Ionic-liquid-like copolymer stabilized nanocatalysts in ionic liquids: II. Rhodium-catalyzed hydrogenation of arenes

Article abstract of DOI:10.1016/j.jcat.2007.05.014

Rhodium nanoparticles stabilized by the ionic-liquid-like copolymer poly[(N-vinyl-2-pyrrolidone)-co-(1-vinyl-3-butylimidazolium chloride)] were used to catalyze the hydrogenation of benzene and other arenes in ILs. The nanoparticle catalysts can endure forcing conditions (75 °C, 40 bar H₂), resulting in high reaction rates and high conversions compared with other nanoparticles that operate in ILs. The hydrogenation of benzene attained record total turnovers of 20,000, and the products were easily separated without being contaminated by the catalysts. Other substrates, including alkyl-substituted arenes, phenol, 4-n-propylphenol, 4-methoxylphenol, and phenyl-methanol, were studied and in most cases were found to afford partially hydrogenated products in addition to cyclohexanes. In-depth investigations on reaction optimization, including characterization of copolymers, transmission electron microscopy, and an infrared spectroscopic study of nanocatalysts, were also undertaken.

Full text of DOI:10.1016/j.jcat.2007.05.014

Journal of Catalysis 250 (2007) 33–40
www.elsevier.com/locate/jcat
Ionic-liquid-like copolymer stabilized nanocatalysts in ionic liquids:
II. Rhodium-catalyzed hydrogenation of arenes
Chen Zhao a, Han-zhi Wang a, Ning Yan a, Chao-xian Xiao a, Xin-dong Mu a, Paul J. Dyson b,∗,
Yuan Kou a,∗
a
PKU Green Chemistry Center, Beijing National Laboratory for Molecular Sciences, College of Chemistry and Molecular Engineering, Peking University,
Beijing 100871, China
b
Institut des Sciences et Ingénierie Chimiques, Ecole Polytechnique Fédérale de Lausanne (EPFL), CH-1015 Lausanne, Switzerland
Received 20 March 2007; revised 18 May 2007; accepted 22 May 2007
Available online 28 June 2007
Abstract
Rhodium nanoparticles stabilized by the ionic-liquid-like copolymer poly[(N-vinyl-2-pyrrolidone)-co-(1-vinyl-3-butylimidazolium chloride)]
◦
were used to catalyze the hydrogenation of benzene and other arenes in ILs. The nanoparticle catalysts can endure forcing conditions (75 C,
40 bar H ), resulting in high reaction rates and high conversions compared with other nanoparticles that operate in ILs. The hydrogenation of
2
benzene attained record total turnovers of 20,000, and the products were easily separated without being contaminated by the catalysts. Other
substrates, including alkyl-substituted arenes, phenol, 4-n-propylphenol, 4-methoxylphenol, and phenyl-methanol, were studied and in most cases
were found to afford partially hydrogenated products in addition to cyclohexanes. In-depth investigations on reaction optimization, including
characterization of copolymers, transmission electron microscopy, and an infrared spectroscopic study of nanocatalysts, were also undertaken.
© 2007 Elsevier Inc. All rights reserved.
Keywords: Rhodium; Nanoparticles; Hydrogenation; Biphasic catalysis; Ionic liquids; Lignins; Copolymer
1. Introduction
structural organization of ILs are believed to be effective for
preparing small nanoparticles and for controlling extended or-
dering of nanoscale structures [5]. In addition, evidence has
been found relating to the fact that imidazolium-based ILs are
good stabilizers for transition-metal nanoparticles [6].
Soluble nanoparticle catalysts with unique properties have
aroused increasing interest in recent years due to their poten-
tially high catalytic efficiency [1,2]. Unlike their counterparts
restricted on solid surfaces, soluble nanoparticles with their ro-
tational freedom and spherically symmetrical geometry are, at
least in principle, more active [3].
Ionic liquids (ILs) provide the opportunity to combine the
advantages of both homogeneous and heterogeneous processes
in a single system [3,4]. Immobilization of nanoparticles by
“supporting” them in an IL rather than on a surface preserves
the rotationally free catalytic centers, in keeping with soluble
nanoparticle systems. Moreover, the low interfacial tension and
Generally, the most pertinent problem encountered in solu-
ble nanoparticle applications is the need to stabilize the parti-
cles against aggregation or agglomeration, which tends to be
done by adding stabilizers. Examples of stabilizers used in con-
junction with catalytically active soluble nanoparticles include
solvents [7,8], soluble polymers [9], quaternary ammonium
salts [10], and polyoxoanions [11]. It must be kept in mind
that very stable nanoparticles are not the ultimate aim in cata-
lysis, because they are generally less active, due to “overpro-
tection” of the catalytically active nanoparticle surface. In our
previous communication, rhodium nanoparticles protected by
“IL-like” copolymers were shown to catalyze benzene hydro-
genation with a record of total turnover (TTO) of 20,000 [12],
*
Corresponding authors. Faxes: +86 10 62751708, +41 21 693 98 85.
