Solvent-free oxidation of benzyl alcohol using Au-Pd catalysts prepared by sol immobilisation

Article abstract of DOI:10.1039/b900151b

We report the preparation of Au-Pd nanocrystalline catalysts supported on TiO₂ and carbon prepared via a sol-immobilisation technique using three different preparation strategies; namely, simultaneous formation of the sols for both metals or initial formation of a seed sol of one of the metals followed by a separate step in which a coating sol of the second metal is added. The catalysts have been structurally characterised using a combination of transmission electron microscopy and X-ray photoelectron spectroscopy. The catalysts have been evaluated for the oxidation of benzyl alcohol under solvent-free conditions. The catalysts prepared using the sol immobilisation technique show higher activity when compared with catalysts prepared by impregnation, particularly as lower metal concentrations can be used. The Au-Pd catalysts were all more active than the corresponding monometallic supported Au or Pd catalysts. For 1 wt% Au-Pd/TiO₂ the order of metal addition in the preparation was not observed to be significant with respect to selectivity or activity. However, the 1 wt% Au-Pd/carbon catalysts are more active but less selective to benzaldehyde than the TiO₂-supported catalysts when compared at iso-conversion. Furthermore, for the carbon-supported catalyst the order of metal addition has a very marked affect on activity. The carbon-supported catalysts are also more significantly affected by heat treatment, e.g. calcination at 400 °C leads to the activity being decreased by an order of magnitude, whereas the TiO₂-supported catalysts show a 50% decrease in activity. Toluene is observed as a by-product of the reaction and conditions have been identified that minimise its formation. It is proposed that toluene and benzaldehyde are formed by competing parallel reactions of the initial benzyl intermediate via an adsorbed benzylidene species that can either be hydrogenated or oxidised. Hence, conditions that maximise the availability of oxygen on the catalyst surface favour the synthesis of benzaldehyde. the Owner Societies 2009.

Full text of DOI:10.1039/b900151b

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
Solvent-free oxidation of benzyl alcohol using Au–Pd catalysts prepared
by sol immobilisation
a
a
a
Nikolaos Dimitratos, Jose Antonio Lopez-Sanchez, David Morgan,
b
a
b
a
Albert F. Carley, Ramchandra Tiruvalam, Christopher J. Kiely, Donald Bethell
a
and Graham J. Hutchings*
Received 5th January 2009, Accepted 20th February 2009
First published as an Advance Article on the web 3rd April 2009
DOI: 10.1039/b900151b
We report the preparation of Au–Pd nanocrystalline catalysts supported on TiO and carbon
2
prepared via a sol-immobilisation technique using three different preparation strategies; namely,
simultaneous formation of the sols for both metals or initial formation of a seed sol of one of the
metals followed by a separate step in which a coating sol of the second metal is added. The
catalysts have been structurally characterised using a combination of transmission electron
microscopy and X-ray photoelectron spectroscopy. The catalysts have been evaluated for the
oxidation of benzyl alcohol under solvent-free conditions. The catalysts prepared using the sol
immobilisation technique show higher activity when compared with catalysts prepared by
impregnation, particularly as lower metal concentrations can be used. The Au–Pd catalysts were
all more active than the corresponding monometallic supported Au or Pd catalysts. For 1 wt%
Au–Pd/TiO
with respect to selectivity or activity. However, the 1 wt% Au–Pd/carbon catalysts are more
active but less selective to benzaldehyde than the TiO -supported catalysts when compared at
2
the order of metal addition in the preparation was not observed to be significant
2
iso-conversion. Furthermore, for the carbon-supported catalyst the order of metal addition has a
very marked affect on activity. The carbon-supported catalysts are also more significantly affected
by heat treatment, e.g. calcination at 400 1C leads to the activity being decreased by an order of
magnitude, whereas the TiO -supported catalysts show a 50% decrease in activity. Toluene is
2
observed as a by-product of the reaction and conditions have been identified that minimise its
formation. It is proposed that toluene and benzaldehyde are formed by competing parallel
reactions of the initial benzyl intermediate via an adsorbed benzylidene species that can either
be hydrogenated or oxidised. Hence, conditions that maximise the availability of oxygen on the
catalyst surface favour the synthesis of benzaldehyde.
Against the background of this diverse catalytic chemistry
that has been discovered for pure gold nanoparticles, we have
recently shown that the combination of gold with palladium
leads to enhanced catalytic activity and selectivity in redox
processes. While gold alone is a very effective catalyst for the
selective oxidation of an alcohol to an aldehyde under solvent
Introduction
There is a continued interest in gold nanoparticles as catalysts
–7
for a broad range of selective chemical transformations.
1
8
These include the low temperature oxidation of CO,
especially when used as part of the processes for purifying a
9
,10
fuel cell feedstock,
1
hydrochlorination of ethyne, the selective oxidation of
the synthesis of vinyl chloride by the
14
free conditions, the alloying of gold with palladium leads to
a twenty five-fold enhancement in activity with simultaneous
1
1
2,13
14
and alcohols to aldehydes as well
alkenes to epoxides
16
retention of selectivity.
While palladium alone is an
exceptionally effective catalyst for the direct hydrogenation
1
5
as selective hydrogenation. Part of this intense fascination
with gold chemistry at the nanoscale is the very wide range of
new applications for gold as a catalyst that are being discovered
on a regular basis. Furthermore, since gold is considered to be
the most noble of metals these observations arouse the interest
of many scientists interested in discovering the origins of this
enhanced catalytic activity at the nanoscale.
17
of molecular oxygen to produce hydrogen peroxide, the
alloying of gold with the palladium significantly enhances
18–23
the activity and selectivity.
These two specific reactions
demonstrate that gold–palladium alloy nanoparticles can give
enhanced reactivity over monometallic gold and palladium
alone, although it should be noted that this enhancement is
not a universal feature for the oxidation of alcohols. In our
16,18–23
a
Cardiff Catalysis Institute, School of Chemistry, Cardiff University,
early studies
wet co-impregnation method to prepare gold–palladium
we used a relatively straightforward
Main Building, Park Place, Cardiff, UK CF10 3AT.
E-mail: mailto:hutch@cardiff.ac.uk; Fax: +44 29 2087 4059;
Tel: +44 29 2087 4059
Center for Advanced Materials and Nanotechnology, Lehigh
University, 5 East Packer Avenue, Bethlehem, PA 18015-3195, USA
23
nanoparticles that were either homogeneous alloys
2
or
2
24
b
core-shell structures. Subsequently, we recently showed
that catalysts with enhanced activity can be prepared using a
5
142 | Phys. Chem. Chem. Phys., 2009, 11, 5142–5153
This journal is ꢀc the Owner Societies 2009

