Pd on carbon nanotubes for liquid phase alcohol oxidation

Article abstract of DOI:10.1016/j.cattod.2009.06.009

Pd nanoparticles supported on carbon nanotubes (CNTs) showed a higher selectivity than Pd nanoparticles supported on activated carbon (AC) in the liquid phase oxidation of benzylic alcohol to benzaldehyde. Under solventless conditions a significant improv

Full text of DOI:10.1016/j.cattod.2009.06.009

Catalysis Today 150 (2010) 8–15
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Pd on carbon nanotubes for liquid phase alcohol oxidation
c
a,
Alberto Villa b, Di Wang b, Nikolaos Dimitratos a,1, Dangsheng Su b, , Valentina Trevisan , Laura Prati
*
*
a
`
Universita di Milano, Dipartimento CIMA ‘‘L.Malatesta’’, via Venezian 21, I-20133 Milano, Italy
b Department of Inorganic Chemistry, Fritz Haber Institute of the Max Planck Society, Faradayweg 4-6, D-14195 Berlin, Germany
c
`
Universita di Venezia, Dipartimento di Chimica, Dorsoduro 2137, I-30123 Venezia, Italy
A R T I C L E I N F O
A B S T R A C T
Article history:
Pd nanoparticles supported on carbon nanotubes (CNTs) showed
a higher selectivity than Pd
Available online 17 July 2009
nanoparticles supported on activated carbon (AC) in the liquid phase oxidation of benzylic alcohol to
benzaldehyde. Under solventless conditions a significant improvement in selectivity was observed for
Pd/CNTs, whereas using Pd/AC a considerable over-oxidation of benzaldehyde was observed. Differently
from other solvents cyclohexane improved significantly the selectivity to benzaldehyde for both
catalysts. Characterisation by means of transmission electron microscopy revealed differences in metal
dispersion between Pd/AC and Pd/CNTs that can be ascribed to textural, chemical and physical
differences between active carbon and carbon nanotubes. The higher activity in the case of Pd on AC than
on CNTs can be attributed to the improved Pd dispersion in the first case.
Keywords:
Alcohol oxidation
Palladium on carbon nanotubes
Gold–palladium catalyst
On recycling Pd/CNTs resulted more stable (activity loss 50% in 7 runs) than Pd/AC (activity loss 70% in
7 runs) even a structural change of catalysts after reaction is observed. The Pd leaching and particle
coalescence are the main reasons for the loss of activity. An extraordinary improving of catalyst life has
been observed by alloying the Pd nanoparticles with Au, When CNTs are used as support the strong Pd
leaching can be greatly limited and the activity/selectivity maintained at least for 8 runs.
ß 2009 Elsevier B.V.. All rights reserved.
1. Introduction
burning off the support. Recently, there is a growing interest in the
use of CNTs as catalyst supports for metal nanoparticles in many
Metal catalyzed alcohol oxidation in the liquid phase using
oxygen as the oxidant represents a well known interesting process,
but its industrial exploitation was limited by the strong deactiva-
tion [1]. Several metals (ruthenium, platinum, palladium and gold)
have been used as monometallic catalysts, in some cases with
modifier (like Bi, Pb) [2–5]. The main requirement for a high
selectivity and long durability has not been yet achieved. More-
over, a variety of experimental conditions have been tested: from
solventless to organic solvents or aqueous conditions. It has been
observed that basic conditions improved the activity and the
durability, probably due to a prolonged catalyst life. However,
under basic conditions the selectivity to aldehyde in the alcohol
oxidation decreased in favour of the corresponding carboxylic acid
(carboxylate).
catalytic applications; for example, CNTs supported platinum
nanoparticles are used as electrocatalyst in proton-exchange
membrane fuel cells [6,7]. Moreover, Ni [8,9] and Pd [10]
nanoparticles decorated CNTs were used for hydrogen storage
and for hydrogen sensing. In addition, Pd [11], Pt and Ru [12] metal
nanoparticles supported on CNTs show a high catalytic activity in
the liquid phase hydrogenation of cinnamaldehyde. However, only
few reports on the use of CNTs instead of the widely used AC can be
found in the literature. For example, Corma et al. [13] showed that
Pd nanoparticles immobilised on CNTs are more active than the
one supported on AC in the Heck reaction of styrene and
iodobenzene and for the Suzuki coupling of phenylboronic and
iodobenzene. They assigned the high activity to the different
particle size obtained for the different supports. They showed that
the Pd/AC is more active than Pd/CNTs for the aerobic oxidation of
cinnamyl alcohol in ethanol. Moreover Karaousis et al. [14]
reported higher activity of CNTs supported metal nanoparticles
with respect to AC supported ones in the hydrogenation of methyl-
9-octadecenoate and 2-methyl-2-pentenal. They concluded that
the high catalytic performance of CNTs supported Pd is due to the
large active surface of metallic palladium.
