Biphasic hydroformylation catalysis in ionic liquid media

Article abstract of DOI:10.1016/j.poly.2004.06.029

A series of ionic liquids have been investigated as reaction media for biphasic hydroformylation catalysis. In the study the rhodium/tppti* (tppi* = tri(m-sulfonyl) triphenyl phosphine 1,2-dimethyl-3-butyl- imidazolium salt) catalyst system was investigated together with hexene-1 as the substrate. The activity of the biphasic ionic liquid system containing the Rh/tppti* catalyst was approximately one order of magnitude lower than that observed for the conventional Rh/PPh₃ catalyst in toluene. The ionic liquids 1-butyl-3-methyl-imidazolium tetrafluoroborate ([bmim][BF ₄]), 1-butyl-3-methyl-imidazolium hexafluorophosphate ([bmim][PF ₆]) and 1,2-dimethyl-3-butyl-imidazolium hexafluorophospate ([bdmim][PF₆]) were investigated as potential media for hydroformylation catalysis in a liquid-liquid biphase reaction environment. The study used the known rhodium/tppti* (tppti* = tri(m-sulfonyl) triphenyl phosphine 1,2-dimethyl-3-butyl-imidazolium salt) system as the catalyst and hexene-1 as the substrate, which enabled the produced heptanal to be conveniently isolated from the ionic liquid by phase separation. All catalyst evaluations were carried out in high-purity ionic liquids to avoid any decomposition of the catalyst, substrate and/or ionic liquid. The activity of the biphasic ionic liquid system containing the Rh/tppti* catalyst was approximately one order of magnitude lower than that observed for the conventional Rh/PPh₃ catalyst in toluene. The Rh/tppti* catalyst system showed a decrease of activity in the semicontinuous hydroformylation of hexene-1 due to catalyst deactivation and/or metal loss. The n/i ratio of the produced heptanal only slightly increased after changing the Rh/P ratio from 1:3 to 1:100 in [bmim][PF₆]. The loss of rhodium metal from the ionic liquid phases appears to be correlated to the solubility characteristics of the ionic liquid and the concentration of the aldehyde in the organic phase. In support of this theory, the ionic liquid [bmim][PF ₆] gave the best results due to its very low miscibility with polar substances. A high-pressure NMR (HP-NMR) study revealed that the solution structure of HRh(¹³CO)(tppts)₃ (tppts = tri(m-sulfonyl)triphenyl phosphine sodium salt) in [bmim][BF₄] was similar to that of HRh(¹³CO)(PPh₃)₃ in toluene-d⁸. The HP-NMR investigation at elevated syngas pressures showed that HRh(¹³CO)(tppts)₃ lies in equilibrium with HRh(¹³CO)₂(tppts)₂.

Full text of DOI:10.1016/j.poly.2004.06.029

Polyhedron 23 (2004) 2679–2688
www.elsevier.com/locate/poly
Biphasic hydroformylation catalysis in ionic liquid media
*
Christian P. Mehnert , Raymond A. Cook,
Nicholas C. Dispenziere, Edmund J. Mozeleski
ExxonMobil Research and Engineering Company, Corporate Strategic Research, Annandale, NJ 08801, USA
Received 15 April 2004; accepted 19 June 2004
Available online 3 September 2004
Dedicated to Prof. M.L.H. Green on the occasion of his retirement from the position of Head of the Inorganic Chemistry
Department at Oxford University
Abstract
The ionic liquids 1-butyl-3-methyl-imidazolium tetrafluoroborate ([bmim][BF4]), 1-butyl-3-methyl-imidazolium hexafluorophos-
phate ([bmim][PF6]) and 1,2-dimethyl-3-butyl-imidazolium hexafluorophospate ([bdmim][PF6]) were investigated as potential media
for hydroformylation catalysis in a liquid–liquid biphase reaction environment. The study used the known rhodium/tppti*
(tppti* = tri(m-sulfonyl) triphenyl phosphine 1,2-dimethyl-3-butyl-imidazolium salt) system as the catalyst and hexene-1 as the sub-
strate, which enabled the produced heptanal to be conveniently isolated from the ionic liquid by phase separation. All catalyst eval-
uations were carried out in high-purity ionic liquids to avoid any decomposition of the catalyst, substrate and/or ionic liquid. The
activity of the biphasic ionic liquid system containing the Rh/tppti* catalyst was approximately one order of magnitude lower than
that observed for the conventional Rh/PPh3 catalyst in toluene. The Rh/tppti* catalyst system showed a decrease of activity in the
semicontinuous hydroformylation of hexene-1 due to catalyst deactivation and/or metal loss. The n/i ratio of the produced heptanal
only slightly increased after changing the Rh/P ratio from 1:3 to 1:100 in [bmim][PF6]. The loss of rhodium metal from the ionic
liquid phases appears to be correlated to the solubility characteristics of the ionic liquid and the concentration of the aldehyde
in the organic phase. In support of this theory, the ionic liquid [bmim][PF6] gave the best results due to its very low miscibility with
polar substances. A high-pressure NMR (HP-NMR) study revealed that the solution structure of HRh(13CO)(tppts)3 (tppts = tri(m-
sulfonyl)triphenyl phosphine sodium salt) in [bmim][BF4] was similar to that of HRh(13CO)(PPh3)3 in toluene-d8. The HP-NMR
investigation at elevated syngas pressures showed that HRh(13CO)(tppts)3 lies in equilibrium with HRh(13CO)2(tppts)2.
ꢀ 2004 Elsevier Ltd. All rights reserved.
Keywords: Hydroformylation; Ionic liquids; Biphase catalysis; High-pressure NMR; Rhodium complexes
1. Introduction
approaches have been explored to overcome this limita-
tion including aqueous [2] and fluorous biphase catalysis
[3], reactions in supercritical media [4] and catalyst
immobilization onto solid support materials [5].
Although some of these novel concepts have successfully
been developed into commercial processes, like the Ruhr
The separation of homogeneous hydroformylation
catalysts from reaction mixtures has been a tremendous
challenge and continues to be the focus of intense re-
search [1]. Over the last three decades, several elegant
ˆ
Chemie Rhone Poulenc Process used in the production
*
Corresponding author at: ExxonMobil Chemical Company,
Baytown Technology and Engineering Complex, 5200 Bayway Drive,
of butyraldehyde [2], catalyst recovery continues to be
one of the key issues of homogeneous catalysis.
More recently, ionic liquids have been used as alter-
native reaction media for homogeneous catalysis [6].
Based on their highly charged nature, ionic liquids are
Baytown, TX 77006, USA. Tel.: +1 281 834 1825; fax: +1 281 834
2371.
