In situ formation of Wilkinson-type hydroformylation catalysts: Insights into the structure, stability, and kinetics of triphenylphosphine- and xantphos-modified Rh/SiO2

Article abstract of DOI:10.1021/cs3007445

An investigation has been carried out to identify the effects of catalyst preparation on the activity, selectivity, and stability of phosphine-modified rhodium/silica catalysts (Rh/SiO₂) for propene hydroformylation. High selectivity to aldehydes was achieved, without the formation of propane or butanol. Catalyst activity and selectivity was found to depend strongly on the nature and concentration of the phosphine ligands and the amount of rhodium dispersed on the silica support. Screening of different ligands showed that a bidentate xantphos (X) ligand was ～2-fold more active than the monodentate phosphine ligand (PPh₃) screened at a ligand-to-rhodium ratio of 15:1. Investigation of the effects of reaction temperature, reactant partial pressures, and phosphine-to-rhodium ratio indicates that the kinetics of propene hydroformylation over X-promoted Rh/SiO₂ is nearly identical to those for sulfoxantphos-modified rhodium-containing supported ionic liquid phase (SX-Rh SILP) catalysts. In-situ FTIR and solid-state ³¹P MAS NMR characterization provide evidence for the formation of HRh(CO) _n(PPh₃)_4-n species on PPh₃-modified Rh/SiO₂, and HRh(CO)₂(X) species on xantphos-modified Rh/SiO₂. The high catalytic activity observed over rhodium-containing silica catalysts is attributed to formation of Rh^(I)(CO)₂ species by the process of corrosive chemisorption of Rh nanoparticles by CO and the subsequent ligation of phosphine ligands to the dicarbonyl species. Evidence is also presented suggesting that the active form of the catalyst resides on the surface of the Rh nanoparticles.

Full text of DOI:10.1021/cs3007445

Research Article
pubs.acs.org/acscatalysis
In Situ Formation of Wilkinson-Type Hydroformylation Catalysts:
Insights into the Structure, Stability, and Kinetics of
Triphenylphosphine- and Xantphos-Modified Rh/SiO2
Sankaranarayanapillai Shylesh, David Hanna, Anton Mlinar, Xu
and Alexis T. Bell*
̈
́ ̌
e-Qian Kong, Jeffrey A. Reimer,
̅
Department of Chemical and Bimolecular Engineering University of California, Berkeley, Berkeley, California 94720, United States
S Supporting Information
*
ABSTRACT: An investigation has been carried out to identify the effects of catalyst
preparation on the activity, selectivity, and stability of phosphine-modified rhodium/silica
catalysts (Rh/SiO ) for propene hydroformylation. High selectivity to aldehydes was
2
achieved, without the formation of propane or butanol. Catalyst activity and selectivity
was found to depend strongly on the nature and concentration of the phosphine ligands
and the amount of rhodium dispersed on the silica support. Screening of different ligands
showed that a bidentate xantphos (X) ligand was ∼2-fold more active than the
monodentate phosphine ligand (PPh ) screened at a ligand-to-rhodium ratio of 15:1.
3
Investigation of the effects of reaction temperature, reactant partial pressures, and phosphine-to-rhodium ratio indicates that the
kinetics of propene hydroformylation over X-promoted Rh/SiO is nearly identical to those for sulfoxantphos-modified rhodium-
2
31
containing supported ionic liquid phase (SX-Rh SILP) catalysts. In-situ FTIR and solid-state P MAS NMR characterization
provide evidence for the formation of HRh(CO) (PPh ) species on PPh -modified Rh/SiO , and HRh(CO) (X) species on
n
3 4−n
3
2
2
xantphos-modified Rh/SiO . The high catalytic activity observed over rhodium-containing silica catalysts is attributed to
2
(I)
formation of Rh (CO) species by the process of corrosive chemisorption of Rh nanoparticles by CO and the subsequent
2
ligation of phosphine ligands to the dicarbonyl species. Evidence is also presented suggesting that the active form of the catalyst
resides on the surface of the Rh nanoparticles.
KEYWORDS: rhodium, xantphos, phosphine ligands, butanal, hydroformylation
1
. INTRODUCTION
near-zero vapor pressure of ILs, which virtually eliminates the
loss of solvent under reaction conditions. Such SILP catalysts
have been demonstrated to be highly effective for propene
hydroformylation. The factors influencing the activity and
stability of Rh-based SILP catalysts for the gas-phase hydro-
The hydroformylation of alkenes to produce aldehydes is
carried out on a large scale using an aqueous solution of a Rh−
tppts complex (tppts = tri(m-sulfonyl)triphenylphosphine
1
−3
trisodium salt).
Although this process is highly effective, it
12
formylation of propene have been discussed by Shylesh et al.
is limited to alkenes with fewer than four carbon atoms because
of the low solubility of higher-molecular-weight alkenes. The
need to separate the reaction products from the catalyst
solution presents an additional challenge for carrying out alkene
Evidence from in situ FTIR and 31P MAS NMR suggests that
interactions of the IL and metal complex with the support are
required to yield a stable catalyst. Catalyst activity is found to
be a strong function of IL composition, as well as ligand-to-Rh
ratio, IL loading, and temperature of support pretreatment.
Ding and co-workers have reported a different approach for
producing an active catalyst for propene hydroformylation. In
this case, silica-supported Rh nanoparticles are promoted with
4
hydroformylation using homogeneous complexes. Heteroge-
nization of homogeneous catalysts on solid supports can easily
mitigate these problems and enable alkene hydroformylation to
be carried out in the gas phase or in the liquid phase without
need for a solvent. Several attempts have been made to
achieve this goal. Davis and co-workers have shown that a
supported aqueous phase catalyst can be prepared by dispersing
tppts to Rh(acac)(CO) dissolved in a thin aqueous layer
within the pores of a high-surface-area silica. Exposure of this
impregnated solid to synthesis gas (H /CO = 1) at atmospheric
pressure and room temperature produced an active catalyst for
propene hydroformylation. Wasserscheid and co-workers have
shown that supported ionic liquid phase (SILP) catalysts can be
prepared by dispersing a solution of the catalyst complex in an
ionic liquid (IL) as a thin film on the internal surface of a
5
,6
1
3,14
PPh3,
and Kim et al. have recently shown that the highest
activity for PPh -modified Rh/SiO is achieved for PPh /Rh =
3
2
3
15
1
5. Because of the similarities in activation energies, partial
2
7
and total pressure dependences of the reaction rates for PPh3-
m
o
d
i
fi
e
d
R
h
/
S
i
O
a
n
d
h
o
m
o
g
e
n
e
o
u
s
H
R
h
(
C
O
)
(
P
P
h
)
2
3
3
2
catalysts suggest that the active species in both systems are
identical.
Received: November 16, 2012
Revised: January 14, 2013
Published: January 15, 2013
8
−11
porous solid.