E-mail addresses: yuankou@pku.edu.cn (Y. Kou), paul.dyson@epfl.ch
(P.J. Dyson).
0021-9517/$ – see front matter © 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2007.05.014

34
C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
demonstrating that an optimum balance between stability and
reactivity to be a critical parameter.
2.1.2. Synthesis of 1-vinyl-3-ethylimidazolium bromide
([VEIM]Br)
The hydrogenation of benzene and arenes, which is ac-
complished using heterogeneous catalysts almost exclusively
[13,14], molecular precatalysts immobilized on heterogeneous
surfaces [15–17], supported nanoparticle [18], or soluble nano-
particle [19,20] systems, sometimes inadvertently derived from
molecular precursors [21,22], represents an important indus-
trial catalytic transformation, particularly for the production of
cleaner-burning, low-aromatic diesel fuels [13]. The partial hy-
drogenation of arenas, which is far more difficult to realize, is
equally important because it can help to simplify many mul-
tistage synthetic procedures and allow the use of alternative
precursors [23]. The catalytic procedure for partial hydrogena-
tion of arenes generally is not simple; for example, the selective
hydrogenation of benzene to cyclohexene performed on an in-
dustrial scale involves a multiphase process using a ruthenium-
based heterogeneous catalyst, which results in 90% conversion
and 60% selectivity [24].
In an extension to our preliminary report on Rh nanoparticles
that catalyze the hydrogenation of benzene in ILs with unprece-
dented lifetimes [12], we now describe their application to the
hydrogenation of benzene and other arenes in greater detail. IL-
like copolymers were evaluated in terms of composition and
average molecular weight and compared with poly(N-vinyl-2-
pyrrolidone) (PVP), a widely used protecting agent [9,25]. The
hydrogenation of arene substrates with various alkyl and other
substituents was then investigated, and activities were corre-
lated to nanoparticle structure. Moreover, we found that these
rhodium nanoparticles are highly active catalysts for the partial
hydrogenation of arene substrates in ILs, with very high selec-
tivity toward the monoene in some cases.
1-Vinylimidazole (5.00 g, 53 mmol) was mixed with ethyl
bromide (6.54 g, 60 mmol), and the mixture was stirred vigor-
ously at 70
◦C for 3 h. After reaction, the ethyl bromide was
removed under vacuum at 60 ◦C for 2 h to afford a pale-yellow
solid [VEIM]Br. Yield 9.80 g, 95%.
1H NMR δ (300 MHz, D2O) 1.35 (t, J = 7.6, 3H, CH3),
4.10 (q, J = 7.6, 2H, CH2), 5.24 (dd, J = 8.6, 2.4, 2H, CH2),
5.62 (dd, J = 15.2, 2.4, 1H, CH), 6.96 (dd, J = 15.2, 8.6, 1H,
CH), 7.30 (s, 1H, NH), 7.59 (s, 1H, NH).
13C NMR δ (75 MHz, D2O) 16.61, 47.50, 111.44, 113.01,
121.66, 124.81, 130.57 ppm.
Anal. Calcd. for C7H11N2Br: C, 41.40; H, 5.46; N, 13.79.
Found: C, 41.47; H, 5.55; N, 13.95.
HRMS Calcd. for C7H11N+2 : 123.0917. Found: 123.0915.
2.1.3. Copolymerization
A mixture of N-vinyl-2-pyrrolidone (NVP) (0.5 g, 4.5
mmol), [VBIM]Cl (0.43 g, 2.3 mmol) or [VEIM]Br (0.46 g,
2.3 mmol), and 2,2ꢀ-azo-bis-isobutyronitrile (AIBN) (6.9 mg,
0.042 mmol) in methanol (5 ml, degassed by 3 freeze/thaw cy-
cles) was stirred in a preheated oil bath at 60 ◦C for 16 h. The
resulting solid was dissolved in methanol (5 ml) and precipi-
tated in diethyl ether (500 ml). The residual monomers were
removed by dialysis in distilled water for 24 h, and the solvent
was removed under reduced pressure. Finally, the copolymers
were dried under vacuum at room temperature. Yield ∼0.8 g.
