Solvent-free aerobic oxidation of alcohols using supported gold palladium nanoalloys prepared by a modified impregnation method

Article abstract of DOI:10.1039/c4cy00387j

The synthesis of stable, supported, bimetallic nanoalloys with controlled size, morphology and composition is of great practical importance. Compared to their monometallic analogues, such materials exhibit remarkable enhancement in functional properties, which can be exploited in various fields including catalysis. Recently, we have reported a simple excess anion modification of the impregnation method to prepare supported gold-palladium catalysts which gives very good control over the particle sizes and the composition without using any stabilizer ligands in the preparation. Here, we report the results from a comparative study of using this modified impregnation catalyst for the solvent-free aerobic oxidation of alcohols in two different reactors: a glass stirred reactor and a micro packed bed reactor under batch and continuous mode respectively. These modified impregnation catalysts are exceptionally active and more importantly, when tested in a micro packed bed reactor under flow conditions, are observed to be stable for several days without any sign of deactivation in contrast to the same catalyst prepared by the sol immobilization method in the presence of stabilizer ligands which showed a 3-5% decrease in conversion over 10-12 h. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cy00387j

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Solvent-free aerobic oxidation of alcohols using
supported gold palladium nanoalloys prepared
by a modified impregnation method†‡
Cite this: Catal. Sci. Technol., 2014,
4, 3120
a
ab
cd
a
Moataz Morad, Meenakshisundaram Sankar, Enhong Cao, Ewa Nowicka,
ae
a
a
a
Thomas E. Davies, Peter J. Miedziak, David J. Morgan, David W. Knight,
ae
cd
a
Donald Bethell, Asterios Gavriilidis and Graham J. Hutchings*
The synthesis of stable, supported, bimetallic nanoalloys with controlled size, morphology and composition
is of great practical importance. Compared to their monometallic analogues, such materials exhibit
remarkable enhancement in functional properties, which can be exploited in various fields including
catalysis. Recently, we have reported a simple excess anion modification of the impregnation method to
prepare supported gold–palladium catalysts which gives very good control over the particle sizes and the
composition without using any stabilizer ligands in the preparation. Here, we report the results from a
comparative study of using this modified impregnation catalyst for the solvent-free aerobic oxidation of
alcohols in two different reactors: a glass stirred reactor and a micro packed bed reactor under batch and
continuous mode respectively. These modified impregnation catalysts are exceptionally active and more
importantly, when tested in a micro packed bed reactor under flow conditions, are observed to be stable
for several days without any sign of deactivation in contrast to the same catalyst prepared by the sol
immobilization method in the presence of stabilizer ligands which showed a 3–5% decrease in conversion
over 10–12 h.
Received 28th March 2014,
Accepted 21st May 2014
DOI: 10.1039/c4cy00387j
www.rsc.org/catalysis
activity of gold nanoparticles, having a particle size above
Introduction
6
,7
10 nm, decreases dramatically. It has been reported that by
The exceptional catalytic activity of gold nanoparticles has
been reported in the literature for more than a quarter of a
combining gold with palladium at the nanoscale, the catalytic
activity of these particles could be increased by several fold
for the selective aerobic oxidation of alcohols. The catalytic
activity for these bimetallic nanoparticles follows a similar
trend to supported monometallic gold particles i.e., catalytic
activity increases with decreasing particle size. Due to this
similarity, many methods of preparing supported monome-
1
,2
8,9
century. Since the crucial pre-requisite of these gold nano-
particles, to be catalytically active, is their size, numerous
attempts have been made to prepare these nanoparticles as
small as possible and with a very narrow particle size
3
–5
distribution. It is important to note here that the catalytic
tallic gold nanoparticles have been directly translated to pre-
a
4,10,11
Cardiff Catalysis Institute, School of Chemistry, Cardiff University, Main Building,
pare supported bimetallic nanoparticles.
However, the
Park Place, Cardiff, CF10 3AT, UK. E-mail: hutch@cardiff.ac.uk;
Fax: +44 (0)2920 874 030
presence of a second metal can impose significant complica-
tions in terms of the particle size distribution, the mode of
the arrangement of these two metals and the compositional
b
Inorganic Chemistry and Catalysis Group, Debye Institute of Nanomaterials and
Science, Utrecht University, Utrecht, The Netherlands
c
12
Department of Chemical Engineering, University College London, Torrington Place,
variation from particle to particle. These added complica-
London WC1E 7JE, UK
tions pose significant challenges to the catalysis chemist to
design methods to control these parameters. In an attempt to
overcome these complications many synthetic methods have
been developed over the years which include (a) wet impreg-
d
Materials Chemistry Centre, University College London, 20 Gordon Street,
London WC1H 0AJ, UK
e
Department of Chemistry, University of Liverpool, Liverpool L69 3BX, UK
†
MM and EN prepared the catalysts and performed catalytic experiments under
9
13
4
the supervision of MS. EC performed experiments in MPBR under the supervision
of AG. DM performed the XPS characterization of all the materials. EN, PJM and
TD performed other characterization of catalysts. DWK and DB helped in the
discussion and interpretation of the data. GJH directed this research program.
All authors commented and contributed in writing this manuscript.
nation, (b) incipient wetness, (c) deposition precipitation,
1
4,15
(
d) sol immobilization or colloidal deposition
and (e)
16
physical grinding of acetate salts. All of these methods have
significant advantages as well as disadvantages; for example
the conventional wet impregnation method (CIm), for preparing
supported gold–palladium nanoparticles, is a very simple
‡
Electronic supplementary information (ESI) available. See DOI: 10.1039/
c4cy00387j
3120 | Catal. Sci. Technol., 2014, 4, 3120–3128
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
method and has widely been practised. However, the resul-
tant catalysts have nanoparticles ranging from sub-nm clus-
ters up to nanoparticles 100 nm in diameter and there is no
effective control over the particle size. To overcome this dis-
advantage, very small gold–palladium nanoparticles were pre-
pared in solution as sols using stabilizers ligands, for example
polyvinyl alcohol (PVA) or polyvinyl pyridine (PVP), which
exhibit very small particle sizes with a narrow distribution.