The advantage of using ILs over water is the
©
2013 American Chemical Society
348
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
Here, we report the characterization and catalytic activity of a
monodentate and bidentate phosphine-modified rhodium/silica
Infrared spectra were acquired using a Thermo Scientific
Nicolet 6700 FTIR spectrometer equipped with a liquid-
nitrogen-cooled MCT detector. Each spectrum was obtained by
averaging 32 scans taken with 1 cm− resolution. For IR studies,
0.05 g of the catalyst was pressed into a 20-mm-diameter pellet
(<1 mm thick) and placed into a custom-built transmission cell
(
L-Rh/SiO L = triphenylphosphine (PPh ) and xantphos (X))
2
,
3
1
catalysts for the gas-phase hydroformylation of propene under
mild reaction conditions. The objective of this work was to
explore how the catalytically active species are formed under
reaction conditions and the effects of how the ligand
composition, the ligand/rhodium ratio, and rhodium loading
affect the catalytic activity, selectivity, and stability of the L-Rh/
equipped with CaF windows, a K-type thermocouple for
2
temperature measurement, and resistive cartridge heaters. All
scans were acquired at 393 K. Experiments at elevated pressure
were carried out by throttling a needle valve located
downstream from the reactor. Pure CO was passed through a
trap packed with 3.2 mm pellets of 3 Å molecular sieve to
remove iron pentacarbonyl formed within the cylinder.
SiO catalyst relative to the silica-supported Wilkinson-type
2
catalyst (HRh(CO)(PPh ) /SiO ). The kinetics of propene
3
3
2
hydroformylation on bidentate xantphos-modified Rh/SiO2
catalyst (X-Rh/SiO ) and PPh -modified Rh/SiO catalyst
2
3
2
31
(
PPh -Rh/SiO ) are also discussed. In situ FTIR character-
Solid-state P MAS NMR experiments were performed on a
3
2
Bruker Avance I-500 MHz spectrometer using a frequency of
ization of the phosphine-modified Rh/SiO2 catalyst was
performed to obtain evidence for the formation of active
species under the reaction conditions. These observations were
2
8
02.5 MHz, 90° pulse in 4.2 μs, and a delay of 60 s relative to
5% H PO . High-resolution transmission electron microscopy
3
4
31
(
HR-TEM) characterizations were conducted on the TEAM
complemented by ex-situ characterization of the catalysts by P
MAS NMR and HR-TEM. Results of the present study provide
additional experimental evidence for in situ formation of
homogeneous catalysts and the interactions of the homoge-
neous catalyst with the rhodium nanoparticles dispersed on the
support surfaces.
.5 high-resolution electron microscope, a modified FEI Titan
0-300 TEM equipped with a gun monochromator, operated at
0 KV at the National Center for Electron Microscopy
operated at 200 KV and incorporated with a gun mono-
chromator.
2
. EXPERIMENTAL SECTION
The preparation of PPh -modified Rh/SiO was as follows: 0.2
3. RESULTS
3
2
wt % Rh/SiO was prepared by incipient-wetness impregnation
2
3.1. Catalytic Activity of Phosphine-Modified Rh/SiO2
Catalysts. 3.1.1. Effect of Ligand Composition. It is known
that the activity, selectivity, and stability of homogeneous
rhodium−phosphine complexes used for hydroformylation
reactions are sensitive to the structure and concentration of
2
−1
of mesoporous SiO (Silicycle, 500 m g , average pore
2
diameter 60 Å) with a solution of rhodium(III) acetyl acetonate
(
Aldrich, 97% pure) dissolved in toluene (Alfa Aesar,
anhydrous 99.8% pure). After impregnation, the solid was
3
−1
1
dried at 393 K for 12 h, then calcined in 100 cm min 10%
O /He (Praxair, certified standard) for 4 h at 673 K (2 K
the ligand. The role of various phosphine ligands was therefore
2
investigated at a fixed ligand-to-rhodium molar ratio of 15 (L/
Rh = 15) and a rhodium loading of 0.2 wt %. The ligands
screened include triphenylphosphine (PPh ), triphenyl phos-
phite (P(OPh) ), fluorinated triphenylphosphine (3F-PPh ),
methoxytriphenylphosphine (OMe-PPh ), and bidentate xant-
phos (X). Table 1 shows that ligand composition had a strong
−1
3
−1
min ) and, finally, reduced in 100 cm min 9% H /He
2
(
Praxair, certified standard) at 673 K for 4 h. Using incipient-
3
wetness impregnation, a solution of PPh (Aldrich, 99% purity)
3
3
3
in toluene was absorbed onto the SiO -supported Rh under
2
3
nitrogen atmosphere. After the impregnation, the solid was
dried for 2 h at ambient temperature in nitrogen and then dried
in a vacuum oven at 353 K for 12 h. The PPh /toluene solution
concentration was varied to obtain final molar ratios of PPh3/
Rh between 5 and 20, and the rhodium loading was varied from
3
Table 1. The Effects of Phosphine Composition on the Rate
of Propene Hydroformylation and the n/iso Ratio of
a
Butanals Formed on Phophine-Modified Rh/SiO2
0.05 to 1 wt %. A similar incipient-wetness impregnation
b
rate (h−1
c
d
catalyst
L/Rh ratio
)
n/iso
procedure using toluene as a solvent was used for the
X-Rh/SiO2
PPh -Rh/SiO2
15
15
15
15
15
15
78
35
31
15
24
83
13
9
preparation of xantphos-modified Rh/SiO catalysts (X-Rh/
2
SiO2).
3
3
F-PPh -Rh/SiO2
12
8
Gas-phase hydroformylation of propene was performed in a
3
3
OMe-PPh -Rh/SiO2
6
.35 mm o.d. (∼4 mm ID) quartz tube containing an expanded
3
section (∼12.7 mm o.d., ∼20 mm length). The reactor was
packed with quartz wool above and below the catalyst bed to
hold the catalyst in place. The feed to the reactor consisted of
propene (Praxair, 3.0 grade), CO (Praxair, 4.0 research grade),
and H (Praxair, 5.0 UHP grade). A reactant ratio of C H /
3
P(OPh) -Rh/SiO2
SX-Rh/SILP
13
a
Reaction conditions: C
H /H /CO = 1:1:1, Ptotal = 2 atm, T = 393 K,
6 2
b
3
t = 8 h, 0.2 wt % Rh metal loading. Molar ligand-to-metal ratio.
c
Turnover frequency in moles of aldehyde per mole of Rh per hour.
2
3
6
d
Linear-to-branched ratio.
CO/H2 of 1:1:1 was used unless specified otherwise.
