Au-Pd Bimetallic Nanoparticles Immobilised on Titanate Nanotubes: A Highly Active Catalyst for Selective Oxidation

Article abstract of DOI:10.1002/cctc.201700851

In this work, a highly active Au-Pd/Ti-NT catalyst has been produced by using colloidal synthesis and immobilisation on essentially sodium-free Ti-nanotubes (NTs). The new catalyst has markedly superior catalytic activity (turnover frequency>19 000 h^?1) for the selective oxidation of benzyl alcohol compared with similar catalysts reported in the literature, and to Au-Pd catalysts supported on Ti-NTs prepared by adsorption, as well as conventional Au-Pd/TiO₂ prepared by impregnation. The superior catalytic activity of the catalyst is shown to be a result of the high metal dispersion on the external surfaces of Ti-NTs, the narrow particle size distribution and the consistent formation of Au-Pd mixed alloy nanoparticles in close to a 1:1 weight ratio. In repeated use, the Au-Pd/Ti-NT catalyst showed only a modest fall in activity, which is shown by FTIR to be associated mainly with the irreversible adsorption of benzoic acid and benzyl benzoate by the catalyst.

Full text of DOI:10.1002/cctc.201700851

Accepted Article
Title: Au-Pd Bimetallic Nanoparticles Immobilised on Titanate
Nanotubes: A Highly Active Catalyst for Selective Oxidation
Authors: Motaz Khawaji and David Chadwick
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201700851
Link to VoR: http://dx.doi.org/10.1002/cctc.201700851
A Journal of
www.chemcatchem.org

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
Au-Pd Bimetallic Nanoparticles Immobilised on Titanate
Nanotubes: A Highly Active Catalyst for Selective Oxidation
Motaz Khawaji[a], David Chadwick[a]*
density ~4.8 –OH/nm2 [11], whereas Ti-NTs possess a surface
In this work, a highly active Au-Pd/Ti-NT catalyst has been
produced using colloidal synthesis and immobilisation on
essentially sodium-free Ti-NTs. The new catalyst has markedly
area ~250 m2/g on average and surface hydroxyl density ~5.8
–OH/nm2 [12]. In general, the metal dispersion increases as the
surface area and functionalities of the support increase[13]
.
superior catalytic activity (TOF = >19,000 hr-1) for the selective
oxidation of benzyl alcohol compared to similar catalysts
reported in literature, and to Au-Pd catalysts supported on Ti-
NTs prepared by adsorption, as well as conventional Au-Pd/TiO2
prepared by impregnation. The superior catalytic activity of the
catalyst is shown to be due to the high metal dispersion on the
external surfaces of Ti-NTs, the narrow particle size distribution,
and the consistent formation of Au-Pd mixed alloy nanoparticles
in close to a 1:1 wt. ratio. In repeated use the Au-Pd/Ti-NT
catalyst showed only a modest fall in activity which is shown by
FT-IR to be associated mainly with the irreversible adsorption of
benzoic acid and benzyl benzoate by the catalyst.
Consequently, the combination of physiochemical properties
possessed by titanate nanotubes (Ti-NTs) such as the open
mesoporous structures, high specific surface area, high ion-
exchange capacity for various metals, and the good stability at
moderately elevated temperatures[14], a single nanotube is
typically 50-200 nm in length and has an internal diameter ca. 3-
8 nm, and an outer diameter ca. 8-15 nm [15], making Ti-NTs a
very attractive support material for highly dispersed Au and Au-
Pd.
In this work, we have attempted to leverage the high
surface area and high population of surface –OH groups in Ti-
NTs, and use them as anchoring sites for Au-Pd NPs. As we will
show, this strategy proved to be successful and resulted in the
creation of a highly active and stable selective oxidation catalyst.
The application of Au-Pd/Ti-NT in the direct synthesis of
hydrogen peroxide from a mixture of molecular oxygen and
hydrogen[16], and the selective oxidation of salicylic alcohol[17]
has been reported previously. In these earlier studies, the Au–
Pd/Ti-NT catalysts were prepared by adsorption of the metal
precursors on to the surface of Ti-NTs, followed by reduction
with molecular hydrogen. These Au–Pd/Ti-NT catalysts were
shown to have markedly higher activities in comparison with the
equivalent catalysts supported on conventional metal oxide
supports (SiO2, Al2O3 and TiO2). Subsequently, the application
Au-Pd/Ti-NT in the liquid-phase selective oxidation of various
alcohols has been investigated[18], where Au-Pd/Ti-NT were
prepared by adsorption on to Na-Ti-NT and again found to be
more active than conventional Au–Pd/TiO2.
Introduction
For a long time, gold was considered to be chemically inert
and inactive. Haruta[1] and Hutchings[2] were the first to show
independently the remarkable catalytic activity of gold
nanoparticles (NPs). These landmark studies ushered a new
age of gold catalysis. In subsequent studies, gold was found to
be active for a range of reactions including the epoxidation of
alkenes[3], selective oxidation of alcohols and carbohydrates[4],
activation of C-H bonds[5], and the direct synthesis of hydrogen
peroxide under mild conditions[6].
It is well-known that the catalytic activity of gold is
enhanced severalfold by alloying gold with palladium[7]. In the
selective oxidation of alcohols to aldehydes, for instance,
alloying Au with Pd can result in up to twenty-five-fold
enhancement in the activity while retaining the selectivity[8]. The
Titanate nanotubes can be synthesised by alkaline
hydrothermal treatment as first described by Kasuga et al.[19]
The efficient preparation of Au-Pd/Ti-NT by the adsorption
method depends on the amount of sodium present in the titanate
nanotubes, the type of metal precursors used and their
concentration, as well as the pH of the solution since both
synergy between Au and Pd has been attributed to the ability of
[8]
Au atoms to draw electron density away from Pd atoms
.
adsorption and exchange processes take place. As
a
Recently, Carter et al. reported that the introduction of Pd to Au
enhances the catalytic activity in selective oxidation (of benzyl
alcohol) while it was found to be detrimental to the activity in
water-gas shift (WGS) reaction, CO oxidation, and formic acid
decomposition[9].
The catalytic activity of gold and gold-palladium catalysts is
highly dependent on the metal dispersion, particle size, extent of
alloying, and nature of the support. Titania has been extensively
and successfully used a support material for Au and Au-Pd
catalysts. Indeed, several oxidation reactions have been shown
to be active and selective with Au and Au-Pd NPs supported on
titania[3, 10]. Titanate nanotubes possess several advantages
over conventional titania. For instance, titania (Aeroxide® P25)
has a surface area of ~50 m2/g and surface hydroxyl group
consequence, it is difficult to control the residual sodium content
in the final catalyst, the metal dispersion, and the degree of alloy
formation reproducibly. Therefore, an alternative preparation
method of Au-Pd bimetallic catalysts that does not depend on
the starting sodium content and gives a greater degree of alloy
formation is highly desirable, and would lead to a more
reproducible catalyst system.
Sol-immobilisation offers
a
major advantage over
conventional preparation methods in that it allows for controlling
the metal particle size irrespective of the support material[20] and
has been shown to give active Au-Pd catalysts on conventional
oxide supports[21], and carbon[22]. In the present work, we show
that a highly active, stable, and reproducible selective oxidation
catalyst based on the use of essentially sodium-free Ti-NTs as
support, can be prepared by colloidal synthesis of Au-Pd NPs,
their immobilisation on the external surface of the Ti-NTs, and
the partial removal of the stabilizing agent by aqueous reflux.