These preformed nanoparticles are subsequently immobilized
onto a solid support, for example a metal oxide; this method
is called colloidal deposition or sol immobilization (SIm).
Such catalysts have an improved catalytic activity for solvent-
In the present work we have studied the supported
0.5 mol%Au–0.5 mol%Pd nanoparticles catalysts in a batch
reactor for the solvent-free aerobic oxidation of benzyl alco-
hol; the catalysts were prepared by the MIm method, and
these were contrasted with catalysts prepared by CIm and
SIm, that have identical Au : Pd ratios and metal content. The
best catalyst was then tested for the oxidation of a range of
alcohols. The long-term stability of equimolar 1%AuPd/TiO2,
prepared by SIm and MIm, was studied using micro-packed
bed reactor (MPBR) for the oxidation of benzyl alcohol.
Results and discussion
Solvent free aerobic oxidation of alcohols in glass stirred
1
0,17
free aerobic oxidation of alcohols,
but pose a different set
of problems: the stabilizer ligands often block the active sites
of such catalysts and these have to be removed either by high
reactor (GSR)
1
8–20
temperature calcination or by solvothermal treatments.
1%AuPd/TiO2 catalyst, having equimolar amounts of gold
and palladium, was prepared by the MIm method using
Another major problem with this method is the size depen-
dent compositional variation; all the smaller particles are gold
21,22
“excess-anion” methodology.
2
For comparison, 1%AuPd/TiO ,
3
,12
rich whereas the larger nanoparticles are palladium rich.
with identical metal content, was prepared by both CIm and
SIm methods whose synthesis procedures and characterisa-
Prüße and co-workers have previously reported the beneficial
effect of using hydrochloric acid to dissolve tetrachloroauric
acid when using the incipient wetness technique for the oxi-
9
,25
tion have been reported in detail elsewhere.
The catalytic
activities of these catalysts were tested for the solvent-free
aerobic oxidation of benzyl alcohol in a glass stirred reactor
(GSR) at 120 °C and the results are presented in Table 1.
From the conversion data it is clear that there are substantial
differences in the catalytic activities of the catalysts prepared
by three different methodologies which follow the order:
MIm > SIm ≫ CIm. After 2 h reaction, CIm, SIm and MIm
catalysts resulted in conversion of 14%, 65% and 88% respec-
tively. This clearly demonstrates the superiority of the current
modified impregnation methodology over the other method-
ologies, especially the SIm method which has been reported
to provide catalysts which are exceptionally active for the oxi-
2
1
dation of glucose. They used a reduction step rather than
calcination to activate the catalysts and reported that the
hydrochloric acid precursor solution led to a final catalyst with
a smaller average particle size when compared to the aqueous
precursor. Recently, we reported an “excess-anion” modifica-
tion of the CIm method to prepare 0.5 wt%Au–0.5 wt.%Pd
2
supported on TiO nanoparticles with precise control over
2
2
the size, composition and nanostructure. The bimetallic
catalysts, prepared by this modified impregnation (MIm)
method, outperformed the identical catalysts prepared by
CIm and SIm methods for the direct synthesis of hydrogen
2
2
27
peroxide from a dilute mixture of hydrogen and oxygen.
Aerobic oxidation of benzyl alcohol has long been used as
dation of alcohols. To test the applicability of this current
MIm methodology for supports other than TiO , we then
2
2
3,24
a model reaction for testing oxidation catalysts.
Solvent-
investigated the use of MgO as the support for gold palladium
free oxidation of benzyl alcohol yields several products includ-
ing toluene, benzoic acid, benzyl benzoate and dibenzyl ether,
besides the desired benzaldehyde. Some of these products
could be eliminated by careful design of the catalysts after
Table 1 Aerobic oxidation of alcohols in a glass stirred reactor using
1
2
%AuPd supported on TiO and MgO catalysts prepared by different
a
2
5,26
methods
understanding the origins of them.
Conventionally, these
reactions are performed either in a glass stirred reactor or in
an autoclave reactor in batch mode. Carrying out the reaction
in a batch mode is not suitable for studying the long-term
stability of the catalyst; continuous reaction is favoured for
this. For multiphase reactions, involving a solid catalyst,
liquid alcohol and gaseous oxidant, packed-bed flow reactor
systems have the advantages of simplifying the process by
eliminating separation of liquid products from the solid cata-
lyst. For these three phase reactions, where mass transfer can
be a limiting factor, micro reactors offer improved reaction
rates and performance when compared to conventional
macro reactors. Additionally, the small channels of the micro
reactor offer efficient gas–liquid mixing along with sufficient
gas–liquid–solid contact resulting in improved mass transfer
in the packed bed of the catalyst.
b
c
Synthesis Conv. Selectivity
TOF
(h
−1
Substrate Catalyst
method
(%)
(%)
)
Benzyl
1%(AuPd)/TiO
2
CIm
SIm
MIm
14
65
88
26
33
8
18
7
12
80
80
77
99
98
42
45
41
47
940
alcohol
4600
6100
1900
2300
1
%(AuPd)/MgO SIm
MIm
SIm
MIm
d
Crotyl
1%(AuPd)/TiO
2
390
920
760
d
alcohol
1
%(AuPd)/MgO SIm
MIm
1200
a
Alcohol: 2 g, catalyst: 0.02 g, temperature: 120 °C, pO : 1 bar,
2
stirring speed: 1000 rpm, time = 2 h. CIm catalysts: calcined at
00 °C/4 h. SIm catalysts: dried at 120 °C/16 h. MIm catalysts: prepared
in a 0.58 M HCl and reduced at 400 °C/4 h. Selectivity to the
corresponding aldehyde. TOF after 2 h of reaction. TOF after 4 h
of reaction.