Experiments were carried out at 393 K, total gas pressures of
3
−1
2
atm, and total gas flow rate of 120 cm min at STP to
effect on the catalyst activity and n/iso ratio. Under identical
3
−1
maintain a constant volumetric flow rate of 60 cm min at
pressure. Under these conditions, the conversion of propene
was always <1%. Reaction products were analyzed using an
Agilent 6890N gas chromatograph containing a bonded and
cross-linked (5% phenyl)-methylpolysiloxane capillary column
reaction conditions, Rh/SiO modified by bidentate xantphos
2
was ∼2-fold more active than the Rh/SiO2 modified by
monodentate phosphine ligands. The catalytic activity and n/
iso ratio observed for X-Rh/SiO catalyst were similar to that
2
observed over sulfoxantphos-containing Rh-SILP (supported
12
(
Agilent, HP-1) connected to a flame ionization detector.
ionic liquid phase) catalysts (see Supporting Information).
3
49
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
Figure 1. The effect of (A) PPh /Rh and (B) X/Rh ratios on the rates of butanal synthesis. P = 2 atm; PC3H6/PCO/P = 1:1:1; T = 393 K; L/Rh
3
total
H
2
3
−1
3
−1
molar ratio = 5−20; Rh = 0.2 wt %; catalyst mass = 0.3 g; total gas flow rate = 60 cm min at pressure, 120 cm min at STP.
Figure 2. The effect of rhodium loading on the rates of butanal synthesis: (A) PPh ligand and (B) Xantphos ligand. P = 2 atm; PC3H6/PCO/PH2
=
3
total
:1:1; T = 393 K; L/Rh molar ratio = 15; total gas flow rate = 60 cm min−1 at pressure, 120 cm min at STP.
3
3
−1
1
The higher activity of Rh complexes involving xantphos over
toward the formation of aldehydes. The results in Figure 1
show that the hydroformylation rate increased rapidly with
increasing L/Rh ratio up to a molar ratio of 15, and above this
value, the activity toward both products declined. The observed
monodentate phosphines can be attributed to the chelating
ability of xantphos, which yields a higher concentration of the
1
coordinated rhodium species. Quite interestingly, an n/iso
ratio of 3 was observed over the phosphite-modified Rh/SiO2
catalysts. Use of derivatives of triphenylphosphine with
patterns were qualitatively similar for PPh and X, but the
3
catalytic activity was roughly 2-fold higher for X. Irrespective of
the L/Rh composition, bidentate xantphos-modified Rh/SiO2
exhibited an almost constant n/iso ratio of 12. It is important to
note that an optimal X/Rh ratio has also been reported in
substituents typically in the para positions can obviate the
problems of simultaneously changing the electronic and steric
16
features of phosphine ligands. For instance, electron-with-
drawing groups in the para position decrease the basicity of the
triphenylphosphine groups, whereas electron-donating groups
decrease the electrophilicity over the rhodium center without
altering the steric confinement around the rhodium site.
Accordingly, it should be noted that the n/iso ratio of the
aldehyde increased as the electron density around the rhodium
was decreased by using electron-withdrawing substituents in
the modifying phosphine ligand (-F vs -OMe, Table 1). In
other words, less basic triaryl phosphines enhance the
formation of linear products, whereas the more basic triaryl
phosphine groups lead to lower n/iso ratios.
studies of HRh(CO) (X) complexes dissolved in various
2
17,18
solvents.
By contrast, for PPh -modified Rh/SiO , the
3 2
rate of linear product formation grew more rapidly than the rate
of branched product formation up to a PPh /Rh ratio of 20,
3
increasing the n/iso ratio from ∼3 to 13 (Figure 1A). The high
n/iso ratio observed in the present study for PPh /Rh ≥ 15 is
3
consistent with studies conducted with HRh(CO)(PPh ) in
3
3
15
homogeneous solutions.
.1.3. Effect of Rhodium Loading. The role of rhodium
3
loading on the phosphine-modified Rh/SiO2 catalyst was
investigated by varying the rhodium concentration on the
silica support from 0.05 to 1 wt % at a fixed L/Rh molar ratio of
3.1.2. Effect of L/Rh Molar Ratio. To investigate the role of
phosphine ligands on the catalytic performance of Rh/SiO2
1
5. Figure 2 shows that, irrespective of ligand composition, the
catalysts, the L/Rh molar ratio (L = PPh and X) was varied
3
rate per rhodium atom increased almost 4-fold as the loading of
rhodium decreased from 1 to 0.05 wt %. It was also found that
the n/iso ratio decreased from ∼9 to 4 with a decrease in the
from 5 to 20 at a fixed rhodium loading of 0.2 wt %. In the
absence of ligands, Rh/SiO is inactive for hydroformylation,
2
clearly showing the importance of phosphine ligands for
promoting hydroformylation. Neither propane nor butanol was
observed in the reaction products, indicating that the
phosphine-modified Rh/SiO2 catalysts are highly selective
rhodium weight loading with the PPh -modified Rh/SiO
3 2
catalysts, whereas the n/iso ratio remained constant at ∼13
for the bidentate xantphos-modified Rh/SiO catalysts.
2
3
50
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
Figure 3. Effect of temperature on the rate of butanal synthesis and regioselectivity over (A) PPh -Rh/SiO and (B) X-Rh/SiO catalysts. P = 2
3
2
2
total
3
−1
3
atm; P H6/PCO:PH2 = 1:1:1; catalyst mass = 0.3 g; X/Rh molar ratio = 15; Rh = 0.2 wt %; total gas flow rate = 60 cm min at pressure, 120 cm
C
3
min− at STP (black line represents n-butanal and red line represents i-butanal).
1
3
.2. Kinetics of Propene Hydroformylation Using L-
Table 2. Comparison of Activation Energies for the
Formation of n- and Isobutanal over Different Phosphine-
Containing Rhodium Catalysts
Rh/SiO Catalysts. Time-on-stream studies demonstrated that
2
the L-Rh/SiO2 catalysts were stable for 30 h, and no
deactivation was observed under the reaction conditions (see
Supporting Information). There is no loss in the rhodium
reaction
phase
EA
pressure
(atm)
catalyst
(kJ/mol) temp (K)
ref
this
content on phosphine-modified Rh/SiO catalysts as analyzed
a
2
X-Rh/SiO2
gas
gas
gas
65.0
353−373
353−373
353−373
2
2
2
from the ICP-OES analysis. The only products observed under
any of the conditions investigated were n- and isobutanal.
Figure 3 shows the effects of reaction temperature on the
catalytic activity and the n/iso ratio of the phosphine-modified
Rh/SiO catalyst (PPh -Rh/SiO ) and xantphos-modified Rh/
work
b
a
7
5.0
PPh -Rh/
65.0
15
3
SiO2
b
a
b
5
7.0
60.0
2.0
2
3
2
SX-Rh-SILP
19
SiO catalyst (X-Rh/SiO ) for a L/Rh molar ratio of 15. For
2
2
7
PPh -Rh/SiO2 catalysts, the rate of n-butanal formation
3
a
b
increased with temperature up to 393 K, after which the rate
decreased. The ratio of n- to isobutanal reached a maximum
value of 24 at 363 K and then decreased sharply to 3 at 413 K.
