Hydrogenation of cinnamaldehyde in the presence of PdAu/C catalysts prepared by the reverse “water-in-oil” microemulsion method

Article abstract of DOI:10.1016/j.apcata.2014.08.036

Liquid phase hydrogenation of cinnamaldehyde was studied on carbon supported (Vulcan XC72) – PdAu catalysts with Au/Pd atomic ratio varying from 0.1 up to 2.1. Colloid-based method, namely the reverse “water-in-oil” microemulsion technique was used to prepare the catalysts. Accordingly, pre-formed metal nanoparticles of controlled size with a narrow size distribution were deposited on carbon support. The BET, XRD, XPS, SEM, TEM, and HRTEM techniques were employed to characterize PdAu/C catalysts and their monometallic 2%Pd/C and 2%Au/C counterparts. The metal particles were spherically shaped, nearly monodispersed and well distributed throughout the carbon support. The particles of low Au-content (Au/Pd < 0.8) were of smaller size than Pd (6.7 nm), and the particle size gradually increased approaching almost the size of Au (8.2 nm) at Au/Pd = 2.1. It was found that the content of Au influenced both activity and selectivity of PdAu/C catalysts for cinnamaldehyde hydrogenation. The C[dbnd]C group of cinnamaldehyde was preferentially hydrogenated compared to the carbonyl (C[dbnd]O) group on the Pd/C catalyst. At Au/Pd < 0.8 the effect of Au-content was relatively weak as reflected in the small decrease in the activity with only slight increase in the reactivity to C[dbnd]O hydrogenation. The effect of Au manifested distinctly on Au-rich catalysts (Au/Pd > 1) giving strongly reduced activity to saturated aldehyde formation accompanied by the selectivity preference to the carbonyl group reduction. These activity/selectivity effects induced by the gold were discussed to be related with the microstructure of metal particles determining their surface composition. Moreover, both geometric and electronic modifications of Pd-sites due to the Pd–Au interactions should be also taken into account. These interactions may also provide new active sites which are able to activate the C[dbnd]O bond of CAL in the PdAu alloy particles-containing catalysts.

Full text of DOI:10.1016/j.apcata.2014.08.036

Applied Catalysis A: General 487 (2014) 1–15
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Hydrogenation of cinnamaldehyde in the presence of PdAu/C catalysts
prepared by the reverse “water-in-oil” microemulsion method
T. Szumełda, A. Drelinkiewicz∗, R. Kosydar, J. Gurgul
Institute of Catalysis and Surface Chemistry, Niezapominajek 8, 30-239 Kraków, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 June 2014
Received in revised form 27 August 2014
Accepted 28 August 2014
Available online 6 September 2014
Liquid phase hydrogenation of cinnamaldehyde was studied on carbon supported (Vulcan XC72) – PdAu
catalysts with Au/Pd atomic ratio varying from 0.1 up to 2.1. Colloid-based method, namely the reverse
“water-in-oil” microemulsion technique was used to prepare the catalysts. Accordingly, pre-formed
metal nanoparticles of controlled size with a narrow size distribution were deposited on carbon support.
The BET, XRD, XPS, SEM, TEM, and HRTEM techniques were employed to characterize PdAu/C catalysts and
their monometallic 2%Pd/C and 2%Au/C counterparts. The metal particles were spherically shaped, nearly
monodispersed and well distributed throughout the carbon support. The particles of low Au-content
(Au/Pd < 0.8) were of smaller size than Pd (6.7 nm), and the particle size gradually increased approaching
almost the size of Au (8.2 nm) at Au/Pd = 2.1. It was found that the content of Au influenced both activity
and selectivity of PdAu/C catalysts for cinnamaldehyde hydrogenation. The C C group of cinnamalde-
hyde was preferentially hydrogenated compared to the carbonyl (C O) group on the Pd/C catalyst. At
Au/Pd < 0.8 the effect of Au-content was relatively weak as reflected in the small decrease in the activity
with only slight increase in the reactivity to C O hydrogenation. The effect of Au manifested distinctly
on Au-rich catalysts (Au/Pd > 1) giving strongly reduced activity to saturated aldehyde formation accom-
panied by the selectivity preference to the carbonyl group reduction. These activity/selectivity effects
induced by the gold were discussed to be related with the microstructure of metal particles determining
their surface composition. Moreover, both geometric and electronic modifications of Pd-sites due to the
Pd–Au interactions should be also taken into account. These interactions may also provide new active
sites which are able to activate the C O bond of CAL in the PdAu alloy particles-containing catalysts.
© 2014 Elsevier B.V. All rights reserved.