View Article Online
sol-immobilisation method, and in this paper we continue
to investigate the advantages of this new alternative
methodology, specifically for the oxidation of benzyl alcohol
as a model substrate.
4 4
prepared solution of NaBH (0.1 M, NaBH /Au (mol/mol) = 5)
was then added to form a red sol. The solution was stirred for
30 minutes. Then, the required amount of the stock aqueous
PdCl solution was added, followed by the desired amount of
2
The oxidation of primary alcohols to aldehydes is an
2
NaBH (NaBH /Pd (mol/mol) = 5), obtaining a dark brown
4
4
5–27
important laboratory and commercial procedure.
Aldehydes are valuable both as intermediates and as
sol. The solution was stirred for a further 30 minutes.
2
5
high value components for the perfume industry. Often
oxidations of this type are carried out using stoichiometric
oxygen donors such as chromate or permanganate, but these
reagents are expensive and have serious toxicity and environ-
(
c) Au@Pd sol. To an aqueous PdCl2 solution of the
desired concentration the required amount of a PVA solution
1 wt%) was added (PVA/(Au + Pd) (wt/wt) = 1.2); a 0.1 M
freshly prepared solution of NaBH (NaBH /Pd (mol/mol) = 5)
was then added to form a brown sol. The solution was stirred
for 30 minutes. Then, the desired amount of the stock HAuCl
aqueous solution was added, followed by the desired amount
of NaBH (NaBH /Au (mol/mol) = 5), obtaining a dark
(
4
4
2
7
mental issues associated with them. Given these limitations,
there is substantial interest in the development of hetero-
geneous catalysts that use molecular O as the oxidant. Au
4
2
nanocrystals have been shown to be highly effective for the
oxidation of alcohols with O in an aqueous base, in particular
4
4
2
brown sol. The solution was stirred for a further 30 minutes.
Two supports have been used, namely, TiO (Degussa P25)
diols and triols; although under these conditions, the product
2
2
8–30
is the corresponding mono-acid, not the aldehyde.
However, gold in the absence of base has recently been shown
and carbon (Darco G60, Aldrich). After 30 min of sol
generation, the colloid was immobilised by adding activated
carbon or titania (acidified at pH 1 by sulfuric acid) under
vigorous stirring conditions. The amount of support material
required was calculated so as to have a total final metal loading
1
4
to be highly effective for the oxidation of alcohols, and when
alloyed with palladium can give enhanced activity under
1
6
solvent free conditions. The observation of high activity
using molecular oxygen under solvent free conditions is of
particular importance for the transformation of alcohols using
a green chemistry approach. In our previous work, we demon-
strated that the combination of gold and palladium could
provide a further pronounced enhancement in activity when
of 1% wt. The metal ratio for the 1 wt% Au + Pd/TiO
was 1:1 molar (we selected the Au + Pd/TiO catalyst with
:1 by wt loading so that it was comparable with catalysts
2
catalyst
2
1
16
prepared previously by impregnation ). For all other catalysts
a 1:1 wt ratio was used (Pd:Au = 2:1 molar). After
2
4
the two metals are deposited together by sol-immobilisation.
2 h the slurry was filtered, the catalyst washed thoroughly
with distilled water (neutral mother liquors) and dried at
In this paper we extend this approach to investigate whether or
not the activity can be further enhanced by modifying the
precise order in which the sols of the two metals are formed
before immobilisation onto the support. In addition, in
the oxidation of benzyl alcohol we observe toluene as a
1
20 1C overnight. The calcined catalysts were pre-treated at
00 1C under static air for 3 h.
4
2
4
Impregnation method. 1 wt% Pd-only, 1 wt% Au-only and
by-product and in this paper we investigate how to minimize
toluene formation.
Au–Pd bimetallic catalysts were prepared by impregnation of
carbon (Darco G60, Aldrich) or TiO (Degussa P25), via an
impregnation method using aqueous solutions of PdCl
Johnson Matthey) and/or HAuCl O (Johnson Matthey).
2
2
Experimental
(
4
Á3H
2
For the 0.5%Au–0.5%Pd/carbon catalyst, the detailed
preparation procedure employed is described below. An
aqueous solution of HAuCl Á3H O (2 ml, 5 g dissolved in
Catalyst preparation
Sol-immobilisation method. Three methods were used to
prepare sol immobilised Au–Pd catalysts based on the
sequence of metal addition and reduction. For the preparation
of (Au–Pd)/support materials an aqueous solution of PdCl2
4
2
water (250 ml)) and an aqueous solution of PdCl (0.83 ml, 1 g
2
in water (25 ml)) were simultaneously added to carbon (3.8 g).
The paste formed was ground and dried at 110 1C for 16 h and
calcined in static air, typically at 400 1C for 3 h.
(
Johnson Matthey) and HAuCl
the desired concentration were prepared. Polyvinylalcohol
PVA) (1 wt% aqueous solution, Aldrich, MW = 10 000,
0% hydrolyzed) and an aqueous solution of NaBH (0.1 M)
were also prepared.
4
Á3H
2
O (Johnson Matthey) of
(
Benzyl alcohol oxidation
8
4
Benzyl alcohol oxidation was carried out in a stirred reactor
100 mL, Autoclave Engineers Inline MagneDrive III). The
(
vessel was charged with alcohol (40 mL) and catalyst (0.1 g).
The autoclave was then purged 5 times with oxygen leaving the
vessel at 10 bar gauge. The stirrer was set at 1500 r.p.m. and
the reaction mixture was raised to the required temperature
and the reaction time was started when the required reaction
time was reached. Samples from the reactor were taken
periodically, via a sampling system. For the analysis of the
products a GC-MS and GC (a Varian star 3400 cx with a 30 m
CP-Wax 52 CB column) were employed. The products were
identified by comparison with known standards. For the
(
a) Au + Pd sol. To an aqueous PdCl2 and HAuCl4
solution of the desired concentration, the required amount
of a PVA solution (1 wt%) was added (PVA/(Au + Pd)
(
(
wt/wt) = 1.2); a freshly prepared solution of NaBH
0.1 M, NaBH /(Au + Pd) (mol/mol) = 5) was then added
4
4
to form a dark-brown sol.
(
b) Pd@Au sol. To an aqueous HAuCl4 solution of the
desired concentration the required amount of a PVA solution
1 wt%) was added (PVA/(Au + Pd) (wt/wt) = 1.2); a freshly
(
This journal is ꢀc the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 5142–5153 | 5143