E-mail address: christian.p.mehnert@exxonmobil.com (C.P.
Mehnert).
0277-5387/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.poly.2004.06.029

2680
C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
well suited for biphasic reactions with organic sub-
strates. While the first hydroformylation reaction in an
ionic liquid (at that time referred to as a molten salt)
was performed by Parshall [7] as earlier as 1972, it was
not until 1995 that Chauvin et. al. [8] used water-soluble
phosphine ligands like the tri(m-sulfonyl)triphenyl phos-
phine sodium salt (tppts) [9] to retain active rhodium
complexes in the ionic liquid phases and used them suc-
cessfully in biphasic hydroformylation catalysis. Since
then, a variety of different metal complexes have been
investigated for liquid–liquid biphasic hydroformylation
reactions in ionic liquid media [10].
were analyzed using a HP-5890 gas chromatograph with
a 30 m Spb-1 capillary column, 0.25 mm i.d. and 0.25
mm film thickness and a HP-6890 GC/MS equipped
with a 30 m HP-1 capillary column, 0.25 mm i.d. and
0.25 mm film thickness. Rhodium analyses were per-
formed on an inductively coupled plasma–atomic emis-
sion spectrometer (ICP/AES), Jarrell-Ash Model 1100,
with argon gas flow and torch alignment. The ICP/
AES instrument was calibrated with 10 ppm rhodium
solution. Elemental analyses were performed by the Gal-
braith Laboratory.
As the production of aldehyde is one of the largest
volume processes [11] in the chemical industry we have
recently attempted to ascertain the viability of ionic liq-
uids for hydroformylation reactions [12]. In this paper,
we report our findings on the biphasic hydroformylation
catalysis using conventional rhodium catalysts in ionic
liquid media, which was encouraged by the initial report
of Chauvin et. al. [8]. In this study we have investigated
the reactivity of the hydroformylation catalyst
HRh(CO)(P)3 (P = phosphine ligand) in imidazolium
based ionic liquid derivatives of tetrafluoroborate and
hexafluorophosphate. In addition, we have compli-
mented our investigation with a high-pressure NMR
(HP-NMR) study of the reaction mixture containing
the active rhodium species.
2.2. Preparation of the ionic liquids
2.2.1. 1-Butyl-3-methyl-imidazolium tetrafluoroborate –
[bmim][BF4]
A 500 ml round bottom flask was charged under N2
with 95 g (0.54 mol) 1-butyl-3-methyl-imidazolium chlo-
ride [14] and 63 g (0.57 mol) sodium tetrafluoroborate in
250 ml of degassed water. After the reaction was stirred
for 15 h at room temperature, the mixture was extracted
with dichloromethane (3· 100 ml). The combined or-
ganic phases were dried over magnesium sulfate and fil-
tered through a bed of activated carbon and alumina.
After the removal of the volatile components under re-
duced pressure the ionic liquid [bmim][BF4] was ob-
tained as a colorless liquid in 73% yield. Elemental
analysis (%) calculated for C8H15N2BF4: C, 42.51; H,
6.69; N, 12.39. Found: C, 42.66; H, 7.12; N, 12.49%.
2. Experimental
2.2.2. 1-Butyl-3-methyl-imidazolium hexafluorophosphate –
[bmim][PF6]
2.1. General experimental details and materials
A 1000 ml round bottom flask was charged under N2
with 175 g (1.0 mol) 1-butyl-3-methyl-imidazolium chlo-
ride and 190 g (1.03 mol) potassium hexafluorophos-
phate in 400 ml degassed water. After the reaction was
stirred for 15 h at room temperature, the mixture was
extracted with dichloromethane (3· 200 ml). The com-
bined organic phases were dried over magnesium sulfate
and filtered through a bed of activated carbon and alu-
mina. After the removal of the volatile components un-
der reduced pressure the ionic liquid [bmim][PF6] was
obtained as a colorless liquid in 80% yield. Elemental
analysis (%) calculated for C8H15N2PF6: C, 33.81; H,
5.32; N, 9.86. Found: C, 33.92; H, 5.64; N, 9.72%.
All operations were performed in a Vacuum Atmos-
pheres or MBraun dry box under purified nitrogen or
using Schlenk techniques under a nitrogen atmosphere.
The following chemicals have been purchased from Ald-
rich and have been used as received: anhydrous heptane,
anhydrous toluene, dichloromethane, hexene-1, 1-chlo-
robutane, 1-methyl-imidazol, 1,2-dimethyl-imidazol,
Celite, alumina, activated carbon, sodium tetrafluorobo-
rate, sodium hexafluorophosphate and potassium hexa-
fluorophosphate. The following compounds have been
purchased from Strem and have been used as received:
tris(m-sulfonylphenyl)phosphine sodium salt (tppts)
and Rh(CO)2(acac) (acac = acetylacetonate). The gases
CO/H2 (1:1, UHP grade) and 13CO (UHP grade) were
purchased from BOC and MG industries, respectively.
NMR spectra were recorded on a Varian Unity 400
spectrometer. Chemical shifts (ppm) were referenced to
either an internal standard (e.g. tetramethylsilane) or
an external standard (e.g. triphenylphoshite). HP-
NMR experiments were performed using single-crystal
sapphire NMR tubes [13]. Toluene-d8 was purchased
from Cambridge Isotopes and was degassed and dried
over potassium metal. Gas chromatography samples
2.2.3. 1,2-Dimethyl-3-butyl-imidazolium hexafluorophos-
phate – [bdmim][PF6]
A 500 ml round bottom flask was charged under N2
with 56.6 g (0.3 mol) 1,2-dimethyl-3-butyl-imidazolium
chloride and 55.4 g (0.33 mol) sodium hexafluorophos-
phate in 250 ml of degassed water. After the reaction
was stirred for 15 h at 40 ꢁC, the mixture was extracted
with dichloromethane (3· 100 ml). The combined or-
ganic phases were dried over magnesium sulfate and fil-
tered through a bed of activated carbon and alumina.

C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
2681
After the removal of the volatile components under re-
duced pressure the ionic liquid [bdmim][PF6] was ob-
tained as a yellow-colored viscous liquid in 70% yield.
Elemental analysis (%) calculated for C9H17N2PF6: C,
36.25; H, 5.75; N, 9.39. Found: C, 36.27; H, 5.83; N,
9.36%.
5 mm above the heavier ionic liquid phase. At the end of
each run, the reactor was cooled to 15 ꢁC, the stirrer was
then stopped and the supernatant product phase was
removed from the reactor. For metal balance, the concen-
tration of Rh in each product was determined by ICP-
AES. For each consecutive run, the catalyst phase was
rapidly heated to the reaction temperature and a fresh
charge of hexene-1 was injected to initiate the reaction.