Keywords:
Pd–Au
Bimetallic
Cinnamaldehyde hydrogenation
1. Introduction
poly(vinylpyrrolidone) (PVP) or poly(vinyl alcohol) (PVA) was the
most frequently used procedure. The first producing colloidal PdAu,
The utilization of bimetallic nanoparticles either in the form
of colloids or supported catalysts has received enormous atten-
tion because they are very effective for the transformation of
organic compounds to highly useful chemical products. Supported
bimetallic catalysts are commonly prepared by conventional meth-
ods, such as co-impregnation and deposition-precipitation. These
methods usually produce relatively large metal particles with a
broad size distribution. Therefore, much attention has been paid
to the colloid-based methods [1,2]. The advantage is that the col-
loidal particles are homogeneous in the composition and the size.
Various methods have already been employed and co-reduction of
metal ions precursors in the presence of the protective polymers,
PtAu, PdPt nanoparticles of true alloy structure [3–5] was also
effectively adopted to synthesize SiO2 and Al2O3 supported PdAu
catalysts [6–8]. The use of method with PVA allowed to prepare
PdAu/C and PdAu/TiO2 catalysts with the particles of alloy struc-
ture with significantly narrower size distribution as compared to
those obtained by the impregnation methods [2,9]. Recently, THPC
(tetrakis(hydroxypropyl)phosphonium chloride), which acted as
both a stabilizer and a reducing agent was applied to synthe-
size SiO2, Al2O3, TiO2 supported PdAu catalysts [2,10]. The reverse
“water-in-oil” microemulsion (ME) is the other colloid-based syn-
thesis method which has already been successfully applied for the
preparation of various nanoparticles, including metals (Pt, Pd, Co,
Ni, etc.), metal oxides (Al2O3, TiO2, CeO2 etc.), salts (BaSO4, ZnS,),
polymers [11–13] and supported monometallic Pd, Pt, Co catalysts
[14–19].
∗
The reverse “water-in-oil” microemulsion consists of nano-
sized water droplets surrounded by an organic phase and stabilized
Corresponding author. Tel.: +48 126395114.
E-mail address: ncdrelin@cyf-kr.edu.pl (A. Drelinkiewicz).
http://dx.doi.org/10.1016/j.apcata.2014.08.036
0926-860X/© 2014 Elsevier B.V. All rights reserved.

2
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
by a surfactant. These water droplets act as nanoreactors and
provide size control for the formed nanoparticles. Thus, the treat-
ment of water droplets containing metal ion precursor with
reducing agent produces metal nanoparticles which are generally
very fine with a narrow size distribution. In general, the par-
ticle size distribution in supported catalysts prepared by ME is
significantly narrower as compared with those synthesized by com-
monly used impregnation methods. Using ME method it is also
possible to achieve a good control of the particle size. The experi-
mental parameters that determine the size of metal nanoparticles
can be shortly summarized as the nature of surfactant, water to
surfactant molar ratio (W), concentration and the type of metal pre-
cursor [PdCl2, Pd(NH3)4Cl2], and the type of reducing agent (NaBH4,
N2H4, H2). In order to prepare supported catalyst, destabilization
of microemulsion-containing pre-formed metal particles is carried
out by slow addition of organic solvent (most frequently tetrahy-
drofurane THF) in the presence of carrier material. This stage plays
an essential role in the dispersion/distribution of metal particles
in final catalysts. The bulk oily phase in the combination with the
hydrophobic character of the support surface determine the sup-
port – microemulsion interaction, and thus the distribution of metal
particles in final catalysts [17,18]. Hydrophobic character of the
support facilitates chemical compatibility between the microemul-
sion and the support surface favoring a better wetting of the support
by organic bulk phase thus improving the dispersion/distribution
of the metal particles.
The advantage of the microemulsion method is its potential
for the preparation of bimetallic particles. Simultaneous reduc-
tion of both metal ion components accomplished in water pool
offers the possibility to synthesize bimetallic particles of uniform
size and well-defined compositions. The kinetics and mechanism
of bimetallic colloidal PdPt, PtAu and PdAu particles formation by
ME method are reported in series of papers by Wu et al. [20–22].
The composition of these bimetallic particles was roughly consis-
tent with that of the precursor solutions and they had essentially
a homogeneous alloy structure. However, their outer surface dif-
fered from the bulk. For instance, the outer surface of colloidal PdAu
nanoparticles was Pd-rich [21].
of carbonyl and olefinic groups [28], PtNi/Al2O3 for 1,3-butadiene
hydrogenation [29].
In the present work the reverse “water-in-oil” microemulsion
method is utilized to prepare a series of carbon black (Vulcan XC72)
supported PdAu catalysts with Au/Pd ratio ranging from 0.1 up to
2.1.
