Ruthenium nanoparticles supported on multi-walled carbon nanotubes: Highly effective catalytic system for hydrogenation processes

Article abstract of DOI:10.1016/j.molcata.2010.09.006

Immobilization of small and homogeneously dispersed ruthenium nanoparticles stabilized by 4-(3-phenylpropyl)pyridine ligand (RuL) on functionalized multi-walled carbon nanotubes (RuL-MWCNT) have been prepared and characterized by elemental analysis and transmission electronic microscopy. A comparative hydrogenation study of unsaturated substrates using non-supported (RuL) and supported catalytic systems (RuL-MWCNT) was carried out. For all the substrates, the activity of the supported catalyst was higher towards the full hydrogenated product than that obtained using the non-supported one. Moreover, the catalytic effect of the support nature was studied using RuL immobilized on silica (RuL-SiO₂), alumina (RuL-Al₂O₃) and activated carbon (RuL-AC). The best activities and selectivities were found for RuL-MWCNT system, maintaining its catalytic behaviour upon recycling.

Full text of DOI:10.1016/j.molcata.2010.09.006

Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Ruthenium nanoparticles supported on multi-walled carbon nanotubes:
Highly effective catalytic system for hydrogenation processes
Mohamad Jahjaha, Yolande Kihnb, Emmanuelle Teumaa, Montserrat Gómeza,∗
a Université de Toulouse, UPS, LHFA, 118 route de Narbonne, F-31062 Toulouse, France, CNRS,Bât 2R1,2ème ètage, LHFA UMR 5069, F-31062 Toulouse, France
b CEMES, UPR 8011 CNRS, 29 rue Jeanne Marvig 31055 Toulouse cedex 4, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 29 April 2010
Received in revised form 31 August 2010
Accepted 4 September 2010
Available online 15 September 2010
Immobilization of small and homogeneously dispersed ruthenium nanoparticles stabilized by 4-(3-
phenylpropyl)pyridine ligand (RuL) on functionalized multi-walled carbon nanotubes (RuL-MWCNT)
have been prepared and characterized by elemental analysis and transmission electronic microscopy. A
comparative hydrogenation study of unsaturated substrates using non-supported (RuL) and supported
catalytic systems (RuL–MWCNT) was carried out. For all the substrates, the activity of the supported
catalyst was higher towards the full hydrogenated product than that obtained using the non-supported
one. Moreover, the catalytic effect of the support nature was studied using RuL immobilized on silica
(RuL–SiO2), alumina (RuL–Al2O3) and activated carbon (RuL–AC). The best activities and selectivities
were found for RuL–MWCNT system, maintaining its catalytic behaviour upon recycling.
© 2010 Elsevier B.V. All rights reserved.
Keywords:
Ruthenium
Nanoparticles
Multi-walled carbon nanotubes
Hydrogenation
Recycling
1. Introduction
have been found as appropriated catalyst support due to their
significant advantages such as: (i) high catalytic activities arising
The use of metallic nanoparticles as catalysts has undergone
an exponential growth during the last decade [1–5]. In relation to
the preparation of these nano-materials from salts or molecular
species (“bottom up” synthetic strategy) [6], their agglomeration
should be avoided using stabilizers (polymers, dendrimers, ionic
liquids, ligands etc.) [7]. On the other hand, these compounds must
allow the presence of active “free” surface sites in order to enhance
the catalytic activity, preventing their plausible poison effect by
saturation of the metallic surface [8]. This kind of nanoparticles
dispersed in a liquid phase (organic, ionic liquid or aqueous solu-
tion) could exhibit an analogous behaviour to that observed using
classical homogeneous catalytic systems with regard to their recy-
cling [9,10]. Consequently, immobilization of nano-catalysts has
been developed, using biphasic systems (e.g. ionic liquids) [11–13]
or solid supports [14–17]. In particular, metal nanoparticles sup-
ported on a solid (commonly, carbon, inorganic oxides –such as
silica, alumina, titanium dioxide – and polymeric materials) have
been applied in different catalytic transformations, being easily
separated from the organic product and therefore recovered to
be reused [18]. It is important to note that the catalytic activity
and/or selectivity could be modified depending on the support
nature. More recently, multi-walled carbon nanotubes (MWCNT)
from its mesoporous nature, avoiding mass transfer limitations
[19]; (ii) high activity/selectivity due to the possibility to oper-
ate in the inner cavity of carbon nanotubes (confinement effect)
[20]; (iii) well-defined and tuneable structure [21]; and (iv) selec-
tive growth of nanocarbons by catalytic chemical vapour deposition
(C-CVD) on defined substrates to build micro-reactors [22]. In par-
ticular, metallic nanoparticles supported on MWCNT have been
widely employed in the last years in hydrogenation reactions of
organic compounds [23]. Palladium nanoparticles supported on
MWCNT are the catalyst of choice for the hydrogenation of C
C
bonds, including aromatic moieties and C C bonds [24]. Bimetal-
lic Pd–Rh nanoparticles supported on carbon nanotubes catalyze
effectively benzene hydrogenation at room temperature [25]. An
advanced study concerning bimetallic Pt–Pd catalysts supported on
MWCNT has evidenced a very good ability to hydrogenate aromatic
rings in contrast to other supports [26]. With ruthenium catalysts,
influence of support nature has been also proved. The reaction
of three types of carbon nanofibers (CNF) has given the corre-
sponding CNF-supported ruthenium nanoparticles and depending
on the nanostructures surface of the CNF, excellent catalytic activity
towards the hydrogenation of toluene has been obtained without
leaching of ruthenium species, leading to a recyclable catalyst [27].