From an industrial point of view carbonaceous supports are
preferable because of the facile recovery of precious metals by
* Corresponding authors.
E-mail addresses: dangsheng@fhi-berlin.mpg.de (D. Su),
Laura.Prati@unimi.it (L. Prati).
1
In the present work we report on the catalytic activity of CNTs
and AC supported Pd nanoparticles prepared by colloidal method
Current address: School of Chemistry, Cardiff University, Main Building, Park
Place, Cardiff CF10 3AT, UK.
0920-5861/$ – see front matter ß 2009 Elsevier B.V.. All rights reserved.
doi:10.1016/j.cattod.2009.06.009

A. Villa et al. / Catalysis Today 150 (2010) 8–15
9
using PVA as protective agent [15] in the liquid phase oxidation of
2.3. Catalytic test
benzyl alcohol with the aim of preserving a high selectivity to
aldehyde but enhancing the catalyst activity and reusability.
Different groups showed that adding Au to Pd catalysts has a
beneficial effect in the alcohol oxidation, improving not only the
catalytic activity and selectivity to the desired product [16], but
also the resistance to the deactivation [17–19]. We recently set up
a two step methodology allowing the preparation of uniformly
alloyed bimetallic Au60–Pd40 catalyst supported on activated
carbon [15]. With the same procedure we prepared a bimetallic
Au–Pd catalyst supported on CNTs.
The reactions were carried out in a thermostatted glass reactor
(30 ml) provided with an electronically controlled magnetic stirrer
connected to a large reservoir (5000 ml) containing oxygen at
2 atm. The oxygen uptake was followed by a mass flow controller
connected to a PC through an A/D board, plotting a flow/time
diagram. The oxidation experiments were carried out in solvent-
less or in the presence of a solvent (0.0125 mol substrate,
substrate/metal = 3000 (mol/mol), 80 8C, pO2 = 2 atm). In the case
of organic solvent, periodic removal of samples from the reactor
was performed, whereas in the case of water solvent, after the end
of the reaction the catalyst was filtered off and the product mixture
was extracted with CH2Cl2. Recoveries were always 98 Æ 3% with
this procedure. For the identification and analysis of the products a
GC–MS and GC (a Dani 86.10 HT Gas Chromatograph equipped with a
capillary column, BP21 30 m  0.53 mm, 0.5 mm Film, made by SGE),
were used by comparison of the authentic samples. For the
quantification of the reactant–products the external calibration
method was used.
2. Experimental
2.1. Materials
Na2PdCl4 and NaAuCl4Á2H2O were from Aldrich (99.99% purity),
activated carbon from Camel (X40S; SA = 900–1100 m2/g;
PV = 1.5 ml/g) and carbon nanotubes from Bayern Materials
Science AG (Baytubes C150P; SA = 30 m2/g; internal diameter
13–16 nm, length 1–10
mm). NaBH4 of purity >96% from Fluka,
polyvinylalcohol (PVA) (Mw = 13,000–23,000; 87–89% hydro-
lysed) from Aldrich were used. Gaseous oxygen from SIAD was
99.99% pure.
2.4. Characterisation
2.4.1. Sol characterisation
UV–vis spectra of sols were performed on HP8452 and HP8453
Hewlett-Packard spectrophotometers in H2O between 190 and
1200 nm, in a quartz cuvette.
2.2. Catalyst preparation
2.2.1. Monometallic catalysts
Pd sol: Na2PdCl4Á2H2O (0.043 mmol) and PVA solution 2 wt%
2.4.2. Catalyst characterisation
(880
(860
m
l) were added to 130 ml of H2O. After 3 min, NaBH4 solution
l) was added to the yellow-brown solution under vigorous
m
(a) The metal content was checked by ICP analysis of the filtrate or
alternatively directly on catalyst after burning off the carbon,
on a Jobin Yvon JY24.
magnetic stirring. The brown Pd(0) sol was immediately formed.