The catalytic performance has been demonstrated by the study
of solvent-less selective oxidation of benzyl alcohol to
benzaldehyde which is widely used as a model selective
[a]
M. Khawaji, Prof. D. Chadwick
Department of Chemical Engineering Imperial College London
South Kensington, London, SW7 2AZ (UK)
E-mail: d.chadwick@imperial.ac.uk
[8,
oxidation reaction for the study of supported Au-Pd catalysts
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
23]. For comparison purposes the catalytic performance of Au–
Pd/Ti-NT prepared by adsorption, and Au-Pd/TiO2 prepared by
conventional impregnation has also been examined. The new
Au–Pd/Ti-NT catalyst exhibits superior catalytic activity to similar
Au-Pd catalysts reported recently in literature, Au-Pd/Ti-NT
catalysts prepared by adsorption, and conventional Au-Pd/TiO2
prepared by impregnation. Furthermore, it is shown that the high
catalytic activity of the new catalyst compared to other titania or
Ti-NT based catalysts is due to a unique combination of
properties: viz the regular, small Au-Pd particle size, the near
complete alloying of Au and Pd in each individual NP, and the
high concentration of evenly distributed Au-Pd NPs on the
external surfaces of Ti-NTs giving good accessibility of the
reactants to the active catalytic centres.
The dependence on activity and selectivity on reaction
conditions has been explored, and the catalyst deactivation and
reuse has been investigated. In particular, the deactivation by
adsorption of benzoic acid related species has been confirmed
by TGA analysis and FT-infra-red spectroscopy of the recovered
catalysts.
than the measured PZC. Upon the addition of Ti-NT-2, the
negatively charged Au-Pd-PVA sol was readily adsorbed by the
positively charged Ti-NT-2 surface.
In colloidal synthesis of metal NPs, the presence of the
stabilising agent (PVA) in the final catalyst could be both
advantageous and harmful. While it provides additional stability
to the catalyst and prevents sintering and coalescence of the
metal NPs, it can also shield the active catalytic centres and
become detrimental to the catalytic activity[28]. Hutchings and co-
workers have developed a technique for the facile removal of the
stabilising agent by refluxing supported sol-derived Au-Pd NPs
in hot water[29]. We applied the same technique to Au-Pd/Ti-NTSI
to give catalysts designated as Au-Pd/Ti-NTSI-R. As expected,
the application of this technique led to some minor leaching of
Au and Pd from the catalyst (see Table 1). Nevertheless, the
treatment proved to be beneficial as it enhanced the surface
exposure of Au-Pd NPs and improved the catalytic performance
of Au-Pd/Ti-NTSI as shown below.
The catalysts prepared in this study have a nominal metal
loading of 2 wt.% with a Au:Pd weight ratio equal to 1.0
corresponding to a Au:Pd atomic ratio = 1:1.86. Au-Pd/Ti-NT
catalysts prepared by adsorption and Au-Pd/TiO2 catalysts
prepared by dry impregnation, were treated with molecular
hydrogen at a relatively moderate temperature of 200°C to avoid
sintering of Au-Pd NPs, and damage to the structure of Ti-NTs.
Previously it was shown that reduction of the metal precursors
under these conditions is sufficient and produces active
catalysts[16],[17]. The actual metal loadings of the reduced
catalysts were determined by ICP-AES analysis (Table 1), and
were found to be close to the nominal loading. Characterisation
of Au-Pd/Ti-NT by N2 adsorption-desorption measurements
showed that no major changes in the textual properties of Ti-
NTs occur due to metal deposition (Table 1).
Results and Discussion
The Ti-NT were prepared by alkaline hydrothermal
synthesis[18]. Two types of Ti-NT were used as catalyst support:
one set in which some residual Na remained and was denoted
as Ti-NT-1, and a second set which was subjected to further
acid treatment to achieve an external surface containing
minimum sodium, and denoted as Ti-NT-2. An Au-Pd/TiO2
catalyst was prepared for comparison by impregnation using
P25 TiO2. Au-Pd/Ti-NT catalysts were characterised by XRD,
HRTEM, XPS, N2 adsorption-desorption measurements, and
analysed for metal content by ICP. The catalysts prepared by
sol-immobilisation and absorption were further characterised by
HAADF-STEM. The characterisation of the different catalysts
prepared in this study is presented in the section below.
The catalysts were also analysed by XRD shown in Figure
DI
1. Catalysts Au-Pd/Ti-NTADS and Au-Pd/TiO2
displayed
relatively sharp peaks at 38.2° corresponding to the pure gold
Au(111) plane. In the case of Au-Pd/Ti-NTSI, and Au-Pd/Ti-NTSI-R
the peaks were slightly shifted with respect to Au(111) to 39.17°.
This shift is due to the change in the lattice constant, and is
indicative of the formation of a Au-Pd alloy phase[30]. No other
clear diffraction peaks corresponding to Au or Pd diffraction
planes were observed owing to the low metal concentration and
relatively small particle size of the metal NPs.
Catalyst characterisation
The formation of Ti-NTs by hydrothermal synthesis was
confirmed
by
XRD,
nitrogen
adsorption-desorption
measurements, and HRTEM. The XRD patterns for Ti-NTs and
the starting material (TiO2, anatase) are shown in the Supporting
Information, Figure S1. The broad peak appearing at ca. 10.2°
2θ in the XRD pattern corresponds to reflection from the (200)
plane, and is characteristic of titanate nanotubes. This peak is
attributed to the interlayer spacing (d-spacing ca. 0.8-0.9 nm) of
the layered titanate phase[24],[25]. The peaks at 24.5°, 28.6° and
48.6° are also indicative of tri-titanate 1D nanomaterials[26]. The
weakening of the peak at 2θ=28.5° and the broadening of the
peak between at 10.2 ° 2θ in Ti-NT-2 is due to the removal of Na
and complete proton exchange[26]. The surface areas, pore
diameters, and pore volumes are given in Table 1 and are
typical of Ti-NTs. The formation of Ti-NTs was also confirmed by
TEM (Figures S2), where the multi-walled tubular structure of
the synthesised Ti-NTs is clearly seen.
The Au-Pd sol was generated as described in the
experimental section. Due to the presence of hydroxyl groups on
the surface of Ti-NTs, the surface of Ti-NTs can be either
positively or negatively charged depending on the pH of the
dispersion medium. Zeta potential measurements for the as-
synthesised “sodium-free” Ti-NT-2 revealed that these Ti-NTs
have a point of zero charge (PZC) of ~5.2 (Figure S3). PVA-
derived gold sols are negatively charged over a large range of
pH values[27]. Hence, Ti-NT-2 were acidified to a pH value lower
Figure 1. X-ray diffraction patterns for the Au-Pd/Ti-NT catalysts prepared.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
Table 1. Metal composition and textural properties of Ti-NTs and the Au-Pd catalysts
[c]
Pd[a],[b]
[wt.%]
-
Au[a],[b]
[wt.%]
-
Na[a]
[wt.%]
5.27
Pd:Au weight ratio
Pd:Au atomic ratio
SBET
Pore volume[c]
Pore diameter[c]
[m2/g]
174
[cm3/g]
0.36
[nm]
7.31
Ti-NT-1 (Na)
-
-
Ti-NT-2 (minimum Na)
Au-Pd/Ti-NTSI
-
-
0.02
0.02
-
-
236
217
217
206
54
0.51
0.52
0.52
0.45
0.28
7.68
8.21
8.21
6.93
18.30
0.81
0.77
0.88
1.10
0.73
0.70
0.64
0.81
1.11
1.10
1.38
1.36
2.05
2.04
2.54
2.51
Au-Pd/Ti-NTSI-R
Au-Pd/Ti-NTADS
-
[d]
2.37
DI
[d]
Au-Pd/TiO2
-
[a] Weight percentage per gram of sample, obtained by ICP-AES analysis.
[b] Nominal loading for Au=1 wt.% and Pd=1 wt.% in all samples.
[c] Determined by nitrogen adsorption-desorption measurements.
[d] Undetected by ICP.