2.2. Preparation of Rh nanoparticles in ILs
Rhodium nanoparticles were prepared by reduction of RhCl3
(1.6 × 10−5 mol) with H2 in 1-butyl-3-methylimidazolium
tetrafluoroborate ([BMIM][BF4]) (6 ml) containing poly(NVP-
co-VBIMCl) ([poly(NVP-co-VBIMCl)]/[Rh] = 5:1). Typically,
the preparation was carried out in a 100 ml autoclave to afford
a black solution after reduction at 60 ◦C and H2 (40 bar) for
30 min.
2. Experimental
2.1. Synthesis of copolymers
2.1.1. Synthesis of 1-vinyl-3-butylimidazolium chloride
([VBIM]Cl)
2.3. Arene hydrogenation reactions
1-Vinylimidazole (5.00 g, 53 mmol) and butyl chloride
(18.50 g, 200 mmol) were mixed and stirred vigorously at 70 ◦C
for 24 h. The residual mixture was cooled to 0 ◦C, and then
the upper liquid layer was removed by decantation. The residue
was then washed with ethyl acetate (3 × 30 ml), evaporated un-
der vacuum, and dried to yield a pale-yellow solid [VBIM]Cl
(7.74 g, 75%).
The arene (32 mmol) and Rh nanoparticle solution (6 ml,
synthesized as described above) were placed in the autoclave. In
a typical experiment, H2 (40 bar) was introduced into the auto-
clave after the reactor was purged 3 times with H2. The mixture
was stirred at 800 rpm at 75 ◦C for 10 h, which afforded a two-
phase system (the lower IL phase-containing catalysts and the
upper organic phase). The upper layer was decanted and ana-
lyzed by GC, GC–MS, IR, and 1H NMR spectroscopy.
1H NMR δ (300 MHz, D2O) 0.98 (t, J = 7.6, 3H, CH3),
1.41 (m, 2H, CH2), 1.95 (m, 2H, CH2), 4.41 (t, J = 7.6, 2H,
CH2), 5.39 (dd, J = 8.7, 3.0, 1H, CH), 6.01 (dd, J = 14.7, 2.7,
1H, CH), 7.55 (dd, J = 14.7, 8.7, 2H, CH2), 7.88 (s, 1H, NH),
11.35 (s, 1H, NH) ppm.
2.4. Mercury poisoning experiments
13C NMR δ (75 MHz, D2O) 13.40, 19.41, 32.00, 50.06,
101.82, 118.92, 122.35, 128.41, 136.87 ppm.
2.4.1. Experiment 1. Hydrogenation without mercury
Anal. Calcd. for C9H15N2Cl: C, 57.90; H, 8.10; N, 15.01.
Found: C, 57.81; H, 8.02; N, 15.10.
To a solution of [BMIM][BF4] (6 ml) containing Rh nano-
particles (0.016 mmol Rh) and poly(NVP-co-VBIMCl)
(0.08 mmol), benzene (2.5 g, 32 mmol) was added in a 100 ml
HRMS Calcd. for C9H15N+2 : 151.1230. Found: 151.1227.

C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
35
autoclave. The reactor was purged 3 times with H2 and the reac-
tor was pressurized to 40 bar. The mixture was reacted at 75 ◦C
and stirred at 800 rpm for 8 h.
In an extension to this work, a series of nanoparticles sta-
bilized by poly(NVP-co-VBIMCl) were evaluated as catalysts
for benzene hydrogenation in ILs under various conditions (Ta-
ble S1 in the supporting information). Various parameters were
explored, including the nature of the metal, the IL used, and the
preparation method of the nanoparticles. In this way, the activ-
ity of the nanoparticles was optimized. It was found that the
Rh0 nanoparticles were considerably more active than the Ru0
and Pd0 nanoparticles in the ILs (Table S1, entries 1–11).
