–Cl).
Experiments in supercritical CO2: solubility measurements
A simple high-pressure apparatus, described elsewhere,11 was used
to carry out the solubility measurements in scCO2. A view cell, with
a volume of 11 ml and equipped with sapphire windows was used
D a l t o n T r a n s . , 2 0 0 4 , 2 5 8 8 – 2 5 9 2
2 5 8 9

View Article Online
Table 1 Synthesised compounds
P(C6H5)3
P(4-Me3SiC6H4)3
P(4-Me3SiC6H4)(C6H5)2
P(4-Me3CC6H4)3
P(4-F3CC6H4)3
L
1a
2a
3a
4a
5a
1b
2b
3b
4b
5b
1c
2c
—
4c
5c
1d
—
3d
4d
—
1e
2e
—
4e
5e
SeL
CpCo(CO)L
CpCoI2L
PdCl2L2
Table 2 31P NMR data of phosphine selenide compounds
for the experiments. Reagents were inserted into the cell under
argon; the reactor was evacuated and then refilled with argon. The
cell was subsequently placed in the high-pressure line and CO2 is
transferred into the cell by using a pneumatic liquid pump. The cell
was heated to the desired temperature in a water-bath. A digital
transducer measured the pressure. For each solubility measure-
ment, a sufficient amount of compound was put into the view cell
to give a saturated solution in scCO2 at 40 °C and at the correspond-
ing pressure. This was visually checked with the presence of the
solid compound in the cell at the given pressure. Before performing
any measurements, the mixtures were efficiently stirred for several
hours in order to guarantee that the saturation of the complex in
the supercritical phase had been achieved. Samples were taken
through a high-pressure sample loop (0.5 ml), by filling it with the
supercritical fluid mixture, depressurising it into a sample vial, and
flushing it with CH2Cl2. Subsequently, the sample loop was dried
with compressed air. The concentration of the CH2Cl2 solutions was
determined by analysing them by UV spectroscopy.
Compounda
J(P–Se)/Hz
(31P)/ppm
SeP(C6H5)3 (2a)
761 (733)b
758
34.7
34.6
34.6
33.1
SeP(4-Me3SiC6H4)3 (2b)
SeP(C6H5)2(4-Me3SiC6H4) (2c)
SeP(4-F3CC6H4)3 (2e)
758
785 (765)c
a NMR in C6D6. b Data taken from ref. 18. c Data taken from ref. 19.
studied by several groups. The first approach for estimating such an
effect was the evaluation of the donor strength of the correspond-
ing selected phosphines. This analysis was carried out from two
complementary viewpoints: experimental and theoretical. On the
experimental side, a simple way of evaluating the -donor ability
of a phosphine is by measuring the magnitude of the coupling con-
1
stant JSeP of the corresponding phosphine selenide.27 To estimate
the donor abilities of the arylphosphines 1b and 1c, we prepared
the corresponding selenides 2b and 2c by reacting the phosphine
with elemental selenium in toluene at 110 °C. In order to compare
the spectroscopic data, we also prepared the known non-substituted
2a18 and CF3-substituted 2e19 derivatives. Selenium satellites gener-
ated by the 77Se isotopomers (7.6% natural abundance) are readily
observed in the 31P NMR spectra (see Table 2).
Computational details
Theelectronicstructureandgeometriesofthephosphinecompounds
1a, 1b, 1c and 1e were computed within the density functional
theory at the B3LYP23,24 level using the 6-31G+(d, p) basis set for all
atoms. The optimised geometries of 1a and 1c were characterized as
local energy minima by digitalisation of the analytically computed
Hessian. In the frequency calculations of 1b and 1e, imaginary
frequencies were obtained corresponding to the rotation of the TMS
and CF3 groups, around the C–Si and C–C bonds, respectively. In
general, imaginary vibrational frequencies of small magnitude are
problematic because recomputation sometimes does not produce
changes in the results.25 For this reason, and also because we assume
that the HOMO energy, the parameter in which we are interested,
would not change significantly after recomputation, no attempts to
eliminate this frequency were performed. The DFT calculations
were performed using the Gaussian 98 suite of programs.26
1
The JSeP values of the TMS-substituted derivatives 2b and 2c
are not much different from those of 2a ([1JSeP(2a) − 1JSeP(2b)] =
3 Hz), which indicates only minor modification of the electronic
properties of the phosphine selenide upon formal addition of the
trimethylsilyl groups. However, these values are significantly lower
than that of 2e ([1JSeP(2a) − 1JSeP(2e)] = 24 Hz), in good agreement
with the increased electron withdrawing character of the aryl sub-
stituent in the trifluoromethyl derivative. Similarly, the 31P NMR
chemical shifts of the selenide phosphine reflect the reduced chemi-
cal effect of the TMS group when attached to the arylphosphine, in
comparison with the effect of the CF3 group.