For the rhodium analysis, aliquots of the product
were ashed with 1 ml of concentrated H2SO4 and left
in a furnace at 530 ꢁC for at least 8 h. The remaining res-
idue was fused with 0.5 g K2S2O7. The resulting mixture
was solubilized in a 20% H2SO4 solution. The rhodium
analysis was performed using ICP/AES instrumentation
with matrix-matched standards (3.35 RSD).
2.3. HP-NMR investigation
A solution of 3 mg (0.013 mmol) Rh(CO)2(acac) and
43 mg (0.075 mmol) tppts in 2 ml of [bmim][BF4] con-
taining 5 wt% of D2O was prepared under N2 and
charged into a sapphire NMR tube. The tube was placed
into a safety shield container and pressurized to 60 psi
with 13CO/H2 (1:1). To ensure the formation of
HRh(CO)(tppts)3 (1a), the tube was shaken at 60 ꢁC
for 1 h. After the pressure was released the 31P{1H}
and 13C{1H} NMR spectra were recorded under 15 psi
of 13CO/H2 (1:1), confirming the presence of 1a. During
the course of the HP-NMR investigation the tube was
pressurize up to 2000 psi with 13CO/H2 (1:1) and further
monitored by NMR spectroscopy.
2.4.2. Preparation of the tri(m-sulfonyl)triphenylphos-
phine 1,2-dimethyl-3-butyl-imidazolium salt (tppti*)
A 50 ml round bottom flask was charged under N2
with 10.1 g (0.017 mol) tri(m-sulfonyl)triphenyl phos-
phine sodium salt (tppts) and 10.0 g (0.053 mol) 1,2-di-
methyl-3-butyl-imidazolium chloride in acetonitrile (25
ml). After the reaction was stirred for 3 days at room
temperature, the mixture was filtered through a bed of
Celite. The filter aid was further washed with acetonitrile
(30 ml). The resulting filtrates were combined and used
as a stock solution for the tppti* ligand.
2.4. Catalyst evaluation
2.4.1. General experimental details for hydroformylation
reactions
In a typical experiment carried out in a 300 ml auto-
clave, catalyst solution was prepared from the ionic liq-
uid, the rhodium precursor Rh(CO)2(acac) and the
phosphine (tppti*) stock solution. The mixture was
charged into the reactor under nitrogen and the volatile
components removed under reduced pressure (10ꢀ2
Torr). The reactor was then flushed and pressurized with
CO/H2 (1:1) to ca. 150 psi. The run olefin was charged in
a glovebox into a 100 ml injection bomb, which was then
mounted into the feed line of the reactor. After the reac-
tor was heated to the run temperature, the reaction was
initiated by injecting the olefin into the reactor and the
pressure adjusted up to 600 psi CO/H2 (1:1). The pres-
sure in the reactor was maintained constant by making
up the consumed CO/H2 (1:1) from the PVT tank. The
conversion was followed by recording the pressure drop
in the PVT tank.
2.4.3. Biphasic Hydroformylation using Rh(CO)2(acac)/
tppti* in [bmim][BF4] (Table 2/Entries 1–10)
A mixture of 26 mg (0.1 mmol) Rh(CO)2(acac) and
3.2 ml (1.0 mmol) tppti* (stock solution) in 60 ml
[bmim][BF4] was stirred at room temperature. After 1
h the yellow solution was charged into a 300 ml auto-
clave and the volatile components removed under re-
duced pressure (10ꢀ2 Torr). The reactor was
pressurized with CO/H2 (1:1) to ca. 150 psi. After the
reactor was heated to 100 ꢁC, the reaction was initiated
by injecting 61 g (720 mmol) hexene-1 with 300 psi of
CO/H2 (1:1) into the reactor. Due to the extended induc-
tion period after the first recycle run the reactor pressure
was increased to 600 psi. All further runs were con-
ducted at this run pressure.
The apparent turnover frequency (TOF) used in the
kinetic analysis is defined as the rate of olefin conversion
by 1 mol of rhodium and is expressed as the moles of
olefin converted to aldehyde per moles of rhodium per
minute: TOF = {(dnolefin/dt)t}/nRh. The rate of olefin
conversion is calculated from the PVT pressure versus
time data. The values for the TOF were determined dur-
ing the initial 30 min of the gas consumption which cor-
responds to an approximate olefin conversion of
between 10% and 20%.
2.4.4.
[bmim][PF6] (Table 2/Entries 11–20)
Reaction
of
Rh(CO)2(acac)/tppti*
in
A mixture of 26 mg (0.1 mmol) Rh(CO)2(acac) and
3.2 ml (1.0 mmol) tppti* stock solution in 60 ml
[bmim][PF6] was stirred at room temperature. After 1
h the yellow solution was charged into a 300 ml auto-
clave and the volatile components removed under re-
duced pressure (10ꢀ2 Torr). The reactor was
pressurized with CO/H2 (1:1) to ca. 300 psi. After the
reactor was heated to 100 ꢁC, the reaction was initiated
by injecting 26 g (307 mmol) hexene-1 in 38 ml of hep-
tane with 600 psi of CO/H2 (1:1) into the reactor. The
For the semicontinuous hydroformylation of hexene-
1, the autoclave was fitted with a dip leg that reached ca.

2682
C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
autoclave studies using Rh/P ratios of 1:3 and 1:100 (Ta-
ble 1/Entries 28, 29) were carried out under the identical
conditions.
substrate hexene-1. The organic phases showed a lower
density and therefore generated the top layers while
the ionic liquids were confirmed as the bottom phases.
After the reaction mixtures were pressurized with CO/
H2 (1:1) to 600 psi the tubes were shaken at 100 ꢁC.
The reaction was stopped after 3 h and the tubes were
cooled to room temperature. During this time the organ-
ic components separated from the ionic liquid phases.
While the reaction without a ligand and with the ligand
tppti* gave colorless organic phases (upper layer), the
experiment using the ligand PPh3 showed a strong yel-
low coloration of the organic phase. As the complex
HRh(CO)(PPh3)3 has a yellow coloration in solution,
it is a good indication that a substantial amount of the
catalyst migrated during the reaction from the ionic liq-
uid into the organic layer. After the pressure was re-
leased from the sapphire NMR tubes the organic
products were analyzed for their composition and rho-
dium metal content. In all three cases a significant
amount of aldehyde was produced. Concerning the issue
of metal leaching, only the reaction mixture using tppti*
exhibited a negligible amount of rhodium loss from the
ionic liquid, while the other ionic liquid phases suffered
from a significant amount of metal loss. On the basis of
these initial experiments, the approach of using highly
charged ligands like for example tppti* seemed to be a
viable method for retaining a rhodium catalyst in an io-
nic liquid.