The decrease in the rate of n-butanal formation above 393 K is
n-Butanal. Isobutanal.
Table 3. Apprent Reaction Order in Propene, CO, and H
2
a
over Phosphine-Modified Rh/SiO Catalysts
2
attributable to decomposition of PPh , which is confirmed by
3
reactant
TGA analysis (see Supporting Information). For X-Rh/SiO2
catalysts, the rates of n- and isobutanal formation increased as
the reaction temperature increased. At 403 K, the rate of n-
butanal formation begins to plateau while the rate of isobutanal
formation continues to rise. Thus, the n/iso ratio decreases
from 15 at 353 K to 8 at 413 K, demonstrating that temperature
can be used to control the n/iso ratio. Above 413 K, a loss in
catalytic activity due to ligand decomposition was noted, as
identified by thermal gravimetric analysis.
catalyst
X-Rh/SiO
C H
CO
H2
3
6
b
0.86
0.76
1.01
0.99
0.93
0.92
−0.47
−0.40
1.12
0.12
0.12
1.02
0.85
0.10
0.00
2
c
b
c
PPh
-Rh/SiO
2
3
0.90
b
c
SX-Rh- SILP
−0.43
−0.61
a
b
c
Error bars were ∼±3 for all the numbers. n-Butanal. Isobutanal.
The apparent activation energies calculated from the
Arrhenius plot for PPh -Rh/SiO in the temperature interval
3
2
containing Rh-SILP catalyst. As shown in Table 3, for X-Rh/
−
1
3
53−413 K are 65 and 56 kJ mol , respectively (Table 2). For
X-Rh/SiO , the apparent activation energy measured over the
interval 353−413 K for n-butanal formation is 65 kJ mol ,
whereas the apparent activation energy for isobutanal formation
is 75 kJ mol . These values are ∼5 kJ mol higher than those
previously reported for a sulfoxantphos ligand containing Rh-
SiO , the partial pressure dependence of H on the rate of n-
2
2
2
butanal formation is 0.12, the partial pressure dependence on
CO is −0.47, and the partial pressure dependence on C H is
0.86. The orders of reaction with respect to the partial pressures
−1
3
6
−1
−1
of H , CO, and C H for n- and isobutanal formation are
2
3
6
similar. The partial pressure dependences observed for X-Rh/
1
9
SILP catalysts.
SiO are similar to those observed for the sulfoxantphos-
2
The dependences of the rate of n- and isobutanal formation
containing Rh-SILP catalysts (Table 3). By contrast, a first-
on the partial pressures of propene, CO, and H were measured
order dependence in C H , CO, and H was noted with the
2
3
6
2
15
by holding the partial pressures of two components at 0.67 atm
while varying the partial pressures of the third reactant and He.
The total gas flow rate and pressure were kept constant at 60
phosphine-modified PPh -Rh/SiO catalysts.
3 2
3.3. Characterization of Phosphine-Modified Rh/SiO2
Catalysts. 3.3.1. In Situ FTIR. In situ FTIR was used to identify
how the phosphine-modified Rh/SiO (L-Rh/SiO , L = PPh
3
−1
3
−1
cm min and 2 atm (120 cm min at STP). Table 3 shows
2
2
3
the effect of varying the partial pressures of C H , CO, and H
and X) changed upon exposure to CO and then CO/H . At
3
6
2
2
partial pressures on the catalytic activity at 363 K of phosphine
very low CO partial pressure (0.2 atm), the IR spectra of PPh3-
ligand-modified Rh/SiO2 catalysts and a sulfoxantphos-
modified 1%Rh/SiO (PPh /Rh = 15) showed a sharp band at
2
3
3
51
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
Figure 4. In situ FTIR of (A) PPh -Rh/SiO and (B) X-Rh/SiO catalysts in the presence of CO and synthesis gas showing the formation of
3
2
2
homogeneous catalysts. L/Rh = 15, Rh = 0.2 wt %, T = 393 K.
Figure 5. Solid-state 31P MAS NMR spectra of (A) PPh -modified Rh/SiO (PPh /Rh = 15) and (B) X-modified Rh/SiO (X/Rh = 15) spent
3
2
3
2
catalyst for different Rh weight loadings: Rh = (a) 0.05, (b) 0.2, (c) 0.5, and (d) 1 wt %.
970 cm− as well as narrow bands at 2000 and 1940 cm−1
1
support) on the silica support. This leads to the formation of
PPh -rich HRh(CO)(PPh ) homogeneous species in 1 wt %
1
(
Supporting Information). As the CO partial pressure was
3
3 3
increased to 1 atm, the intensity of these bands increased
further, with nearly constant relative intensities. The broad peak
at 1970 cm− arises from linear Rh(CO), whereas the bands at
Rh samples; the 0.2% Rh sample leads to ligand-lean
HRh(CO) (PPh ) type species. This leads to the difference
3
3
1
in the IR spectra of both samples. As seen in Figure 4A,
contacting the catalyst with CO in He (1 atm) produced sharp
−1
2
000 and 1940 cm are due to Rh(CO) species red-shifted as
2
20
−1
a result of the presence of adsorbed PPh3. Coordination of
CO bands at 2090, 2030, 2000, and 1940 cm . As noted above,
the bands at 2000 and 940 cm− are due to Rh(CO) species
1
electron-donating PPh ligand enhances the electron density
3
2
around the rhodium center, resulting in substantial back-
donation to CO ligands, thereby lowering the CO vibration
red-shifted due to the presence of adsorbed PPh , whereas the
3
intense bands at 2090 and 2030 cm− arise from the geminal-
1
frequency. Addition of H to CO in the gas stream resulted in
the appearance of bands at 2045, 2010, 1990, and 1950 cm .
The positions of these bands are identical to those observed for
dicarbonyl rhodium species. Addition of H to CO in the gas
stream produced peaks at 2045, 2010, 1990, and 1940 cm
2
2
−1
−1
associated with HRh(CO) (PPh ) species. Remarkably, there
2
3 2
soluble HRh(CO) (PPh )
complexes. The peaks at 2045
is no evidence for phosphine rich HRh(CO)(PPh ) species, as
n
3 4−n
3 3
and 1990 cm− are attributable to ee-HRh(CO) (PPh ) ,
1
were observed in the spectrum of PPh -modified 1% Rh/SiO .
2
3 2
3
2
−1
whereas the peaks at 2010 and 1950 cm are attributable to
More interestingly, this catalyst showed the presence of a broad
21,22
−1
−1
ea-HRh(CO) (PPh ) complexes.
The band at 1900 cm
peak at 1820 cm , which is characteristic of bridging CO
2
3 2
can be assigned to HRh(CO)(PPh ) species downshifted
present in rhodium dimers.