The rhodium nanoparticles protected by the IL-like copoly-
mer were also found to be active catalysts in the hydrogenation
of other arenas, and because these nanoparticles can endure
comparative forcing conditions, faster reaction rates and higher
conversions were obtained for the hydrogenation of arenes
compared with those catalyzed by nanoparticles protected only
by imidazolium ILs (see Table 1). Furthermore, the products
could be separated from the IL (catalyst) phase by decantation,
and loss of rhodium was below the detection limit of the ICP
instrument (<1 µg/ml). The Rh nanoparticles immobilized in
ILs are excellent catalysts for arene hydrogenation reactions ir-
respective of the presence of functional groups. In the present
study, only arenes with oxygen-containing functional groups
were examined, because these substrates correspond to model
lignin compounds (the hydrogenation of lignins is important
in paper production [26] and the production of fuels from bio-
mass [27]). As described elsewhere [28], the TOF for benzene
hydrogenation is higher than that of arenes under the same reac-
tion conditions, with conversions and TOFs of alkyl derivatives
decreasing with increasing size or number of substitutes. This
is believed to be due to a balance between the coordination
properties of a particular arene (a prerequisite for hydrogena-
tion) and electronic stabilization/destabilization of the aromatic
character by the substituent groups, with the two effects often
working against each other. The hydrogenolysis product ob-
served from the hydrogenation of anisole is negligible under the
given conditions ([substrate]/[rhodium] = 2000 mol/mol). Re-
markably, whereas simple arene substrates yield the expected
hydrogenated product with >99% selectivity, reduction of cer-
tain substrates results in the formation of significant quantities
of partially hydrogenated products. (See Tables 2 and S2–S6
in supporting information for details of reaction optimization.)
The unsaturated bond in the partially hydrogenated products
is, as expected, stabilized by electron-donating substituents at-
tached to the aromatic ring. Previously, it has been shown
that substrates such as 1,2,4,5-tetramethylbenzene and anisole
may be converted into their corresponding monoenes by Rh0
nanoparticles in aqueous phases [29,30]. In addition, partial
hydrogenation of linear and cyclic dienes has previously been
accomplished in IL biphases using molecular catalysts [31–33]
and soluble nanoparticles [7,8]. It was shown that the increased
levels of partially hydrogenated products were due in part to the
higher solubility of the diene relative to the monoene in the IL,
which results in its spontaneous elimination from the catalyst
phase [34]. Partial reduction of benzene also has been reported
using nanoparticle catalysts in ILs [35,36]. Because far more
active catalysts are required to hydrogenate arenes compared
with alkenes, by virtue of the need to break the aromaticity,
2.4.2. Experiment 2. Hydrogenation with addition of mercury
The reaction procedure was the same as described in ex-
periment 1, except, the reaction was stopped when conversion
reached 50% and mercury (3.21 g, 16 mmol, 1000 equiv.) was
added to the autoclave. Then the mixture was stirred for 1 h un-
der protection of nitrogen, the reactor was repressurized with
H2 (40 bar), and the reaction continued. A graph showing the
effect of mercury poisoning experiments is given in Fig. S3.
2.5. Characterization
2.5.1. ICP measurements
Rhodium loss after catalytic reactions was measured by ICP.
In this procedure, the product phase was mixed with 40 ml of
aqua regia and refluxed at 100 ◦C for 18 h. The resulting so-
lution was diluted to 50 ml with water and used for the ICP
measurement (PROFILE SPEC, LEEMAN LABS, detection
limit ∼µg–mg/ml).
2.5.2. High-resolution transmission electron microscopy
(HRTEM)
Transmission electron microscopic micrographs were re-
corded on a Hitachi H-9000 HRTEM at 300 keV. Rh nanoparti-
cles immobilized in [BMIM][BF4] were diluted in methanol.
Samples were prepared by the dropwise addition of the Rh
nanoparticle suspension onto Cu grids coated with holey car-
bon. More than 200 particles were counted to determine the
size distribution.
2.5.3. Fourier-transform infrared spectroscopy (FTIR)
A flow of 1 bar CO was bubbled into [BMIM][BF4] (1 ml)
solution containing Rh nanoparticles for 2 h. The solution was
then spread into a liquid film on KBr windows, and IR spectra
were recorded immediately on a Bruker Tensor 27 Fourier-
transform infrared spectrometer with a resolution of 1 cm−1 at
20 ◦C.
3. Results and discussion
3.1. Hydrogenation of benzene and arene derivatives
As reported previously, the rhodium nanoparticles stabilized
by an IL-like copolymer [poly(NVP-co-VBIMCl)] showed un-
precedented lifetime and activity for benzene hydrogenation
under relatively mild conditions with a TTO of 20,000 (from
five recyclings of 4000 TTOs per batch) and TOFs exceeding
200 h−1 [12]. It was shown that the Rh0 nanoparticles could
be used at least 5 times without deterioration in activity (i.e.,
quantitative conversion under the conditions used) for benzene
hydrogenation, demonstrating that the combination of ILs with
IL-like stabilizers represents an excellent strategy for providing
highly stable, catalytically active nanoparticles.