2.4.5.
Reaction
of
[bdmim][PF6] (Table 2/Entries 21–27)
Rh(CO)2(acac)/tppti*
in
A mixture of 26 mg (0.1 mmol) Rh(CO)2(acac) and
3.2 ml (1.0 mmol) tppti* stock solution in 60 ml
[bdmim][PF6] was stirred at room temperature. After 1
h the yellow solution was charged into a 300 ml auto-
clave and the volatile components removed under re-
duced pressure (10ꢀ2 Torr). The reactor was
pressurized with CO/H2 (1:1) to ca. 300 psi. After the
reactor was heated to 100 ꢁC, the reaction was initiated
by injecting 25 g (295 mmol) hexene-1 in 38 ml of hep-
tane with 600 psi of CO/H2 (1:1) into the reactor.
2.4.6. Reaction of Rh(CO)2(acac)/PPh3 in toluene
(Table 2/Entries 30)
A mixture of 6.5 mg (0.025 mmol) Rh(CO)2(acac)
and 66 mg (0.25 mmol) PPh3 was dissolved in 60 ml tol-
uene and stirred at room temperature. The yellow solu-
tion was charged into a 300 ml autoclave and
pressurized with CO/H2 (1:1) to ca. 300 psi. After the
reactor was heated to 100 ꢁC, the reaction was initiated
by injecting 61 g (720 mmol) hexene-1 at 600 psi of CO/
H2 (1:1) into the reactor.
One interesting result was that although the organic
phase in the reaction using no ligand was colorless, a
large amount of rhodium metal was found in this phase.
Furthermore, a significant amount of aldehyde was pro-
duced in this reaction which indicates that species other
than HRh(CO)(PPh3)3 are active for the hydroformyla-
tion of the olefin. Most likely the aldehyde was pro-
duced by rhodium-hydrido-carbonyl complexes which
tend to form under hydroformylation reaction condi-
tions in the absence of suitable ligands and have been
observed in other areas of oxo-catalysis [15]. For this
reason, all of our produced aldehyde samples were ana-
lyzed for their rhodium content.
3. Results and discussion
3.1. Catalyst screening in HP-NMR tubes
We started our study by investigating hydroformyla-
tion reactions in ionic liquids under different reaction
environments. This initial study was conveniently car-
ried out in sapphire NMR tubes which enabled us to vis-
ually follow phase separations and determine solubility
limitations. In this investigation the well-established cat-
alyst precursor Rh(CO)2(acac) (acac = acetylacetonate)
was dissolved in the ionic liquid [bmim][PF6] with (a)
no phosphine, (b) triphenylphoshine (PPh3) and (c) the
tri(m-sulfonyl)triphenyl phosphine 1,2-dimethyl-3-bu-
tyl-imidazolium salt (tppti*). The resulting solutions
were loaded into HP-NMR tubes and treated with the
Another potentially complicating factor in the evalu-
ation of these ionic liquid media for hydroformylation
catalysis are the solubility limitations of the highly
charged phosphine ligand. In an early HP-NMR study
Table 1
Selected 31P{1H} and 13C{1H} NMR data for the Rh/P catalyst system (P = PPh3, tppts) in [bmim][BF4], toluene-d8 and water under varying
pressure conditions
Solvent
Pressurea (psi)
Complex
d(13CO) (ppm)
JC–Rh (Hz)
d(31PR3) (ppm)
JP–Rh (Hz)
[bmim][BF4]
[bmim][BF4]
Toluene-d8
Toluene-d8
H2O
15
HRh(13CO)(tppts)3
HRh(13CO)2(tppts)2
HRh(13CO)(PPh3)3
HRh(13CO)2(PPh3)2
HRh(13CO)(tppts)3
195.2
199.2
206.6
199.9
204.9
62
61
62
63
55
32.2
39.1
37.8
33.8
42.8
126
137
156
139
156
2000
15
1000
15/3000
a
Syngas 13CO/H2 (1:1).

C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
2683
we realized that not all the water-soluble phosphine lig-
ands like the tri(m-sulfonyl)triphenyl phosphine sodium
salt (tppts) are also automatically soluble in ionic liquid
phases. Although the ligand showed excellent solubility
in water, it only dissolved in [bmim][BF4] after the addi-
tion of small amounts of water. Even more surprising
was the fact that it has negligible solubility in
[bmim][PF6]. To verify this low solubility we charged
[bmim][PF6] together with Rh(CO)2(acac) and tppts
(Rh/P ratio 1:10) into a HP-NMR tube. The resulting
mixture was stepwise pressurized with CO/H2 (1:1) to
2000 psi and heated up to 150 ꢁC. During this process
we recorded the 31P{1H} NMR spectra of the solution
and were unable to detect neither coordinated nor free
tppts. On the basis of this investigation we synthesized
a more soluble derivative of the phoshine ligand. After
exchanging the sodium cation of tppts with 1,2-di-
methyl-3-butyl-imidazolium we prepared the tri(m-sul-
3.2. In situ investigation of the rhodium catalyst in ionic
liquids
Despite the fact that the mechanism for hydroformy-
lation catalysis is well established in organic [16] and
aqueous [17] solvents, we wanted to study the active rho-
dium species in an ionic liquid environment (Scheme 1).
To build upon a recent study focused on the in situ
investigation of the rhodium-sulfoxantphos hydro-
formylation catalyst [18] in ionic liquids, we set out to
investigate the rhodium-tppts catalyst system using
HP-NMR techniques. In these experiments a mixture
consisting of Rh(CO)2(acac), tppts (Rh/P ratio 1:6)
and [bmim][BF4] was pressurized with 60 psi of 13CO/
H2 (1:1) in a sapphire NMR tube. To ensure the forma-
tion of the catalyst species HRh(13CO)(tppts)3 (1a) the
solution was shaken at 60 ꢁC for 1 h. After the pressure
was released the 31P{1H} and 13C{1H} NMR spectra
were recorded (Fig. 1). The 31P{1H} NMR spectrum
of the catalyst solution showed a doublet at 32.2 ppm
for the rhodium bound tppts ligand and a minor reso-
nance at 29.1 ppm for the uncoordinated phosphine
oxide species tri(m-sulfonyl)triphenylphosphine oxide
trisodium (tppts-oxide) (Table 1). A strong signal at
ꢀ3.7 ppm was apparent for uncoordinated tppts. The
corresponding 13C{1H} NMR spectrum of the catalyst
solution showed a doublet at 195.2 ppm for the metal
coordinated carbon monoxide. For a better comparison
we also investigated HRh(13CO)(PPh3)3 1b in toluene-d8
under 13CO/H2 (1:1) pressure. Furthermore, we used the
fonyl)triphenyl
phosphine
1,2-dimethyl-3-butyl-
imidazolium salt (tppti*) which showed good solubility
characteristics in [bmim][PF6] and was used in all the
following catalyst evaluations.