From the theoretical point of view, Senn and co-workers have
recently shown, through DFT calculations, that there is a good cor-
relation between the energy of the phosphine lone-pair (HOMO)
and the donor strength of the ligand.28 By using the same approach,
we have calculated the HOMO energies of the following phosphines
1a, 1b, 1c and 1e. Their molecular structures (see ESI†) have been
optimised at the same level of theory (B3LYP/6-31G+(d, p)) for
an adequate comparison. The calculated structural parameters
Results and discussion
The aim of this work is to demonstrate that the introduction of one
or several TMS groups on the 4-position of the phenyl group of
triphenylphosphine produces an increase of the solubility of the
corresponding metal complex in scCO2, without significantly af-
fecting its electronic properties. With this goal in mind, we selected
several phosphines and prepared their corresponding selenides and
new cobalt and palladium complexes containing these phosphines
as ligands (Table 1). Two TMS-containing phosphines, P(4-
Me3SiC6H4)3 1b and P(C6H5)2(4-Me3SiC6H4) 1c, were chosen and,
for comparative purposes, several of the parent, non-substituted,
triphenylphosphine derivatives, a, were also synthesised. Further-
more, two other arylphosphines were selected, containing two com-
pletely different substituents, 4-tert-butyl and 4-trifluoromethyl, in
terms of their donor–acceptor properties. Thus, compounds d and
e (Table 1), respectively, were also prepared and characterised.
Selected spectroscopic data (NMR and IR) of the corresponding
metal compounds were studied for each series, which were used as
indicators for electronic variations at the metal centre.
29
are in agreement with those experimentally found for P(C6H5)3
and compares well with those observed in the few complexes
10
containing coordinated P(4-Me3SiC6H4)3 and P(4-F3CC6H4)3.19,30
To our knowledge, no structural data are known for P(C6H5)2(4-
Me3SiC6H4).
Similar consequences to those obtained from 31P NMR of
selenides can be derived from the calculated data. The donor
strength of the TMS-containing phosphines 1b and 1c (measured by
the HOMO energies, −5.95 and −5.98 eV, respectively) are close to
that of P(C6H5)3, 1a (HOMO energy of −6.00 eV) and the formal H
substitution by TMS introduced in the 4-position does not generate
a noteworthy effect in the donor strength of the corresponding
phosphine. By contrast, P(4-F3CC6H4)3, 1e, shows an important
stabilization of the HOMO (−6.89 eV) and subsequently a lower
donorstrength. Both, theexperimentalandthetheoreticalapproaches
are interrelated. In fact, a good correlation (R = 0.994) can be found
Donor strengths of free phosphine ligands
1
The effect produced upon formal substitution of the hydrogen atom
on the 4-position of the phenyl ring of triphenylphosphine was
when the coupling constant JSeP is compared with the energy of
the lone pair (HOMO) and, additionally, an acceptable correlation
2 5 9 0
D a l t o n T r a n s . , 2 0 0 4 , 2 5 8 8 – 2 5 9 2

View Article Online
Table 3 Spectroscopy data of cobalt derivatives 3 and 4
CpCo(CO)L, 3
(31P)a/ppm
CpCoI2L, 4
(31P)b/ppm
Ligand, L
(13C) (CO)a/ppm
(CO)/cm−1
P(C6H5)3, a
70.0
69.3
208.7
208.2
1913
1924
31.1
31.0
30.8
27.7
32.1
P(4-Me3SiC6H4)3, b
P(C6H5)2(4-Me3SiC6H4), c
P(4-Me3CC6H4)3, d
P(4-F3CC6H4)3, e
66.1
209.7
1924
a Solvent, C6D6. b Solvent, CDCl3.
Table 4 Spectroscopy data of PdCl2L2 derivatives
(R = 0.983) is obtained when the latter energy is compared with the
31P chemical shift of the selenide (see graphs in ESI†).