An UV–vis spectrum of the palladium sol was recorded for ensuring
the complete reduction of Pd(II).
Within few minutes from their generation, the colloids
(acidified at pH 2, by sulphuric acid) were immobilised by adding
carbon under vigorous stirring. The catalyst has been filtered
washed for several times to ensure the removal of the material
arising from the reduction treatment. The samples were dried at
353 K for 2 h. The amount of support was calculated to obtain a
final metal loading of 1 wt%.
(b) Morphology and microstructures of the catalysts were
characterised by TEM. The powder samples of the catalysts
were ultrasonically dispersed in ethanol and mounted onto
copper grids covered with holy carbon film. A Philips CM200
FEG electron microscope, operating at 200 kV and equipped
with a Gatan imaging filter, GIF Tridiem, was used for TEM
observation. EDX analysis was performed in the same
microscope using a DX4 analyzer system (EDAX).
(c) Z-potential measurements were performed on Zetasizer
nanoZS/MPT-2 from Malvern. 10 mg of support was dispersed
in 50 ml of distilled water and sonicated for 10 min. The
support was then removed by filtration and the solution was
ready for zeta potential measurements. For the colloidal metal
solution no pre-treatment was used before the measurement.
2.2.2. Bimetallic catalyst
Solid NaAuCl4Á2H2O (0.072 mmol) was dissolved in 140 ml of
water (final 10À4 M) and 0.706 ml of PVA (2%, w/w) was added (Au/
PVA 1:1 w/w). The yellow solution was stirred for 3 min and 2.9 ml
of 0.1 M NaBH4 (Au/NaBH4 1:3 mol/mol) was added under
vigorous magnetic stirring. The ruby red Au(0) sol was immedi-
ately formed. An UV–vis spectrum of the gold sol was recorded to
3. Results and discussion
À
check the complete AuCl4 reduction and the formation of
Plasmon peak. Within few minutes of sol generation, the gold
sol was immobilised by adding carbon nanotubes (acidified until
pH 2 by sulphuric acid) under vigorous stirring. The amount of
support was calculated as having a gold loading of 0.73 wt% when
Au/Pd was prepared. After 2 h the slurry was filtered, the catalyst
was thoroughly washed with distilled water (neutral mother
liquors). ICP analyses were performed on the filtrate using a Jobin
Yvon JV24 to verify the total metal loading on the support. The Au/
CNTs was then dispersed in 140 ml of water; Na2PdCl4 (0.048 mol)
and 0.225 ml of PVA solution (0.2%, w/w) (Au/PVA 1:1 w/w) were
added. H2 has been bubbled (50 ml/min) under atmospheric
pressure and room temperature for 2 h. After additional 18 h, the
slurry was filtered, the catalyst was thoroughly washed with
distilled water. ICP analyses were performed on the filtrate using a
Jobin Yvon JV24 to verify the metal loading on the support. 1 wt%
was the total metal loading.
3.1. Catalytic activity of Pd supported on AC and CNTs
AC is a widely used catalyst support for liquid phase reaction.
However, recent literature disregards AC as the support for alcohol
oxidation; the most used supporting materials are oxides despite
their disadvantage in precious metal recovery. We concentrated on
benzylic alcohol to benzaldehyde selective oxidation which is an
important reaction from an industrial point of view and has been
studied under a variety of reaction conditions using different
catalytic systems.
We synthesized 1 wt% Pd/AC or CNT via the sol immobilisation
method (NaBH4/PVA). The palladium metal sol was generated by
NaBH4 reduction of Na2PdCl4 salt in the presence of a protective
agent (polyvinyl alcohol, PVA) which provides electrostatic as well
as sterical stabilisation of the nanoparticles. Pd-sol appeared
negatively charged at any pH of the solution (Fig. 1). The AC and

10
A. Villa et al. / Catalysis Today 150 (2010) 8–15
CNT auto-generates different basic pHs (9.18 and 10.28, respec-
tively) but acidic groups determined by the Boehm method [20]
revealed similar density of carboxylic groups (9.74 and
a
9.89 mmol/g). Thus the support suspension was previously
acidified (pH 2) in order to create a positive electrostatic attraction
between the support and the protected metal particles (Fig. 1). In
principle the immobilisation of preformed Pd nanoparticles
assured a similar particle size distribution on different support.