1a-4a:P=tppts
1b-4b: P = PPh3
HRh(13CO)(P)3
HRh(13CO)2(P)2
1
2
- P
+ CO
- CO
- P
+ P
+ P
{HRh(13CO)(P)2}
{HRh(13CO)2(P)3}
3
4
´
Further Steps into the
Catalysis Cycle
HP-NMR study by Horvath et. al. [17] for hydroformy-
lation catalysis in aqueous phases as a reference. Thus,
similar to the results of our catalyst investigation in
[bmim][BF4], the 31P{1H} NMR spectrum of 1b in tolu-
Scheme 1. Initial steps of the hydroformylation mechanism.
31P{1H}-HP-NMR
13C{1H}-HP-NMR
tppts
13CO
2000 psi
2000 psi
HRh(13CO)2(tppts)2
HRh(13CO)2(tppts)2
tppts-oxide
tppts
15 psi
15 psi
HRh(13CO)(tppts)3
HRh(13CO)(tppts)3
tppts-oxide
50
40
30
20
10
0
PPM
210
205
200
195
190
185
PPM
Fig. 1. High-pressure NMR investigation of a rhodium hydroformylation catalyst in the ionic liquid [bmim][BF4] at 15 and 2000 psi 13CO/H2 (1:1).

2684
C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
ene-d8 showed a doublet at 37.8 ppm for the rhodium
bound PPh3 (Table 1). The corresponding 13C{1H}
NMR spectrum of the solution exhibited a doublet at
206.6 ppm for the coordinated carbon monoxide ligand.
On the basis of these spectroscopic results it can be as-
sumed that the structure of HRh(13CO)(tppts)3 in
sible to suppress most of the side reactions through the
use of high-purity ionic liquids, we did sometimes ob-
serve the formation of unknown species during the
investigation of 1a in ionic liquids.
3.3. Preparation of high-purity ionic liquids
[bmim][BF4]
is
very
similar
to
that
of
HRh(13CO)(PPh3)3 in toluene-d8.
In addition to being important for catalyst lifetime,
we also discovered that the purity of the ionic liquid is
critical for the stability of the ionic liquid phase itself.
In several biphasic hydroformylation experiments we
observed the partial decomposition of [bmim][BF4]
and [bmim][PF6]. After carefully analyzing the purity
of these ionic liquids we found that trace amounts of
acid were responsible for their decomposition. For this
reason, we have used a synthesis procedure which
produces high-purity ionic liquids [14]. In our synthesis
1-butyl-3-methyl-imidazolium chloride and 1,2-dimethyl-
3-butyl-imidazolium chloride were either treated with
NaBF4 or KPF6 giving [bmim][BF4], [bmim][PF6] and
[bdmim][PF6], respectively. The reactions were carried
out in water followed by extraction with dichlorometh-
ane. The work-up procedure involved the filtration of
the extract through a bed of activated carbon and alu-
mina which produced high-purity ionic liquids after
the removal of the solvent. The ionic liquids prepared
via this route showed a neutral pH value and did not
suffer from any substantial decomposition during hydro-
formylation reactions.
Following this we studied the rhodium catalyst in io-
nic liquids under elevated syngas pressure. Mechanistic
studies [16] of HRh(13CO)(PPh3)3 (1b) and
HRh(13CO)2(PPh3)2 (2b) in toluene have suggested the
presence of two coordinatively unsaturated catalyst
intermediates
{HRh(13CO)(PPh3)2}
(3b)
and
{HRh(13CO)2(PPh3)} (4b). While both of these species
are responsible for the production of aldehyde, it is be-
lieved that the normal to iso ratio (n/i) of the aldehyde is
largely controlled by the competitive reactions of the
olefin with the species 3b and 4b resulting in high or
low n/i ratios, respectively. Although we are not able
to record the NMR spectra of the reactive intermediates
3b and 4b, we can observe the resting state of the cata-
lyst in form of complexes 1b and 2b. As the equilibrium
between these two complexes is strongly pressure
dependent, complex 1b can readily be converted into
complex 2b at a 13CO/H2 (1:1) pressure of less than
300 psi in a toluene-d8 solution. In contrast, the investi-
gation of HRh(13CO)(tppts)3 1a in an aqueous solution
showed that applied pressures of 13CO/H2 (1:1) as high
as 3000 psi did not lead to the formation of
HRh(13CO)2(tppts)2 (2a) [17]. Based on this informa-
tion, we pressurised the complex 1a in [bmim][BF4] step-
wise with 13CO/H2 and recorded the corresponding
NMR spectra. The increasing pressure resulted in the
formation of new doublets in both the 31P{1H} NMR
spectrum (39.1 ppm) and the 13C{1H} NMR spectrum
(199.2 ppm). By going to higher syngas pressures, the
intensities of the NMR resonances for 1a continued to
decrease. Finally, at a 13CO/H2 pressure of 2000 psi
the NMR signals for 1a disappeared. We believe that
the newly formed species in [bmim][BF4] has the solu-
tion structure of HRh(13CO)2(tppts)2 (2a) and corre-
sponds to 2b which has been observed in toluene-d8
under elevated pressure. After the ionic liquid sample
was depressurized the NMR resonances of 1a were de-
tected again. It is interesting to note that the NMR sig-
nal of the high-pressure species 2a in [bmim][BF4]
exhibited a down field shift, while the related complex
2b in toluene-d8 showed an upfield shift for the corre-
sponding resonances. Although the ionic liquid and tol-
uene samples showed the formation of the high-pressure
rhodium species 2a and 2b, the chemical environment on
the metal center in [bmim][BF4] seems to be different to
that in the toluene-d8. Finally, it should be noted that
the in situ formation of 1a in [bmim][BF4] can be accom-
panied by a series of side reactions. Although it was pos-
3.4. Evaluation of biphasic hydroformylation catalysis
using ionic liquid phases
As a control experiment in the beginning of our study
we investigated the reactivity of Rh(CO)2(acac) in
[bmim][PF6] without a ligand for the hydroformylation
of hexene-1. The formed catalyst species produced C7-
aldehyde with a TOF between 21 and 2 minꢀ1 and
showed an n/i ratio for the aldehyde of 0.5–1.5 in the five
respective runs. The low n/i ratio between 0.5 and 1.0 for
the produced aldehyde is typically observed for rhodium
catalysts in the absence of any coordinating ligand [15].