Compound
(M–Cl)/cm−1
(31P)a/ppm
Synthesis of triarylphosphine metal complexes and their
spectroscopic characterization
PdCl2[P(C6H5)3]2 (5a)
358 (360)b
360
362
22.2
22.7
22.6
23.9c
PdCl2[P(4-Me3SiC6H4)3]2 (5b)
PdCl2[P(C6H5)2(4-Me3SiC6H4)]2 (5c)
PdCl2[P(4-F3CC6H4)3]2 (5e)
To ascertain the electronic properties of the TMS-containing
phosphine ligands, their coordination and organometallic chemistry
have been investigated through the preparation of several cobalt
and palladium complexes. Selected spectroscopic properties of the
latter compounds have been compared with that of the complexes
364
a Solvent, CDCl3. b Data taken from ref. 21. c Data taken from ref. 22.
Table 5 Solubility of cobalt complexes in scCO2 at 40 °C and different
pressures
t
containing the non-substituted, and the Bu- or CF3-substituted
phosphines.
The cobalt complexes (5-C5H5)Co(CO)[P(4-Me3SiC6H4)3],
3b, (5-C5H5)Co(CO)[P(4-Me3CC6H4)3], 3d, were prepared by
reaction of [(5-C5H5)Co(CO)2] with the corresponding phosphine,
following an experimental procedure similar to that reported for
compound (5-C5H5)Co(CO)[P(C6H5)3], 3a.20 Reaction of [(5-
C5H5)Co(CO)2] with the trifluoromethylphosphine 1e did not afford
isolable materials. Selected spectroscopic properties of the new
derivatives were used to compare the electronic effects produced
on the complexes by varying the coordinated phosphine. As
convenient tools were chosen the 31P NMR and 13C NMR chemical
shifts because the C–O stretching frequency in the IR spectra is
not a good indicator of the phosphine donor strength in these
compounds.31 The NMR data for 3b are not much different from
those of 3a (see Table 3), which indicates only minor modification
of the electronic properties of the metal-complex upon formal
addition of the trimethylsilyl groups. The NMR data of 3a slightly
differ from those corresponding to the cobalt complex with the tert-
butyl-substituted arylphosphine 3d, which reflects the expected
electron-donating effect of the alkyl groups. In fact, despite the
inductive effect of the Me3Si group, it appears to be only a modest
electron donor, poorer than the alkyl groups. This phenomenon
can be attributed to an electron-withdrawing component in the
behaviour of the silicon, in which charge from the aromatic ring is
delocalized through backdonation into the antibonding orbitals of
the SiC3 framework, cancelling the electron-donating effect of the
methyl groups in the TMS group.32
CpCo(CO)[P(C6H5)3] (3a)
CpCo(CO)[P(4-Me3SiC6H4)3] (3b)
Solubility (g ml−1 104 CO2) P/bar Solubility (g ml−1 104 CO2) P/bar
1.59
1.83
1.88
1.99
160
178
188
205
5.95
6.21
6.1
166
188
204
However, the 31P NMR chemical shifts show the same trend previ-
ously observed for the arylphosphines.
Solubility of triarylphosphine-complexes in scCO2
The newly synthesised TMS-substituted cobalt complex 3b was
used to study the solubilising effects of the TMS groups in scCO2.
For comparison, the measurements were performed in parallel to
that of the well-known non-substituted derivative 3a. It should be
noted that complexes 4 and 5 are not soluble in scCO2, so it was
not possible to evaluate the effect that the presence of TMS groups
have in the solubility properties of these complexes. Solubility mea-
surements in scCO2 were carried out by taking samples through a
high-pressure sample loop followed by UV spectroscopy analysis,
as described elsewhere.11 As summarised in Table 5, the concentra-
tion of 3b in scCO2 was estimated to be more than three times higher
than that of 3a (see graph in ESI†).
Enhancement of the solubility of the arylphosphine cobalt com-
plex in scCO2 upon formal substitution with Me3Si groups is low,
but it is interesting to emphasize that these solubilisers do not inter-
fere with the chemical properties of the coordinated metal centre.
This idea could be used in the design of new catalysts soluble in
scCO2 just by adding Me3Si groups to the metal complexes used in
conventional homogeneous catalysis. The resulting systems should
be more soluble in the supercritical media while maintaining their
catalytic activities. This approach has been recently adopted in the
rhodium-catalysed hydroformylation of olefins in scCO2.13
Similar tendencies are maintained in the CpCoI2L series. These
derivatives are easily obtained by reaction of CpCoI2(CO) with the
corresponding arylphosphine, following the same procedure de-
scribed for the synthesis of the PPh3 derivative 4a.20 The 31P NMR
data of the diiodo derivatives revealed the same trend observed
above for the series of cobalt carbonyl complexes (see Table 3).