4
b
c
d
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3120–3128 | 3121

View Article Online
Paper
Catalysis Science & Technology
nanoparticles. We specifically made this choice because
MgO-supported AuPd catalysts have previously been reported
to be effective in switching off toluene formation in the
solvent-free aerobic oxidation of benzyl alcohol to yield
preparation, the selectivity to benzaldehyde decreases from
85% with zero HCl addition to 75% with HCl addition to give
a final concentration of 0.58 M in the preparation. A further
increase in the HCl concentration from 0.58 M to 2 M
improves the selectivity from 75% to 84%. It is clear from
these data that increasing the concentration of HCl leads to
both an increase in the conversion of the benzyl alcohol and
the selectivity of the desired product, benzaldehyde. However,
we have previously reported that catalysts prepared with con-
centrations of greater than 0.58 M HCl are not stable under
reaction conditions for the direct synthesis of hydrogen
2
5
benzaldehyde with approximately 99% molar selectivity.
The results from the oxidation of benzyl alcohol using
%AuPd/MgO catalysts prepared by both the SIm and MIm
1
methods are also presented in Table 1. Both catalysts show
the equivalent selectivity towards the formation of benzalde-
hyde (98–99%) but the conversion for the MIm catalyst was
found to be 33% whereas the SIm catalyst resulted in 26%
conversion. This set of results show that this MIm methodol-
2
2
peroxide. The presence of chloride ions, used in MIm meth-
odology, plays a crucial role in controlling the size and com-
ogy delivers superior AuPd catalysts not only with TiO as the
2
2
2
support, but also with MgO. In our initial experiments, we
have used only benzyl alcohol as the substrate for the oxida-
tion reaction and our next objective was to test the tolerance
of this catalyst to other substrates. To this end we evaluated
the catalysts for the oxidation of crotyl alcohol (an aliphatic,
primary allylic alcohol). These results are also presented in
Table 1; again, the supported AuPd catalysts prepared by MIm
methodology exhibited superior catalytic activity to those
materials prepared by the SIm method.
position of supported AuPd nanoalloys. Supported AuPd
nanoalloys prepared by CIm and SIm methodologies exhibit
3
a complex size-dependent compositional variation. For cata-
lysts prepared by the CIm methodology, all the larger parti-
cles are either exclusively gold or gold-rich, whereas all the
2
8,29
smaller particles are palladium-rich.
Catalysts prepared
by the SIm methodology display a reverse trend, i.e., the
smaller particles are Au-rich and the larger particles are
Pd-rich. In the new MIm methodology, the presence of excess
chloride ions eliminates the larger “gold-rich” particles and
Effect of HCl concentration on MIm catalysts
22
increases the Au concentration in the bimetallic particles.
−
It is considered that Cl concentration plays a crucial role in
The concentration of Au in the bimetallic particles increases
with the increase in the chloride ion concentration. In
addition to this, there was no size-dependent compositional
variation observed for the catalysts prepared by this MIm
methodology i.e., bimetallic particles of all sizes showed
keeping the metal ions separated during the impregnation
2
2
stage, as discussed in our earlier study. Thus the effect of
HCl concentration on the catalyst performance was studied.
1
%AuPd/TiO catalysts prepared by MIm with HCl concentra-
2
1
2
tion of 0, 0.58, 1 and 2 M respectively were prepared and eval-
uated for the oxidation of benzyl alcohol to benzaldehyde
at 120 °C in a GSR. The results are given in Fig. 1. It can be
observed that in general the conversion of benzyl alcohol
increases with HCl concentration from 76% at 0 M HCl (CIm)
to 86% at 2 M HCl. With increasing HCl addition during the
almost the same concentration of Au and Pd. The superior
catalytic activity for the materials prepared by this MIm meth-
odology, observed in this work, is due to this precise control
over the composition of Au and Pd in the nanoalloy particle.
Effect of thermal treatment conditions on MIm catalysts
The effect of thermal treatment of the material during the
catalyst preparation was studied by varying the temperature
and the gas environment of this step. A 1%AuPd/TiO catalyst
2
prepared by the MIm method was reduced at temperatures of
2
50, 300, 400 and 500 °C in flowing H /Ar. The aerobic oxida-
2
tion of benzyl alcohol in the GSR was carried out to examine
the performance of the reduced catalysts. The results, given
in Table 2, show that the best performance is from the cata-
lyst which is treated at 400 °C in H /Ar flow, achieving a con-
2
version of 88% with selectivity 77% to benzaldehyde and 21%
to toluene. The enhancement in activity as the reduction tem-
perature increases could in part be due to both a decrease in
the level of chloride on the surface coupled with surface oxi-
dation of the Pd leading to the formation of PdO. Further
increasing the thermal treatment temperature to 500 °C
causes a marked decrease in conversion to ~26% and there-
fore we heat-treated catalysts at lower temperatures to main-
tain activity. With the thermal treatment at 400 °C in flowing
air, the conversion of benzyl alcohol decreases to almost half
Fig. 1 Effect of HCl amount, during catalyst preparation, on the
catalytic performance of 1%AuPd/TiO
glass stirred reactor (GSR). Reaction conditions: benzyl alcohol = 2 g,
catalyst = 0.02 g, temperature = 120 °C, pO = 1 bar, stirring speed =
000 rpm, time = 2 h.
2
(MIm reduced at 400 °C) in a
of that obtained at a same temperature with H /Ar flow, to
2
2
1
45%, although the selectivity to benzaldehyde increases to 85%.
3122 | Catal. Sci. Technol., 2014, 4, 3120–3128
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 2 Aerobic oxidation of benzyl alcohol in a glass stirred reactor using 1%AuPd supported on TiO
2
with different catalyst high temperature
a
treatment procedures
b
Selectivity (%)
Heat treatment
temperature °C
Heat treatment
atmosphere
Catalyst (synthesis method)
Conv. (%)
Aldehyde
Toluene
1
%AuPd/TiO
2
(MIm)
250
H
H
H
H
2
2
2
2
/Ar
/Ar
/Ar
/Ar
79.7
78.5
88
25.7
45
77
80
77
92
85
85
80
22
19
21
6
11
11
18
3
4
5
4
00
00
00
00
Air
/Ar
Air
1%AuPd/TiO
2
(CIm)
400
00
H
2
76
14
4
a
2
Substrate = 2 g, catalyst = 0.02 g, temperature = 120 °C, pO = 1 bar, stirring speed = 1000 rpm, time = 2 h. All the MIm catalysts were
b
prepared in a 0.58 M HCl solution. Other products include benzyl benzoate and benzoic acid.