(Figure 2a and b) highlight an important and distinctive feature,
which is the noticeably high surface loading density. Although all
catalysts have nearly the same overall metal loading, Au-Pd/Ti-
NTSI and Au-Pd/Ti-NTSI-R display the largest population of Au-Pd
NPs. Catalyst sample Au-Pd/Ti-NTSI-R was further characterised
with STEM (Figure 3) in which the highly dispersed, regular
nature of the Au-Pd NPs is clearly seen. In this work and in
previous work[17] it was observed that in the case of Au-Pd/Ti-
NTADS a significant number of NPs appear to have a rod-like
shape and be elongated along the long axis of the Ti-NTs, and
have a width equivalent to the inner diameter of a single
nanotube (i.e. 3-8 nm). It was previously suggested that this is
due to the constraint from the radial curvature of the NTs and
the axial degree of freedom. However, it might imply that some
of the NPs are confined inside the tubular structure of Ti-NTs.
Furthermore, because of the layered structure of Ti-NTs, it is
also possible for metal ions to occupy the interstitial cavities
(typically 0.7-0.9 nm) between the different layers that make up
a single nanotube[14]. Hence, the extremely small nanoparticles
(< 1.0 nm) observed in Au-Pd/Ti-NTADS may well be located in
the interstitial sites of Ti-NTs.
Figure 2. HRTEM images and the corresponding particle size distribution for
a) Au-Pd/Ti-NTSI, b) Au-Pd/Ti-NTSI-R, c) Au-Pd/Ti-NTADS
HRTEM analysis for the Au-Pd/Ti-NT catalysts are shown
in Figure 2. These HRTEM images revealed information about
the structure and morphology of the catalysts, as well as the
particle size distribution (PSD) of Au-Pd NPs on the surface of
Ti-NTs. The catalyst sample prepared by sol-immobilisation
(Figure 2a) exhibited the narrowest particle size distribution and
smallest mean particle size. The refluxed catalyst (Figure 2b)
exhibited a slightly larger mean particle size, which is in
agreement with previous studies concerning the effect of
reflux[29],[31]
. Catalysts prepared by adsorption (Figure 2c)
exhibited a small mean particle size, and a slightly wider PSD.
Significantly, TEM images of Au-Pd/Ti-NTSI and Au-Pd/Ti-NTSI-R
Figure 3. HAADF-STEM images for Au-Pd/Ti-NT SI-R
.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
To investigate the surface chemistry and atomic
presence of some oxidised gold species arising from drying and
transfer[9, 34]. However, some have also attributed this peak
(Auδ+) to the very small gold nanoparticles[9, 34]. The XPS and
XRD results are consistent, therefore, and confirm that the Au-
Pd NPs are present mostly as alloys in Au-Pd/Ti-NTSI and Au-
Pd/Ti-NTSI-R. No chloride was detected by XPS in any of the
catalyst samples demonstrating the full reduction of the metal
precursors and removal of chlorides. The corresponding Pd(3d)
spectra are given in Supporting Information Figure S4. There is
again evidence of surface oxidation during the drying and
transfer. This is more apparent in Au-Pd/Ti-NTADS which is
consistent with the presence of a significant fraction of unalloyed
particles (see below).
composition of Au-Pd/Ti-NT, the catalysts were also
characterised by XPS. The surface atomic concentrations are
given in Table 3; measured peak area ratios are given in Table
S1. It is striking that the catalyst prepared by sol-immobilisation
(Au-Pd/Ti-NTSI) showed a markedly higher surface concentration
than Au-Pd/Ti-NTADS. Furthermore, Au-Pd/Ti-NTSI-R exhibited
even higher Au and Pd surface concentrations than Au-Pd/Ti-
NTSI, which is clear evidence that the refluxing technique does
indeed improve the surface exposure of the Au-Pd NPs on the
surface of Ti-NTs (Table 3). The combined Au-Pd surface
concentration in Au-Pd/Ti-NTSI-R is 10 times that of Pd/Ti-NTADS
.
The low Au and Pd surface concentration observed in Au-Pd/Ti-
NTADS could be ascribed to the confinement of some small NPs
(<5 nm) inside the tubular morphology of the titanate nanotubes.
Because of the limited escape depth of electrons emitted in XPS
and relatively thick walls of Ti-NTs (ca. 4-8 nm), Au-Pd NPs
entrapped inside the nanotubes would not contribute
significantly to the measured spectral intensity. The XPS
analysis depth for Au(4f) was recently reported to be between
5.4 and 5.8 nm; for Pd(4d) XPS signal the analysis depth was
found to be between 4.6-5.0 nm[32]. Interestingly, therefore,
although Au-Pd/Ti-NTADS exhibited a small mean particle size
(3.8 nm) and a reasonably high dispersion, it displayed relatively
weak Au and Pd XPS intensities and low surface concentration,
Table 3. This might be taken as confirming the hypothesis that in
the case of Au-Pd/Ti-NTADS a significant fraction of the smallest
Au-Pd NPs is located inside the pores and interstitial sites of Ti-
NTs. On the other hand, Au-Pd NPs prepared by colloidal
synthesis are essentially wrapped by the stabilising polymer and
hence are more likely to be on the exterior surfaces of Ti-NTs
due to the spatial constraint imposed by the size of the inner
diameter of the nanotubes. It is also plausible that during the
immobilisation of the Au-Pd sol the negatively charged acetate
groups in the partially hydrolysed PVA chains adsorb strongly to
the positively charged Ti-HO2+ species on the external surface of
Ti-NTs, and prevent the dispersion of the Au-Pd NPs inside the
pores. Additionally, it is important to highlight that the atomic
Pd:Au ratio in Au-Pd/Ti-NTSI-R and Au-Pd/Ti-NTSI is very close to
the actual ratio given in Table 1, which is indicative of the
homogenous distribution and alloying of both Au and Pd on the
external surface of Ti-NTs.
Table 3. Surface concentration as determined by XPS
Catalyst
Surface content [atomic%]
Figure 4. XPS spectra of Au(4f) in a) Au-Pd/Ti-NTADS, b) Au-Pd/Ti-NTSI, and c)
Au-Pd/Ti-NTSI-R
.
Au(4f)
1.0
Pd(3d)
2.0
C(1s)
12.0
Ti(2p)
23.7
O(1s)
60.8
Pd:Au
1.98
Au-Pd/Ti-NTSI-R catalyst was further analysed by STEM-
EDX, Figure 5. The elemental mapping of a single Au-Pd NP
revealed that these particles are truly mixed random alloys of Au
and Pd. In order to determine the extent of alloying, 69 NPs
were randomly selected and analysed by STEM-EDX (Figure
S5). Remarkably, the vast majority (83%) of the analysed NPs
were found to be mixed Au-Pd alloys with a Au:Pd weight ratio
equal to 1.0 ± 0.2. Therefore, the analysis carried out on Au-
Pd/Ti-NTSI-R using XRD, XPS and STEM-EDX establishes that
the overwhelming majority of the NPs are undoubtedly mixed
alloys of Au and Pd, and by implication the same is true of Au-
Pd/Ti-NTSI. The relatively low degree of alloying or lack of it in
catalyst Au-Pd/Ti-NTADS was further confirmed by STEM-EDX. A
few NPs were selected and analysed with STEM-EDX, and
showed to be either monometallic in nature or poorly alloyed
(Figure S6), which is in agreement with the XRD and XPS
results.
Au-Pd/Ti-NTSI
Au-Pd/Ti-NTSI-R
Au-Pd/Ti-NTADS
1.86
0.22
4.09
0.39
10.1
5.9
30.9
31.5
53.0
62.0
2.20
1.77
Catalyst sample Au-Pd/Ti-NTADS exhibited a clear Au(4f)
peak at ~ 84 eV, Figure 4, which corresponds to the pure
monometallic reduced phase of gold (Au0). In comparison, Au-
Pd/Ti-NTSI, Au-Pd/Ti-NTSI-R exhibited peaks at slightly lower
binding energies. This peak shift from ~ 84.0 eV to ~83.5 eV
could be ascribed to the electronic modification of Au species by
Pd, and is indicative of a close interaction between the Au and
Pd atoms, and the formation of Au-Pd alloys[9, 33]. The small
peaks (Auδ+) appearing at B.E.~ 85.4 eV are likely due to the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
constant, that the reaction was kinetically controlled at these
conditions. Conversion versus reaction time for 18 mg of catalyst
was a good fit to a first order reaction up 1 h, but deviated
thereafter, Figure S8, probably due to the onset of some
deactivation as discussed below. From these initial
investigations, 12 mg of catalyst and a reaction time of 1 h were
selected to compare catalyst performance free of the effects of
any deactivation and in the kinetic regime, and minimise the
effect of heat-up time which was about 8 min.