Typically under these reaction conditions, as mentioned
earlier, rhodium forms an active rhodium-hydrido-car-
bonyl complex which readily converts olefin to alde-
hyde. The active species has no charge and dissolves
preferentially in the organic phase. As a result, a sub-
stantial amount of the rhodium metal leaches from the
ionic liquid into the organic layer. The loss of metal in
our control experiment was also confirmed by rhodium
metal analysis of the isolated product phases. After the
first catalyst run an amount of 119 ppm of rhodium me-
tal was found in the product phase. This amount corre-
sponded to almost 75% of the initial rhodium charge to
the reactor. Further runs showed continuous leaching of
the metal until >90% of the initial rhodium charge was

C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
2685
removed with the product phase. Corresponding to the
loss of active metal species, the conversion of olefin
was significantly reduced after every consecutive run.
It is therefore not viable to run hydroformylation reac-
tions in ionic liquids with rhodium catalysts which can-
not be sufficiently retained in the ionic layer.
for the loss in activity. The rhodium leaching from the
ionic liquid phase into the organic layer may be associ-
ated with the solubility characteristics of [bmim][BF4].
The ionic liquid has the ability to readily dissolve in
water and other polar solvents. Depending on the alde-
hyde content in the product phase, small amounts of io-
nic liquid and rhodium complex can become soluble and
leach into the organic layer. In one of our investigations
(Entry 3) 2.7 ppm rhodium metal was detected in the or-
ganic phase at an aldehyde content of 61%. However,
when the aldehyde level in the product phase was only
21% the amount of rhodium leaching was reduced to
only 0.3 ppm (Entry 10). This trend indicated that the
increasing polarity of the organic phase is partially
responsible for the metal loss from the ionic liquid
phase.
The next investigation focused on the evaluation of
[bmim][PF6]. All reactions were carried out at 600 psi
CO/H2 (1:1) and 100 ꢁC using Rh(CO)2(acac) and tppti*
(Rh/P ratio of 1:10) (Entries 11–20). The TOF values of
the catalyst system decreased over the course of the
investigation from 70 to 23 minꢀ1. Again the n/i ratios
of the produced aldehydes were observed from 2.6 to
1.8, supporting a phosphine-stabilized rhodium center
as the catalytically active species. Although the initial
rhodium retention was better than in [bmim][BF4], after
the 6th recycle run the metal leaching became more
prominent and surpassed the values obtained for the
[bmim][BF4] system.
Our next investigation focused on the evaluation of
Rh(CO)2(acac) with tppti* in [bmim][BF4] (Fig. 2/Table
2/Entries 1–10). The initial two runs (Entries 1 and 2)
were carried out at a CO/H2 (1:1) pressure of 300 psi.
After we observed an extended induction period of 450
minutes in the first recycle we increased the reactor pres-
sure to 600 psi CO/H2 (1:1) for all of the consecutive
runs. The activity of the catalyst decreased in the follow-
ing eight runs (Entries 3–10) with their TOF values
decreasing from 14 to 4 minꢀ1. The n/i ratio of the alde-
hyde remained between 2.6 and 2.1. This ratio is typi-
cally observed for a rhodium based hydroformylation
catalyst in the presence of a coordinating phosphine lig-
and (Rh/P ratio 1:10). The analysis of the resulting or-
ganic products revealed the presence of low amounts
of soluble rhodium species. During the 10 runs almost
10% of the total rhodium charged was removed from
the reactor with the product phase. Although the rho-
dium leaching contributed towards the drop in activity,
other catalyst deactivation processes like the formation
of rhodium carbonyl clusters are most likely the reason
To investigate the impact of a changing metal to
phosphine ratio we also investigated the [bmim][PF6] io-
nic liquid system with a Rh/P ratio of 1:3 and 1:100. As
expected the TOF for the reaction carried out with a
lower amount of phosphine ligand (1:3) was slightly in-
creased to 75 minꢀ1 (Entry 28), while the TOF for the
catalysis with the higher loading (1:100) was reduced
to 30 minꢀ1 (Entry 29). Under both sets of reaction con-
ditions the n/i ratios remained unchanged at 2.5 and 2.6.
However, the higher amount of phosphine ligand helped
to better retain the rhodium catalyst and reduced the
metal loss to levels below the detection limit.
TOF
(min-1
)
70
60
50
40
30
20
10
0
Entries
Entries
Entries
1-10
11-20
20-27
1
2
3
4
5
6
7
8
9
10
Rh in
Reactor
Recycle Runs
To broaden the scope of the studied ionic liquids we
also investigated [bdmim][PF6] (Entries 21–27).
Although the TOF values were slightly higher than
those obtained for [bmim][BF4], ranging from 37 to 3
minꢀ1, the rhodium leaching was significantly increased.
This may indicate that the less polar cation is partially
responsible for the higher rhodium loss. Furthermore,
the catalyst system showed a drastically reduced activity
after the 5th recycle run and we concluded the campaign
after the 7th run due to low activity.
In our final study we investigated the conventional
homogeneous rhodium catalyst under similar reaction
conditions. The reaction was carried out in toluene using
Rh(CO)2(acac) together with PPh3 (Rh/P ratio of 1:10).
The homogeneous catalyst in a monophasic system
(µmol)100
95
90
85
80
75
1
2
3
4
5
6
7
8
9
10
Recycle Runs
Fig. 2. TOF for biphasic hydroformylation reactions in different ionic
liquid phases and rhodium metal content in the reactor during recycle
runs.

Table 2
Evaluation of hydroformylation reactions in ionic liquid media
Entry
Cyclea
Ionic liquidb
Pc/Rh ratio
Rh reactor (lmol)
Rh product (ppm)
Timed (min)
Induct. (min)
TOFe (minꢀ1
)
n/i
Selectivity (%)
1 ꢀ C6
Ald.