Table 4 Comparison of fresh and spent catalysts (1%AuPd/TiO
by CIm, SIm and MIm methodologiesa
2
) prepared
For comparison, the catalyst prepared by the conventional
impregnation technique was thermally treated at 400 °C in
2
H /Ar flow. This catalyst gave a conversion of 76% (Table 2)
Synthesis
methodology
Nature of
the catalyst
b
c
demonstrating that the thermal treatment of the catalysts
plays an important role in their activity and furthermore that
reduction treatment should be considered as an alternative
to oxidation treatments when preparing gold palladium bime-
tallic catalysts for oxidation reactions. The high-temperature
thermal treatment plays an important role in deciding the
final nanostructure of the bimetallic nanoalloys. High-
temperature calcination leads to the formation of a structure
with a Au-core Pd-shell, whereas high temperature reduction
Particle size
Pd/Au ratio
CIm
SIm
Fresh
Spent (GSR)
Fresh
4.7
3.9
2.5
2.5
3.8
3.0
3.0
4.2
4.0
4.5
3.8
8.1
3.8
1.1
3
2.4
3.7
4.8
1.8
1
Spent (GSR)
Spent (MPBR)d
Spent (MPBR)
e
Spent (MPBR)
Fresh
Spent (GSR)
Spent (MPBR)
Spent (MPBR)
MIm
1.2
1.1
2
8,29
results in a homogeneous random alloy structure.
This is
f
reflected in the Pd/Au surface ratio determined by XPS mea-
surements (Tables 3 and 4). Core–shell nanostructures always
give higher Pd/Au ratio compared to homogeneous random
alloy structure (Tables 3 and 4). As the reduction temperature
increases from 250 °C to 500 °C the surface Pd/Au ratio
The MIm catalysts were prepared in a 0.58 M HCl solution.
a
b
Corresponding histograms are presented in ESI. The particle sizes
c
were measured using TEM. The Pd/Au ratios were measured by XPS.
d
e
Oxidation in O
2
at 120 °C for 1 h. Activation in reaction mixture
overnight and continuing
the reaction the next day. Activation in reaction mixture from 80 to
20 °C, before reaction.
from 80 to 120 °C, kept at 80 °C in O
2
f
2
+
0
decreases from 1.5 to 0.9. The Pd : Pd ratio decreases with
increase in reduction temperature. Based on our previous
1
STEM and XPS results in the case of 0.5 wt.%Au–0.5 wt.Pd/TiO
and our current XPS results, we confidently conclude that in
the current catalytic system (0.5 mol%Au–0.5 mol%Pd/TiO
the gas-phase reduced catalysts have a homogeneous random
alloy structure and the calcined catalysts have a Aucore–Pdshell
nanostructure. This difference in nanostructures is one of
the contributing factors for the observed difference in the
2
catalytic activities of the materials that underwent different
thermal treatments. For alcohol oxidation, the homogeneous
random alloy structure is more active compared to the
core–shell structure. It is important to note that the catalysts
prepared by SIm methodology always have a homogeneous
random alloy structure.
2
)
Table 3 Quantified XPS results for 0.58 M HCl MIm prepared samples. The Cl, Pd and Au atom% compositions are mean values for the surface region
samples by XPS. The CIm H /Ar sample is included as a comparison
2
MIm
CIm
a
b
2+
0c
a
b
2+
0c
Treatment
% Cl
% Pd
% Au
Pd/Au
Pd : Pd
% Cl
% Pd
% Au
Pd/Au
Pd : Pd
d
H
H
H
H
2
2
2
2
/Ar 250 °C
/Ar 300 °C
/Ar 400 °C
/Ar 500 °C
0.53
0.35
0
0
0
0.28
0.31
0.25
0.25
0.24
0.14
0.12
0.11
0.18
0.09
1.5
2.1
1.8
0.9
2.2
—
—
0.53
—
—
—
—
0.49
—
—
—
—
0.06
—
—
—
—
7.7
—
—
—
—
2.23 : 1
—
—
1.1 : 1
0.75 : 1
0.32 : 1
1.43 : 1
Air 400 °C
a
b
Calculated from the apparent Pd(3d) intensity, which incorporates intensity from the Au(4d5/2) component. Ratio corrected for the overlap
c
d
of the Pd(3d) doublet and the Au(4d5/2) component. Ratio of Pd species determined by curve-fitting analysis. Due to the weak Pd signal, the
ratio cannot be ascertained, however the most intense spectral intensity is at a binding energy for Pd .
0
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3120–3128 | 3123

View Article Online
Paper
Catalysis Science & Technology
Catalytic stability of 1%AuPd/TiO MIm catalysts
(ca. 80% conversion) which is very similar to that found for
the reaction in a GSR at the same space time and exhibited
very good stability over the whole time period, here the space
times in MPBR and GSR.
2
In the literature, there are reports describing catalysts which
are extremely active but lack stability over longer reaction
3
0,31
times.
To test the long-term stability of the present cata-
For comparison, a SIm catalyst of the same wt% composi-
tion was tested under exactly the same reaction conditions in
a MPBR and the results are presented in Fig. 2a. The SIm
catalyst showed a lower reactivity (ca. 55% conversion on
average) by comparison to MIm catalyst at a same catalyst
loading. Deactivation of the catalyst was observed as the con-
version declined ca. 3–5% in 10–12 h. The catalyst appeared
to be reactivated after refilling the syringe which caused an
interruption to the reaction and a very short period of expo-
sure of the catalyst in O2 under the reaction temperature.