View Article Online
quantification of the amounts of reactants consumed and
products generated, the external calibration method was used.
Catalyst characterisation
Samples for examination by transmission electron microscopy
(
TEM) were prepared by dispersing the catalyst powder in
high purity ethanol, then allowing a drop of the suspension to
evaporate on a holey carbon film supported by a 300 mesh
copper TEM grid. Samples were then subjected to bright field
diffraction contrast imaging experiments in order to determine
particle size distributions. The instrument used for this
analysis was a JEOL 2000FX TEM operating at 200 kV.
X-Ray photoelectron spectra were recorded on a Kratos
Axis Ultra DLD spectrometer employing a monochromatic
AlK(X-ray source (75–150 W) and analyser pass energies of
1
60 eV (for survey scans) or 40 eV (for detailed scans).
Samples were mounted using double-sided adhesive tape and
binding energies referenced to the C(1s) binding energy of
adventitious carbon contamination which was taken to be
2
84.7 eV.
Results and discussion
Fig. 1 Effect of catalyst loading on benzyl alcohol oxidation with
Oxidation of benzyl alcohol with Au + Pd/TiO
2
1 wt% Au + Pd/TiO
E benzaldehyde, ’ toluene. Reaction conditions: 120 1C, PO2
0.2 bar, stirring rate 1500 rpm, 2 h reaction time.
2
: (a) conversion and (b) selectivity, Key:
=
The effect of the reaction conditions on the oxidation of benzyl
alcohol was investigated using 1 wt% Au + Pd/TiO . First the
1
2
effect of catalyst mass was investigated using the same reaction
conditions (120 1C, PO2 10.2 bar, stirring rate 1500 r.p.m., 2 h
reaction time). In the absence of catalyst no conversion was
observed. Increasing the catalyst mass up to 100 mg had a
linear effect on the benzyl alcohol conversion and turn over
frequency (TOF) remained constant (Fig. 1), indicating that in
this range no diffusion limitations were apparent. Product
selectivity is affected by catalyst loadings above 50 mg, and
benzaldehyde and toluene are the major products, while
benzoic acid and benzyl benzoate are formed in minor
amounts. Benzaldehyde selectivity decreases as the catalyst
mass is increased, and this is coupled with a higher selectivity
to toluene. This is confirmed by data at iso-conversion of
benzyl alcohol (10%) which show these trends clearly
benzaldehyde was higher at the lower stirring speed and the
selectivity to benzaldehyde steadily increased with stirring
speed above 500 rpm (Fig. 2). Data at iso-conversion confirm
this observation (Table 2). By combining the selectivities at
different stirring speeds with the TOF-values one is able to see
the contributions of the two pathways to the total turnover
as a function of stirring. This reveals that the toluene
À1
contribution quickly rises to ca. 4000 h at 500 rpm and
stays constant thereafter, while the TOF for benzaldehyde
pathway increases continuously and, interestingly, in
proportion to the square root of the stirring speed. We
consider this behaviour to be similar to that observed for
forced convection to a planar surface (cf. the square root
dependence of current on rotation speed using a rotating disc
electrode). We therefore propose that toluene formation arises
wholly from reaction of adsorbed species on the surface, while
oxidation involves oxygen that is not adsorbed but reacts with
a suitable intermediate at the surface.
(
Table 1). At higher catalyst mass a slight decrease in TOF
was apparent indicating that diffusion limitations were
present, hence in subsequent experiments a catalyst mass of
1
00 mg was used as standard. We then investigated the effect
of stirring rate and oxygen pressure. The reactions are carried
out in a stirred autoclave reactor and involve a three phase
system (solid catalyst, liquid and gas phase reactants) and the
oxygen has to dissolve in the benzyl alcohol and the solubility
will be influenced by these parameters. Increasing the stirring
rate will increase the mixing between the liquid and gas phases
in the reaction mixture and this will increase the availability of
oxygen at the catalyst surface. Consequently, increasing the
stirring rate increased the rate of reaction (Fig. 2) and at the
highest stirring rate possible with the autoclave reactor we
have used, i.e. 1500 rpm, the selectivity of benzaldehyde was
highest. The highest increase in rate was observed when the
We subsequently investigated the effect of oxygen partial
pressure. There was no effect on rate but a marked increase in
benzaldehyde selectivity was observed (Fig. 3). Data at
iso-conversion of benzyl alcohol confirm these trends
(Table 3). The oxygen partial pressure should, presumably,
affect the equilibrium oxygen concentration in solution, so the
absence of any effect implies that oxygen, whose availability is
controlled by the stirring speed, very rapidly reacts with
surface-adsorbed species. Hence the reaction is zero order in
oxygen and the initial activation of benzyl alcohol on the
catalyst surface does not depend on the partial pressure of
oxygen; however, the subsequent formation of either
benzaldehyde or toluene is heavily dependent on the concen-
tration of oxygen at the surface. The formation of the products
À1
stirring rate was increased from 250 rpm (TOF 5850 h ) to
À1
5
00 rpm (TOF 10 000 h ). Interestingly, the selectivity to
5
144 | Phys. Chem. Chem. Phys., 2009, 11, 5142–5153
This journal is ꢀc the Owner Societies 2009