36
C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
Table 1
a
b
Hydrogenation of arenes by IL-like copolymer stabilized rhodium nanoparticles in [BMIM][BF ]
4
e
Entry
Substrate
Conversion
(%)
Product
(%)
Product
(%)
Product
(%)
TOF
1
2
3
c
d
−1
(h )
1
96
95
82
22
160
158
137
130
100%
>99%
>99%
81%
2
3
4
<1%
<1%
19%
5
37
62
62%
38%
29%
<1%
26%
6
7
8
37
84
15
247
140
100
71%
>99%
74%
9
33
55
68%
88%
90%
16%
6%
16%
6%
10
11
12
42
41
11
70
68
25
10%
32%
49%
19%
f
13
21
70
71%
29%
a
−5
◦
−1
Conditions: 1.6 × 10 mol Rh in 6 ml [BMIM][BF ], [S]/[Rh] = 2000, 40 atm H , 75 C, stirred at 800 min for 12 h.
4
2
b
c
d
e
f
Poly(NVP-co-VBIMCl) (M = 50,400, M = 75,300), IL-copolymer/Rh = 5:1 (mol/mol).
n
w
Determined by GC analysis.
Product proportions.
TOF based on mole substrate conversion per total mol metal per hour during the first 10 h.
[S]/[Rh] = 1000.

C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
37
Table 2
are hydrogenated further. Such solvent effects also influence
the conversion and turnover frequency, because mass transport
between the phases becomes important. It is noteworthy that
whereas hydrogen gas solubility is low in ILs, the rate of mass
transfer of hydrogen into ILs appears to be very fast [37] and is
unlikely to be rate-limiting under the reaction conditions.
a
Characterization of copolymers
b
c
Entry
M
M
M /M
w
n
w
n
1
2
3
17,900
25,000
50,400
21,500
32,000
75,300
1.20
1.28
1.49
a
The concentration of NVP (mol%) in the poly(NVP-co-VBIMCl) is 57%,
which was determined by NMR with an average molecular weight of the re-
peating unit of 143.6 g/mol.
3.2. Characterization and influence of the copolymers
b
Numeral-average molecular weight.
Weight-average molecular weight.
c
The method used to prepare the IL-like copolymers is il-
lustrated in Scheme S1 in supporting information. Two IL-
like copolymers, poly[NVP-co-VBIM]Cl and poly[NVP-co-
VEIM]Br, and the homopolymer (poly[VBIM]Cl) (Figs. 1
and 2) were prepared for comparison purposes. The polymer
concentration and ratio of [copolymer]/[Rh] (with the mol
amount of polymer calculated by the amount of equivalent
monomer) were 1.33 × 10−5 mol/ml and 5:1 mol/mol, respec-
tively. The numeral-average molecular weight of the polymers
suppressing further hydrogenation is a considerable challenge
and results in somewhat lower selectivity toward the partial
hydrogenation product compared with dienes; for example, cy-
clohexadiene can be converted to cyclohexene with a selectivity
of 96% with high conversion [31]. It is interesting to note that
dienes are not observed, in keeping with the solubility effect de-
scribed; that is, the dienes are soluble in the IL phase and thus
(a)
(b)
(c)
Fig. 1. Structure of (a) poly[(N-vinyl-2-pyrrolidone)-co-(1-vinyl-3-butylimidazolium chloride)], abbreviated as poly[NVP-co-VBIM]Cl; (b) poly[(N-vinyl-
2-pyrrolidone)-co-(1-vinyl-3-ethylimidazolium bromide)], abbreviated as poly[NVP-co-VEIM]Br, (c) poly(1-vinyl-3-butylimidazolium chloride), abbreviated as
poly[VBIM]Cl.
1
Fig. 2. H NMR spectrum of poly[VBIM]Cl, spectra of poly[NVP-co-VBIM]Cl and poly[NVP-co-VEIM]Br are shown in Figs. S1 and S2. Solvent D O.
2

38
C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
(a)
(b)
(c)
Fig. 3. (a) TEM micrograph of rhodium nanoparticles protected by ionic copolymer (from entry 3 in Table 2) in [BMIM][BF ] before catalysis (scale bar = 10 nm).
4
(b) Histogram showing the particle size distribution. (c) TEM micrograph of rhodium nanoparticles protected by ionic copolymer in [BMIM][BF ] after hydrogena-
4
tion of p-xylene (scale bar = 20 nm).