The benzyl alcohol conversion and product selectivity for
the different Au-Pd/Ti-NT catalysts are given in Table 4. Each
reaction was repeated three times, and the average values were
reported. In the absence of any catalyst, a small conversion (ca.
4.4 mol%) was obtained most probably due to the autoxidation
of benzyl alcohol with molecular oxygen. The two catalyst
samples prepared by sol-immobilisation (i.e. Au-Pd/Ti-NTSI and
Au-Pd/Ti-NTSI-R) were significantly more active and gave higher
yields of benzaldehyde. In terms of turnover frequency (moles of
benzyl alcohol converted per mole of metal), Au-Pd/Ti-NTSI-R
displays nearly double the catalytic activity of Au-Pd/Ti-NTADS
,
Figure 5. a) HAADF−STEM image for Au-Pd/Ti-NTSI-R
, b) reconstructed
DI
and more than triple the activity of the Au-Pd/TiO2 reference
catalyst. The high degree of Au-Pd alloy formation achieved in
Au-Pd/Ti-NTSI and Au-Pd/Ti-NTSI-R is a major factor in the
superior activity of these catalysts. In addition, in Au-Pd/Ti-NTSI
and Au-Pd/Ti-NTSI-R all the metal NPs were consistently in the
range 1-10 nm in diameter and were alloyed, whereas Au-Pd/Ti-
NTADS presents some larger particles. Only small metal NPs that
are ≤10 contribute significantly to the catalytic activity, whereas
overlay image of the STEM-EDX elemental maps shown in c), d) and e).
Catalytic activity
The solvent-less selective oxidation of benzyl alcohol to
benzaldehyde has traditionally been used as a model oxidation
reaction for studying Au-Pd supported catalysts. While the
primary product in the selective oxidation of benzyl alcohol is
benzaldehyde, other products such as toluene, benzene,
benzoic acid and benzyl benzoate can also be produced in often
competing reaction pathways [35]. Consequently, the reaction is a
good test of the selective oxidation performance of Au-Pd
supported catalysts.
larger NPs >10 nm are essentially catalytically inactive as
[31, 36]
reported by previous studies of alcohol oxidation
.
Furthermore, the high catalytic activity displayed by Au-Pd/Ti-
NTSI and Au-Pd/Ti-NTSI-R is also enhanced by the location of the
Au-Pd NPs on the external surface of the titanate nanotubes
rather than in the inner pores or the inertial sites, which makes
all the NPs accessible to the reactants. In sum, the superior
catalytic activity of Au-Pd/Ti-NTSI-R in comparison to the other
catalysts prepared in this study can be attributed to: 1) the
narrow PSD and small mean particle size, 2) the consistent
formation of mixed Au-Pd alloyed NPs, 3) the enhanced
accessibility to the active metal sites by the reactants, and 4) the
high surface concentration of Au and Pd.
Although catalyst Au-Pd/Ti-NTSI-R was reduced in-situ with
NaBH4 at room temperature, as a check it was also further
treated under the same reduction conditions used for the other
catalysts, and then tested in the selective oxidation of benzyl
alcohol. The performance of the so-treated Au-Pd/Ti-NTSI-R
catalyst remained virtually unchanged, which confirms that the
superior catalytic activity of Au-Pd/Ti-NTSI-R is due to the distinct
structural and morphological properties of the catalyst and not to
the reduction conditions.
It is noteworthy to highlight that catalyst Au-Pd/Ti-NTSI-R
achieved a TOF comparable or higher than those reported for
Au-Pd/NaTi-NT[18] (318 h-1), Au-Pd-PVA/TiO2[23] (10,300 h-1), and
the recently reported Au-Pd-GO/TiO2 composites[23] (10,400 h-1).
Indeed, when Au-Pd/Ti-NTSI-R was tested under exactly the
same reaction conditions as used in the latter study[23] (benzyl
alcohol = 1.0 g, catalyst = 10 mg, T = 120°C, PO2 = 1 bar, stirring
rate = 1,000 rpm, reaction time = 0.5 h), a TOF exceeding
14,000 h-1 was obtained, and under our standard conditions at
0.5 h reaction time the turnover frequency can exceed 19,000 h-1.
Figure 6. TOF (h-1) and benzaldehyde yield obtained for the different Au-
Pd/Ti-NT catalysts. Reaction conditions: T= 120°C, PO2 =4 bar, stirring
rate=1,000 rpm, reaction time= 1 hour, catalyst=12 mg, benzyl alcohol= 2.5 g.
All the catalysts were tested for benzyl alcohol selective
oxidation under the same reaction conditions (T= 120°C, PO2 = 4
bar, stirring rate = 1,000 rpm, reaction time = 1 h, catalyst weight
= 12 mg). In a preliminary study, conversion was found to be
independent of stirrer speed in the range 600-1000 rpm. A study
of conversion versus catalyst weight at a constant reaction time
of 1 h was essentially linear up to 12 mg and overall was a good
Effect of reaction conditions
Using the most active catalyst (Au-Pd/Ti-NTSI-R), the effect of
varying the reaction conditions on the activity and selectivity for
benzyl alcohol selective oxidation has been investigated. As
noted above, benzaldehyde is formed by the sequential
fit to
a first order reaction, Figure S7 and Table S2,
demonstrating, as the pressure of oxygen was maintained
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
oxidative dehydrogenation of benzyl alcohol, while further
oxidation produces benzoic acid. However, there are several
other by-products from competing pathways which are sensitive
to the reaction conditions. Benzyl ether is produced by the
dehydration of benzyl alcohol, and benzyl benzoate is formed by
either of two pathways: the esterification of benzoic acid by
benzyl alcohol, or via hemi-acetal from benzaldehyde. Benzene
is formed by the decarbonylation of benzaldehyde[37]. The origin
Table 4. Benzyl alcohol conversion and product selectivity for the different catalysts[a]
Catalyst
Conversion
[%]
TOF[b]
[1/h]
Product Selectivity [mol%]
Benzaldehyde
Toluene
Benzene
Benzyl ether
Benzoic acid
Benzyl benzoate
Blank
4.4
-
83.7
79.5
80.4
86.1
78.6
2.6
12.8
11.5
8.2
13.7
2.4
2.4
0.8
7.9
0
0
0
Au-Pd/Ti-NTSI
Au-Pd/Ti-NTSI-R
Au-Pd/Ti-NTADS
61.6
66.5
36.1
14.0
9,096
10,172
5,693
2,260
1.9
2.0
0.7
1.1
1.3
1.2
1.9
0.2
2.2
2.5
2.2
1.7
DI
Au-Pd/TiO2
10.5
[a] Reaction conditions: T= 120°C, PO2 =4 bar, stirring rate=1,000 rpm, reaction time= 1 hour, catalyst=12 mg, benzyl alcohol= 2.5 g.
[b] Turnover frequency is expressed per total metal loading as determined by ICP-AES.
of toluene has been debatable for a long time[38], some
suggested that it is produced as a result of the hydrogenolysis of
benzyl alcohol[39]. However, in light of more recent studies,
disproportionation of benzyl alcohol appears to be the most
TOF > 19,000 h-1 and benzyl alcohol conversion ca. 33%. The
catalyst was marginally more selective to benzaldehyde at short
reaction times as expected, displaying benzaldehyde selectivity
ca. 83.2 mol% at 30 minutes. As the reaction time is increased,
higher benzyl alcohol conversion and benzaldehyde yield are
obtained, while reaction selectivity towards benzaldehyde
decreased to 79.7 mol% at 120 minutes and remained virtually
constant thereafter.
likely source of toluene [38b]
.