The chemical shifts for 4b and 4c are close to that of 4a (maximum
change of 0.3 ppm), while those of 4d and 4e are shifted by 3.4 and
1 ppm, respectively. These facts confirm again the small effect cre-
ated by the TMS groups on arylphosphine properties.
The series of trans-[PdCl2L2] (L = P(4-Me3SiC6H4)3 5b,
P(C6H5)2(4-Me3SiC6H4), 5c) complexes were prepared by the reac-
tion of PdCl2 with two equivalents of the arylphosphine ligand. The
already known 4-H (5a)21 and 4-CF3 (5e)22 substituted triarylphos-
phine palladium complexes were also prepared in order to compare
their spectroscopic properties with the TMS derivatives. The Pd–Cl
stretching frequency of the palladium series was not too sensitive
to the electronic influence of the 4-substituted groups (see Table 4).
Conclusion
The Me3Si groups, introduced as substituents in triarylphosphines,
can enhance the solubility of metal complexes in scCO2 without
greatly affecting the electronic properties of the metal centre.
These observations agree with our previous work on transition
metal catalysts with TMS-substituted cyclopentadienyl ligand.
Nevertheless, the increase in the solubility of the metal complexes
D a l t o n T r a n s . , 2 0 0 4 , 2 5 8 8 – 2 5 9 2
2 5 9 1

View Article Online
18 D. W. Allen, I. W. Nowell and B. F. Taylor, J. Chem. Soc., Dalton
Trans., 1985, 2505.
in scCO2 using either TMS-substituted triarylphosphine or
cyclopentadienyl ligands is quite modest to be used as a systematic
solution in order to solubilise metal-complexes in the supercritical
medium. New approaches have to be made in order to design new
scCO2 homogeneously catalysed processes based on traditional
catalytic systems. With respect to this, we are presently attempting
to increase the solubility of selected catalysts in scCO2 by
introducing polysiloxane chains pendant to the ligands.
19 J. A. S. Howell, N. Fey, J. D. Lovatt, P. C. Yates, P. McArdle,
D. Cunningham, E. Sadeh, H. E. Gottlieb, Z. Goldschmidt, M. B.
Hursthouse and M. E. Light, J. Chem. Soc., Dalton Trans., 1999, 3015.
20 H. G. Schuster-Woldan and F. Basolo, J. Am. Chem. Soc., 1966, 88,
1657.
21 W. A. Herrmann and A. Salzer, Synthetic Methods of Organometal-
lic and Inorganic Chemistry, Herrmann/Brauer. Stuttgart, New York,
Thieme, 1996, vol. 1, p. 160.
22 J. P. Neilan, R. M. Laine, N. Cortese and R. F. Heck, J. Org. Chem.,
1986, 41, 3455; D. Hancu, J. Green and E. J. Beckman, Ind. Eng. Chem.
Res., 2002, 41, 4466.
Acknowledgements
The authors thank the financial support from European Commission
(Contracts HPMF-CT-1999-00373 and HPMF-CT-2002-01609)
and the Spanish MCyT (Ramón y Cajal Contract).
23 A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
24 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785.
25 See similar comments in: Y. Xie, J. H. Jang, R. B. King and H. F.
Schaefer III, Inorg. Chem., 2003, 42, 5219.
26 M. J. Frisch,
G. W. Trucks,
H. B. Schlegel,
G. E. Scuseria,
References
M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery,
Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam,
A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo,
S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui,
K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari,
J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara,
C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen,
M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle and
J. A. Pople, GAUSSIAN 98 (Revision A.7), Gaussian, Inc., Pittsburgh,
PA, 1998.
1 D. J. Cole-Hamilton, Science, 2003, 299, 1702, and references therein.
2 B. Cornils and W. A. Herrmann, Applied Homogeneous Catalysis with
Organometallic Compounds, VCH, Weinheim, Germany, 1996.
3 C. J. Li and T. G. Chan, Organic Reactions in Aqueous Media, Wiley,
New York, 1997.