Fig. 2a and b shows the TEM overview images of the as prepared
Pd/AC and Pd/CNTs, respectively. The histograms of the particle
size distribution for AC and CNTs supported Pd particles are shown
in Fig. 2c and d, respectively. Fitting by Log-Normal function, the
Pd/AC was found to have median size of 3.9 nm and the Pd/CNTs of
3.4 nm (Table 1). However, though Pd particles on CNTs have
smaller size, Pd particles on AC exhibit a more homogeneous
distribution. In fact for Pd/CNTs, some CNTs were found without Pd
particles, while many Pd particles are observed in aggregates. This
difference could be attributed to a different and inhomogeneous
distribution of chemical groups on the CNTs surface respect to AC
ones.
Fig. 1. Zeta potential measurement.
Fig. 2. The TEM overview images of (a) Pd/AC and (b) Pd/CNTs, (c) and (d) the Pd particle size distribution of the two catalysts, respectively.

A. Villa et al. / Catalysis Today 150 (2010) 8–15
11
Table 1
Statistical median and standard deviation of particle size analysis for Pd and Au/Pd catalysts.
Pd/AC fresh
Pd/AC after 7 runs
Pd/CNTs fresh
Pd/CNTs after 8 runs
Au–Pd/AC fresh
Au–Pd/CNTs fresh
Statistical median (nm)
Standard deviation,
3.94
1.15
5.27
1.42
3.35
0.87
6.44
2.56
3.41
0.69
3.53
0.78
s
Scheme 1. Possible reaction scheme for benzyl alcohol oxidation.
The catalysts were tested in the liquid phase oxidation of benzyl
alcohol. This reaction has been widely studied in the literature
using a variety of experimental conditions. In the present study we
tested the most promising solventless reaction, the most benign
reaction in water although in the literature toluene represents the
most common solvent used [21]. However, as toluene is a possible
by-product in the benzyl alcohol oxidation (Scheme 1), we avoided
its use and checked a different organic solvent (cyclohexane).
Table 2 reports the results of Pd supported on AC and CNT using
different amounts of each solvent. The experimental conditions
were maintained constant (0.0125 mol substrate, alcohol/metal
Table 2
Benzyl alcohol oxidation in the presence of supported 1 wt% Pd.
b
Solvent
Selectivity
toluenea
to
Selectivity
benzaldehydea
to
TOF (hÀ1
)
Pd/AC
Pd/CNT
Pd/AC
Pd/CNT
Pd/AC
Pd/CNT
Neat
33
25
30
36
29
23
8
17
25
41
37
24
3
61
71
67
62
67
74
88
78
73
57
59
74
92
88
2980
1485
1227
840
1412
730
25% water
50% water
506
75% water
554
20% cyclohexane
50% cyclohexane
80% cyclohexane
2846
1482
490
1462
1340
493
8
Reaction condition: alcohol/metal 3000/1 (mol/mol), 80 8C, pO2 2 atm, 1250 rpm.
a
Fig. 3. (a) Conversion/selectivity on recycling 1%Pd/AC. (b) Conversion/selectivity
Selectivity at 90% conversion.
b
on recycling 1%Pd/CNTs.
TOF calculated after 15 min of reaction.

12
A. Villa et al. / Catalysis Today 150 (2010) 8–15
3000/1 (mol/mol), 80 8C, pO2 2 atm, 1250 rpm) just varying the
amount of solvent.
enhancement in aldehyde selectivity and a consistent decreasing
in the toluene one.
The screening of activity of Pd on AC (Table 2) revealed that,
Comparing the TOFs of Pd/CNTs with that of Pd/AC (Table 2) we
could generally observe that Pd on CNTs showed a lower activity
than Pd/AC. However, the most interesting point came from the
referring to solventless conditions, water addition causes
a
decrease in activity. This becomes most significant when 75% of
water was added (TOF drops from 2980 to 840 hÀ1). The effect on
selectivity, however, was negligible. 61–62% selectivity to ben-
zaldehyde at 90% conversion (S90) was obtained in both cases.