The catalytic activity of Au-Pd/Ti-NTSI-R as a function of
reaction time is shown in Figure 7a and the corresponding
product selectivities are given in Table 5. We found that Au-
Pd/Ti-NTSI-R was remarkably active after 15 min, exhibiting a
Figure 7. Effect of (a) time, (b) O2 partial pressure, and (c) temperature on benzyl alcohol conversion, and benzaldehyde selectivity for Au-Pd/Ti-NTSI-R
.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
[a]
Table 5. Benzyl alcohol conversion and product selectivity at different reaction times for Au-Pd/Ti-NTSI-R
Time
[min]
Conversion
[%]
Product Selectivity [mol%]
Benzaldehyde
Toluene
Benzene
Benzyl ether
Benzoic acid
Benzyl benzoate
15
30
33.1
65.2
80.9
83.2
80.2
80.7
79.7
79.5
9.6
9.8
1.2
1.5
2.3
2.1
2.7
2.1
2.5
1.8
2.3
2.1
1.9
1.8
2.1
1.1
1.2
1.4
1.6
2.0
3.7
2.6
2.5
3.1
3.0
3.0
60
85.3
89.8
91.7
94.0
11.1
10.6
11.1
11.6
90
120
150
[a] Reaction conditions: T= 120°C, PO2 =4 bar, stirring rate=1,000 rpm, catalyst=18 mg, benzyl alcohol= 2.5 g.
The effect of varying the oxygen partial pressure while
keeping all other parameters constant is shown in Figure 7b and
Table 6. The oxygen partial pressure profoundly affected the
activity and product selectivity. High conversion of benzyl
alcohol (>98%) was achieved at 10 bar, compared to only 65.6%
at 1 bar. The selectivity towards benzaldehyde was enhanced at
higher oxygen partial pressures, reaching 86.7 mol% at 10 bar.
The highest yield of benzaldehyde (ca. 85.3 mol%) was
achieved at 10 bar. Increasing the oxygen partial pressure
appears to have a near linear effect on benzaldehyde selectivity
and yield. Surprisingly, higher oxygen pressures did not lead to
significant formation of benzoic acid, but contributed more to
suppression of the formation of toluene. The solubility of O2 in
benzyl alcohol increases with pressure, which leads to an
increase in the oxygen available at the catalyst surface. More O2
availability enhances the reaction rate, and favours the oxidative
dehydrogenation reaction pathway in the formation of
benzaldehyde at the expense of the disproportionation pathway,
which is responsible for toluene production.
The effect of varying the reaction temperature is given in
Figure 7c and Table 7. The conversion and TOF increased
considerably with temperature, while the selectivity towards
benzaldehyde decreased. The selectivity towards toluene is
enhanced markedly with temperature, increasing from less than
1 mol% at 80°C to 23.6 mol% at 140°C. High reaction
temperatures, therefore, enhance the rate of benzyl alcohol
disproportionation. The yield of benzaldehyde increased with
temperature up to 120°C, and started to decline at 140°C as
more toluene and heavy compounds are formed. Hence, it
appears that 120°C is the optimum temperature for maximum
yield of benzaldehyde over these catalysts.
[a]
Table 6. Benzyl alcohol conversion and product selectivity at different oxygen partial pressures for Au-Pd/Ti-NTSI-R
Pressure
[bar]
Product Selectivity [mol%]
Conversion
[%]
Benzaldehyde
Toluene
Benzene
Benzyl ether
Benzoic acid
Benzyl benzoate
1
65.6
85.3
87.2
98.4
77.7
80.2
84.8
86.7
14.9
11.1
6.9
1.4
2.3
2.2
1.9
3.0
2.3
1.5
0.3
0.9
1.2
1.4
2.7
2.1
2.5
3.0
2.2
4
8
10
6.2
[a] Reaction conditions: T= 120°C, stirring rate=1,000 rpm, reaction time= 1 hour, catalyst=18 mg, benzyl alcohol= 2.5 g.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
[a]
Table 7. Benzyl alcohol conversion and product selectivity at different temperatures for Au-Pd/Ti-NTSI-R
Temperature
[°C]
Conversion
[%]
Product Selectivity [mol%]
Benzaldehyde
Toluene
Benzene
Benzyl ether
Benzoic acid
Benzyl benzoate
80
21.2
51.8
85.3
92.5
90.0
89.2
80.2
67.4
0.8
3.8
6.3
2.5
2.3
2.1
0.2
0.8
2.3
4.4
1.1
1.2
1.2
1.1
1.7
2.5
2.5
1.5
100
120
140
11.1
23.6
[a] Reaction conditions: PO2 =4 bar, stirring rate=1,000 rpm, reaction time= 1 hour, catalyst=18 mg, benzyl alcohol= 2.5 g.
Metal leaching, catalyst deactivation and re-use
The stability and reusability were investigated using catalyst Au-
Pd/Ti-NTSI-R. Leaching of active components into the reaction
medium was investigated as follows. The oxidation reaction was
carried out for 30 minutes over Au-Pd/Ti-NTSI-R at the normal
reaction condition (T = 120°C, PO2 = 4 bar, stirring rate = 1,000
rpm). The catalyst was subsequently removed and the liquid
aliquot was returned to the reactor, and the reaction was run for
an additional 60 minutes. Very small increase in the conversion
was observed (Figure 8). This marginal increase in conversion is
due to autoxidation of benzyl alcohol under the reaction
conditions. Indeed, a similar conversion (~ 4.4 %) was observed
in blank runs (Table 4). Furthermore, no traces of Au or Pd were
detected when the reaction liquid was analysed by ICP-AES.
Figure 9. Benzaldehyde yield and TOF for recycled Au-Pd/Ti-NTSI-R. Reaction
conditions: T= 120°C, PO2 =4 bar, stirring rate=1,000 rpm, reaction time= 1
hour, catalyst=18 mg, benzyl alcohol= 2.5 g.
Table 9. Metal loading and weight loss for recycled Au-Pd/Ti-NTSI-R
.
Catalyst
Pd
loading[a
Au
loading[a]
Pd/Au
weight
ratio
Weight
loss[b]
[wt.%]
0.77
[wt.%]
0.70
[wt.%]
4.09
Au-Pd/Ti-NTSI-R
(fresh catalyst)
0.91
1.13
1.13
Au-Pd/Ti-NTSI-R
(after 1st use)
0.74
0.72
0.65
0.64
18.9
8.16
Figure 8. Catalyst leaching test for Au-Pd/Ti-NTSI-R
.
Au-Pd/Ti-NTSI-R
(after 2nd use)
The catalytic performance of Au-Pd/Ti-NTSI-R was
investigated with repeated usage. After each catalytic test, Au-
Pd/Ti-NTSI-R was recycled by washing the recovered catalyst
several times with ethanol. A mild loss in activity and a slight
drop in TOF and benzaldehyde yield was observed after usage:
the TOF dropped by 12% after the first use, and 8% in the third
use, Figure 9. Despite this, the Au-Pd/Ti-NTSI-R catalyst appears
to be more stable than similar Au-Pd-PVA/TiO2 catalysts
prepared by sol-immobilisation for which an activity drop of
~30% was reported after the first use, and ~60% after the
[a] Weight percentage per gram of sample as obtained by ICP-AES analysis.
[b] Weight loss calculated based on TGA results (200-550 °C).
Catalyst deactivation and activity loss can be due to one or
more of several factors: leaching of active components from the
support, sintering of metal particles, and poisoning of active sites.
In the present case, no metal leaching into the reaction liquid
was observed during the runs. However, a small amount of Au
second use[23]
.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
and Pd (with only a small change in the Au/Pd ratio) was lost
during the catalyst washing cycles between each test, Table 9,
which probably contributes to the observed drop in the catalytic
activity with re-use.