The sensitivity of the SIm catalyst to reaction conditions was
also observed in a study of the effect of pretreatment on the
catalyst stability, where a set of experiments were designed to
treat the catalyst in the reaction mixture at 80–120 °C or in
lyst, it was evaluated for the continuous aerobic oxidation of
benzyl alcohol in a micro-packed bed reactor (MPBR) at
1
1
20 °C over 75 h. During the test, samples were taken at
–2 h intervals and analysed during the daytime, while
overnight only one sample was taken and the analysed data
was recorded as an average over the night period. Since a
syringe pump was used for the delivery of benzyl alcohol to
the MPBR, the syringe (2.5 ml) needed to be refilled every
1
2 hours which caused about a 5 minute interruption to the
reaction. The details about MPBRs and experimental proce-
3
2,33
dure for this reaction have been reported elsewhere
and
are also available in the experimental section of this paper.
The conversion and selectivity of the reaction on the MIm
catalyst against the reaction time are shown in Fig. 2b. It is
observed that the MIm catalyst showed catalytic activity
O at 120–140 °C before starting the reaction. Fig. 3 shows
2
the conversions of benzyl alcohol on the SIm and the MIm
catalysts at different pretreatment stages. It can be seen that
for the SIm catalyst, pretreating in the reaction mixture from
8
0 to 120 °C produces catalysts with poor stability. The cata-
lyst stability is much improved when the catalyst is pretreated
in O at 120 °C. However, for the MIm catalyst, the stability
2
of the catalyst is almost unaffected by the pre-treatments.
This study of the catalyst stability in micro-packed bed
reactors (MPBRs) clearly showed the long term stability of
the MIm catalyst and its much improved catalytic activity.
The enhanced stability of the MIm catalyst we observe in
these studies is considered to be due to the combination of
gold-rich bimetallic particles with almost no size-dependent
compositional variation coupled with the high-temperature
thermal treatment.
Fig. 2 Oxidation of benzyl alcohol in micro-packed bed reactors
(
MPBRs) using 1%(AuPd)/TiO
2
catalysts prepared by a) the sol immobiliza-
Fig. 3 Effect of pre-treatment on the stability of the SIm and the MIm
(0.58 M HCl, reduced at 400 °C) catalysts. Reaction conditions: SIm
0.0062 g, MIm 0.005 g; catalyst particle size 53–63 μm; benzyl alcohol
tion (SIm) and b) the modified impregnation (MIm, 0.58 M HCl, reduced
at 400 °C) methods. Reaction conditions: catalyst weight 0.004 g, catalyst
particle size 53–63 μm, benzyl alcohol flow 0.003 mL min , O
0
−
1
−1
−1
2
flow
flow 0.003 mL min , O
2
flow 0.6 mL min ; 120 °C. Pre-treatment:
, C—120 °C in O
−1
.6 mL min , 120 °C.
A—80 to 120 °C in reaction mixture, B—140 °C in O
2
2
.
3124 | Catal. Sci. Technol., 2014, 4, 3120–3128
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
X-ray Photoelectron Spectroscopy (XPS) analysis
Conclusions
XPS analysis of the MIm samples subjected to the different
treatment regimes, revealed similar Au(4f) and Pd(3d) bind-
ing energies throughout the series of treatments, however,
appreciable differences in the Pd /Pd and Pd/Au ratios were
noted (Table 3).
A modified impregnation methodology for preparing supported
gold–palladium nanoalloy catalysts in the presence of an excess
amount of chloride ligands in the impregnation medium has
been developed. Catalysts prepared by this method are found
to be extremely active for the aerobic oxidation of alcohols
and superior to similar composition catalysts prepared by
other methodologies such as sol-immobilization and conven-
tional impregnation. The catalysts prepared by this new meth-
odology are found to be more active and stable for the
oxidation of alcohols to corresponding aldehyde than those
made by conventional impregnation and sol-mobilisation
methods. This provides both the academic research commu-
nity and the industrial community, with a very convenient
and reproducible methodology for preparing supported Au–
Pd nanoalloy catalysts with high activity and stability without
using any exotic ligands or stabilizers, while at the same time
not compromising the catalytic activity.
2
+
0
Considering first the Au(4f) signals, binding energies ca.
8
2.5 eV were observed which are much lower than those
expected for bulk-like metallic gold (ca. 84.0 eV), and even
those of small particles where a shift of ca. 0.5 eV to lower
binding energy is noted as a consequence of the lower sur-
3
4
face atom coordination. Clearly, the values reported here
are a further 1 eV lower and are consistent with those noted
1
8,35
previously for samples treated in H
2
.
This effect has been
attributed to an increased initial-state contribution to the
photoemission process, which has been calculated to mani-
3
6
fest as a lowering of the expected energy by 1.1–1.5 eV. Such
changes have also been ascribed to charge transfer from Pd
to Au, increasing the Au s-state occupancy during alloy for-
1
8
mation or electron transfer from support to nanoparticle.
Pd(3d) energies, determined through curve-fitting, are con-
Experimental section
Catalyst preparation
2
+
sistent with both metallic Pd (334.1 eV) and Pd (336.1 eV)
3
7
states, with the latter ascribed to PdO; with increasing tem-
Modified Impregnation (MIm) method. For the preparation
of 1%AuPd supported catalyst, HAuCl ·3H O (Sigma Aldrich)
peratures under the reducing H atmosphere, a steady rise of
2
4
2
the metallic Pd signal is observed (Table 3). The presence of
Pd is unsurprising avoided since samples are air handled
before insertion in to the spectrometer.
and PdCl (Sigma Aldrich) were used as a metals precursors.
2
2
+
In a typical synthesis run, the requisite amount of aqueous
−
1
−1
gold (8.9 ml ml ) and palladium solution (6 mg ml ; HCl
concentration: 0, 0.58, 1, 2 M) were charged into a clean
50 mL round-bottom flask fitted with a magnetic stirrer bar.