View Article Online
Table 1 Effect of catalyst loading in the benzyl alcohol oxidation at iso-conversion with the dried 1 wt% (Au + Pd)/TiO
immobilisation
2
catalyst prepared by sol
a
Selectivity (%)
À1b
Catalyst weight/mg
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
TOF/h
2
5
00
00
4
0
10
10
10
10
3.3
1.1
6.3
1.3
22.4
23.1
25.9
36.1
72.8
72.8
66.6
60.2
0.8
1.3
0.6
1.2
0.7
1.7
0.6
1.2
0
0
0
0
14 860
14 330
14 280
11 630
1
2
a
b
= 150 psi, stirring rate 1500 rpm. Calculation of TOF (h ) after
À1
Reaction conditions: benzyl alcohol, 0.1–0.2 g of catalyst, T = 120 1C, pO
2
0.5 h of reaction. TOF numbers were calculated on the basis of total loading of metals.
Fig. 2 Effect of stirring speed on benzyl alcohol oxidation with
wt% Au + Pd/TiO : (a) turn over frequency and (b) conversion
1
2
Fig. 3 Effect of oxygen pressure on benzyl alcohol oxidation with
wt% Au + Pd/TiO : (a) turn over frequency and (b) selectivity, Key:
and selectivity, Key: Â conversion, E benzaldehyde and ’ toluene.
Reaction conditions: 120 1C, PO2 = 10.2 bar, 100 mg of catalyst, 2 h
reaction time.
1
2
E benzaldehyde and ’ toluene. Reaction conditions: 120 1C, stirring
rate 1500 rpm, 100 mg of catalyst, 2 h reaction time.
can, therefore, be rationalised in terms of a series of pathways
(Fig. 4). It is considered that benzyl alcohol initially adsorbs
dissociatively on the surface of the metal nanoparticles giving
benzyl and hydroxyl intermediates. The benzyl surface species
is further dehydrogenated to give a surface benzylidene species
and surface adsorbed hydrogen atoms. The benzylidene then
Table 2 Effect of stirring speed in the benzyl alcohol oxidation at iso-conversion with the dried 1 wt%(Au + Pd)/TiO
immobilisation
2
catalyst prepared by sol
a
Selectivity (%)
À1b
Stirring speed/rpm
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
TOF/h
2
5
000
500
50
00
10
10
10
10
0.7
0.3
0.0
0.5
23.8
34.4
29.7
22.2
73.3
63.5
68.1
76.3
1.3
0.8
0.6
0.4
0.9
1.0
1.6
0.7
0
0
0
0
5850
10 030
12 830
16 060
1
1
a
b
= 150 psi, stirring rate 250–1500 rpm. Calculation of TOF (h ) after
À1
Reaction conditions: benzyl alcohol, 0.05 g of catalyst, T = 120 1C, pO
.5 h of reaction. TOF numbers were calculated on the basis of total loading of metals.
2
0
This journal is ꢀc the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 5142–5153 | 5145

View Article Online
Table 3 Effect of pressure in the benzyl alcohol oxidation at iso-conversion with the dried 1 wt%(Au + Pd)/TiO
immobilisation
2
catalyst prepared by sol
a
Selectivity (%)
À1b
Pressure/psi
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
TOF/h
5
00
50
0
10
10
10
0.4
0.6
0.5
33.0
24.8
22.0
65.0
70.1
76.2
0.7
1.8
0.5
0.9
2.7
0.8
0
0
0
16 010
16 250
16 420
1
1
a
b
= 50–150 psi, stirring rate 1500 rpm. Calculation of TOF (h ) after
À1
Reaction conditions: benzyl alcohol, 0.05 g of catalyst, T = 120 1C, pO
2
0.5 h of reaction. TOF numbers were calculated on the basis of total loading of metals.
situation but cannot be the main route to benzaldehyde. It
seems plausible to suggest that, if the carbonyl oxide is indeed
an intermediate, it is probably surface-bound, subsequently
fragmenting to yield free benzaldehyde and an oxo-species on
the surface, capable of generating a further molecule of
benzaldehyde.
Hence the key factor controlling selectivity is the availability
of oxygen at the catalyst surface. We tested this proposition by
reacting benzyl alcohol in the absence of oxygen with a
2
Au + Pd/TiO catalyst prepared by impregnation. Toluene
was observed when the reaction was carried out under equiva-
lent reaction conditions and also toluene and benzaldehyde
were observed in approximately equal amounts. In addition, if
the catalyst has a marked dehydrogenation activity then the
amount of surface hydrogen will be increased leading to higher
toluene formation. The catalyst prepared from the addition of
a Pd sol on a pre-formed Au sol (Pd@Au sol) resulted in a
catalyst with enhanced hydrogenation/dehydrogenation activity
and hence higher yields of toluene can be expected. This is
consistent with the nanoparticle morphology as the surface of
these particles can be expected to be rich in Pd which has a
higher hydrogenation activity than Au. The catalysts prepared
by sol immobilisation are more active than the 5 wt%
Fig. 4 Schematic representation of chemisorption of benzyl alcohol
and subsequent formation of toluene in competition with oxidation to
benzaldehyde.
2
reacts with O to give benzaldehyde or the surface hydrogen
reacts with either benzyl or benzylidene to give toluene. It is
worth noting that the following exothermic chemical equation
can be written:
2
Au + Pd/TiO catalysts prepared by co-impregnation that
1
6
we investigated previously. However, the selectivity observed
for toluene is significantly higher with the very active
sol-immobilised catalysts, and this, we consider, is related to
the higher hydrogenation activity for these catalysts.
2
PhCH OH - PhCH + PhCHO + H O
2 3 2
In the absence of added oxygen therefore, toluene and
benzaldehyde should be produced in equimolar amounts. This
situation is most nearly achieved using catalysts supported on
An important factor that has to be considered for hetero-
geneous catalysts operating in a three phase system concerns
2
carbon that are not calcined. For TiO -supported catalysts,
3
2
equimolar toluene and benzaldehyde is most closely attained
at low oxygen pressures and low but, perhaps surprisingly, not
the lowest stirring speeds.
catalyst stability. One major problem is that the metallic
components can leach into solution and this can lead either to
a loss in catalyst activity on subsequent uses, or can enable the
formation of an active homogeneous catalyst. Previously, we
The chemistry of the formation of benzaldehyde detailed in
Fig. 4 deserves comment. Benzylidene generated in its ground
triplet state, for example in the gas or liquid phases, in the
presence of oxygen would be expected rapidly to form the
carbonyl oxide, PhC–O–O (often represented as an ylid,
have shown that Au + Pd/TiO
2
catalysts prepared by
co-impregnation are not stable when used as catalysts for
the direct synthesis of hydrogen peroxide if they are not
2
2
calcined at 400 1C. This is due to significant loss of the Au
and Pd which did not produce an active homogeneous species
but did lead to significant loss in catalyst activity, and this was
+
À
31
PhCQO –O ). Carbonyl oxides are very reactive species;
in competition with isomerisation to the dioxirane and
dimerisation to the 1,2,4,5-tetroxane, reaction with benzyl
alcohol would be expected to yield the hemiperacetal 1 and
thence benzyl benzoate by dehydration, and reaction with
benzaldehyde would yield stilbene ozonide 2, a precursor of
benzoic acid (Fig. 5). The presence of both benzyl benzoate
and benzoic acid among the products of the reactions over
supported Au–Pd catalysts, albeit with low selectivities,
indicates the relevance of this chemistry in the heterogeneous
2
3
related to the structure of the Au–Pd alloy nanoparticles.
The effect of catalyst re-use was therefore investigated for
the non calcined Au + Pd/TiO catalyst (Table 4). These
2
experiments were performed by initially reacting a 200 mg
batch of the catalyst for 30 min (120 1C, 1500 r.p.m., PO2 10.2 bar),
recovering the material by filtration, washing the solid with
acetone three times followed by drying (20 1C, 16 h).
Unfortunately the amount of catalyst recovered was always
5
146 | Phys. Chem. Chem. Phys., 2009, 11, 5142–5153
This journal is ꢀc the Owner Societies 2009