The modification of Pd-sites reactivity by Au is a well known
phenomenon observed in a number of catalytic reactions, such
as oxidation of alcohols [30] and glycerol [31], direct synthesis of
hydrogen peroxide from H2 and O2 [9], hydrodesulfurization [6],
and hydrodehalogenation [32]. The reactivity of PdAu catalysts in
the hydrogenation reactions was definitively less extensively stud-
ied. Nevertheless, there are reports showing that the addition of
Au has an impact on the reactivity of Pd in the hydrogenation
of aromatic compounds [8], acetylene [33,34], butadiene [2,35],
and 2-hexyne [36]. This effect has been also recently studied for
the hydrogenation of cinnamaldehyde (CAL) and citral, the exam-
ples of ␣,␤-unsaturated aldehydes. The hydrogenation of CAL can
proceed either at C C or C O groups (Scheme 1) [37,38]. The
former gives saturated aldehyde (hydrocinnamaldehyde, HCAL),
the latter leads to unsaturated alcohol (cinnamalcohol, COL). These
products have industrial (pharmaceutic, perfurmery) and biolog-
ical applications. Recently, hydrocinnamaldehyde was found to
be an important intermediate in the synthesis of pharmaceut-
icals used in the treatment of HIV. The hydrogenation of C
bond in ␣,␤-unsaturated aldehydes is thermodynamically favored
thus making much more difficult the hydrogenation of the C
C
O
bond yielding unsaturated alcohols. Yang et al. [39] reported that
insertion of a small amount of Au leading to Au/Pd = 0.1–0.2 pro-
duced mesoporous supported alloy PdAu particles of smaller size
than that of pure Pd but enhanced activity for cinnamaldehyde
hydrogenation. This activity improvement was attributed to syn-
ergistic Pd–Au effect, which however had only slight impact on
the selectivity control. High selectivity to C C hydrogenation yield-
ing saturated aldehyde typical for initial Pd catalyst (ca. 88%) only
slightly increased (ca. 90%) on these PdAu catalysts with low Au
contents [39]. Similar activity/selectivity effects were also observed
for Al2O3, SiO2 and C supported nanostructured Pt-on-Au catalysts
[40,41] in which a small amount of fully dispersed Pt entities were
deposited onto the surface of pre-formed colloidal Au particles.
The Pt entities resulted in dramatic enhancement of Au activity
without altering the selectivity of Au-catalyst. The high selectiv-
ity to C C hydrogenation (70–80%) observed on Au/SiO2 catalyst
[42] was preserved on these nanostructured Pt-on-Au catalysts. In
these catalysts the Pt entities were acting as the activity promoter
of Au only, since the selectivity propensity of Au for the hydro-
genation of CAL (and other carbonyl reactants) was not changed by
Most publications devoted to bimetallic catalysts prepared by
ME method concern electrocatalysts (PtRu, PtNi, PtCo, PtFe, PdAu,
AuAg, PtAu) for the polymer membrane fuel cell, methanol oxida-
tion and oxygen reduction reactions [23–25]. The suitability of the
microemulsion method is also reflected by an increasing number
of published works dealing with the supported bimetallic catalysts
for various organic transformations such as PtIr/boehmite for the
ring opening reaction in indan [26], PdPt/Al2O3 for hydrogenol-
ysis and isomerization [27], NiB nanoparticles for hydrogenation
O
H
H2
H2
HCAL
O
OH
H
HCOL
PB
CAL
H2
H2
OH
COL
H2
CH3
Scheme 1. Hydrogenation pathways of cinnamaldehyde.

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
3
THF solvent at a very slow rate 0.25 cm3/min with the aid of the
automatic syringe pump. The system was vigorously stirred. The
volume of THF added was 3–4 times the volume of the microemul-
sion. Upon addition of THF, the color of liquid gradually changed
from black to gray and finally the liquid was colorless, showing
complete deposition of metal particles on the support. The obtained
catalyst was separated by filtration, dried in air for ca. 24 h and
washed with copious amount of methanol and acetone six times
and filtrated in between. Finally, the catalysts were washed with
water up to the removing of chloride ions, dried overnight in air
and then at 120 ◦C for 16 h.
the presence of Pt. The Pt sites were responsible for the activation
of H2 molecules while activation of aldehyde occurred on the Au
sites and the activity improvement was due to H2 spillover effect.
The Pt entities activated hydrogen which migrated by spillover
toward nearby Au sites on which aldehyde molecule is adsorbed.