Here toluene was the main by-product due to a hydrogen transfer
process. Indeed Pd is a well known catalyst for this type of reaction
[22]. Prior to perform the reaction in organic solvent (cyclohexane)
we carried out blank experiment that proved the inertness of
cyclohexane under the actual reaction condition. As reported in
Table 2, a 20 vol.% addition of cyclohexane did not modified the
result of solventless experiment, but when the amount of solvent
increased, the expected decreasing of the reaction rate was
accompanied by a huge improving in selectivity. By increasing the
cyclohexane amount from 50 to 80%, we observed a strong
selectivity. In the solventless reaction Pd/CNTs exhibited
a
considerably higher selectivity to aldehyde with respect to the
one obtained with Pd/AC (S90 78% versus 61%). When water was
added, a decrease of selectivity to aldehyde was observed. This
trend became more serious when water content was at 50–75%.
Meanwhile the toluene formation increased. On the contrary,
when cyclohexane was added we obtained a remarkable enhance-
ment of benzaldehyde production with selectivity higher than 90%
with 50 vol.% solvent. In this latter case it should be noted that the
activity of the catalyst is quite good (1340 hÀ1) being comparable
with the one of Pd/AC (TOF 1482 hÀ1).
On the basis of these results we can conclude that performances
of both catalysts (Pd/AC and Pd/CNTs) are very sensitive to the
Fig. 4. The TEM overview image for (a) Pd/AC and (b) Pd/CNTs after reaction. Pd particle size distribution of (c) Pd/AC and (d) Pd/CNTs after reaction.

A. Villa et al. / Catalysis Today 150 (2010) 8–15
13
solvent employed and that in both cases the addition of a solvent
decreased the catalyst activity compared to solventless conditions.
Cyclohexane with respect to water appeared to be more able to
preserve the activity. However, the main positive effect of solvent
is to improve the selectivity to aldehyde, this effect being slightly
higher for Pd on CNTs than on AC.
be found in the characterisation of the recovered catalysts which
have been observed by TEM after washing with water and dried. In
both cases a growing in size of Pd particles was observed which are
expected to show lower activity than the one of smaller, highly
dispersed particles (Fig. 4a and b and Table 1). Fig. 4c and d shows
the particle size distribution of the Pd/CNTs and Pd/AC after
reaction, in which the median particle sizes are 6.4 and 5.3 nm,
respectively (Table 1), considerably larger than those of the
3.2. Catalyst durability (recycling test)
corresponding fresh catalysts. Apart from the aggregates,
a
From an applicative point of view, one of the most appreciable
properties of a heterogeneous catalyst is its durability. Thus we
tested the Pd on AC and on CNTs catalysts on recycling, checking
the overall leaching of the Pd or any other modification on the used
catalyst after several runs. Each run was carried out under identical
conditions (alcohol/metal 3000/1 (mol/mol), 80 8C, pO2 2 atm,
1250 rpm, cyclohexane 50 vol.%) and reusing the catalyst recov-
ered by filtering off the solution of the previous run after 1 h. After
7 runs, despite the selectivity did not change so far, conversion
after 1 h has declined about 70% in the case of Pd/AC (Fig. 3a) and
47% in the case of Pd/CNTs (Fig. 3b). A reason of this behaviour can
considerable number of the particles were observed to undergo
coalescence to form big and irregularly shaped particles. Moreover,
we performed ICP analysis on the used catalysts, after burning off
the carbonaceous material, detecting a consistent leaching of Pd
(28% for Pd/AC and 25% for Pd/CNTs). All these factors lead to the
drastic reduction of the accessible reaction sites on Pd. We can then
conclude that these systems are not stable under the reaction
conditions used and need to be stabilised for an industrial
exploitation. Moreover, considering that the selectivity has not
changed during the recycled runs, we also concluded that the
selectivity is independent on the particle size range we have
Fig. 5. HRTEM of (a) Au–Pd/AC and AC and (b) Au–Pd/CNTs. Particle size distribution of (c) Au–Pd/AC and (d) Au–Pd/CNTs.