Iso.f
¼
1
2
1g
2g
3
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][BF4]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bmim][PF6]
[bdmim][PF6]
[bdmim][PF6]
[bdmim][PF6]
[bdmim][PF6]
[bdmim][PF6]
[bdmim][PF6]
[bdmim][PF6]
[bmim][PF6]
[bmim][PF6]
Toluene
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
3
100.00
97.79
96.51
94.62
93.75
93.07
92.54
92.34
92.03
91.79
100.00
99.59
98.69
97.77
96.57
93.87
92.57
91.07
89.17
86.57
100.00
98.5
4.1
1.9
2.7
1.3
1.0
0.8
0.3
0.5
0.4
0.3
0.7
1.4
1.8
1.7
3.9
1.9
2.2
2.9
3.9
3.7
2.3
2.2
2.5
2.7
2.9
3.2
2.1
0.68
0.02
–
230
700
330
255
265
360
410
365
310
275
100
120
100
100
120
265
240
300
300
300
180
180
210
180
250
360
500
150
200
120
40
450
35
0
23
9
2.2
1.9
2.2
2.4
2.1
2.5
2.5
2.5
2.6
2.6
2.6
2.5
2.4
2.3
2.3
1.9
1.9
1.8
1.8
1.9
2.3
2.2
2.2
2.3
2.3
2.4
2.6
2.5
2.6
2.6
5
58
54
61
46
47
44
39
34
30
21
63
73
67
62
55
59
54
52
49
44
37
57
53
45
41
39
17
81
87
95
37
37
34
25
25
24
22
19
15
13
5
9
3
14
14
13
11
9
4
4
4
29
28
32
39
47
55
66
32
10
8
5
5
0
6
6
0
7
7
0
8
8
0
8
9
9
0
7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30i
a
10
1
0
4
12
26
13
19
18
25
28
41
50
75
20
20
50
35
50
105
250
60
13
0
70
68
68
61
52
44
35
32
26
23
37
37
33
28
20
12
3
2
17
25
33
37
40
45
47
50
54
59
38
40
42
50
47
34
17
4
3
4
5
5
8
6
1
7
1
8
1
9
1
10
1
2
4
2
5
3
97.1
95.4
8
4
13
9
5
93.6
91.7
6
14
49
2
7
89.6
1
100
100
75
30
400
1
100
10
9
1
25.0
2
3
All reactions were carried out in a 300 ml autoclave (2200 rpms) at 100 ꢁC and 600 psi (CO/H2 ratio 1:1).
Ionic liquid [bmim][BF4] 1-butyl-3-methyl-imidazolium tetrafluoroborate, [bmim][PF6] 1-butyl-3-methyl-imidazolium hexafluorophosphate and [bdmim][PF6] 1,2-dimethyl-3-butyl-imidazolium
b
hexafluorophosphate.
c
Phosphine ligand tri(m-sulfonyl)triphenylphosphine tris(1,2-dimethyl-3-butyl-imidazolium) salt (tppti*).
Full reaction time (including induction period).
d
e
TOF defined as mol(aldehyde) per mol(rhodium) per minute (determined during the initial 30 min of gas consumption).
¼
f
¼
¼
Isomerized 1 ꢀ C6 to 2 ꢀ C6 and 3 ꢀ C6 .
Reactor pressure at 300 psi.
Run ligand – triphenylphosphine.
g
i

C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
2687
exhibited a significantly higher TOF value of 400 minꢀ1
References
with an n/i ratio of 2.6 for the produced aldehyde. For
comparison the best TOF values obtained for any of
the biphasic hydroformylation runs in ionic liquids ran-
ged between 70 and 75 minꢀ1 under comparable reac-
tion conditions. Finally the homogeneous catalyst
system showed a significantly higher reactivity, however,
the biphasic ionic liquid catalyst systems offer the possi-
bility of convenient product separation and easy catalyst
recovery.
[1] C.D. Frohning, C.W. Kohlpainter, in: B. Cornils, W.A. Herr-
mann (Eds.), Applied Homogeneous Catalysis with Organome-
tallic Compounds, Wiley–VCH, Weinheim, 1996, p. 29.
[2] (a) B. Cornils, E.G. Kuntz, in: B. Cornils, W.A. Herrmann
(Eds.), Aqueous-Phase Organometallic Catalysis, Wiley–VCH,
Weinheim, 1998, p. 271;
(b) J.P. Arhancet, M.E. Davis, J.S. Merola, B.E. Hanson, Nature
339 (1989) 454.
[3] I.T. Horvath, J. Rabai, Science 266 (1994) 72.
[4] (a) P.G. Jessop, T. Ikariya, R. Noyori, Chem. Rev. 99 (1999)
475;
(b) P.G. Jessop, W. Leitner, in: P.G. Jessop, W. Leitner (Eds.),
Leitner, Chemical Synthesis using Supercritical Fluids, Wiley–
VCH, Weinheim, 1999, p. 351.
4. Conclusion
[5] (a) F.R. Hartley, Supported Metal Complexes. A New Generation
of Catalysts, Kluwer Academic Publishers, Dordrecht, 1985;
(b) M.T. Reetz, G. Lohmer, R. Schwickardi, Angew. Chem., Int.
Ed. Engl. 36 (1997) 1526;
The ionic liquids [bmim][BF4], [bmim][PF6] and
[bdmim][PF6] were successfully used for the immobiliza-
tion of rhodium based hydroformylation catalysts and
have been evaluated for the production of aldehyde in
a liquid–liquid biphase environment. The purity of the
ionic liquids was strictly monitored to avoid any decom-
position of the catalyst, substrate and/or ionic liquid. A
complimentary HP-NMR study confirmed that the solu-
tion structure of HRh(13CO)(tppts)3 (1a) in [bmim][BF4]
is similar to that of HRh(13CO)(PPh3)3 (1b) in toluene-
d8. Furthermore, the NMR investigation at elevated
syngas pressures showed that 1a lies in equilibrium with
the high-pressure species HRh(13CO)2(tppts)2 (2a). As
expected the activity of the biphasic ionic liquid system
containing the Rh/tppti* catalyst is approximately one
order of magnitude lower than the conventional Rh/
PPh3 catalyst in toluene. The Rh/tppti* catalyst system
showed a decrease of activity in the semicontinuous hyd-
roformylation of hexene-1 due to catalyst deactivation
and/or metal loss. The n/i ratio of the produced heptanal
only slightly increased after changing the Rh/P ratio
from 1:3 to 1:100 in [bmim][PF6]. The loss of rhodium
metal from the ionic liquid phases seemed to be depend-
ent on the solubility characteristics of the ionic liquid
and on the concentration of the aldehyde in the organic
phase. In this respect, the ionic liquid [bmim][PF6] was
preferred due to its very low miscibility with polar sub-
stances. Although ionic liquids offer tremendous oppor-
tunities for the development of novel separation
technologies, further progress will be necessary to fully
establish their potential as a large-volume reaction
medium.
(c) W. Keim, B. Driessen-Ho¨lscher, in: G. Ertl, H. Kno¨zinger, J
Weitkamp (Eds.), Handbook of Heterogeneous Catalysis, vol. 1,
Wiley–VCH, Weinheim, 1997.