Conclusions
In this work, a highly active Au-Pd catalyst for selective oxidation
of benzyl alcohol has been produced on a titanate nanotube
support, which was synthesised to have as low a sodium content
as possible, by a sol-immobilisation method. The new catalyst
was shown to be significantly more active, achieving higher
TOFs, than similar Au-Pd/TiO2 catalysts reported in the literature
and a catalyst prepared by conventional impregnation of
standard P25 titania support. It was also shown to be more
active in comparison with a Au-Pd catalyst prepared by the
adsorption method reported previously. Through detailed
characterisation of the catalyst structure, morphology and
composition, the high catalytic activity of the Au-Pd/Ti-NTSI-R
catalyst was attributed to the regular, small Au-Pd particle size,
the near complete alloying of Au and Pd in each individual NP,
and the high dispersion of Au-Pd NPs on the external surfaces
of Ti-NTs leading to good accessibility by the reactants to the
active catalytic particles.
A study of catalyst deactivation and reuse showed that
metal leaching during reaction was negligible, and that
irreversible adsorption of the deep oxidation products benzioc
acid and benzyl benzoate was the main cause of deactivation in
agreement with previous work. Despite this, Au-Pd/Ti-NTSI-R
catalyst suffered only a 19% loss in activity over three use
cycles, which compares favourably in terms of stability with Au-
Pd catalysts reported in the literature.
Figure 10. FT-IR spectra for as-synthesised Ti-NT, fresh and recycled Au-
Pd/Ti-NTSI-R catalysts.
The catalytic performance of Au-Pd/Ti-NTSI-R in the
selective oxidation of benzyl alcohol was dependent on the
reaction conditions, and in particular the side reaction of
disproportionation of benzyl alcohol to benzaldehyde and
toluene was promoted at high temperatures and low oxygen
partial pressures.
The catalyst described here can be expected to be active
in several other selective oxidation reactions. In principle, the
approach used here could be applied to synthesis catalysts with
other precious metals supported on Ti-NTs.
As for sintering and metal particle growth, previous studies
have shown that the particle size of PVA-derived Au-Pd NPs
remains unchanged with repeated benzyl alcohol oxidation
reaction cycles[23]. Therefore, blocking of the catalytic sites by
adsorbed products (i.e. product inhibition) was investigated
using thermogravimetric analysis (TGA) and FT-IR. The TGA of
the used catalysts, albeit being washed and dried three times,
exhibited a higher weight loss (200-550 °C) than the fresh
catalyst: 4 wt. % for the fresh catalyst and 19% and 8% after the
1st and 2nd use, respectively (Table 9). The weight loss
observed in this temperature range (200-550 °C) is due to the
decomposition of carbon-based species (i.e. adsorbed products
+ residual PVA). To confirm the nature of the absorbed species,
the fresh and the washed and recycled catalysts were analysed
with FT-IR. In the washed and recycled catalyst, several new
bands were observed, Figure 10. The new and strong bands at
1524 cm-1 (νa COO-), 1415 (νs COO-) are characteristic of
carboxylate functionality[40]. The bands appearing at 1602 and
1495 cm-1 are associated with the aromatic ν(C=C) + δ(C-H)
modes [41]. These results are in agreement with recent studies
which showed that benzoic acid and benzoate species tend to
Experimental Section
Materials
All metal precursors and chemical reagents were purchased
from Sigma Aldrich and used as received: NaOH (99.99% trace
metals basis), H2SO4(≥97.5% purity), TiO2 (Aeroxide® P25),
poly(vinyl alcohol) (PVA) (Mw 9,000-10,000, 80% hydrolysed),
HAuCl4.3H2O (99.999% purity), PdCl2 (5 wt. % in 10 wt. % HCl),
NaBH4 (Aldrich ≥98.0%), Palladium(II) acetate (≥99.9% purity),
Benzyl alcohol (99.8% purity). O2 (100% pure) for catalytic tests
was supplied by BOC.
[41]
be strongly adsorbed on the surface of Au-Pd-PVA/TiO2 even
when heated to 280°C, and that the addition of a small amount
of benzoic acid at the beginning of the reaction led to a marked
drop in the activity of Au-Pd-PVA/TiO2 in the selective oxidation
of benzyl alcohol[23]. Wang et al.[40] carried out a detailed
spectroscopic study on the interaction of benzoic acid with Ti-
NTs found that benzoic acid can disperse spontaneously onto
the surface of titanate nanotubes in a sub-monolayer state, and
concluded that carboxylate species were formed by chemical
reaction of the surface hydroxyl groups on Ti-NTs with
carboxylic acid functionalities. These studies and the present
findings reinforce the conclusion that the mild fall in activity
observed with repeated usage of Au-Pd/Ti-NTSI-R is associated
mainly with the seemingly irreversible adsorption of benzoic acid
and benzyl benzoate by the catalyst.
Synthesis of titanate nanotubes (Ti-NTs)
Titanate Nanotubes (Ti-NTs) were synthesised by the alkaline
hydrothermal treatment method as reported by Kasuga et al. [19]
.
In a typical synthesis, 11 g of TiO2 (anatase nanopowder, 99.7%
Sigma Aldrich) was added to 185 mL of 10 M NaOH in a 200 mL
PTFE-liner and stirred for two hours. The PTFE-lined steel
autoclave was then placed in an air-circulating oven at 140°C for
24 hours. The obtained slurry was then washed with DI water,
filtered and dried overnight at 120°C. The dried powder was
washed once with 0.1 M H2SO4 and several times with distilled
water until pH 7 is reached. The obtained powder was
subsequently filtered dried. These Ti-NTs contained some
residual Na and was denoted to as Ti-NT-1. A portion of Ti-NT-1
was acid treated once more, and denoted to by Ti-NT-2.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
Catalysts prepared by sol-immobilisation (Au-Pd/Ti-NTSI)
20 mL of acetonitrile and then centrifuged at 5,000 rpm for 15
minutes to separate out the catalyst. A sample from the
supernatant (600 μL) was taken out and added to 100 μL of n-
butanol (internal standard) prior to product analysis. Analysis of
the reaction products was carried out using a GC (Shimadzu
GC-2014) fitted with a Flame Ionization Detector (FID), and a
wax column (Agilent CP WAX 52 CB UltiMetal, L= 25m, ID=
0.53 mm, film thinness =2.0 µm). Carbon balance was 96 ± 3%.
Catalysts were prepared in manner similar to the sol-
immobilisation procedure previously reported[21]. First, the Ti-NT-
2 were acidified to a pH below the point of zero charge (PZC).
The PZC of Ti-NTs was measured and found to be ~5.2.
Typically, 1.0 g of the support was added to 75 mL of water and
a 1 M solution of H2SO4 was added drop-wise to the slurry until
the pH dropped to ~1.5. After 2 hours of vigorous stirring, the
slurry was filtered and the acidified support was collected. The
metal colloid or sol was prepared by adding calculated amounts
of the gold precursor (HAuCl4.3H2O) and palladium precursor
(PdCl2,5 wt.% in HCl) to 100 mL of DI water at 5°C while stirring
vigorously at 1,500 rpm. Subsequently, 2,400 mg of 1.0 wt.%
PVA solution was added to the solution and stirred for 15
minutes. PVA was used as a stabiliser herein and the weight
ratio of PVA/(Au+Pd) was kept at 1.2. The metal precursors
were reduced in-situ by the addition of an excess amount of
NaBH4 such that the molar ratio of NaBH4:(Au+Pd) was 5:1.