The volume of the solution was adjusted using deionized
water to a total volume of 16 mL. The flask was then
immersed into an oil bath sitting on a magnetic stirrer hot
plate. The solution was stirred vigorously at 1000 rpm and the
temperature of the oil bath was raised from room temperature
to 60 °C over a period of 10 min. At 60 °C, 1.98 g of the metal
It is interesting to see how the Pd/Au ratio is affected by
the different thermal treatments. The theoretical Pd : Au ratio
in the Au–Pd catalysts is 1 : 1 by weight; equivalent to a Pd : Au
4
atom ratio of 1.9 : 1, with the ratio for the best catalyst in the
present study not too dissimilar from this theoretical value.
Higher temperature treatment (500 °C) decreases this ratio to
0
.9 : 1, possibly indicating Au enrichment of the surface of an
Au–Pd alloy. It may be this which is responsible for the
higher selectivity towards benzaldehyde (albeit at much lower
conversion), with the Au minimising the sites for dispropor-
oxide support material [TiO (Degussa Evonik P25) or MgO
2
(BDH)] was added slowly over a period of 8–10 min with con-
stant stirring. After the completion of addition of the support
material, the slurry was stirred at 60 °C for an additional
15 min. Following this, the temperature of the oil bath was
raised to 95 °C, and the slurry was stirred at that temperature
overnight until all the water evaporated leaving a dry solid.
Subsequently the solid powder was transferred into a mortar
and pestle and was ground thoroughly to form a uniform mix-
ture. This was stored and designated as a “dried-only” sample.
For calcination/reduction, a portion of 400 mg of the uncalcined
sample was transferred and spread out over a glass calcination
boat (30 cm in length). This boat was then placed inside a cal-
3
tionation to toluene. By comparison, the CIm sample treated
at 400 °C under H /Ar, revealed a Pd/Au ratio of 7.7. Such a
2
large increase over that expected for homogeneous alloys, or
equally dispersed monometallic Au and Pd particles, may
reflect particle size changes, and would be consistent with the
weak signal, and therefore apparent low concentration, of Au.
2
Interestingly for the MIm H /Ar treated samples, Cl is detected
for samples treated below 400 °C, whereas for 400 °C and
above, no Cl is observed within the detection limits of the
XPS spectrometer employed in this study. Whilst it is known
that halides can have a promoting effect on the formation of
3
8
Pd–O bonds, and is demonstrated herein by the equivalent
cination furnace fitted with an inlet and outlet valve. Heating
2
+
−1
CIm sample which has a high level of Pd , with larger Cl
concentration (comparable to the MIm sample treated at
rate of 10 °C min under a steady flow of gas (5%H
2
/Ar or
Air) was used for calcination or reduction respectively in range
of different temperatures (250–500 °C). After the furnace was
cooled, the sample was used and denoted MIm catalyst.
Conventional Impregnation (CIm) method. Detailed synthesis
procedure for the CIm method is typically carried out as follows.
2
50 °C) it should not detract from the observation that with
increasing temperature and reductive atmosphere, the rela-
tive amount of metallic Pd increases together with the
removal of Cl in the formation of active and selective catalysts.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3120–3128 | 3125

View Article Online
Paper
Catalysis Science & Technology
To prepare 1%AuPd/TiO the requisite amount of the metal
a specific time, the stirring was stopped and the reactor was
immediately cooled in an ice bath. After cooling for 10 min,
the reactor was opened slowly and the contents were
centrifuged. An aliquot of the clear supernatant reaction mix-
ture (0.5 mL) was then diluted with mesitylene (0.5 mL) for
GC analysis. It was also verified that no reaction occurred in
the absence of the catalyst or in the presence of the catalyst
support alone. This experimental method is identical to our
standard reaction protocols for solvent-free benzyl alcohol
2
precursors (HAuCl ·3H O for Au and PdCl for Pd) were wet
4
2
2
impregnated on to the solid support [TiO
2
(Degussa P25)]. The
requisite amount of solid PdCl was added to a predetermined
2
volume of an aqueous solution of HAuCl ·3H O and stirred
4
2
vigorously at 80 °C until the palladium salt completely
dissolved. The metal concentrations were maintained in such
a way that the Pd/Au molar ratio equals 1 and the total weight
percentage of the metal was kept as 1%. After that, the
requisite amount of the support was added to this solution
under vigorous stirring conditions. The solution was agitated
at 80 °C in this way until it formed a paste, which was then
dried at 120 °C for 16 h and then calcined in static air, typically
at 400 °C for 4 h. This catalyst is labelled as CIm catalysts.
Sol Immobilization (SIm) method. Catalysts were prepared
using a well-established sol immobilization methodology
2
2,25
oxidation reported elsewhere.
The turnover frequency
(TOF) was calculated from the moles of substrate consumed
and the total number of moles of metal present in the cata-
lyst. Here, we assume that all kinds of metal sites are catalyti-
cally active.
Solvent-free aerobic oxidation of benzyl alcohol in micro
packed bed reactor (MPBR)
3
,18
whose procedure is reported elsewhere in detail.
cal catalyst synthesis (1%AuPd/TiO ) aqueous solutions of
PdCl (Sigma Aldrich) and HAuCl ·3H O (Sigma Aldrich) of
In a typi-
2
The overall size of the silicon/glass microreactor chip used
was 23 mm × 23 mm with a reaction channel dimension of
2
4
2
the desired concentrations were prepared. The metal concen-
trations were maintained in such a way that the Pd/Au molar
ratio equals 1 and the total weight percentage of the metal
was kept as 1%. Polyvinyl alcohol (PVA) (1 wt% aqueous solu-
tion, Aldrich, MW = 10 000, 80% hydrolyzed) and an aqueous
0
.6 mm (W) × 0.3 mm (H) × 190 mm (L). A pillar structure
(
small rectangular posts of 60 μm (W) × 1 mm (L)) was
arranged at 40 μm intervals near the outlet of the reaction
channel to retain the catalyst. A schematic representation
and the detailed procedure of operating MPBR are presented
elsewhere.