In the hydrogenation of glucose to sorbitol, ruthenium catalysts
immobilized on MWCNT also showed higher catalytic activity than
that obtained with the corresponding Ru catalysts immobilized on
Al2O3 or SiO2 [28].
∗
Corresponding author. Tel.: +33 561557738; fax: +33 561558204.
E-mail address: gomez@chimie.ups-tlse.fr (M. Gómez).
1381-1169/$ – see front matter © 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.molcata.2010.09.006

M. Jahjah et al. / Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
107
In a recent work, we have prepared MWCNT grafted by ionic
liquids, which could be further utilized as supports of ionic liquid
catalytic phase (SILCP), using a molecular rhodium precursor [29].
The resulting rhodium catalyst exhibited a significant higher activ-
ity than the corresponding SILCP supported on oxide supports and
activated carbon, in the hydrogenation of 1-hexene. This catalytic
behaviour could be attributed to a good dispersion of the MWCNT
support in the reaction mixture that minimizes the negative effects
of mass transport.
ature overnight. The solvent was evaporated under vacuum. The
isolated particles were further washed with pentane (3 × 10 mL).
Organic phase was concentrated and analysed by 1H NMR proving
the absence of free ligand. The black solid was dried under reduced
pressure. Yield: 93% (26.0 mg). Mean diameter = 1.31 0.43 nm. IR
absorptions ꢀmax/cm−1 2917 and 2841 (C–H), 1612 and 1550 (C N),
1402 and 1384 (C C), 1258, 1089, 1019 and 798 (CO and CC of the
THF).
2.3. Supported ruthenium nanoparticles
In this paper, we describe the immobilization of ruthenium
nanoparticles (RuL), preformed by organometallic compound
decomposition in the presence of a pyridine derivative stabilizer,
and onto different supports (functionalized multi-walled carbon
nanotubes, silica and activated carbon). These supported catalysts
were used in the catalytic hydrogenation of several unsaturated
substrates. A comparative catalytic study in terms of activity,
selectivity and recycling of supported and non-supported RuL
nanoparticles was purposely studied.
95 mg of support (MWCNT, SiO2, Al2O3 or AC) were added to
5 mg (0.032 mmol) of RuL dissolved in 1 mL of THF under Ar atmo-
sphere. The system was stirred under sonication for 2 h to form a
stable black suspension. The solvent was then evaporated under
vacuum. The resulting black solid was dried under reduced pres-
sure.
2.4. Hydrogenation catalytic processes
2. Experimental
1 mg (0.0064 mmol) of RuL or 5 mg of supported catalyst was
dissolved in 1 mL of degassed ethanol. After stirring under Ar atmo-
sphere at room temperature for 5 min, 0.64 mmol of the desired
substrate was added. The resulting solution was then transferred
to the autoclave, which was previously degassed three times with
argon and twice with hydrogen. The catalytic mixture was pres-
surized with hydrogen and stirred for the reaction time. At the end
of the reaction time, the reactor was degassed and the reaction
mixture filtered throught Celite before GC analysis.