View Article Online
Fig. 5 Chemistry of carbonyl oxides in the gas or liquid phases.
Table 4 Recycling catalytic tests in the liquid phase oxidation of benzyl alcohol with a 1% (Au + Pd)/TiO
immobilisation method
2
catalyst prepared by the sol
a
Selectivity (%)
À1b
Number of uses
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
TOF/h
1
2
3
4
st use
nd use
rd use
th use
18.0
14.3
12.9
12.5
2.4
2.0
3.2
3.4
35.8
32.5
32.4
30.0
60.5
64.1
62.9
64.9
0.6
0.7
0.7
0.6
0.7
0.8
0.8
1.1
0
0
0
0
10 020
7970
7210
6960
a
b
= 150 psi. Calculation of TOF (h ) after 0.5 h of reaction. TOF
À1
Reaction conditions: benzyl alcohol, 0.1–0.2 g of catalyst, T = 120 1C, pO
2
numbers were calculated on the basis of total loading of metals.
lower than the initial amount used due to experimental losses
in recovery/filtration and hence different masses of catalyst
were used (Table 4). However, the TOF was constant within
experimental error and consequently it is concluded that the
uncalcined catalysts prepared using the sol-immobilisation
method are stable and can be readily reused.
at 0.5 h reaction (Table 5) and the order of activity was observed
to be Pd@Au/carbon 4 Au + Pd/carbon 4 Au@Pd/carbon.
The carbon-supported catalysts were much more active by
about a factor of two when compared with the TiO -supported
2
catalysts (Tables 5 and 6), however, they were much less
selective to benzaldehyde.
Investigation of the effect of order of metal addition
Effect of calcination on catalyst performance
There are three ways in which the sol-immobilised catalysts
can be prepared, i.e. the metals can be co-immobilised or either
metal can be deposited first and the second in a subsequent
step. Such methods can be expected to have significant effects
on activity and selectivity as a result of the different structures
The sol-immobilised catalysts were calcined at 400 1C and the
activity was investigated for the oxidation of benzyl alcohol
(catalyst 100 mg, 120 1C, 1500 r.p.m., PO2 10.2 bar). The results
show that both sets of catalysts are markedly affected by
calcination, but the effect is far more significant for the carbon-
supported catalysts, which decrease in activity by about an
that can be made. We have investigated this for TiO - and
2
carbon-supported catalysts. For the TiO -supported catalysts
2
order of magnitude, whereas the TiO -supported catalysts halve
2
(
Table 5) the initial rate, as determined by TOF at 0.5 h
reaction was highest for Pd@Au/TiO , with the order of
activity being Pd@Au/TiO 4 Au@Pd/TiO 4 Pd + Au/TiO
When compared at 50% conversion all catalysts gave very
similar selectivities but Pd + Au/TiO did give a slightly
in activity (Table 7). The calcined carbon-supported catalysts
are all more selective for the synthesis of benzaldehyde, and
comparison at iso-conversion shows that this is not just an
effect of lower conversion and hence these calcined catalysts are
more selective for partial oxidation (Table 8).
2
2
2
2
.
2
higher toluene selectivity and this catalyst was the most active
at 2 h reaction time (Table 6). For the carbon-supported
catalysts the order of addition had a marked effect on the
initial rate (Table 5), although all showed similar activity at 2 h
reaction time (Table 5). Pd@Au/carbon was significantly
more active than the other two catalysts based on TOF
The activity of the 1 wt% Au + Pd/carbon catalyst
prepared by sol-immobilisation and calcined at 400 1C could
be compared with the Au–Pd catalysts prepared by
1
6
co-impregnation that we have reported previously. The
sol-immobilised catalysts have a significantly higher activity
À1
À1
(TOF 1780 h ) than the co-impregnated catalyst (TOF 490 h ).
This journal is ꢀc the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 5142–5153 | 5147