[6] (a) T. Welton, Chem. Rev. 99 (1999) 2071;
(b) J.D. Holbrey, K.R. Seddon, Clean Prod. Process. 1 (1999)
223;
(c) P. Wasserscheid, W. Keim, Angew. Chem., Int. Ed. 39 (2000)
3772;
(d) R. Sheldon, Chem. Commun. (2001) 2399;
(e) J. Dupont, R.F. de Souza, P.A.Z. Suarez, Chem. Rev. 102
(2002) 3667;
(f) H. Olivier-Bourbigou, L. Magna, J. Mol. Catal. A 182 (2002)
419;
(g) P. Wasserscheid, in: P. Wasserscheid, T. Welton (Eds.), Ionic
Liquids in Synthesis, Wiley–VCH, Weinheim, 2003, p. 213.
[7] G.W. Parshall, J. Am. Chem. Soc. 94 (1972) 8716.
[8] Y. Chauvin, L. Mussmann, H. Olivier, Angew. Chem., Int. Ed. 34
(1995) 2698.
[9] E.G. Kuntz, Chemtech (1987) 570.
[10] (a) J.F. Knifton, J. Mol. Catal. 43 (1987) 65;
(b) J.F. Knifton, J. Mol. Catal. 47 (1988) 99;
(c) N. Karodia, S. Guise, C. Newlands, J.-A. Andersen, Chem.
Commun. (1998) 2341;
(d) W. Keim, D. Vogt, H. Waffenschmidt, P. Wasserscheid, J.
Catal. 186 (1999) 481;
(e) P. Wasserscheid, H. Waffenschmidt, J. Mol. Catal. A 164
(2000) 61;
(f) C.C. Brasse, U. Englert, A. Salzer, H. Waffenschmidt, P.
Wasserscheid, Organometallics 19 (2000) 3818;
(g) P. Wasserscheid, H. Waffenschmidt, P. Machnitzki, K.W.
Kottsieper, O. Stelzer, Chem. Commun. (2001) 451;
(h) M.F. Sellin, P.B. Webb, D.J. Cole-Hamilton, Chem. Commun.
(2001) 781;
(i) F. Favre, H. Olivier-Bourbigou, D. Commereuc, L. Saussine,
Chem. Commun. (2001) 1360;
(j) J. Dupont, S.M. Silva, R.F. de Souza, Catal. Lett. 77 (2001)
131;
(k) R.P.J. Bronger, S.M. Silva, P.C.J. Kamer, P.W.N.M. van
Leeuwen, Chem. Commun. (2002) 3044;
Acknowledgements
(l) O. Stenzel, H.G. Raubenheimer, C. Esterhuysen, J. Chem.
Soc., Dalton Trans. (2002) 1132.
[11] S.N. Bizzari, S. Fenelon, M. Ishikawa-Yamaki, Chemical Eco-
nomics Handbook, SRI International, Menlo Park, 1999,
682.7000A.
We thank Dr. H. Freund for carrying out a thermal
stability study and Drs. C.A. Costello, J.L. Robbins
and R.H. Schlosberg for valuable discussions and their
continued support.
[12] C.P. Mehnert, R.A. Cook, N.C. Dispenziere, M. Afeworki, J.
Am. Chem. Soc. 124 (2002) 12932.

2688
C.P. Mehnert et al. / Polyhedron 23 (2004) 2679–2688
´
[13] I.T. Horvath, J.M. Millar, Chem. Rev. 91 (1991) 1339.
ˆ
[14] (a) P. Bonhote, A.P. Dias, N. Papageorgiou, K. Kalyanasunda-
(i) A.A. Oswald, J.S. Merola, E.J. Mozeleski, R.J. Kastrup, R.V.
Reisch, in: L.D. Quin, J.G. Verkade (Eds.), Advances in Chemistry
Series, vol. 171, American Chemical Society, Washington (DC),
1981, p. 503;
ram, M. Gra¨tzeb, Inorg. Chem. 25 (1996) 1168;
(b) J.D. Holbrey, K.R. Seddon, J. Chem. Soc., Dalton Trans.
(1999) 2133;
(j) R.J. Kastrup, J.S. Merola, A.A. Oswald, in: E.C. Alyca, D.W.
Mcck (Eds.), Advances in Chemistry Series, vol. 196, American
Chemical Society, Washington (DC), 1982, p. 43;
(k) A.A. Oswald, D.E. Hendriksen, R.J. Kastrup, J.S.
Merola, Prepr. Pap., Am. Chem. Soc., Div. Pet. Chem. 27
(1982) 292;
(c) J.H. Davis, C.M. Gordon, C. Hilger, P. Wasserscheid, T.
Welton (Eds.), Wiley–VCH, Weinheim, 2003, p. 7.
[15] B. Cornils, in: J. Falbe (Ed.), New Syntheses with Carbon
Monoxide, Springer, Berlin, 1980, p. 1.
[16] (a) C. OÕConnor, G. Yagupsky, D. Evans, G. Wilkinson, J.
Chem. Soc., Chem. Commun. (1968) 420;
(l) J.M. Brown, L.R. Canning, A.G. Kent, P.J. Sidebottom,
J. Chem. Soc., Chem. Commun. (1982) 721;
(m) J.M. Brown, A.G. Kent, J. Chem. Soc., Chem. Commun.
(1982) 723;
(b) D. Evans, G. Yagupsky, G. Wilkinson, J. Chem. Soc. (A)
(1968) 2660;
(c) D. Evans, J.A. Osborn, G. Wilkinson, J. Chem. Soc. (A)
(1968) 3133;
(n) J.M. Brown, A.G. Kent, J. Chem. Soc., Perkin Trans. 2
(1987) 1597;
(d) C.K. Brown, G.J. Wilkinson, J. Chem. Soc. (A) (1970) 2753;
(e) R.L. Pruett, J.A. Smith, J. Org. Chem. 34 (1969) 327;
(f) R.L. Pruett, Adv. Organomet. Chem. 17 (1979) 1;
(g) G. Gregoria, G. Montrasi, M. Tampierei, P. Cavaliezi dÕOro,
G. Pagani, A. Andreetta, Chim. Ind. (Milan) 62 (1980) 389;
(h) P. Cavaliezi dÕOro, L. Raimondi, G. Pagani, G. Montrasi,
G. Gregorio, A. Andreetta, Chim. Ind. (Milan) 62 (1980) 572;
(o) W.R. Moser, C.J. Papile, D.A. Brannon, R.A. Duwell,
S.J. Weininger, J. Mol. Catal. 41 (1987) 271.
[17] I.T. Horva´th, R.V. Kastrup, A.A. Oswald, E.J. Mozeleski, Catal.
Lett. 2 (1989) 85.
[18] S.M. Silva, R.P.J. Bronger, Z. Freixa, J. Dupont, P.W.N.M. van
Leeuwen, New J. Chem. 27 (2003) 1294.