2.1. General
Catalysts were prepared under Ar atmosphere conditions using
Schlenk and Fisher–Porter bottle techniques and stocked in glove-
box. THF was distilled over sodium benzophenone. Solvents were
degassed under vacuum at liquid nitrogen temperature by three
vacuum-argon cycles before being used. The metallic precursor
[Ru(cod)(cot)] and Ru/C were purchased from NanoMeps and
Aldrich, respectively. Ruthenium nanoparticles [30] and multi-
walled carbon nanotubes [31] were prepared following our
previously published methodologies. Al2O3 (Brockmann I, standard
grade, neutral, 150 mesh, pore volume = 0.89 cm3 g−1, Aldrich),
mesoporous silica (EP10X amorphous silica, Crosfield Ltd., pore
volume = 1.81 cm3 g−1, particle size ≈100 ␮m, INEOS Silicas) and
activated carbon (Sigma Aldrich) were used as supports. Other
commercial chemicals were used as supplied. All hydrogenation
tests were carried out in a stainless steel Top Industrie autoclave
(Top 45) of 100 mL volume and capacity of 70 bar pressure, with
magnetic stirring and electric heating collar. Elemental analyses
were performed in a Perkin Elmer 2400 series II microanalyzer in
the “Service d’Analyses du LCC” at the CNRS in Toulouse for car-
bon, nitrogen and hydrogen determinations. Ruthenium analysis
were carried out at the “Service Central d’Analyse” in Lyon by ICP
technique. Samples were dispersed in ethanol and sonicated for
5 min before dropped them onto a holey carbon copper grid and
the solvent was allowed to evaporate. TEM images were obtained
using a Philips CM12 microscope operating at 120 kV. GC analyses
were performed on a Hewlett-Packard 5890 Series II gas chromato-
graph (50 m Ultra 2 capillary column) with a FID detector. Substrate
conversions and products selectivity were determined by gas chro-
matography based on the relative area of GC-signals referred to an
internal standard, calibrated to the corresponding pure compound.
Enantiomeric excesses were determined by GC on FS-cyclodex-␤-
I/P and FS-cyclodex-␣-I/P columns.
3. Results and discussion
3.1. Synthesis and characterization of supported ruthenium
nanoparticles
The supported ruthenium nanoparticles were prepared in two
steps: first, preparation of ruthenium nanoparticles stabilized by
4-(3-phenylpropyl)pyridine (RuL); second, immobilization of these
particles on several solid supports (RuL-support).
The ruthenium nanoparticles RuL were synthesized by decom-
position of [Ru(cod)(cot)] in the presence of the ligand and under
hydrogen atmosphere, giving small (mean size: 1.31 0.43 nm,
Fig. 1a and b) and homogenously dispersed nanoparticles, as pre-
viously described (Scheme 1) [30]. It is important to note that full
decomposition of the organometallic precursor ([Ru(cod)(cot)]) is
achieved after ca. 18 h at room temperature under 3 bar of hydro-
gen, checked by NMR (organic solution analysis) in agreement with
the high yield to give RuL (more than 90% for each ruthenium
nanoparticles synthesis). In addition, kinetic measurements mon-
itored by 1H NMR under hydrogen pressure, evidenced the total
hydrogenation of coordinated olefins to give cyclooctane [30].
Once the particles were isolated, they were supported on com-
mercial supports (alumina, silica and activated carbon: RuL–Al2O3,
RuL–SiO2, RuL–AC) and also on functionalized multi-walled carbon
nanotubes (RuL–MWCNT). The MWCNT used in this work were
prepared by C-CVD [31] and further functionalized by nitric acid
treatment, giving mainly carboxylic acid groups but also carbonyl,
quinine and phenol groups on their surface [32]. Full charac-
terization of these MWCNT has been described in our previous
contribution [29]. The RuL immobilization on the support was car-
ried out by addition of ruthenium nanoparticles dispersed in THF
on the solid support, followed by ultrasound treatment. The solvent
was then removed under reduced pressure to isolate the material
as a black powder. For RuL–MWCNT, several loads of RuL in rela-
tion to the support were used (5, 10 or 40 mg of RuL per 100 mg of
MWCNT). For the commercial supports, 10 mg of RuL per 100 mg
2.2. Synthesis of ruthenium nanoparticles
The preparation of RuL was based on the work previously
reported [30]. A solution of 4-(3-phenylpropyl)-pyridine in THF
(2 mL of a solution 0.02 M of ligand in THF, 0.04 mmol) was intro-
duced in a Fischer–Porter bottle. The solvent was then evaporated
and a solution of [Ru(cod)(cot)] (60 mg, 0.2 mmol) in THF (80 mL)
was introduced under argon atmosphere. The system was next
pressurized with hydrogen (1 bar) and stirred at room temper-

108
M. Jahjah et al. / Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
Fig. 1. TEM micrographs of RuL (a) and their size distribution (dmean = 1.31 0.43 nm for 1417 counted particles) (b) and supported RuL–MWCNT materials for differ-
ent RuL weight loads (c–e): (c) 5% (dmean(RuL) = 2.01 0.39 nm for 30 counted particles); (d) 10% (dmean(RuL) = 1.87 0.29 nm for 30 counted particles); and (e) 40%
(dmean(RuL) = 2.09 0.39 nm for 150 counted particles) and the corresponding size distribution (f).
of support was only used. After supporting, the stabilizing ligand
L remained in the composition of the ruthenium nanoparticles, as
evidenced by elemental analysis (Table 1).