4 Y. Chauvin, L. Mussmann and H. Olivier, Angew. Chem., Int. Ed. Engl.,
1996, 34, 2698; M. F. Sellin, P. B. Webb and D. J. Cole-Hamilton,
Chem. Commun., 2001, 8, 781.
5 I. T. Horvath, Acc. Chem. Res., 1998, 31, 641.
6 P. G. Jessop and W. Leitner, in Chemical Synthesis Using Supercritical
Fluids, Wiley-VCH, 1999, special issue devoted to supercritical fluids;
Chem. Rev., 1999, 99, 2.
27 D. W. Allen and B. F. Taylor, J. Chem. Soc., Dalton Trans., 1982, 51;
S. M. Socol and J. G. Verkade, Inorg. Chem., 1984, 23, 3487; T. S.
Barnard and M. R. Mason, Organometallics, 2001, 20, 206.
28 H. M. Senn, D. V. Deubel, P. E. Blöchl, A. Togni and G. Frenking,
J. Mol. Struct. (Theochem), 2000, 506, 233.
7 S. Kainz, D. Koch, W. Baumann and W. Leitner, Angew. Chem., Int. Ed.
Engl., 1997, 36, 1628.
8 S. Kainz, A. Brinkmann, W. Leitner and A. Pfaltz, J. Am. Chem. Soc.,
1999, 121, 6421; M. Carroll and A. B. Holmes, Chem. Commun., 1998,
1395.
29 B. J. Dunne and A. G. Orpen, Acta Crystallogr., Sect. C: Cryst.
Struct. Commun., 1991, 47, 345; A. N. Chekhlov, Kristallografiya,
1993, 38, 79; J. Bruckmann, C. Kruger and F. Lutz, Z. Naturforsch.,
Teil B: Chem. Sci., 1995, 50, 351; H. Kooijman, A. L. Spek, K. J. C.
van Bommel, W. Verboom and D. N. Reinhoudt, Acta Crystallogr.,
Sect.C: Cryst. Struct. Commun., 1998, 54, 1695; B. Ziemer, A. Rabis
and H.-U. Steinberger, Acta Crystallogr., Sect.C: Cryst. Struct. Com-
mun., 2000, 56, e58.
30 K. Biradha, V. M. Hansen, W. K. Leong, R. K. Pomeroy and M. J.
Zaworotko, J. Cluster Sci., 2000, 11, 285; D. H. Farrar, E. V. Grachova,
A. Lough, C. Patirana, A. J. Poe and S. P. Tunik, J. Chem. Soc.,
Dalton Trans., 2001, 2015; S. A. Serron, L. Luo, C. Li, M. E. Cucullu,
E. D. Stevens and S. P. Nolan, Organometallics, 1995, 14, 5290.
31 See, for example: A. J. Hart-Davis and W. A. G. Graham, Inorg. Chem.,
1970, 9, 2658.
9 D. R. Palo and C. Erkey, Organometallics, 2000, 19, 81; T. Davis and
C. Erkey, Ind. Eng. Chem. Res., 2000, 39, 3671.
10 B. Richter, A. L. Spek, G. Van-Koten and B. J. Deelman, J. Am. Chem.
Soc., 2000, 122, 3945; B. Richter, E. de Wolf, B. J. Deelman and
G. Van-Koten, J. Org. Chem., 2000, 65, 3885.
11 F. Montilla, V. Rosa, C. Prevett, T. Avilés, M. Nunes da Ponte, D. Masi
and C. Mealli, Dalton Trans., 2003, 11, 2170.
12 K. C. Frisch and H. Lyons, J. Am. Chem. Soc., 1953, 75, 4079.
13 M. F. Sellin, I. Bach, J. M. Webster, F. Montilla, V. Rosa, T. Avilés,
M. Poliakoff and D. J. Cole-Hamilton, J. Chem. Soc., Dalton Trans.,
2002, 4569.
14 H. A. Brune, M. Falck, R. Hemmer, G. Schmidtberg and H. G. Alt,
Chem. Ber., 1984, 117, 2791.
15 K. C. Eapen and C. Tamborski, J. Fluorine Chem., 1980, 15, 239.
16 M. D. Rausch and R. A. Genetti, J. Org. Chem., 1970, 35, 3888.
17 R. B. King, Inorg. Chem., 1966, 5, 82.
32 M. A. Weiner and M. Lattman, J. Org. Chem., 1975, 40, 1292.
2 5 9 2
D a l t o n T r a n s . , 2 0 0 4 , 2 5 8 8 – 2 5 9 2