(Mn) was approximately 100,000, and the NVP concentration
was ca. 57 mol% in the copolymers. RhCl3 was reduced in
30 min at 60 ◦C in the presence of the IL-like copolymers
but could not be reduced even at 75 ◦C for several hours in
the presence of homopolymer, indicating the superior stabi-
lization effect of IL-like copolymer compared with the simple
homopolymer.
The effect of two IL-like copolymer stabilizers on the
rhodium nanoparticles was compared in the hydrogenation of
benzene in [BMIM][BF4] under the same conditions (75 ◦C,
40 bar H2, for 10 h). Conversions of 46% with a TOF of
115 h−1 and 99% with a TOF of 243 h−1 were obtained us-
ing the poly[NVP-co-VBIM]Br- and poly[NVP-co-VBIM]Cl-
stabilized nanoparticles, respectively. This difference in activity
is almost certainly due to the poisoning effect of the Br− anions
on nanoparticles, which, as noted elsewhere [38], leads to re-
duced catalytic activity.
(see Figs. S5 and S6 in supporting information). Benzene
was hydrogenated in [BMIM][BF4] containing Rh nanoparti-
cles stabilized by various ratios of poly[NVP-co-VBIM]Cl and
an optimized [copolymer]/[Rh] ratio of 5:1 mol/mol afforded
rhodium nanoparticles that achieved a TTO of 20,000 with
a TOF of 250 h−1. In contrast, a [copolymer]/[Rh] ratio of
1:1 mol/mol afforded nanoparticle catalysts that maintained ac-
tivity for only 15 min with a very low TOF of ca. 5 h−1. More-
over, the catalytic activity of the Rh nanoparticles decreased by
20% when the [copolymer]/[Rh] was raised to 10:1 mol/mol,
reflecting the need for an optimum balance between the stability
and activity of the nanocatalyst. Rhodium aggregates prepared
in the absence of a copolymer additive showed no activity in
the reduction of benzene.
The polydispersity of the three poly(NVP-co-VBIM)Cl
copolymers with different molecular weights [weight-average
molecular weight (Mw)/numeral-average molecular weight
(Mn)] ranged from 1.20 to 1.49 (Table 2). All of the copolymers
effectively stabilized the nanoparticles in the hydrogenation of
The copolymer metal ratio also affected the ability of the
rhodium nanoparticles to catalyze the hydrogenation reaction

C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
39
Table 3
Infrared data of carbon monoxide adsorbed on IL-like copolymer stabilized
rhodium nanoparticles before and after catalysis in [BMIM][BF ]
only in low yield. Completely deactivated catalysts were ob-
tained from benzene hydrogenation after several runs, and par-
ticles from both the upper and lower phases were examined
using the CO probe (Table 3, entries 5 and 6, respectively). Inte-
gration of terminal to bridged (T/B) adsorption was compared
with fresh nanoparticle catalysts (Table 3, entry 1), the group
of partially deactivated catalysts (Table 3, entries 2–4), and the
completely deactivated catalysts (Table 3, entries 5 and 6). Ta-
ble 3 clearly shows that the carbonyls bound to newly prepared
rhodium nanoparticles had the highest terminal to bridged ratio
of 1.95, which steadily decreased to 0.66, with the stepwise de-
activation corresponding to a reduction in active reaction sites.
a
4
b
Entry Rhodium
nanoparticles
Fresh
Terminal CO
Bridged CO
T/B
−1
−1
)
ν
C=O
(cm
)
ν
C=O
(cm
1
2
2115
2115
1819
1817
1.95
1.66
After hydrogenation
of m-xylene
3
4
5
After hydrogenation
of p-xylene
2117
2119
1817
1818
1817
1.81
1.83
1.09
After hydrogenation
of isobutylbenzene
c
Deactivated (after hydro- 2121
genation of benzene—
organic phase)
3.4. Mercury poisoning experiments
Mercury poisoning experiments are frequently performed to
ascertain whether catalysts are truly homogeneous or heteroge-
neous (including soluble nanoparticle catalysts) [40]. Mercury
poisoning experiments applied here revealed that hydrogen was
no longer consumed after mercury was added to the reac-
tion, demonstrating that the catalysts are heterogeneous (see
Fig. S3).
d
a
6
Deactivated (after hydro- 2119
genation of benzene—
ionic liquid phase)
1817
0.66
Poly(NVP-co-VBIMCl) (M = 50,400, M = 75,300), IL-copolymer/
n
w
Rh = 5:1 (mol/mol).
b
Integration ratio of terminal per bridged adsorptions.