Concerning RuL–MWCNT systems, the ruthenium nanoparticles
did not exhibit important difference with the RuL load increase in
relation to the support (mean diameter ca. 2.0 nm), showing spher-
ical shapes and a quite homogeneous dispersion on the support
without aggregates formation (Fig. 1c and e); but their mean size
Table 1
Elemental analysis (weight percentage of Ru, C, H and N) of RuL-support materials.
RuL-supporta
Ru
C
H
N
RuL–MWCNT (5%)
RuL–MWCNT (10%)
RuL–MWCNT (40%)
RuL–Al2O3 (10%)
RuL–SiO2 (10%)
2.38
4.63
21.65
6.20
6.62
3.11
86.0
86.4
65.7
2.12
2.33
79.1
0.25
0.30
1.10
0.54
0.20
1.20
0.30
0.28
0.73
0.14
0.17
0.56
RuL–AC (10%)
a
In brackets, weight percentage of RuL in relation to the support.
Scheme 1. Synthesis of Ru nanoparticles stabilized by L, RuL.

M. Jahjah et al. / Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
109
Fig. 2. TEM micrographs of RuL–SiO2 (left) and RuL–Al2O3 (right).
is rather higher than that observed before immobilization (Fig. 1a
3.2. Catalytic hydrogenation
comparative hydrogenation study using supported
(RuL–MWCNT) and non-supported (RuL) catalysts was car-
ried out for several substrates (Scheme 2), in order to prove the
support influence in the catalytic reaction. The results are collected
in Table 2.
For the non-activated olefin methyl oleate (substrate 2), slight
differences were observed between both kind of catalysts (entries
3 and 4, Table 2). However, for the other functionalized substrates,
Ru-supported nanoparticles were definitively more active (entries
2, 6, 8, 10 and 12 versus 1, 5, 7, 9 and 11, respectively, Table 2).
In particular, for acetophenone (3) and styrene (6) hydrogena-
tion, total substrate conversion was obtained using both types of
catalytic systems (entries 5 and 6 and 11 and 12, Table 2), but for
the supported catalyst only the full hydrogenated product (3a and
6a) was observed (entries 6 and 12, Table 2), in contrast to the
behaviour of the non-supported catalyst which gave a mixture of
hydrogenated compounds. When p-methylanisole (5) was used as
substrate, RuL was clearly less active than RuL-MWCNT, giving only
42% substrate conversion after 18 h (entries 9 and 10, Table 2). As
and b), although the relative population remains distant. For the
nanocomposite containing 40% RuL load, two particles populations
could be distinguished at ca. 1.8 nm and 2.4 nm (Fig. 1e and f); for
the other RuL–MWCNT systems (5% and 10% RuL loads), few num-
bers of particles could be counted for the determination of their
mean size, giving in these two cases more uncertainty. Considering
a compact packing arrangement (as presented by the bulk metal)
of spherical nanoparticles, a cluster estimated composition of Ru85
(mean diameter ca. 1.3 nm) can be proposed for non-supported
nanoparticles RuL and of Ru309 (mean diameter ca. 2.0 nm) for
the supported ones (RuL–MWCNT). By means of “magic number”
approach [33], we can guess that up to 62% (for 3-shell nanoclus-
ters, 147 atoms) and 52% (for 4-shell nanoclusters, 309 atoms) of
metal atoms are placed on the metallic surface.
Ruthenium nanoparticles supported on silica and alumina
exhibited an important tendency to be agglomerated as observed
by the corresponding TEM micrographs (Fig. 2); for both sup-
ports, the mean diameter of RuL is close to that observed for the
nanoparticles supported on multi-walled carbon nanotubes (ca.
2 nm).
A
Scheme 2. Ru-catalyzed hydrogenation of unsaturated substrates (1-6) using non-supported (RuL) and supported (RuL-MWCNT) catalysts.

110
M. Jahjah et al. / Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
Table 2
Ru-catalyzed hydrogenation of 1–6 substrates, using non-supported (RuL) and MWCNT supported (RuL–MWCNT) precursors.a
Entry
Substrate
Catalyst
Time (h)
%Conv.b (Sel.)