and then crushed to obtain the desired particle size fraction
by sieving. The catalyst was introduced into the reaction
channel through the gas inlet with the help of a vacuum
pulled at the outlet of the reaction channel. The packed
microreactor was then assembled with a heating and temper-
ature control unit. During the reaction, benzyl alcohol was
delivered into the reactor by a syringe pump. Gases were reg-
ulated by mass flow controllers (Brooks 5850TR) and directed
to the gas inlet of the reactor. The gas flow rate reported here
solution of NaBH (0.1 M) were also prepared. To an aqueous
32,33
4
The prepared catalyst powder was pelletized
2 4
mixture of PdCl and HAuCl solution of the desired concen-
tration, the required amount of a PVA solution (1 wt%) was
added (PVA/(Au + Pd) (w/w) = 1.3). A freshly prepared solution
−
1
4 4
of NaBH (0.1 M, NaBH /(Au + Pd) (mol mol ) = 5) was then
added to form a dark-brown sol. After 30 min of sol genera-
tion, the colloid was immobilized by adding the support
material (TiO
prepared with TiO
H SO was added under vigorous stirring. The amount of
2
, Degussa P25; MgO, BDH). For the catalysts
2
as the support, 1 drop of concentrated
2
4
support material required was calculated so as to have a total
final metal loading of 1 wt%. The metal ratio for the Au + Pd
bimetallic catalyst was 1 : 1 by weight. After 2 h, the slurry
was filtered, and the catalyst was washed thoroughly with 2 L
of distilled water (until the mother liquor was neutral) and
then dried at 120 °C overnight under static air.
−1
is in mL min at Standard Temperature and Pressure (STP,
2
into a small glass vial (2 ml) which was located in a cold trap
73 K and 1.01325 bar). The effluent from the reactor passed
(
ice-water bath), where gas and liquid were separated.
Experiments were normally started by treating the catalyst
with O
introducing benzyl alcohol at a specified flow rate. After
0 min, liquid product was collected and analyzed at
2
at 393 K for 60 min. The reaction was started by
3
Solvent-free alcohol oxidation in a batch rector
pre-determined time intervals during day time. Over the
night period, only one sample was taken and the analysed
data was recorded as an average over the night period. The
syringe (2.5 ml) needed to be refilled every 12 hours which
caused about a 5 min interruption to the reaction.
Alcohols oxidation was carried out in a Radleys carousel reac-
tor using a 50 mL glass stirred reactor. In a typical reaction,
the requisite amounts of catalyst and substrate were charged
into the reactor at room temperature which was then purged
2
with the required gas (O ) three times before the reactor was
The space times (ST) in the MPBR and GSR are defined as:
sealed using a Teflon screw threaded cap. The reactor was
always connected to the gas-line to ensure the consumed gas
was replenished. The pressure was monitored using the pres-
sure gauge fitted in the inlet line.
mass of catalyst (gcat
)
1
In MPBR, ST 
gcat s galc ,
1
alcohol mass flow-rate (galc s )
The reactor with the reaction mixture was loaded into a
preheated heating block, which was maintained at the reac-
tion temperature. The reaction started by commencing stir-
ring inside the reactor with a magnetic bar at 1000 rpm. After
mass of catalyst (gcat
)
1
In GSR, ST 
 reaction time (s) gcat s galc .
mass of alcohol (galc )
3126 | Catal. Sci. Technol., 2014, 4, 3120–3128
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Product analysis
8 A. Villa, N. Janjic, P. Spontoni, D. Wang, D. S. Su and
L. Prati, Appl. Catal., A, 2009, 364, 221–228.
For the analysis of the products, GC-MS (Waters, GCT
Premier) and GC (a Varian star 3800 cx with a 30 m CP-Wax
9
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, 362–365.
5
2 CB column) were employed. The products were identified
by comparison with known standards. For the quantification
of the amounts of reactants consumed and products gener-
ated an external calibration method, using mesitylene as the
external standard, was used.
1
0 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.,
2
008, 10, 1921–1930.
1
1
1 J. M. Campelo, D. Luna, R. Luque, J. M. Marinas and
A. A. Romero, ChemSusChem, 2009, 2, 18–45.
X-ray Photoelectron Spectroscopy (XPS)
A Kratos Axis Ultra DLD system was used to collect XPS spec-
tra using monochromatic Al Kα X-ray source operating at
2 R. C. Tiruvalam, J. C. Pritchard, N. Dimitratos,
J. A. Lopez-Sanchez, J. K. Edwards, A. F. Carley, G. J. Hutchings
and C. J. Kiely, Faraday Discuss., 2011, 152, 63–86.
3 C. Baatz and U. Pruesse, J. Catal., 2007, 249, 34–40.
4 D. Wang, A. Villa, F. Porta, L. Prati and D. Su, J. Phys. Chem.
C, 2008, 112, 8617–8622.
5 P. Haider, B. Kimmerle, F. Krumeich, W. Kleist, J. D. Grunwaldt
and A. Baiker, Catal. Lett., 2008, 125, 169–176.
6 S. A. Kondrat, G. Shaw, S. J. Freakley, Q. He, J. Hampton,
J. K. Edwards, P. J. Miedziak, T. E. Davies, A. F. Carley,
S. H. Taylor, C. J. Kiely and G. J. Hutchings, Chem. Sci., 2012,
3, 2965–2971.
1
20 W. Data was collected in the Hybrid mode of operation,
using a combination of magnetic and electrostatic lenses,
over an area approximately 700 × 300 μm at pass energies of
1
1
2
4
0 and 160 eV for high resolution and survey spectra respec-
tively. Magnetically confined charge compensation was used
to minimize sample charging and the resulting spectra were
calibrated to the C(1s) line at 284.7 eV.