More specifically, 7.5 mL of 0.1 M NaBH4 was added to the
solution. The solution turned brown immediately upon the
addition of NaBH4, which confirmed the formation of the sol. The
sol was left stirring for 1 hour before the acidified Ti-NTs were
added to form a slurry. The pH of the slurry was maintained at
~3 during the immobilisation of the Au-Pd sol on the Ti-NTs. The
slurry was left stirring overnight before it was filtered and
washed with DI water several times until the pH of the mother
liquor reached ~7.0. The obtained catalyst was finally dried
overnight at 120 °C. A portion of the dried catalyst was refluxed
in hot water (90°C) for 60 minutes, filtered and dried overnight at
120 °C. This treated catalyst portion was denoted to as Au-
Benzyl alcohol (BA) conversion (푋퐵퐴 ), product yield (푌 ), and
푖
selectivity (푆푖) were calculated using the following equations:
푤푖
푤퐵°퐴
(
)
푌 푤푡. % =
× 100%
푖
푛
푋퐵퐴(%) = ∑ 푌 (푤푡. %)
푖
푖=1
푛푖
(
)
푆푖 푚표푙 % =
× 100%
× 100%
푛
∑
푖=1 푛푖
푛퐵퐴퐿
푛퐵°퐴
(
)
푌
푚표푙 % =
퐵퐴퐿
Where: 푤퐵°퐴 is the initial weight of benzyl alcohol, 푤푖 is the final
weight of product 푖, 푛퐵°퐴 is the initial number of moles of benzyl
alcohol, and 푛푖 is the final number of moles of product 푖, 푛퐵퐴퐿 is
the final amount of benzaldehyde. Conversion and selectivity
values reported represent the average of three experiments.
The %standard deviation for the conversion was ca. 2.5% for the
conversion and less than 1% for the selectivities.
Pd/Ti-NTSI-R
.
Catalysts prepared by adsorption (Au-Pd/Ti-NTADS
)
The required amounts of (HAuCl4.3H2O) and (PdCl2, 5 wt.% in
HCl) corresponding to the desired loadings were added to 100
mL of DI water at 80°C. Under vigorous stirring conditions, 1.0 g
of the support (Ti-NT-1) was added to the solution. After 18
hours, the solution was filtered and dried overnight at 120 °C.
The catalyst was then reduced for 2 hours at 200°C in a tube
furnace under a continuous flow of diluted hydrogen (5 vol.% H2
/Ar). The furnace ramp rate was set to 5°C/min, and gas flow
rate was fixed at ~50 mL/min.
Metal leaching and catalyst re-use
In order to quantify any leaching of active components occurring
during the course of the reaction, the oxidation of benzyl alcohol
was carried out for 30 minutes in the presence of Au-Pd/Ti-NTSI-
R, and then terminated. The reaction slurry was subsequently
centrifuged to separate out the catalyst, and the clean
supernatant was filtered and returned to the reactor. After one
hour, the reaction was terminated and the liquid was analysed
with GC. In order to examine catalyst reusability, the used
catalysts were washed with an excess amount of ethanol and
centrifuged. The supernatant was removed, and the washing
procedure was repeated three times. Next, the catalyst was
dried overnight at 80 °C. The catalytic activity of the recycled
catalysts was tested under the same reaction conditions used
for the fresh catalyst.
DI
Catalysts prepared by dry Impregnation (Au-Pd/TiO2
)
The requisite amounts (HAuCl4.3H2O) and (PdCl2, 5 wt.% in
HCl) were dissolved in a predetermined volume of DI water to
give a surface loading of ~200,000 m2/L. The precursor solution
was added drop-wise to the support (TiO2, Aeroxide® P25) to
form a fine paste. The paste was dried overnight at 120°C, and
then reduced under the same conditions described in the
previous section.
Catalyst characterisation
Catalysts were analysed by PANalytical X’Pert Pro Multi-
Purpose Diffractometer using Cu-Kα radiation. The analysis was
performed over a scan angle of 5-70° 2ϴ, and a step size of
0.0167°/s. Nitrogen adsorption-desorption measurements were
carried out at -196°C using a Micromeritics TriStar II. Prior to
carrying out any measurements, all samples were degased at
130°C for 12 hours. Brunauer–Emmett–Teller (BET) and Barrett-
Joyner-Halenda (BJH) were used to measure the specific
surface area (SBET), pore volume and pore diameter.
The final Au and Pd loadings were determined by inductively
coupled plasma atomic emission spectroscopy (ICP-AES, PE
Optima 2000 DV). Typically, 15 mg of catalyst was dissolved in
15 mL of aqua regia, sonicated for 2 hours, and left overnight in
order to completely dissolve any remaining solids. The samples
were diluted with 20 mL of DI water before analysis. The
Reaction procedure
Benzyl alcohol oxidation was carried out in batch mode using 25
mL glass-lined minclaves (Buchiglas, Switerland). In a typical
run, benzyl alcohol (2.50 g) and the requisite amount of catalyst
were charged to the reactor before it was purged five times with
O2. The reactor was subsequently pressurised with O2 to the
required pressure at room temperature. The reactor was heated
in an oil bath set at the desired temperature and stirred at 1,000
rpm. An oxygen reservoir was connected to the reactor to
ensure that any oxygen consumed by the reaction was
replenished. The reactor pressure was controlled using a
forward-pressure regulator and a series of check-valves. At the
end of each run, the reactor was cooled down to room
temperature and vented. The reaction mixture was dissolved in
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
D. V. Bavykin, F. C. Walsh, Titanate and Titania Nanotubes :
Synthesis, 2009.
A. Villa, D. Wang, G. M. Veith, F. Vindigni, L. Prati, Catalysis
Science & Technology 2013, 3, 3036-3041.
D. V. Bavykin, A. A. Lapkin, P. K. Plucinski, L. Torrente-Murciano,
J. M. Friedrich, F. C. Walsh, Topics in Catalysis 2006, 39, 151-160.
Á. Kukovecz, K. Kordás, J. Kiss, Z. Kónya, Surface Science
Reports 2016, 71, 473-546.
L. Torrente-Murciano, Q. He, G. J. Hutchings, C. J. Kiely, D.
Chadwick, ChemCatChem 2014, 6, 2531-2534.
L. Torrente-Murciano, T. Villager, D. Chadwick, ChemCatChem
2015, 7, 925-927.
D. Nepak, S. Darbha, Catalysis Communications 2015, 58, 149–
153.
T. Kasuga, M. Hiramatsu, A. Hoson, T. Sekino, K. Niihara,
Langmuir 1998, 14, 3160-3163.
F. Porta, L. Prati, M. Rossi, S. Coluccia, G. Martra, Catalysis
Today 2000, 61, 165-172.
N. Dimitratos, J. A. Lopez-Sanchez, D. Morgan, A. F. Carley, R.
Tiruvalam, C. J. Kiely, D. Bethell, G. J. Hutchings, Phys Chem
Chem Phys 2009, 11, 5142-5153.
N. Dimitratos, F. Porta, L. Prati, Applied Catalysis A: General
2005, 291, 210-214.
J. Wang, S. A. Kondrat, Y. Wang, G. L. Brett, C. Giles, J. K.
Bartley, L. Lu, Q. Liu, C. J. Kiely, G. J. Hutchings, ACS Catalysis
2015, 5, 3575-3587.
S. K. Pradhan, Y. Mao, S. S. Wong, P. Chupas, V. Petkov,
Chemistry of Materials 2007, 19, 6180-6186.
T. Gao, H. Fjellvåg, P. Norby, Inorganic Chemistry 2009, 48, 1423-
1432.
E. Morgado, M. A. S. de Abreu, G. T. Moure, B. A. Marinkovic, P.
M. Jardim, A. S. Araujo, Chemistry of Materials 2007, 19, 665-676.
L. Prati, A. Villa, Acc Chem Res 2014, 47, 855-863.
M.-C. Daniel, D. Astruc, Chemical Reviews 2004, 104, 293-346.
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. Knight, C. J. Kiely, G. J. Hutchings, Nat Chem 2011, 3,
551-556.
B. Pawelec, A. M. Venezia, V. La Parola, E. Cano-Serrano, J. M.
Campos-Martin, J. L. G. Fierro, Applied Surface Science 2005,
242, 380-391.