TTOc
TONd
1
2
3
4
5
6
7
8
1
1
2
2
3
3
4
4
5
5
6
6
RuL
RuL–MWCNT
RuL
RuL–MWCNT
RuL
RuL–MWCNT
RuL
RuL–MWCNT
RuL
RuL–MWCNT
RuL
2.5
2.5
2
40
100
90
100
40
100
90
100
74.25
100
20
100
31.5
70
64.5
192
145
192
120
192
32
192
51
135
121
192
2
16
16
14
14
18
18
6
99 (75/15/9)e
100 (100/0/0)e
20
100
9
42 (75/25)f
100 (70/30)f
100 (75/25)g
100 (100/0)g
10
11
12
75
100
RuL–MWCNT
6
a
See Scheme 2 for the products labelling. Results obtained from duplicated tests. For each substrate, the reaction time was established monitoring the consumption of
hydrogen. Reaction conditions: 0.64 mmol of substrate, 0.0064 mmol of RuL, 40 bar hydrogen pressure, 50 ◦C.
b
Conversions and selectivities determined by GC.
Total Turnover = moles of substrate converted to full hydrogenated product per mole of catalyst (based on total metal).
Turnover = moles of substrate converted to full hydrogenated product per mole of catalyst (based on ruthenium atoms on the surface, see above in the text for its
c
d
determination).
e
Chemoselectivity towards the hydrogenated products as a ratio of 3a/3b/3c.
Chemoselectivity towards the hydrogenated products as a ratio of 5a/5b.
Chemoselectivity towards the hydrogenated products as a ratio of 6a/6b.
f
g
expected, no reduction of the ester group was observed for the
substrates 1 and 2.
In order to study the relative rate to hydrogenate the exocyclic
double bond and the aromatic group, the styrene hydrogenation
was examined under smooth conditions. At room temperature and
under 1 bar hydrogen pressure for a substrate/ruthenium ratio of
100, the styrene hydrogenation monitoring showed the exclusive
formation of ethylbenzene (6b) during the first twenty minutes of
reaction; once the substrate was completely converted into ethyl-
benzene, the hydrogenation of the aromatic ring took place, giving
exclusively ethylcyclohexane after one hour of reaction (Fig. 4).
This fact is due to the different hydrogenation rate between the
exocyclic double bond and the aromatic ring, being faster the
hydrogenation of the olefinic function.
Effectively, these results evidence a higher catalyst activity
towards the hydrogenation of the exocyclic olefin (TOF = 577 h−1
[35]) than that corresponding to the reduction of the aromatic ring
(TOF = 288 h−1 [35]), different enough to permit selectively the for-
mation of hydrogenated products, 6a and 6b.
Therefore, the total turnover numbers (based on the total metal
amount) towards the full hydrogenated products are up to 100 for
RuL–MWCNT and 75 for RuL (except for the non-activated substrate
2, entry 3, Table 2), for a constant ruthenium load. Because of only
the ruthenium atoms placed on the surface are available for cat-
alytic reactivity, these relative activities represent the lower limit
and the differences between both kind of catalysts are higher taking
into account the exposed ruthenium atoms; therefore, the turnover
numbers (based on the ruthenium atoms on the surface) are found
in the range of 135–192 for RuL–MWCNT and 32–145 for RuL. The
catalytic activity enhancement of the supported (RuL–MWCNT)
system in relation to the unsupported one (RuL) can be due to the
substrate adsorption not only at the metal surface but also at the
interface between the metal and the catalyst support as suggested
by several research groups [34].
With the purpose of evaluating the chemoselectivity using
RuL–MWCNT as catalyst, temperature, hydrogen pressure and
substrate/metal ratio parameters were optimized for the styrene
hydrogenation. RuL–MWCNT was still active under smooth condi-
tions, at room temperature and low hydrogen pressures (even at
1 bar). Concerning the hydrogen pressure, full substrate conver-
sion was also attained at room temperature under 25 bar hydrogen
after three hours of reaction, with exclusive formation of ethylcy-
clohexane (6a). At lower hydrogen pressures (Fig. 3), the catalytic
system remained active with total consumption of styrene using
1 bar of hydrogen, but affording a mixture of ethylbenzene and
ethylcyclohexane, 6b/6a = 70/30, indicating that the catalytic sys-
tem is less active under smooth conditions, favouring the faster
hydrogenation that corresponds to the exocyclic C–C double bond.
The pressure increase favoured the formation of the full hydro-
genated product (6a), obtaining essentially ethylcyclohexane under
20 bar. These facts are in agreement with a low hydrogen solubil-
ity working at low pressures which leads to a low hydrogenation
rate of styrene and in consequence the formation of ethylbenzene
is favoured at 1 bar of hydrogen, in contrast to the results obtained
working at higher pressure (20 bar), only obtaining the full hydro-
genated product (ethylcyclohexane).
At constant substrate/ruthenium ratio (3 0 0), an increase of
ruthenium mass in relation to the support (5, 10 or 40 mg of RuL
per 100 mg of MWCNT) did not improve the activity.