Upper layer of catalytic phase.
Lower layer of catalytic phase.
c
d
4. Conclusion
benzene and toluene with no appreciable differences in activity
(see Table S7 and Fig. S4). The copolymer with Mn = 50,400
(entry 3 in Table 2) gave slightly higher conversions for benzene
and toluene and thus were selected for evaluation in the hydro-
genation of a range of arenes.
Rhodium nanoparticles protected by IL-like copolymers
are highly active catalysts for the hydrogenation of arenes.
The nanoparticle catalysts protected by the IL-like copolymer
[poly(NVP-co-VBIMCl)] immobilized in ILs can endure forc-
ing reaction conditions, resulting in high reaction rates and high
conversions. The solubility of the substrates in the reaction me-
dia and the steric/electronic properties of the substituents on
the aromatic ring influence the rate of catalytic hydrogenation
and the nature of the reaction product, that is, partially hydro-
genated versus fully hydrogenated systems.
Our findings also demonstrate the delicate balance between
activity and stability of IL-like stabilized nanoparticles. IR
spectroscopy showed a strong correlation between the number
of CO ligands chemisorbed on the rhodium nanoparticles and
catalytic activity; that is, a decrease in ratio of the terminal to
bridged adsorptions occurs as the catalyst is deactivated, which
corresponds to a decrease in available reaction sites.
3.3. Characterization of the nanoparticle catalysts by TEM
and FT-IR spectroscopic
The rhodium nanoparticles were analyzed by TEM and
found to have a mean diameter of 3.0 nm and a standard de-
viation of 0.8 nm (Fig. 3). Because nanoparticle catalysts are
usually not uniform, with the particles observed in the TEM
representing only part of the system with respect to dispersity
and size, the nanoparticles were studied in solution by infrared
spectroscopy using CO as a probe.
As nanoparticles aggregate or increase in size, their surfaces
undergo a change in geometry. Using CO as a probe, which
coordinates with the vacant sites of the nanoparticles, allows
tracking of such changes. IR spectra of CO chemisorbed on
copolymer-stabilized rhodium nanoparticles in [BMIM][BF4]
(prepared by merely passing CO gas through the solution for
2 h) exhibited two main peaks centered at ca. 2115–2121 and
1817–1819 cm−1 (see Table 3). The peaks in the range 2115–
2121 cm−1 can be assigned to terminally bound CO on the
rhodium surface, and the peaks in the range 1817–1819 cm−1
are characteristic of bridging CO ligands [39].
Acknowledgments
This work was supported by the National Science Foun-
dation of China (projects nos. 20533010 and 20473002). The
authors thank Professor Zi-Chen Li and Dr. Yong-Quan Dong,
Department of Polymer Science and Engineering, College
of Chemistry and Molecular Engineering, Peking University,
for supplying the poly(NVP-co-VBIMCl) sample used in this
study.
The relative ratio of terminal and bridged CO ligands on
rhodium nanoparticles was estimated before and after catalysis
to examine the potential coordination sites provided by the cat-
alysts. We speculated that the rhodium catalysts were partially
deactivated during the hydrogenation of m-xylene, p-xylene,
and isobutylbenzene, because these substrates were converted
Supporting information
Supporting information for this article may be found on Sci-
enceDirect, in the online version.
Please visit DOI: 10.1016/j.jcat.2007.05.014.

40
C. Zhao et al. / Journal of Catalysis 250 (2007) 33–40
References
[22] L. Vieille-Petit, G. Süss-Fink, B. Therrien, T.R. Ward, H. Stoeckli-Evans,
G. Labat, L. Karmazin-Brelot, A. Neels, T. Bürgi, R.G. Finke, C.M. Ha-
gen, Organometallics 24 (2005) 6104.
[23] T.J. Donohoe, R. Garg, C.A. Stevenson, Tetrahedron: Asymmetry 7 (1996)
317.
[24] H. Nagahara, Rev. J. Surf. Sci. Technol. Avant-Garde 30 (1992) 951.
[25] X.D. Mu, D.G. Evans, Y. Kou, Catal. Lett. 97 (2004) 151.
[26] B.R. James, Y. Wang, T.Q. Hu, Chem. Ind. 68 (1996) 423.
[27] L. Petrus, M.A. Noordermeer, Green Chem. 8 (2006) 861.