IR spectra of bidentate xantphos-modified 0.2% Rh/SiO2
3
3
21
because of the presence of three PPh groups. The growth of
3
bands associated with HRh(CO)(PPh3)3 and HRh-
catalyst (X-Rh/SiO , X/Rh = 15) in the presence of synthesis
2
(
CO) (PPh ) with increasing partial pressure provides
gas are presented in Figure 4B. Upon contacting the catalyst
with CO in He, carbonyl bands appeared at 2005 and 1940
2
3 2
evidence for in situ formation of homogeneous catalysts. The
nearly linear growth of these bands with increasing synthesis
gas pressure also indicates that an equilibrium exists between
gaseous CO, Rh metal, and homogeneous HRh-
−1
cm . The positions of these bands are attributed to the
12
formation of the rhodium carbonyl phosphine complexes.
Addition of H to CO to the gas stream resulted in the
2
15
(
CO) (PPh )
species. Notably, no bands were observed
formation of four carbonyl bands at 2035, 1995, 1970, and 1940
n
3 4−n
−1
−1
in the region of 1850−1700 cm , characteristic of bridging CO
ligands, suggesting the absence of dimeric rhodium complexes.
The IR spectra of PPh -modified 0.2% Rh/SiO (PPh /Rh =
cm . These features are identical to those observed for Rh
complexes formed with xantphos dissolved in organic solvents,
ionic liquids, and supported ionic liquid phase (SILP) catalytic
3
2
3
22
1
5) differed from those of PPh -modified 1% Rh/SiO as a
systems (Supporting Information). The bands at 2035 and
3
2
1970 cm− are attributable to ee-HRh(CO) (X), in which both
1
result of the different amounts of PPh ligands (per gram of
3
2
3
52
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
Figure 6. (a, b) High-resolution transmission electron micrographs (HR-TEM) and (c, d) high-angle annular dark field scanning tunneling electron
microscopy (HAADF-STEM) of (a, c) 1% Rh/SiO and (b, d) 0.2% Rh/SiO samples. Red circles represent rhodium nanoparticles of ∼1 nm
2
2
diameter.
phosphorus atoms occupy equatorial coordination sites, and the
bands at 1995 and 1940 cm are attributable to ea-
By contrast, X-Rh/SiO exhibited peaks at 36, −16, and −22
2
−
1
ppm, similar to the features observed for sulfoxantphos-
12
HRh(CO) (X), in which one phosphorus atom is equatorial
modified rhodium SILP catalysts. The peak at −16 ppm is
attributable to xantphos weakly physisorbed on silica; the broad
peak at −22 ppm is attributable to phosphorus atoms in either
singly protonated or nonprotonated, strongly physisorbed
xantphos ligands. The peak at 35 ppm is assigned to the
phosphine groups of xantphos interacting with Rh. These
results are indicative of in situ formation of homogeneous
2
and the other is apical. No bands were observed in the region
−1
characteristic of CO bridging ligands (1800−1700 cm ),
suggesting the absence of dimeric [Rh(X)(CO)(μ-CO)]2
species for catalysts prepared with xantphos ligand on Rh/SiO .
2
3
.3.2. 31P MAS NMR Spectroscopy. Further evidence for in
situ formation of homogeneous complexes over the phosphine-
31
HRh(CO) (L)
species on phosphine-modified Rh/SiO2.
modified Rh/SiO catalysts was obtained by solid-state
P
n
4−n
2
3.3.3. HR-TEM Measurements. HR-TEM image of 1% Rh/
MAS NMR spectroscopy. The spectrum of a PPh -modified
3
SiO2 showed rhodium nanoparticles ∼5 nm in diameter,
silica support (PPh /SiO ) exhibits a band at −5.6 ppm
3
2
13
whereas the 0.2% Rh/SiO catalyst showed particles 1.5 nm in
ascribed to PPh adsorbed on the silica surface. By contrast,
2
3
diameter. In both cases, the particles are well distributed over
the support surfaces, as shown in Figure 6. Further
confirmation for the high dispersion of rhodium nanoparticles
on the silica support surfaces was obtained by the high angle
annular dark-field-scanning tunneling electron microscopy
(HAADF-STEM). A spent catalyst showed no difference in
the size of the rhodium nanoparticle, highlighting that the
particle remains unchanged under the reaction conditions.
PPh -Rh/SiO showed the presence of physisorbed phosphine
at −5.6 ppm and a broad, less intense, peak centered at 33 ppm
due to phosphine adsorbed on rhodium nanoparticles. The
3
2
position of this peak is similar to the peak observed at 36 ppm
in the 31P MAS NMR spectra of HRh(CO)(PPh ) /SiO ,
3
3
2
+
which is attributed to the coordination of the Rh ion and the
PPh of the rhodium complex (Supporting Information).
Figure 5 shows the P MAS NMR spectrum of the spent PPh -
13
3
3
1
3
modified Rh/SiO2 and the xantphos-modified Rh/SiO2
4
. DISCUSSION
catalysts. The spent PPh -0.05% Rh/SiO catalyst showed a
3
2
sharp peak at ∼38 ppm, and with increasing rhodium loading, a
progressive upfield shift of this band is observed. Free
phosphine characterized by a peak at −5.6 ppm was observed
The strong similarities in the rate parameters for hydro-
formylation of propene over phosphine- and xantphos-modified
Rh/SiO and the corresponding homogeneous species of the
2
1
5
in the spent PPh -1% Rh/SiO catalyst, whereas free phosphine
type HRh(CO) (L)
suggest that Rh complexes are
3
2
n
4−n
was absent for catalysts prepared with lower rhodium loadings.
This result suggests that PPh -1% Rh/SiO contains free
produced in situ. The manner in which this occurs is as
follows: Nanoparticles of Rh supported on silica or alumina are
known to undergo corrosive chemisorption in the presence of
3
2
phosphines and homogeneous species, mainly as HRh(CO)-
PPh ) . Indeed, a phosphine-modified HRh(CO)(PPh ) /
(I)
(
CO to produce Rh(I) gem-dicarbonyl species (Rh (CO) )
2
20
3
3
3 3
−1
SiO (Rh = 1%) spent catalyst also showed a peak at 30 ppm.
exhibiting bands at 2090 and 2030 cm . It has been
2
3
53
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
proposed that such species are formed from supported Rh
nanoparticles by the adsorption of CO on the surface of the
particle, which leads to a weakening of Rh−Rh bonds due to
the higher strength of Rh−CO bonds (185 kJ) than Rh−Rh
hydroformylation cycle in which each intermediate contains
two phosphines per rhodium derived from HRh(CO)(PPh ) .
3
3
The decreased activity for PPh /Rh > 15 is related to the
3
difficulty of HRh(CO) (PPh ) 18-electron complexes to lose
n
3 4−n
2
3
bonds (44.5 kJ). Neutral Rh dicarbonyl species formed via
this process can then interact with the hydroxyl groups of the
a PPh ligand to form the 16-electron species or to the
3
15
formation of HRh(L) type species. By contrast, the bidentate
4
(
I)
support to release H and produce Si−O−Rh (CO)
xantphos ligand favors the formation of HRh(CO) (X) 18-
2
2
2
species, as shown in Scheme 1. Evidence for the formation of
electron species, which then dissociates a CO ligand to form
the catalytically active, regioselective, 16-electron HRh(CO)-
(X) species.