14
A. Villa et al. / Catalysis Today 150 (2010) 8–15
Table 3
to be an active catalyst under basic conditions. Moreover, we were
quite disappointed to observe that Pd and Au–Pd showed, under
these experimental conditions, a similar behaviour from both
activity and selectivity point of view. In fact only a slight variation
of TOFs was obtained: from 1482 to 1350 hÀ1 for AC and from 1340
to 1490 hÀ1 for CNTs. Selectivity to benzaldehyde only improved in
the case of Au–Pd on AC compared to Pd/AC.
Comparison of mono and bimetallic catalyst activities in benzyl alcohol oxidation.
b
Selectivity
to toluenea
Selectivity
TOF (hÀ1
)
to benzaldehydea
Pd/AC
23
–
74
–
1482
<1
Au/AC
Au–Pd/AC
Pd/CNTs
Au/CNTs
Au–Pd/CNTs
9
89
92
–
1350
1340
<1
3
On the contrary a completely different catalytic behaviour was
highlighted on recycling. Under the same conditions as for the test
of monometallic Pd, Au/Pd on AC losses in 8 runs about 50% of its
activity whereas CNTs only 4% (Fig. 6a and b). ICP measurements
revealed a leaching of palladium of 10% for Au–Pd/AC whereas only
4% in the case of Au–Pd/CNTs. In both cases leaching of Au resulted
negligible. Considering the normal loss of the catalyst i.e. the
expected decreasing in conversion due to the reduced amount of
catalyst recovered from any subsequent run (3–4%), the leaching of
Pd is inconsistent with both of catalytic results. However, we
already observed that the leaching of Pd depends on particle size
[23]. This different leaching could create the presence of different
alloy composition that could possibly show different activity. This
was already observed in the case of aqueous glycerol oxidation
where the maximum of activity was reached for Au95Pd05
composition [24]. At this stage we could only conclude that gold
stabilises the palladium more efficiently on CNTs than on active
carbon. Further studies are needed to determine the nature of this
stabilisation.
–
3
89
1490
Reaction condition: alcohol/metal 3000/1 (mol/mol), 80 8C, solvent: cyclohexane
50 vol.%, pO2 2 atm, 1250 rpm.
a
Selectivity at 90% conversion.
b
TOF calculated after 15 min of reaction.
investigated. Therefore, the difference in selectivity between Pd/AC
and Pd/CNTs is expected to relate with the support properties and
their influences on the Pd sites.
We already experienced the beneficial effect of alloying gold to
Pd nanoparticles [23]. A two step procedure that ensures the
formation of uniform alloyed Pd/Au nanoparticles on AC has been
published [15] and proved to enhance both activity and durability
of the catalyst in glycerol oxidation [23]. As we were interested to
improve the catalyst stability, we now extended the same
technique to prepare bimetallic Au–Pd nanoparticles on CNTs
and thus allowing comparing the behaviour of Au–Pd on AC with
that of Au–Pd on CNTs. The resulted catalysts show similar particle
size, 3.4 nm for Au–Pd/AC and 3.5 nm for Au–Pd/CNTs (Table 1),
even differently distributed (Fig. 5c and d). Preliminary tests for
evaluating the real effect of the adding of Au to Pd were performed
in cyclohexane (50 vol.%) that, as outlined before, strongly differs
from water as solvent and in that also Au–Pd/AC was never used in
the past [23]. Table 3 reports the activities of monometallic
catalysts compared to bimetallic Au–Pd ones. Under these
conditions monometallic gold was inactive both supported on
AC and on CNTs. This behaviour did not surprise as gold is expected
4. Conclusions
We compared the activity of Pd nanoparticles supported on AC
and CNTs in the liquid phase oxidation of benzylic alcohol to
benzaldehyde. Although we used a procedure (sol immobilisation)
that in principle assures a similar distribution on different
supports, we obtained a slight different metal distribution in the
two cases. Pd nanoparticles on CNTs are smaller in size but less
dispersed than on AC. Catalytic test revealed that Pd supported on
CNTs behaves differently from Pd on AC. In fact, the expected lower
activity of Pd/CNTs according to the lower metal dispersion was
related to a higher selectivity toward benzaldehyde strongly
dependent on the solvent used. The maximum of selectivity (92%)
was obtained for a 50% volume of cyclohexane as solvent. The use
of Pd on both AC and CNTs is however limited by the high Pd
leaching revealed on recycling (28 and 25%, respectively). A
consistent improving in the long-term use of the catalysts has been
obtained by modifying the monometallic Pd with gold. With this
purpose the procedure set up for producing Au–Pd alloyed
nanoparticles on AC was extended to CNTs. On recycling this
latter catalyst showed limited Pd leaching and stable catalytic
performances over 8 runs. Under similar conditions Au–Pd on AC
was less stable.