[28] P.J. Dyson, Dalton Trans. (2003) 2964.
[1] R.G. Finke, J.A. Widegren, J. Mol. Catal. A Chem. 191 (2003) 187.
[2] D.B. Zhao, M. Wu, Y. Kou, E.Z. Min, Catal. Today 74 (2002) 157.
[3] P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed. 39 (2000) 3772.
[4] J. Dupont, R.F. de Souza, P.A.Z. Suarez, Chem. Rev. 102 (2002) 3667.
[5] J. Dupont, P.A.Z. Suarez, Phys. Chem. Chem. Phys. 8 (2006) 2441.
[6] P. Migowski, J. Dupont, Chem. Eur. J. 13 (2007) 32.
[7] J. Dupont, G.S. Fonseca, A.P. Umpierre, P.F.P. Fichtner, S.R. Teixeira,
J. Am. Chem. Soc. 124 (2002) 4228.
[8] L.S. Ott, R.G. Finke, Inorg. Chem. 45 (2006) 8382.
[9] Y. Wang, N. Toshima, J. Phys. Chem. B 101 (1997) 5301.
[10] K.S. Weddle, J.D. Aiken III, R.G. Finke, J. Am. Chem. Soc. 120 (1998)
5653.
[11] J.D. Aiken, R.G. Finke, J. Am. Chem. Soc. 121 (1999) 8803.
[12] X.D. Mu, J.Q. Meng, Z.C. Li, Y. Kou, J. Am. Chem. Soc. 127 (2005) 9694.
[13] A. Stanislaus, B.H. Cooper, Catal. Rev. Sci. Eng. 36 (1994) 75.
[14] P. Giannoccaro, M. Gargano, A. Fanizzi, C. Ferragina, M. Aresta, Appl.
Catal. A Gen. 284 (2005) 77.
[15] M.S. Eisen, T.J. Marks, J. Am. Chem. Soc. 114 (1992) 10358.
[16] H. Gao, R.J. Angelici, J. Am. Chem. Soc. 119 (1997) 6937.
[17] C. Bianchini, V. Dal Santo, A. Meli, S. Moneti, M. Moreno, W. Ober-
hauser, R. Psaro, L. Sordelli, F. Vizza, Angew. Chem. Int. Ed. 42 (2003)
2636.
[29] J.A. Widegren, R.G. Finke, Inorg. Chem. 41 (2002) 1558.
[30] J. Blum, I. Amer, K.P.C. Vollhardt, H. Schwartz, G. Hohne, J. Org. Synth.
52 (1987) 2804.
[31] Y. Chauvin, L. Missmann, H. Olivier, Angew. Chem. Int. Ed. Engl. 34
(1996) 2698.
[32] P.A.Z. Suarez, J.E.L. Dullius, S. Einloft, R.F. de Souza, J. Dupont, Inorg.
Chim. Acta 255 (1997) 207.
[33] D. Zhao, P.J. Dyson, G. Laurenczy, J.S. McIndoe, J. Mol. Catal. A
Chem. 214 (2004) 19.
[34] P.J. Dyson, Appl. Organomet. Chem. 16 (2002) 495.
[35] J.B. Ning, J. Xu, J. Liu, F. Lu, Catal. Lett. 109 (2006) 175.
[36] G.S. Fonseca, A.P. Umpierre, P.F.P. Fichtner, S.R. Teixeira, J. Dupont,
Chem. Eur. J. 9 (2003) 3263.
[37] P.J. Dyson, G. Laurenczy, C.A. Ohlin, J. Vallance, T. Welton, Chem. Com-
mun. (2003) 2418.
[18] K.H. Park, K. Jang, H.J. Kim, S.U. Son, Angew. Chem. Int. Ed. 46 (2007)
1152.
[38] J.D. Aiken III, Y. Lin, R.G. Finke, J. Mol. Catal. A Chem. 114 (1996) 29.
[39] A. Duteil, R. Quéau, B. Chaudret, Chem. Mater. 5 (1993) 341.
[40] J.A. Widegren, M.A. Bennett, R.G. Finke, J. Am. Chem. Soc. 125 (2003)
10301.
[19] L.N. Lewis, Chem. Rev. 93 (1993) 2693.
[20] J. Schulz, A. Roucoux, H. Patin, Chem. Eur. J. 6 (2000) 618.
[21] C.M. Hagen, L. Vieille-Petit, G. Laurenczy, G. Süss-Fink, R.G. Finke,
Organometallics 24 (2005) 1819.