At room temperature and under 25 bar hydrogen pressure,
substrate/metal ratio could be increased up to 1,000 to give full
conversion of substrate with a 6b/6a ratio of 60/40 after 3 h of
reaction.
Fig. 3. Chemoselectivity for the RuL–MWCNT styrene hydrogenation, depending on
the hydrogen pressure at full conversion of substrate. In black, relative percentage
of ethylbenzene (6b) and in gray, ethylcyclohexane (6a).

M. Jahjah et al. / Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
111
Table 4
Ru-catalyzed hydrogenation of styrene, using RuL–MWCNT and RuL–SiO2 supported
precursors.a
Entry
Catalyst
Styrene/Ru ratio
Run
%Conv.b
6a/6b ratiob
1
2
3
4
5
6
RuL–MWCNT
RuL–MWCNT
RuL–MWCNT
RuL–SiO2
RuL–SiO2
RuL–SiO2
300
300
300
200
200
200
1st
100
100
100
100
100
100
100/0
100/0
100/0
100/0
92/8
2nd
3rd
1st
2nd
3rd
40/60
a
See Scheme 2 for the products labelling. Results obtained from duplicated tests.
Reaction conditions: 1 mmol of substrate, 25 bar hydrogen pressure, rt, 3 h.
b
Substrate conversion and hydrogenated products ratio determined by GC.
for both silica and multi-walled carbon nanotubes supports (see
Supplementary material), in agreement with the lower activity of
RuL–Al2O3 catalyst compared with RuL–MWCNT and RuL–SiO2.
When commercial Ru/C was used for the hydrogenation of p-
methylanisole, the conversion was low (entry 13, Table 3) and for
the styrene hydrogenation, only ethylbenzene was formed (entry
14, Table 3), showing lower performance of the classical heteroge-
neous catalyst than the supported ruthenium nanoparticles, RuL.
Contrary to RuL–MWCNT, RuL–SiO2 catalyst was not active
for the styrene hydrogenation under smooth conditions (room
temperature and under 1 bar hydrogen pressure for a sub-
strate/ruthenium ratio of 100). Hydrogenation took place at
room temperature but at 25 bar hydrogen pressure with a sub-
strate/metal ratio of 200. Three consecutive additions of styrene
were carried out under these conditions. A deactivation of the cat-
alytic system was observed after the first run (entries 4–6, Table 4)
by a decrease of the ethylcyclohexane/ethylbenzene selectivity,
in contrast to the robustness of the catalytic system supported
on multi-walled carbon nanotubes, which maintained its catalytic
activity after three consecutive runs (entries 1–3, Table 4). Taking
into account that the ruthenium metal content on the organic prod-
ucts is in all cases less than 2 ppm (determined by ICP–MS) and that
the ligand used as stabilizer was not detected (checked by 1H NMR),
this different behaviour could be associated to an agglomeration of
the nanoparticles in the case of the catalyst supported on silica.
Fig. 4. RuL–MWCNT-catalyzed styrene hydrogenation monitoring.
In order to evaluate the effect of the support nature in the hydro-
genation of C C bonds, the ruthenium nanoparticles RuL were
immobilized on silica, alumina and activated carbon. These mate-
rials were tested in the hydrogenation of substrates 3, 5 and 6
(Table 3). The catalytic system supported on AC gave low hydro-
genation rate for the all substrates tested (entries 4, 8 and 12,
Table 3). This behaviour can be associated to the high degree of
microporosity of the support, leading to more important transfer
mass limitations than those observed for the other supports, in spite
of its large surface area. For acetophenone (3), p-methylanisole (5)
and styrene (6), alumina was less efficient catalysts towards the
full hydrogenated products (entries 3, 7 and 11, Table 3) comparing
with multi-walled carbon nanotubes and silica supports (entries 1
and 2 vs. entry 3; 5 and 6 vs. entry 7; 9 and 10 vs. entry 11, Table 3).
Effectively, the ethylcyclohexane/ethylbenzene ratio points to the
relative activity of the different catalytic systems due to the differ-
ent hydrogenation rate between both exocyclic C–C double bond
and aromatic moiety. TEM analyses after styrene hydrogenation
evidenced that RuNPs appear more agglomerated for alumina than
Table 3
Ru-catalyzed hydrogenation of 3, 5 and 6 substrates, using supported precursors
(RuL–MWCNT, RuL–SiO2, RuL–Al2O3, RuL–AC)a.
Entry
Substrate
Catalyst
Time (h)
%Conv.b (Sel.)