Scheme 1. Proposed Mechanism for the Formation of
Homogenous Catalysts on Phosphine-Modified Rh/SiO2
The dependence of propene hydroformylation activity and
stability on rhodium loading can be interpreted in the following
manner. It is known that the ratio of surface atoms located at
edges and vertices relative to those in planes increases
significantly as the size of the nanoparticle becomes small.
Our TEM observations show that the size of the Rh particles
supported on SiO decreases with decreasing Rh loading. Since
2
the Rh atoms located at the particle corners and edges are more
coordinatively unsaturated, they are more susceptible to
corrosive chemisorption than Rh atoms located on the planar
surfaces of the particles. Thus, we attribute the higher specific
activity per Rh site observed with decreasing Rh weight loading
to an increase in the facility with which corrosive chemisorption
releases Rh atoms from Rh nanoparticles, which in turn
such species can be drawn from the results of in situ IR
spectroscopy. The IR spectra for both 0.2 wt % and 1 wt % Rh/
SiO exposed to CO at 120 °C showed bands for gem-rhodium
2
−
1
dicarbonyl bands at 2090 and 2030 cm (Supporting
Information) and the concurrent decrease in the intensity of
(
I)
increases the concentration of Rh (CO) species. Thus, the
2
−
1
highly dispersed and smaller nanoparticles (preferably cluster
sized) leads to an ease of formation of the homogeneous
complexes in the presence of syngas than larger particles. This
result is clearly confirmed from the comparison between the
size of the rhodium nanoparticles that the highly dispersed and
smaller nanoparticles showed a higher rate than the larger Rh
nanoparticles (Figure 2).
the band at 3660 cm due to Si−O−H vibrations.
(I)
Interaction of H with the SiO−Rh (CO) species leads
2
2
to the formation of mobile HRh(CO) species and a
2
24
restoration of hydroxyl groups. We envisage that, under
reaction conditions, the phosphine ligands interact with these
species to generate in situ HRh(CO) (X) species in the
2
presence of xantphos ligand and HRh(CO) (PPh ) species
n
3 4−n
The higher catalytic activity of bidentate xantphos-modified
in the presence of PPh (Scheme 1).
3
Rh/SiO compared with monodentate PPh -modified Rh/SiO
The effect of the PPh /Rh ratio on the rate of n- and
2
3
2
3
can be related to the chelating ability of xantphos, which yields
isobutanal formation can be attributed to the equilibrium
shown in Scheme 2. Because of the low CO partial pressures
1
a higher concentration of the coordinated Rh species. This, in
turn, results in increased complexation of the rhodium center
by the diphosphine X ligand, and the rigid backbone of the
ligand compells the phenyl groups of xantphos to exert a
greater degree of steric hindrance on the alkene entering the
coordination sphere, favoring linear products. Usage of different
ligands suggests that as the electron density on the rhodium
atom is reduced by increasing the electron-withdrawing
used, formation of HRh(CO) species is unlikely under our
experimental conditions, which explains the absence of catalytic
4
activity without PPh ligands. The increase in catalytic activity
3
with increasing PPh /Rh ratio is due to an increase in the
3
concentration of HRh(CO) (PPh ) species (Figure 1). The
n
3 4−n
need for excess PPh arises from the relatively weak Rh−PPh
3
3
2
5
binding. Hence, to maintain a high activity and regioselec-
tivity, excess PPh is required to drive the equilibrium in favor
functionality in the phosphine ligand (3F-PPh
-Rh/SiO vs
3 2
3
of HRh(CO)(PPh ) . Consequently, as the ligand concen-
OMe-PPh -Rh/SiO ), the selectivity to linear aldehydes
3
2
3
2
tration increases, the equilibria in Scheme 2 shift to the left, and
the steric bulk of the complexes increases, contributing to the
formation of complexes that produce higher n/iso ratios. Stated
another way, the enhancement in linear aldehyde selectivity
increased with a significant change in the activity. Since para-
substituted aryl phosphines were utilized in the present study,
the steric environment around the active rhodium center is
16
nearly constant. Hence, the catalytic activity differences can
observed at high PPh /Rh ratios may be due to the
be correlated nicely with the electronic properties of the
3
Scheme 2. Illustration of Species That Might Form HRh(CO) (PPh ) under Reaction Conditions
2
3 2
3
54
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Research Article
phosphine ligands. If the ligand has good back-bonding ability,
the equilibrium in Scheme 2 will shift to the left to favor
18-electron Rh(III) species H, followed by the reductive
elimination of butanal. For X-Rh/SiO2 catalysts, hydro-
formylation of propene is initiated by the loss of a CO ligand
(
I)
HRh(CO)(L) to prevent excess charge building up on Rh
3
cations. These results suggest that electron-withdrawing
phosphine ligands tend to form the 18-electron HRh-
from HRh(CO) (X) (species A) to form the 16-electron
2
species HRh(CO)(X) (B) (Scheme 4B). Isobutanal is formed
in a catalytic cycle by an analogous mechanism, in which
species D−G involve a secondary, rather than a primary, alkyl
species.
(
CO) (PPh )
species and then rapidly lose a ligand to
n
3 4−n
form the coordinatively unsaturated, active 16-electron HRh-
CO) (PPh ) species. Consistent with this interpretation,
(
n
3 3−n
electron-withdrawing substituents in phosphine ligands increase
the carbonyl stretching frequencies, which provide direct
evidence that the M-CO bond is weakened as the electron
The observed reaction orders at 363 K for PPh -Rh/SiO
3
2
catalyst were first order in C H , CO and H , suggesting that
3
6
2
the rate-determining step (RDS) is the oxidative addition of
hydrogen. Under this assumption, the rate of butanal formation
can be written as follows:
16
withdrawing ligands reduce M−CO dπ−pπ interactions. The
low n-aldehyde selectivity observed with the phosphite ligands
suggests that the electronegativity difference of the groups
attached to the phosphorus will change the σ donor and π
acceptor properties of the ligands, leading to a change in the
6
i=1
∏
K k P P P /[PPh ]
i 7 C H CO H 3
3 6 2
rbutanal
=
[Rh]
1
+ K P /[PPh ]
(1)
25
1 CO
3
overall performance of the catalysts.
The catalytic activity, n/iso ratio, and stability of phosphine-
where k is the rate coefficient for reaction i, K is the
i
i
modified Rh/SiO (PPh -Rh/SiO ) and HRh(CO)(PPh ) /
equilibrium constant for reaction i, P is the partial pressure of
2
3
2
3 3
j
SiO differ significantly. Phosphine-modified Rh/SiO is stable
species j, [PPh ] is the concentration of PPh species, and [Rh]
2
2
3
3
under reaction conditions, whereas HRh(CO)(PPh ) /SiO
is the total moles of Rh. If it is assumed that the second term in
the denominator of eq 1 is small relative to unity, then the rate
expression is consistent with the observed kinetics at 363 K.