View Article Online
Table 5 Benzyl alcohol oxidation results after 2 h reaction with Au–Pd catalysts supported on carbon and titania prepared by immobilisation
method: effect of the order of metal reduction in the preparation
a
Selectivity (%)
À1b
Catalyst
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
TOF/h
1
1
1
1
1
1
% (Au + Pd)/TiO
2
61.2
57.8
48.7
81.1
70.4
79.2
0.5
0.2
0.4
0.5
0.4
0.7
26.7
18.1
23.0
40.9
35.1
28.8
69.2
77.1
72.3
55.0
62.6
65.0
1.7
1.8
1.9
1.3
2.0
2.7
1.9
2.8
2.5
2.1
0.0
2.7
0
0
0
0.2
0
15 360
19 250
17 360
35 400
41 930
24 310
% (Pd@Au)/TiO
% (Au@Pd)/TiO
% (Au + Pd)/C
% (Pd@Au)/C
% (Au@Pd)/C
2
2
0
a
b
= 150 psi, stirring rate 1500 rpm. Calculation of TOF (h ) after 0.5 h
À1
Reaction conditions: benzyl alcohol, 0.1g of catalyst, T = 120 1C, pO
2
of reaction. TOF numbers were calculated on the basis of total loading of metals.
Table 6 Benzyl alcohol oxidation results at iso-conversion with Au–Pd catalysts supported on carbon and titania prepared by immobilisation
method: effect of the order of metal reduction in the preparation
a
Selectivity (%)
Catalyst
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
1
1
1
1
1
1
% (Au + Pd)/TiO
2
50
50
50
50
50
50
0.5
0.2
0.4
0.3
0.3
0.6
26.9
23.0
23.0
44.0
41.6
38.9
69.1
72.8
72.2
51.3
54.6
57.1
1.5
1.5
1.9
1.7
1.6
1.6
2.0
2.5
2.5
2.7
1.8
1.7
0
0
0
0.1
0.1
0.1
% (Pd@Au)/TiO
% (Au@Pd)/TiO
% (Au + Pd)/C
% (Pd@Au)/C
% (Au@Pd)/C
2
2
a
Reaction conditions: benzyl alcohol, 0.1g of catalyst, T = 120 1C, pO
2
= 150 psi.
Table 7 Benzyl alcohol oxidation results after 2 h reaction with Au–Pd catalysts supported on carbon and titania prepared by immobilisation
a
method and calcined at 400 1C in air: effect of the order of metal reduction in the preparation
Selectivity (%)
À1b
Catalyst
Conv. (%)
Benzene
Toluene
Benzaldehyde
Benzoic acid
Benzyl benzoate
Acetal
TOF/h
1
1
1
1
1
1
% (Au + Pd)/TiO
2
22.7
32.8
33.5
6.7
6.0
7.2
0.8
0.6
0.4
5.5
5.2
1.7
22.2
21.3
25.9
2.4
2.9
4.7
69.7
71.6
68.7
78.7
75.1
79.1
1.9
1.9
1.3
3.6
4.3
3.6
5.5
4.6
3.7
9.8
12.6
10.8
0
0
0
0
0
0
3940
8650
8780
2490
2430
3410
% (Pd@Au)/TiO
% (Au@Pd)/TiO
% (Au + Pd)/C
% (Pd@Au)/C
% (Au@Pd)/C
2
2
a
b
= 150 psi, stirring rate 1500 rpm. Calculation of TOF (h ) after 0.5 h
À1
Reaction conditions: benzyl alcohol, 0.1g of catalyst, T = 120 1C, pO
2
of reaction. TOF numbers were calculated on the basis of total loading of metals.
Table 8 Benzyl alcohol oxidation results at iso-conversion with Au–Pd catalysts supported on carbon and titania prepared by immobilisation
method: effect of the order of metal reduction in the preparation and catalyst pre-treatment
a
Selectivity (%)
Catalyst
Catalyst pre-treatment Conv. (%) Benzene Toluene Benzaldehyde Benzoic acid Benzyl benzoate Acetal
1
1
1
1
1
1
1
1
1
1
1
1
% (Au + Pd)/C
% (Pd@Au)/C
% (Au@Pd)/C
% (Au + Pd)/C
%Pd@Au/C
Dried
Dried
Dried
Calcined 400 1C/Air
Calcined 400 1C/Air
Calcined 400 1C/Air
Dried
Dried
Dried
Calcined 400 1C/Air
Calcined 400 1C/Air
Calcined 400 1C/Air
5
5
5
5
5
0.2
0.3
0.6
6.1
6.8
2.2
0.4
0.3
0.4
0.8
0.5
0.5
44.2
40.1
35.2
3.6
3.4
4.3
21.3
33.1
28.8
22.3
23.8
28.3
54.9
56.8
59.5
76.1
73.7
85.1
75.3
63.8
66.9
69.8
71.0
66.9
0.6
1.0
1.4
3.5
4.1
2.1
1.0
0.9
0.9
1.8
1.3
1.1
0.1
1.8
3.3
10.6
12.0
6.3
2.0
1.9
3.0
5.4
3.3
3.2
0
0
0
0
0
0
0
0
0
0
0
0
%Au@Pd/C
%Au + Pd/TiO
5
2
2
20
20
20
20
20
20
%Pd@Au/TiO
%Au@Pd/TiO
2
2
%Au + Pd/TiO
%Pd@Au/TiO
%Au@Pd/TiO
2
2
a
Reaction conditions: benzyl alcohol, 0.1g of catalyst, T = 120 1C, pO
2
= 150 psi.
5
148 | Phys. Chem. Chem. Phys., 2009, 11, 5142–5153
This journal is ꢀc the Owner Societies 2009

View Article Online
Characterisation of sol immobilised catalysts
the Pd@Au and AuPd particle size distributions on carbon are
a little larger at B5.5 nm, and the Au@Pd on carbon are
slightly smaller at 3.8 nm in diameter. It also interesting to
note the Pd@Au, Au@Pd and AuPd alloy particles on TiO2
all have a significantly narrower size distribution than their
direct counterparts on carbon.
TEM results. TEM has been used to characterise the particle
size distributions for the 1 wt% Pd@Au, 1 wt% Au@Pd and
1
wt% AuPd sol-immobilised specimens on TiO2, and carbon
supports and the data are shown in Fig. 6 and 7, respectively.
In all cases, the sol-immobilised particles are homogeneously
dispersed over the support surfaces. Fig. 6 shows that the
Fig. 8 shows the Pd@Au particle size distribution data on
the TiO2 and carbon supports after calcinations for 3 h at
400 1C. It is clear that the thermal stability of the Pd@Au
alloy particles shows a strong dependence on support identity.
On TiO2 the Pd@Au particles exhibit some very limited
Pd@Au, Au@Pd and AuPd sols immobilised on TiO have
2
rather similar particle size distributions with median diameter
values all in the 4.0–4.5 nm range. By comparison, (see Fig. 7)
Fig. 6 Particle size distribution data, as determined from bright field TEM micrographs for the (a,b) 1 wt% Pd@Au/TiO
2
,(c,d) 1 wt%
Au@Pd/TiO and (e,f) 1 wt% AuPd/ TiO samples dried overnight at 120 1C.
2
2
This journal is ꢀc the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 5142–5153 | 5149