TTOc
1
2
3
4
5
6
7
8
3
3
3
3
5
5
5
5
6
6
6
6
5
6
RuL–MWCNT
RuL–SiO2
RuL–Al2O3
RuL–AC
RuL–MWCNT
RuL–SiO2
RuL–Al2O3
RuL–AC
16
16
16
16
18
18
18
18
6
6
6
6
18
6
100 (100/0/0)d
100 (100/0/0)d
100 (75/25/0)d
78 (20/30/50)d
100 (70/30)e
100 (43/57)e
30 (72/28)e
100
100
75
15.6
70
43
21.6
44.6
100
100
35
43
12
100
62 (72/28)e
9
RuL–MWCNT
RuL–SiO2
RuL–Al2O3
RuL–AC
100 (100/0)f
100 (100/0)f
100 (35/65)f
100 (43/57)
10 (80/20)e
10
11
12
13
14
Ru/Cg
Ru/Cg
100 (0/100)f
a
See Scheme 2 for the products labelling. Results obtained from duplicated tests.
Reaction conditions: 0.64 mmol of substrate, 0.0064 mmol of RuL, 40 bar hydrogen
pressure, 50 ◦C.
b
Conversions and selectivities determined by GC.
Total Turnover = moles of substrated converted to full hydrogenated product per
c
mole of catalyst (based on total metal).
Fig. 5. Catalytic activity of RuL–MWCNT and RuL–SiO2 for styrene hydrogenation in
relation to the styrene/metal ratio at full conversion of substrate. In white, relative
percentage of ethylbenzene (6b) and in gray, ethylcyclohexane (6a). (For interpre-
tation of the references to color in the text, the reader is referred to the web version
of the article.)
d
Chemoselectivity towards the hydrogenated products as a ratio of 3a/3b/3c.
Chemoselectivity towards the hydrogenated products as a ratio of 5a/5b.
Chemoselectivity towards the hydrogenated products as a ratio of 6a/6b.
e
f
g
Ru, 5% on activated carbon.

112
M. Jahjah et al. / Journal of Molecular Catalysis A: Chemical 332 (2010) 106–112
The catalytic activity of both systems RuL–MWCNT and
References
RuL–SiO2 was examined depending of the styrene/ruthenium ratio.
RuL–MWCNT was more active for substrate/metal ratios in the
range of 200–300 than RuL–SiO2, showing comparable activities for
more diluted systems; in particular for styrene/Ru = 1,000, both cat-
alysts exhibited identical behaviour (Fig. 5). The catalytic systems
are less active at lower metal load as evidenced by the increase of
ethylbenzene in relation to the full hydrogenated product, ethylcy-
clohexane, favouring in consequence the product exhibiting a faster
hydrogenation.
The reactivity differences observed between MWCNT and oxide
supports, in particular with silica, could be due to the metal-
lic nanoparticles arrangement on the support. While for silica,
they would be preferentially located inside the pores, for multi-
walled carbon nanotubes, the particles remain on the support outer
surface, favouring in this case the mass transfer under catalytic
conditions.
[1] J.D. Aiken III, R.G. Finke, J. Mol. Catal. A: Chem. A 145 (1999) 1–44.
[2] M.A. El-Sayed, Acc. Chem. Res. 34 (2001) 257–264.
[3] H. Bönnemann, R.M. Richards, Eur. J. Inorg. Chem. (2001) 2455–2480.
[4] A. Roucoux, J. Schulz, H. Patin, Chem. Rev. 102 (2002) 3757–3778.
[5] A. Roucoux, K. Philippot, Hydrogenation with Noble Metal Nanoparticles in
Handbook of Homogeneous Hydrogenations, Wiley-VCH, Weinheim, 2006.
[6] G. Schmid, V.C.H. Weinheim, Clusters and Colloids, from Theory to Applications,
1994.
[7] L.S. Ott, R.G. Finke, Coord. Chem. Rev. 251 (2007) 1075–1100.
[8] Y. Li, M.A. El-Sayed, J. Phys. Chem. B 105 (2001) 8938–8943.
[9] C. Burda, X. Chen, R. Narayanan, M.A. El-Sayed, Chem. Rev. 105 (2005)
1025–1102.
[10] J. Durand, E. Teuma, M. Gómez, Eur. J. Inorg. Chem. (2008) 3577–3586.
[11] P. Migowski, J. Dupont, Chem. Eur. J. 13 (2007) 32–39.
[12] H. Olivier-Bourbigou, L. Magna, J. Mol. Cat. A: Chem. 182 (2002) 419–437.
[13] L. Duran Pachón, C.J. Elsevier, G. Rothenberg, Adv. Synth. Catal. 348 (2006)
1705–1710.
[14] M. Studer, H.-U. Blaser, C. Exner, Adv. Synth. Catal. 345 (2003) 45–65.
[15] H. Bönnemann, G.A. Braun, Chem. Eur. J. 3 (1997) 1200–1202.