Similar positive order kinetics have been reported earlier for
hydroformylation reactions carried out at pressures lower than
those described as typical hydroformylation conditions (343−
3
3
2
progressively loses activity under identical reaction conditions
Supporting Information). These observations suggest that the
(
interaction of the Rh complex with the rest of the catalyst is
different in the two samples. As discussed earlier, we envisage
that the formation of geminal rhodium dicarbonyl species
provides a favorable stabilizing effect, as proposed by Augustine
1
5,28
393 K, 10−30 atm).
26
and co-workers using heteropolyacid as linkers (Scheme 3).
The apparent reaction orders observed at 363 K for
xantphos-modified Rh/SiO2 catalyst were first-order in
propene, negative-order in CO, and nearly zero-order in H2.
These results suggest that the rate-determining step (RDS)
now becomes alkene insertion, in good accordance with the
kinetics reported for hydroformylation over sulfoxantphos-
Scheme 3. Proposed Stabilization of HRh(CO) (PPh ) on
n
3 4−n
Rh/SiO2
19
containing Rh-SILP catalysts. Further support for alkene
insertion as the RDS for diphosphine ligands can be drawn
29
from the work of van Leeuwen and co-workers. Upon the
basis of the hydroformylation reaction mechanism and
considering alkene insertion to be the irreversible, rate-limiting
step, the rate of butanal formation is given by the following
relation:
Comparison of the IR spectra of PPh -Rh/SiO and HRh-
3
2
K K k P [Rh]
−1
1 2 3 C3H6
(
CO)(PPh ) /SiO shows a greater red shift (∼1945 cm ) for
3
3
2
rbutanal = P
+ K + K K P
the CO resonances of homogeneous species in PPh -Rh/SiO
CO
1
1
2 C H
6
(2)
3
2
3
than the Wilkinson-catalyst-supported silica sample (∼1975
−1
where k is the rate coefficient for reaction i, K is the
i i
equilibrium constant for reaction i, P is the partial pressure of
species j, and [Rh] is the total moles of Rh.
Equation 2 is consistent with the observed kinetics at 363 K.
The rate of butanal synthesis is zero-order in the partial
pressure of H and less than first-order in the partial pressure of
C H . As the partial pressure of C H increases, the K K P
cm ) (see Supporting Information). The sharp decrease in
activity of the HRh(CO)(PPh ) /SiO catalyst thus relates to
j
3
3
2
the weak interaction of the homogeneous catalyst with the silica
surface and thereby to the easy formation of inactive dimers/
1
2,13
clusters.
The kinetics of butanal formation observed for monodentate
2
PPh -modified and bidentate xantphos-modified Rh/SiO2
3
6
3
6
1 2 C H
3 6
3
catalysts can be interpreted using the mechanism proposed
term in the denominator becomes increasingly significant,
thereby reducing the effective order from unity. The
experimentally observed, inhibitory effect of CO is also
effectively represented by eq 2. Since the order of CO is
−0.4, the remaining terms in the denominator are significant in
27
by Heck and Breslow. HRh(CO)(PPh ) (A) and HRh-
3
3
(
CO) (PPh ) (B) species exist in equilibrium for phosphine-
2 3 2
modified PPh -Rh/SiO catalysts, and hydroformylation is
3
2
initiated with the loss of a CO ligand to form the 16-electron
species, HRh(CO)(PPh ) (C) (Scheme 4A). Propene then
magnitude compared with P .
3
2
CO
2
coordinates to species C to form a η -Rh complex (species D).
Migratory insertion into the Rh−H bond forms a propyl-Rh
complex (species E). CO then adds to species E to form an 18-
electron intermediate species F. Migratory insertion of the
propyl to a carbonyl group leads to the formation of an acyl
The decrease in the n/iso ratio from low temperature to high
temperature over the phosphine-modified Rh/SiO catalysts
2
can be explained as follows: PPh -Rh/SiO catalysts show a
3
2
very high n/iso ratio of 22 at 353 K, which decreases to 3 at 413
K. Preference for forming linear versus branched aldehydes in
PPh -Rh/SiO catalysts can be attributed to the preferred
intermediate (species G). Oxidative addition of H forms the
2
3
2
3
55
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
Scheme 4. Proposed Mechanism for the Hydroformylation of Propene on (i) PPh -Rh/SiO and (ii) X-Rh/SiO
2
Research Article
3
2
isomerization of HRh(CO)(PPh ) to the structure in which
SiO . A first-order dependence on the partial pressures of
3
2
2
both phosphine groups are coordinated in the equatorial
propene, CO, and H2 were observed for PPh -Rh/SiO ,
3
2
1
5
plane. The high n/iso ratio observed with PPh -Rh/SiO at
whereas X-Rh/SiO catalysts showed first-order partial pressure
3
2
2
low temperatures is consistent with studies conducted with
HRh(CO)(PPh ) in homogeneous solutions. For X-Rh/SiO
dependence in propene, negative order in CO, and zero order
in H . Differences in the kinetics between the monodentate and
3
3
2
2
catalysts, a decrease in the n/iso molar ratio from 15 at 353 K
to ∼8 at 413 K was observed. For the bidentate xantphos
ligand, the regioselectivity is determined by alkene insertion
into the Rh−H bond to form either primary or secondary
the bidentate phosphine ligands are related to the difference in
the rate-limiting steps. Oxidative addition of hydrogen is the
rate-limiting step for PPh -Rh/SiO , whereas alkene insertion is
3
2
3
1
the rate-limiting step over the X-Rh/SiO . In situ FTIR and P
2
1
alkyls. The apparent activation energies measured for n- and
MAS NMR indicate the in situ formation of HRh(CO) (X)
2
−1
isobutanal formation are 65 and 75 kJ mol , respectively.
Therefore, the resulting n/iso ratio is high because primary
alkyls are favored kinetically over secondary alkyls. Indeed,
theoretical studies demonstrate that the transition for the
species on the xantphos-modified Rh/SiO2 catalyst and
HRh(CO) (PPh ) species on the phosphine-modified Rh/
n
3 4−n
(
I)
SiO . In each case, isolated Rh (CO) species are produced via
2
2
corrosive chemisorption, as shown in Scheme 1.
−1
formation of secondary alkyls is ∼9 kJ mol greater than that
for the formation of primary alkyls because of the steric effects
ASSOCIATED CONTENT
30
■
exerted from the phenyl groups of the phosphine ligand. The
decrease in the n/iso ratio at high temperature is due to a shift
from the thermodynamically favorable primary alkyl complex to
the less favorable secondary alkyl complex. These results are in
agreement with previous deuteroformylation studies, which
found that Rh−alkyl formation is irreversible at low temper-
*
S
Supporting Information
Seventeen figures are provided as Supporting Information. This
material is available free of charge via the Internet at http://
pubs.acs.org.