View Article Online
Fig. 7 Particle size distribution data, as determined from bright field TEM micrographs for the (a,b) 1 wt% Pd@Au/C, (c,d) 1 wt% Au@Pd/C
and (e,f) 1 wt% AuPd/ C samples dried overnight at 120 1C.
sintering with the median Pd@Au particle size increasing
from 4.4 nm to 5.2 nm. In stark contrast, the sintering
observed on carbon is much more pronounced with an
identical heat treatment causing the Pd@Au median size
to increase to more than 20 nm. This difference in sintering
behaviour correlates directly with the differing extent in
loss of catalytic activity observed on the two supports
following calcinations at 400 1C (i.e. a factor of 2 for
XPS results
Titania supported catalysts. Fig. 9 and 10 show the Au(4f)
and Pd(3d) spectra for the series of titania supported Au–Pd
catalysts. Also displayed on the Pd(3d) spectra are the Pd:Au
molar ratios derived from the integrated XPS spectra.
The Pd:Au molar ratio is as expected from the preparation
2
methods, i.e. for the 1 wt% Au + Pd/TiO catalyst is 1:1 and
is close to 2:1 for all other samples. Calcination at 400 1C does
not affect this ratio, in contrast to Au–Pd samples prepared
by impregnation of titania where calcination results in a
Pd@Au on TiO
2
and an order of magnitude for Pd@Au on
carbon).
5
150 | Phys. Chem. Chem. Phys., 2009, 11, 5142–5153
This journal is ꢀc the Owner Societies 2009

View Article Online
Fig. 8 Particle size distribution data, as determined from bright field TEM micrographs for (a,b) the 1 wt% Pd@Au/TiO
Pd@Au/C calcined at 400 1C for 3 hours.
2
and (c,d) the 1 wt%
Fig. 9 Au(4f) spectra for the TiO
2
supported Au–Pd catalysts—
AuPd: (a) dried and (b) calcined at 400 1C; Pd@Au: (c) dried,
Fig. 10 Pd(3d) spectra for the TiO
2
supported Au–Pd catalysts—
(
d) calcined at 200 1C (e) calcined at 400 1C; Au@Pd: (f) dried,
g) calcined 400 1C; AuPd: (h) dried, (i) used for benzyl alcohol
AuPd: (a) dried and (b) calcined at 400 1C; Pd@Au: (c) dried,
(
(
d) calcined at 200 1C (e) calcined at 400 1C; Au@Pd: (f) dried,
g) calcined 400 1C; AuPd: (h) dried, (i) used for benzyl alcohol
oxidation.
(
oxidation. Pd:Au atom ratios are shown on the spectra.
significant increase in the Pd:Au ratio due to the formation of
contain Pd predominantly in the metallic state which trans-
2+
forms almost completely to Pd after calcination at 400 1C.
2
a Pd(shell)–Au(core) morphology. The uncalcined catalysts
2
This journal is ꢀc the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 5142–5153 | 5151

View Article Online
intensities, a high metal dispersion for a given amount of
deposited metal resulting in a higher signal than for a lower
dispersion.
Conclusions
The oxidation of benzyl alcohol under solvent-free conditions
has been investigated using supported gold palladium
bimetallic nanoparticles and comparing their activity and
performance with mono-metallic catalysts. The catalysts
prepared using the sol immobilisation technique show
higher activity when compared with catalysts prepared by
impregnation. In particular, much lower concentrations of
the metals can be used and also the method can be tuned by
preparing samples with gold or palladium seed sols. The
Au–Pd catalysts were all more active than the corresponding
monometallic supported Au or Pd catalysts. For the
Fig. 11 Au(4f) spectra for the G60 supported Au–Pd catalysts—
AuPd: (a) dried and (b) calcined at 4001C; Pd@Au: (c) dried,
TiO -supported catalysts the order of metal addition
2
(d) calcined at 200 1C (e) calcined at 300 1C, (f) calcined at 4001C;
Au@Pd: (g) dried, (h) calcined 400 1C.
(i.e. whether a gold seed sol or a palladium seed sol was used
initially) in the preparation was not observed to be significant
with respect to selectivity. For the carbon-supported catalyst
the order of metal addition has a very marked affect on
activity and the carbon-supported catalysts are also more
significantly affected by heat treatment, e.g. calcination at
4
00 1C leads to the activity being decreased by an order of
magnitude, whereas the TiO -supported catalysts show a 50%
2
decrease in activity. The TEM analysis shows that the carbon-
supported catalysts are much more sensitive to sintering and
this is the cause of the lower activity observed. Toluene is
observed as an unwanted by-product of the reaction and
conditions have been identified that minimise its formation.
It is proposed that toluene and benzaldehyde are formed by
competing parallel reactions of the initial benzyl intermediate
via an adsorbed benzylidene species that can either be
hydrogenated or oxidised. In this way we identified reaction
conditions that maximised the availability of oxygen on the
catalyst surface and thereby favour the synthesis of benzaldehyde
which is the required selective oxidation product.
Fig. 12 Pd(3d) spectra for the G60 supported Au–Pd catalysts—
AuPd: (a) dried and (b) calcined at 400 1C; Pd@Au: (c) dried,
(d) calcined at 200 1C (e) calcined at 300 1C, (f) calcined at 400 1C;
Au@Pd: (g) dried, (h) calcined 400 1C. Pd:Au atom ratios are shown
on the spectra.
Acknowledgements
This work formed part of the EU AURICAT project
Contract HPRN-CT-2002-00174) and the EPSRC/Johnson
(
Carbon supported catalysts. Fig. 11 and 12 show the Au(4f)
and Pd(3d) spectra for the series of carbon supported Au–Pd
catalysts. Also displayed on the Pd(3d) spectra are the Pd:Au
molar ratios derived from the integrated XPS spectra. In
marked contrast to the titania-supported catalysts, we observe
a significant increase in the Pd:Au molar ratio upon
calcination. This is illustrated most clearly by the Pd@Au
series of catalysts comprising the uncalcined, material, and
samples calcined at 200 1C, 300 1C and 400 1C. The Pd:Au
ratio increases monotonically from 2.2 to 5.7 and this is
Matthey funded ATHENA project and we thank them for
funding this research. CJK and RT would also like to
acknowledge the financial support of the NSF Nano-
manufacturing program under contract # CMMI-0457602.
References
1
G. C. Bond and D. T. Thompson, Catal. Rev. Sci. Eng., 1999, 41,
19.
3
2 G. C. Bond and D. T. Thompson, Gold Bull., 2000, 33, 41.
3 M. Haruta, Gold Bull., 2004, 37, 27.
4 A. S. K. Hashmi, Gold Bull., 2004, 37, 51.
0
accompanied by the increasing oxidation of Pd species to
2
+
Pd . Interestingly, both the Au@Pd and Pd@Au samples
behave identically in this respect i.e. the order in which the
metals are deposited is not important with regard to the Pd:Au
ratio. The increasing Pd:Au ratio is consistent with the
5
R. Meyer, C. Lemaire, Sh. K. Shaikutdinov and H.-J. Freund,
Gold Bull., 2004, 37, 72.
6 G. J. Hutchings, Gold Bull., 2004, 37, 37.
7
A. S. K. Hashmi and G. J. Hutchings, Angew. Chem., Int. Ed.,
006, 45, 7896.
2
2
2
formation of core-shell particles but we must also bear
in mind the effect of particle size on detected XPS signal
8
M. Haruta, T. Kobayashi, H. Sano and N. Yamada, Chem. Lett.,
1987, 16, 405.
5
152 | Phys. Chem. Chem. Phys., 2009, 11, 5142–5153
This journal is ꢀc the Owner Societies 2009