1
1
Transmission Electron Microscopy (TEM)
Analysis was performed in a JEOL 2100 TEM instrument. The
powdered samples were dispersed in deionised water before
being added to holey carbon film on copper grids. EDX analy-
sis was performed using Oxford Instruments SDD detector
XmaxN 80T and analysed using Aztec software. Particle size
distributions were determined using ImageJ software and
17 N. Dimitratos, F. Porta and L. Prati, Appl. Catal., A, 2005,
291, 210–214.
18 J. A. Lopez-Sanchez, N. Dimitratos, C. Hammond,
G. L. Brett, L. Kesavan, S. White, P. Miedziak, R. Tiruvalam,
R. L. Jenkins, A. F. Carley, D. W. Knight, C. J. Kiely and
G. J. Hutchings, Nat. Chem., 2011, 3, 551–556.
1
50 particles were counted for this.
1
9 R. M. Rioux, H. Song, M. Grass, S. Habas, K. Niesz,
Notes
J. D. Hoefelmeyer, P. Yang and G. A. Somorjai, Top. Catal.,
2
006, 39, 167–174.
The authors declare no competing financial interests.
2
0 S. O. Blavo, E. Qayyum, L. M. Baldyga, V. A. Castillo,
M. D. Sanchez, K. Warrington, M. A. Barakat and J. N. Kuhn,
Top. Catal., 2013, 56, 1835–1842.
Acknowledgements
2
2
1 C. Baatz, N. Decker and U. Prüße, J. Catal., 2008, 258(1),
The authors acknowledge the EPSRC for funding and M.M.
thanks the Umm Al-Qura University, Saudi Arabia for Ph.D.
studentship. Authors thank the Research Complex at Harwell
for allowing us to use the TEM facility.
1
65–169.
2 M. Sankar, Q. He, M. Morad, J. Pritchard, S. J. Freakley,
J. K. Edwards, S. H. Taylor, D. J. Morgan, A. F. Carley,
D. W. Knight, C. J. Kiely and G. J. Hutchings, ACS Nano,
2
012, 6, 6600–6613.
References
2
3 T. Mallat and A. Baiker, Chem. Rev., 2004, 104, 3037–3058.
1
2
G. J. Hutchings, J. Catal., 1985, 96, 292–295.
M. Haruta, T. Kobayashi, H. Sano and N. Yamada, Chem.
Lett., 1987, 16, 405–408.
J. Pritchard, L. Kesavan, M. Piccinini, Q. He, R. Tiruvalam,
N. Dimitratos, J. A. Lopez-Sanchez, A. F. Carley, J. K. Edwards,
C. J. Kiely and G. J. Hutchings, Langmuir, 2010, 26, 16568–16577.
P. J. Miedziak, Q. He, J. K. Edwards, S. H. Taylor,
D. W. Knight, B. Tarbit, C. J. Kiely and G. J. Hutchings,
Catal. Today, 2011, 163, 47–54.
24 W. B. Hov, N. A. Dehm and R. W. J. Scott, J. Catal., 2008,
253, 22–27.
25 M. Sankar, E. Nowicka, R. Tiruvalam, Q. He, S. H. Taylor,
C. J. Kiely, D. Bethell, D. W. Knight and G. J. Hutchings,
Chem. – Eur. J., 2011, 17, 6524–6532.
26 G. Kovtun, T. Kameneva, S. Hladyi, M. Starchevsky,
Y. Pazdersky, I. Stolarov, M. Vargaftik and I. Moiseev,
Adv. Synth. Catal., 2002, 344, 957–964.
27 N. Dimitratos, F. Porta and L. Prati, Appl. Catal., A, 2005,
291, 210–214.
3
4
5
6
M. Haruta, Catal. Today, 1997, 36, 153–166.
N. Lopez, T. V. W. Janssens, B. S. Clausen, Y. Xu,
M. Mavrikakis, T. Bligaard and J. K. Norskov, J. Catal., 2004,
28 P. Paalanen, B. M. Weckhuysen and M. Sankar, Catal.
Sci. Technol., 2013, 3, 2869–2880.
2
23, 232–235.
29 G. J. Hutchings and C. J. Kiely, Acc. Chem. Res., 2013, 8,
1759–1772.
7
M. C. Daniel and D. Astruc, Chem. Rev., 2004, 104, 293–346.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 3120–3128 | 3127

View Article Online
Paper
Catalysis Science & Technology
3
3
0 S. Hermans and M. Devillers, Appl. Catal., A, 2002, 235, 253–264.
1 M. Comotti, P. C. Della, R. Matarrese and M. Rossi, Angew.
Chem., Int. Ed., 2004, 43, 5812–5815.
2 E. Cao, M. Sankar, E. Nowicka, Q. He, M. Morad,
P. J. Miedziak, S. H. Taylor, D. W. Knight, D. Bethell,
C. J. Kiely, A. Gavriilidis and G. J. Hutchings, Catal. Today,
34 J. R. Radnik, C. Mohr and P. Claus, Phys. Chem. Chem. Phys.,
2003, 5, 172–177.
35 F. W. Chang, H. Y. Yu, L. S. Roselin, H. C. Yang and
T. C. Ou, Appl. Catal., A, 2006, 302, 157–167.
36 Z. Yang, R. Wu and D. Goodman, Phys. Rev. B: Condens.
Matter, 2000, 61, 14066–14071.
3
2
013, 203, 146–152.
37 M. Brun, A. Berthet and J. C. Bertolini, J. Electron Spectrosc.
Relat. Phenom., 1999, 104, 55–60.
3
3 E. Cao, M. Sankar, S. Firth, K. F. Lam, D. Bethell, D. W. Knight,
G. J. Hutchings, P. F. McMillan and A. Gavriilidis, Chem. Eng. J.,
38 M. Arenz, V. Stamenkovic, T. J. Schmidt, K. Wandelt,
P. N. Ross and N. M. Markovic, Surf. Sci., 2003, 523, 199–209.
2
011, 167, 734–743.
3128 | Catal. Sci. Technol., 2014, 4, 3120–3128
This journal is © The Royal Society of Chemistry 2014