1
9,31
ature but reversible at higher temperatures.
AUTHOR INFORMATION
■
Corresponding Author
5. CONCLUSIONS
*
Tel. +1 510 642 1536. Fax +1 510 642 477. E-mail: bell@
cchem.berkeley.edu.
The results of this study provide a detailed understanding of the
factors influencing the activity, selectivity, and stability of
Notes
phosphine-modified Rh/SiO catalysts for the hydroformyla-
2
The authors declare no competing financial interest.
tion of propene. Catalyst performance and stability are found to
be strong functions of the nature and concentration of ligand
and rhodium loading. The catalysts produced are very stable
and exhibit 100% selectivity to butyraldehydes; no propane or
alcohols were detected. Higher catalytic activity and n/iso ratios
ACKNOWLEDGMENTS
■
This work was supported by the XC2 program funded by BP.
The authors thank Dr. Chris Canlas (Berkeley NMR facility)
for his assistance in acquiring the NMR spectra, Eric Bloch for
the TGA results, Dr. Zhengmeng Peng for the HR-TEM
results.
were observed over xantphos-modified Rh/SiO than over
2
phosphine-modified Rh/SiO . The reaction temperature has a
2
strong effect on the activity and regioselectivity of PPh -Rh/
3
3
56
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357

ACS Catalysis
REFERENCES
Research Article
■
(
1) Rhodium Catalyzed Hydroformylation; van Leeuwen, P. W. N. M.,
Claver, C., Eds.; Kluwer, Dordrecht, 2000.
(
2) Cornils, B.; Hermann, W. A.; Rasch, M. Angew. Chem., Int. Ed.
1
(
(
994, 33, 2144−2163.
3) Cole-Hamilton, D. J. Science 2003, 299, 1702−1706.
4) Beller, M.; Cornils, B.; Frohning, C. D.; Kohlpainter, C. W. J.
Mol. Catal. A: Chem. 1995, 104, 17−85.
5) Sandee, A. J.; Reek, J. N. H.; Kamer, P. C. J.; van Leeuwen, P. W.
N. M. J. Am. Chem. Soc. 2001, 123, 8468−8476.
6) Mehnert, C. P.; Cook, R. A.; Dispenziere, N. C.; Afeworki, M. J.
Am. Chem. Soc. 2002, 124, 12932−12933.
7) Arhanchet, J. P.; Davis, M. E.; Hanson, B. E. J. Catal. 1991, 129,
4−99.
8) Riisager, A.; Wasserscheid, P.; Van Hal, R.; Fehrmann, R. J. Catal.
003, 219, 452−455.
9) Riisager, A.; Fehrmann, R.; Flicker, S.; van Hal, R.; Haumann, M.;
Wasserscheid, P. Angew. Chem., Int. Ed. 2005, 44, 815−819.
10) Riisager, A.; Fehrmann, R.; Haumann, M.; Gorle, B. S. F.;
Wasserscheid, P. Ind. Eng. Chem. Res. 2005, 44, 9853−9859.
11) Riisager, A.; Fehrmann, R.; Haumann, M.; Wasserscheid, P. Eur.
J. Inorg. Chem. 2006, 695−706.
12) Shylesh, S.; Hanna, D.; Werner, S.; Bell, A. T. ACS Catal. 2012,
, 487−493.
13) Yan, L.; Ding, Y. J.; Zhu, H. J.; Xiong, J. M.; Wang, T.; Pan, Z.
D.; Lin, L. W. J. Mol. Catal. A: Chem. 2005, 234, 1−7.
14) Zhu, H. J.; Ding, Y. J.; Yan, L.; Xiong, J. M.; Lu, Y.; Lin, L. W.
Catal. Today 2004, 93, 389−393.
15) Kim, T. J.; Celik, F. E.; Hanna, D. G.; Shylesh, S.; Werner, S.;
Bell, A. T. Top. Catal. 2011, 54, 299−305.
16) Moser, W. R.; Papile, C. J.; Brannon, D. A.; Duwell, R. A. J. Mol.
Catal. 1987, 41, 271−292.
17) Sandee, A. J.; Slagt, V. F.; Reek, J. N. H.; Kamer, P. C. J.; van
Leeuwen, P. W. N. M. Chem. Commun. 1999, 1633−1634.
18) Dupont, J.; Silva, S. M.; De Souza, R. F. Catal. Lett. 2001, 77,
31−133.
19) Hanna, D.; Shylesh, S.; Werner, S.; Bell, A. T. J. Catal. 2012,
92, 166−172.
20) Yang, A. C.; Garland, C. W. J. Phys. Chem. 1957, 61, 1504−
512.
21) Evans, D.; Yagupsky, G.; Wilkinson, G. J. Chem. Soc. 1968,
660−2663.
22) Silvia, S. M.; Bronger, R. P. J.; Freixa, Z.; Dupont, J.; van
Leeuwen, P. W. N. M. New J. Chem. 2003, 27, 1294−1296.
23) Vant Blik, H. F. J.; Van Zon, J. B. A. D.; Huizinga, T.; Vis, J. C.;
Koningsberger, D. C.; Prins, R. J. Am. Chem. Soc. 1985, 107, 3139−
145.
24) Basu, P.; Panayotov, D.; Yates, J. T., Jr. J. Am. Chem. Soc. 1988,
(
(
(
9
(
2
(
(
(
(
2
(
(
(
(
(
(
1
(
2
(
1
(
2
(
(
3
(
1
(
(
10, 2074−2081.
25) Pruett, R. L.; Smith, J. A. J. Org. Chem. 1969, 34, 327−330.
26) Augustine, R.; Tanielyan, S.; Anderson, S.; Yang, H. Chem.
Commun. 1999, 1257−1258.
27) Heck, R. F.; Breslow, D. S. J. Am. Chem. Soc. 1962, 84, 2499−
503.
28) Deshpande, R. M.; Bhanage, B. M.; Kanagasabapathy, S.;
Chaudhari, R. V. Ind. Eng. Chem. Res. 1998, 37, 2391−2396.
29) Zuidema, E.; Escorihuela, L.; Eichelsheim, T.; Carbo, J. J.; Bo,
C.; Kamer, P. C. J.; van Leeuwen, P. W. N. M. Chem.Eur. J. 2008,
4, 1843−1851.
30) Carbo, J.; Maseras, F.; Bo, C.; van Leeuwen, P. W. N. M. J. Am.
Chem. Soc. 2011, 123, 7630−7638.
31) Casey, C. P.; Petrovich, L. M. J. Am. Chem. Soc. 1995, 117,
007−6013.
(
2
(
(
1
(
(
6
3
57
dx.doi.org/10.1021/cs3007445 | ACS Catal. 2013, 3, 348−357