I. Gandarias, P. J. Miedziak, E. Nowicka, M. Douthwaite, D. J.
Morgan, G. J. Hutchings, S. H. Taylor, ChemSusChem 2015, 8,
473-480.
instrument was calibrated with authenticated standards (Sigma
Aldrich) containing predetermined amounts of each metal. The
average particle size and distribution of the metal NPs were
determined from high resolution transmission electron
microscopy (HRTEM) images. Scanning transmission electron
microscopy (STEM) and HRTEM images were collected using a
JEOL JEM-2100F microscope operating at 200 kV.. Samples
were dispersed in ethanol and sonicated for 30 minutes before
they were placed on copper grids with a holey carbon film (300
mesh size). The particle size distribution and average particle
size were determined by analysing 100 randomly selected
nanoparticles in each catalyst sample. Energy Dispersive X-Ray
Analysis (EDX) provided elemental analysis and was carried out
using a X-MaxN 80 T Silicon Drift Detector from Oxford
Instruments.
X-ray
photoelectron
spectroscopy
(XPS)
measurements were recorded using
a
Thermo K-Alpha
Spectrometer equipped with Al Kα source gun. Samples were
mounted on double-sided adhesive tape, and the spectra were
collected using an X-ray spot size of 400 μm and a pass energy
of 20 eV with 0.1 eV increments. The binding energies (BE)
were referenced to the C(1s) peak of adventitious carbon at
284.8 eV. Data analysis and peak fitting was performed using
Avantage software from Thermo Scientific. Thermogravimetric
analysis (TGA) was carried under flowing air (60 mL/min) using
TGA Q500 (TA Instruments). Samples were heated at a rate of
10°C/min to a final temperature of 550°C.The point of zero
charge for the as-synthesised Ti-NTs was determined by
measuring the zeta potential of Ti-NTs as a function of pH using
Brookhaven ZetaPALS Potential Analyzer. Ti-NTs were
dispersed in solutions of HNO3 and KOH at varying pH values
ranging from 2 to 11. FT-IR measurements were performed
using PerkinElmer Spectrum 100.
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
Acknowledgements
C. M. Olmos, L. E. Chinchilla, E. G. Rodrigues, J. J. Delgado, A. B.
Hungría, G. Blanco, M. F. R. Pereira, J. J. M. Órfão, J. J. Calvino,
X. Chen, Applied Catalysis B: Environmental 2016, 197, 222-235.
aA. M. Venezia, L. F. Liotta, G. Pantaleo, V. La Parola, G.
Deganello, A. Beck, Z. Koppány, K. Frey, D. Horváth, L. Guczi,
Applied Catalysis A: General 2003, 251, 359-368; bP. A. P.
Nascente, S. G. C. de Castro, R. Landers, G. G. Kleiman, Physical
Review B 1991, 43, 4659-4666.
Motaz Khawaji gratefully acknowledges the financial support of
Saudi Aramco. This work was funded in part by EPSRC
(EP/K014749/1). We are grateful to Dr. Ecaterina Ware for her
assistance with TEM.
[34]
[35]
C. M. Olmos, L. E. Chinchilla, J. J. Delgado, A. B. Hungría, G.
Blanco, J. J. Calvino, X. Chen, Catalysis Letters 2016, 146, 144-
156.
Keywords: Gold-palladium • titanate nanotubes • selective
oxidation • sol-immobilisation • benzyl alcohol
S. Meenakshisundaram, E. Nowicka, P. J. Miedziak, G. L. Brett, R.
L. Jenkins, N. Dimitratos, S. H. Taylor, D. W. Knight, D. Bethell, G.
J. Hutchings, Faraday Discussions 2010, 145, 341-356.
C. H. Bartholomew, R. J. Farrauto, Fundamentals of Industrial
Catalytic Processes, John Wiley & Sons, 2011.
aE. Cao, M. Sankar, S. Firth, K. F. Lam, D. Bethell, D. K. Knight,
G. J. Hutchings, P. F. McMillan, A. Gavriilidis, Chemical
Engineering Journal 2011, 167, 734-743; bS. S. Hladyi, M. K.
Starchevsky, Y. A. Pazdersky, M. N. Vargaftik, I. I. Moiseev,
Mendeleev Communications 2002, 12, 45-46.
aG. Kovtun, T. Kameneva, S. Hladyi, M. Starchevsky, Y.
Pazdersky, I. Stolarov, M. Vargaftik, I. Moiseev, Advanced
Synthesis & Catalysis 2002, 344, 957-964; bE. H. 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, G. J. Hutchings,
Catalysis Today 2013, 203, 146-152.
D. M. Meier, A. Urakawa, A. Baiker, The Journal of Physical
Chemistry C 2009, 113, 21849-21855.
W. Wang, J. Zhang, H. Huang, Z. Wu, Z. Zhang, Applied Surface
Science 2007, 253, 5393-5399.
E. Nowicka, J. P. Hofmann, S. F. Parker, M. Sankar, G. M. Lari, S.
A. Kondrat, D. W. Knight, D. Bethell, B. M. Weckhuysen, G. J.
Hutchings, Physical Chemistry Chemical Physics 2013, 15, 12147-
12155.
[1]
M. Haruta, T. Kobayashi, H. Sano, N. Yamada, Chemistry Letters
1987, 16, 405-408.
[36]
[37]
[2]
[3]
G. J. Hutchings, Journal of Catalysis 1985, 96, 292-295.
T. Hayashi, K. Tanaka, M. Haruta, Journal of Catalysis 1998, 178,
566-575.
[4]
[5]
L. Prati, M. Rossi, Journal of Catalysis 1998, 176, 552-560.
R. Zhao, D. Ji, G. Lv, G. Qian, L. Yan, X. Wang, J. Suo, Chem
Commun (Camb) 2004, 904-905.
[38]
[6]
[7]
[8]
P. Landon, P. J. Collier, A. J. Papworth, C. J. Kiely, G. J.
Hutchings, Chemical Communications 2002, 2058-2059.
A. Villa, D. Wang, D. S. Su, L. Prati, Catal. Sci. Technol. 2015, 5,
55-68.
D. I. Enache, J. K. Edwards, P. Landon, B. Solsona-Espriu, A. F.
Carley, A. A. Herzing, M. Watanabe, C. J. Kiely, D. W. Knight, G.
J. Hutchings, Science 2006, 311, 362-365.
[39]
[40]
[41]
[9]
J. H. Carter, S. Althahban, E. Nowicka, S. J. Freakley, D. J.
Morgan, P. M. Shah, S. Golunski, C. J. Kiely, G. J. Hutchings, ACS
Catalysis 2016, 6, 6623-6633.
[10]
[11]
D. I. Enache, D. Barker, J. K. Edwards, S. H. Taylor, D. W. Knight,
A. F. Carley, G. J. Hutchings, Catalysis Today 2007, 122, 407-411.
R. Mueller, H. K. Kammler, K. Wegner, S. E. Pratsinis Langmuir
2003, 19, 160-165.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700851
ChemCatChem
FULL PAPER
FULL PAPER
A highly active Au-Pd/Ti-NT catalyst
has been produced using colloidal
synthesis and immobilisation on
Motaz Khawaji, Prof. David Chadwick*
Page No.12 – Page No.12
essentially
sodium-free
titanate
Au-Pd Bimetallic Nanoparticles
Immobilised on Titanate Nanotubes:
A Highly Active Catalyst for Selective
Oxidation
nanotubes (Ti-NT). The new catalyst
has markedly superior catalytic activity
(TOF = >19,000 hr-1) for the selective
oxidation of benzyl alcohol compared to
similar catalysts reported in literature,
and to Au-Pd catalysts supported on Ti-
NTs prepared by adsorption, as well as
conventional Au-Pd/TiO2 prepared by
impregnation.
This article is protected by copyright. All rights reserved.