View Article Online
9
P. Landon, J. Ferguson, B. E. Solsona, T. Garcia, A. F. Carley,
A. A. Herzing, C. J. Kiely, S. E. Golunski and G. J. Hutchings,
Chem. Commun., 2005, 3385.
20 B. E. Solsona, J. K. Edwards, P. Landon, A. F. Carley, A. Herzing,
C. J. Kiely and G. J. Hutchings, Chem. Mater., 2006, 18, 2689.
21 J. K. Edwards, B. Solsona, P. Landon, A. F. Carley, A. Herzing,
M. Watanabe, C. J. Kiely and G. J. Hutchings, J. Mater. Chem.,
2005, 15, 4595.
22 J. K. Edwards, B. E. Solsona, P. Landon, A. F. Carley, A. Herzing,
C. J. Kiely and G. J. Hutchings, J. Catal., 2005, 236, 69.
23 J. K. Edwards, A. F. Carley, A. A. Herzing, C. J. Kiely and
G. J. Hutchings, Faraday Discuss., 2008, 138, 225.
24 J. A. Lopez-Sanchez, N. Dimitratos, P. Miedziak, E. Ntainjua,
J. K. Edwards, D. Morgan, A. F. Carley, R. Tiruvalam,
C. J. Kiely and G. J. Hutchings, Phys. Chem. Chem. Phys, 2008,
10, 1921.
1
0 P. Landon, J. Ferguson, B. E. Solsona, T. Garcia, S. Al-Sayari,
A. F. Carley, A. Herzing, C. J. Kiely, M. Makkee, J. A. Moulijn,
A. Overweg, S. E. Golunski and G. J. Hutchings, J. Mater. Chem.,
2006, 16, 199.
1
1
1 G. J. Hutchings, J Catal., 1985, 96, 292.
2 A. K. Sinha, S. Seelan, S. Tsubota and M. Haruta, Angew. Chem.,
Int. Ed., 2004, 43, 1546.
3 M. D. Hughes, Y.-J. Xu, P. Jenkins, P. McMorn, P. Landon,
D. I. Enache, A. F. Carley, G. A. Attard, G. J. Hutchings, F. King,
E. H. Stitt, P. Johnston, K. Griffin and C. J. Kiely, Nature, 2005,
1
4
37, 1132.
25 R. A. Sheldon and J. K. Kochi, Metal-Catalyzed Oxidations of
Organic Compounds, Academic Press, New York, 1981.
26 G. ten Brink, I. W. C. E. Arends and R. A. Sheldon, Science, 2000,
287, 1636.
1
4 A. Abad, P. Conception, A. Corma and H. Garcia, Angew. Chem.,
Int. Ed., 2005, 44, 4066.
5 A. Corma and P. Serna, Science, 2006, 313, 332.
1
1
6 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 and G. J. Hutchings, Science, 2006, 311, 362.
7 V. R. Choudhary and C. Samanta, J. Catal., 2006, 238, 28.
8 P. Landon, P. J. Collier, A. J. Papworth, C. J. Kiely and
G. J. Hutchings, Chem. Commun., 2002, 2058.
9 P. Landon, P. J. Collier, A. F. Carley, D. Chadwick,
A. J. Papworth, A. Burrows, C. J. Kiely and G. J. Hutchings,
Phys. Chem. Chem. Phys., 2003, 5, 1917.
27 M. Pagliaro, S. Campestrini and R. Ciriminna, Chem. Soc. Rev.,
2005, 34, 837.
28 L. Prati and M. Rossi, J. Catal., 1998, 176, 552.
29 F. Porta, L. Prati, M. Rossi and G. Scari, J. Catal., 2002, 211, 464.
30 F. Porta and L. Prati, J. Catal., 2004, 224, 397.
31 H. Tomioka, in Reactive Intermediates in Chemistry,
ed. R. A. Moss, M. S. Platz and M. Jones, Wiley, 2004, p. 375.
32 R. A. Sheldon, M. Wallau, I. W. C. E. Arends and U. Schuchardt,
Acc. Chem. Res., 1998, 31, 485.
1
1
1
This journal is ꢀc the Owner Societies 2009
Phys. Chem. Chem. Phys., 2009, 11, 5142–5153 | 5153