Acknowledgements
Authors thank DAAD, Programma Vigoni and Fondazione
Cariplo for financial support.
References
[1] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037.
[2] M. Besson, P. Gallezot, Catal. Today 57 (2000) 127.
[3] R.A. Sheldon, I.W.C.E. Arends, G.J. ten Brink, A. Dijksman, Acc. Chem. Res. 35 (2002)
774.
[4] B.A. Steinhoff, S.R. Fix, S.S. Stahl, J. Am. Chem. Soc. 124 (2002) 766.
[5] T. Nishimura, Y. Maeda, N. Kaikiuchi, S. Uemura, J. Chem. Soc., Perkin Trans. 1
(2000) 4301.
Fig. 6. (a) Conversion/selectivity on recycling of 1%Au–Pd/AC. (b) Conversion/
[6] Z. Liu, X. Lin, J.Y. Lee, W. Zhang, M. Han, L.M. Gan, Langmuir 18 (2002) 4054.
[7] Y. Xing, J. Phys. Chem. B 108 (2004) 19255.
selectivity on recycling of 1%Au–Pd/CNTs.

A. Villa et al. / Catalysis Today 150 (2010) 8–15
15
[8] H.-S. Kim, H. Lee, K.-S. Han, J.-H. Kim, M.-S. Song, M.-S. Park, J.-Y. Lee, J.-K. Kang, J.
Phys. Chem. B 109 (2005) 8983.
[9] J.-Y. Lee, H. Lee, H.-S. Kim, Mater. Sci. Forum 475–479 (2005) 2463.
[10] Kong, M.G. Chapline, H. Dai, Adv. Mater. 13 (2001) 1384.
[11] J.P. Tessonier, L. Pesant, G. Ehret, M.J. Ledoux, C. Pham-Huu, Appl. Catal. A 288
(2005) 203.
[16] S. Demeril, K. Lehnert, M. Lucas, P. Claus, Appl. Catal. B 70 (2007) 637.
[17] D.I. Enache, J.K. Edwards, P. Landon, B. Solsona-Espriu, A.F. Carley, A.A. Herzing, M.
Watanabe, C.J. Kiely, D.W. Knight, G.J. Hutchings, Science 11 (2006) 362.
[18] N. Dimitratos, A. Villa, D. Wang, F. Porta, D. Su, L. Prati, J. Catal. 244 (2006) 113.
[19] W. Hou, N.A. Dehm, R.W.J. Scott, J. Catal. 253 (2008) 22.
[20] H.P. Boehm, Advances in Catalysis, 16, Academic Press, New York, 1966.
[21] C. Della Pina, E. Falletta, L. Prati, M. Rossi, Chem. Soc. Rev. 37 (2008) 2077 (and
references cited therein).
[12] H. Vu, F. Goncalves, R. Philippe, E. Lamouroux, M. Corrias, Y. Kihn, D. Plee, P. Kalck,
P. Serp, J. Catal. 240 (2006) 18.
[13] A. Corma, H. Garcia, A. Leyva, J. Mol. Catal. A 230 (2005) 97.
[14] N. Karaousis, G. Tsotsou, F. Evangelista, P. Rudolf, N. Ragoussis, N. Tagmatarchis, J.
Phys. Chem. C 112 (2008) 13463.
[22] J.M. Thomas, W.J. Thomas Principles, Practice of Heterogeneous Catalysis, VHC
Ed, Weinheim, Germany, 1997.
[23] L. Prati, A. Villa, F. Porta, D. Wang, D. Su, Catal. Today 122 (2007) 386.
[24] D. Wang, A. Villa, F. Porta, L. Prati, D. Su, J. Phys. Chem. C 112 (2008) 8617.
[15] D. Wang, A. Villa, F. Porta, D. Su, L. Prati, Chem. Commun. (2006) 1956.