[16] X. Zuo, H. Liu, D. Guo, X. Yang, Tetrahedron 55 (1999) 7787–7804.
[17] J.U. Köhler, J.S. Bradley, Langmuir 14 (1998) 2730–2735.
[18] B. Corain, P. Centomo, S. Lora, M. Kralik, J. Mol. Catal. A: Chem. 204–205 (2003)
755–762.
4. Conclusions
[19] H. Vu, F. Gonc¸ alves, R. Philippe, E. Lamouroux, M. Corrias, Y. Kihn, D. Plee, P.
Kalck, P. Serp, J. Catal. 240 (2006) 18–22.
[20] X. Pan, Z. Fan, W. Chen, Y. Ding, H. Luo, X. Bao, Nat. Mater. 6 (2007) 507–511.
[21] P. Serp, M. Corrias, P. Kalck, Catal. Appl. A. 253 (2003) 337–358.
[22] M. Ruta, I. Yuranov, P.J. Dyson, G. Laurenczy, L. Kiwi-Minsker, J. Catal. 247 (2007)
269–276.
[23] B.F.G. Johnson, Top. Catal. 24 (2003) 1–4.
[24] E.V. Starodubtseva, M.G. Vinogradov, O.V. Turova, N.A. Bumagin, E.G. Rakov, V.I.
Sokolov, Catal. Commun. 10 (2009) 1441–1442.
[25] B. Yoon, H.-B. Pan, C.M. Wai, J. Phys. Chem. C 113 (2009) 1520–1525.
[26] B. Pawelec, V. La Parola, R.M. Navarro, S. Murcia-Mascarós, J.L.G. Fierro, Carbon
44 (2006) 84–98.
[27] M. Takasaki, Y. Motoyama, K. Higashi, S.-H. Yoon, I. Mochida, H. Nagashima,
Chem. Asian J. 2 (2007) 1524–1533.
[28] J. Pan, J. Li, C. Wang, Z. Yang, React. Kinet. Catal. Lett. 90 (2007) 233–242.
[29] L. Rodríguez-Pérez, E. Teuma, A. Falqui, M. Gómez, P. Serp, Chem. Commun.
(2008) 4201–4203.
[30] I. Favier, S. Massou, E. Teuma, K. Philippot, B. Chaudret, M. Gómez, Chem. Com-
mun. (2008) 3296–3298.
In summary, small and well-dispersed ruthenium nanoparticles
supported on MWCNT were successfully used in catalytic hydro-
genation reactions of unsaturated substrates. The beneficial effect
of the support was proved comparing the catalytic behaviour with
that corresponding to the non-supported catalyst, obtaining for all
the substrates better activities using RuL–MWCNT system. More-
over, the support nature was evaluated. While Ru nanoparticles
supported on alumina and activated carbon led to lower catalytic
activities and chemoselectivities, silica and multi-walled carbon
nanotubes supported catalysts gave preferentially the full hydro-
genated products. However, only RuL–MWCNT could be reused
without activity loss exclusively leading to ethylcyclohexane, in
contrast to the analogous RuL–SiO2 catalyst.
[31] M. Corrias, B. Caussat, A. Ayral, J. Durand, Y. Kihn, P. Kalck, P. Serp, Chem. Eng.
Sci. 58 (2003) 4475–4482.
Acknowledgements
[32] A. Solhy, B.F. Machado, J. Beausoleil, Y. Kihn, F. Gonc¸ alves, M.F.R. Pereira, J.J.M.
Órfão, J.L. Figueiredo, J.L. Faria, P. Serp, Carbon 46 (2008) 1194–1207.
[33] B.K. Teo, H. Zhang in, D.L. Feldheim, C.A. Foss Jr. (Eds.), Metal Nanoparticles:
Synthesis, Characterization and Applications, Marcel Dekker, New York, 2002
(Chapter 3).
The authors thank ANR (Nanoconficat, project no. ANR-05-
NANO-030), CNRS and Université Paul Sabatier for their financial
support, and Prof. Philippe Serp and his team for the fruitful collab-
oration.
[34] (a) G. St-Pierre, A. Chagnes, N.-A. Bouchard, P.D. Harvey, L. Brossard, H. Ménard,
Langmuir 20 (2004) 6365–6373;
(b) F. Laplante, L. Brossard, H. Ménard, Can. J. Chem 81 (2003) 258–264;
(c) G. Nery, A.M. Visco, A. Donato, C. Milone, M. Malentacchi, G. Gubitosa, Appl.
Catal. A: Gen. 110 (1994) 49–59.
Appendix A. Supplementary data
[35] Turnover frequency = mol of converted substrate per mol of catalyst per hour
(based on Ru atoms on the surface).
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molcata.2010.09.006.