By hydrogen spillover effect also the enhanced activity of physically
mixed less active Au/TiO2 with Pd/TiO2 catalysts in hydrogenation
of citral was explained [43]. Simultaneously the selectivity to C
C
hydrogenation was increased. Catalytic performance of ordered
mesoporous carbon containing separate Au and Pd nanoparticles
was also explained by hydrogen spill-over effect [44]. It resulted in
promotion of C C selectivity giving HCAL accompanied by much
higher activity compared to monometallic Pd.
In the present work, the hydrogenation of cinnamaldehyde
(CAL) has been chosen for exploring the catalytic performance
of PdAu/C catalysts synthesized by the microemulsion method.
The catalysts with fixed Pd loading (2 wt%) and the content of Au
ranging from very small as in Au-poor (Au/Pd = 0.1) catalyst up
to as high as Au/Pd = 2.1 in Au-rich catalysts are prepared. Their
monometallic Pd and Au counterparts are also synthesized. Carbon
black Vulcan XC72 is used as the support because of its advanta-
geous textural characteristics, among them extended mesoporous
structure and hydrophobous character [45]. Our intention is to
exploit reactivity of these systems for CAL hydrogenation and in
2.2. Methods of characterization
The X-ray diffraction (XRD) patterns were obtained with a Philips
X’PERT diffractometer using Cu K␣ radiation (40 kV, 30 mA). The
average diameter of Pd particles was calculated on the basis of the
Pd (1 1 1) peak broadening according to Scherrer equation (taking
into account instrumental broadening).
The X-ray Photoelectron Spectroscopy (XPS) measurements were
carried out with a hemispherical analyzer (SES R4000, Gammadata
Scienta). The unmonochromatized Al K␣ and Mg K␣ X-ray source
with the anode operating at 12 kV and 20 mA current emission was
applied to generate core excitation. All binding energy values were
corrected to the carbon C 1s excitation at 285.0 eV. The samples
were pressed into indium foil and mounted on a holder. All spec-
tra were collected at pass energy of 100 eV except survey scans
which were collected at pass energy of 200 eV. Intensities were
estimated by calculating the integral of each peak, after subtraction
of the Shirley-type background, and fitting the experimental curve
with a combination of Gaussian and Lorentzian lines of variable
proportions (70:30).
Electron Microscopy (SEM, TEM) studies were performed by
means of Field Emission Scanning Electron Microscope JEOL
JSM–7500F equipped with the X-ray energy dispersive (EDS) sys-
tem. Two detectors were used and the images were recorded in
two modes. The secondary electron detector provided SEI images,
and back scattered electron detector provided BSE (COMPO) micro-
graphs. The catalyst sample was mixed with methanol and the
obtained slurry was further treated with ultrasounds (5–10 min).
Then, a droplet of the solution was put onto a grid, dried in air and
used in the TEM instrument.
particular the capability of the Au-presence to control the C C/C
selectivity.
O
2. Experimental
2.1. Catalysts preparation
Carbon support Vulcan XC72 (supplied by CABOT) was used as
the support. The catalysts consisting of the same palladium load-
ing equal to 2 wt.% Pd and the growing Au content ranging from
0.4 wt% up to 6 wt% were prepared by means of reverse “water-in-
oil” microemulsion method according to our previously published
procedure [18,19]. In the PdAu/C catalysts the molar ratio of Pd:Au
ranges from 1:0.1 up to 1:2.1 and they are designated PdAu-y,
where y represents the Au/Pd molar ratio (the number of Au moles
per 1 mol of Pd). The monometallic 2 wt% Pd/C and 2 wt% Au/C cat-
alysts were also prepared using the same preparation procedure.
The reverse micellar solutions were prepared using
polyoxyethylene(7-8)octylphenyl ether (Triton-X 114) (Aldrich)
as the surfactant and cyclohexane (Aldrich) as the oil phase.
The aqueous solutions of PdCl2 (molar ratio of NaCl:PdCl2 = 2,
Pd2+ ions concentration of 0.2 mol/dm3 and HAuCl4 (0.05 mol/dm3)
were used as the precursors, and hydrazine hydrate (N2H4 × H2O,
N2H4 64–65%, reagent grade 98%, Aldrich) was used as reducing
agent. Taking into account our previous studies [18] the value of
parameter W equal to 5 was applied.
The solution of TRITON-X-114 in cyclohexane (surfactant con-
centration of 0.62 mol/dm3) was prepared. In order to prepare 1 g
of 2%Pd/C catalyst 0.94 cm3 of PdCl2 solution was added to 16.8 cm3
cyclohexane-surfactant solution and the obtained suspension was
vigorously stirred up to the formation of transparent, clear dark
orange solution. In the preparation of PdAu catalysts a mixture of
PdCl2 and HAuCl4 solutions of appropriate volume ratio and the
cyclohexane solution of higher surfactant concentration giving the
parameter W = 5 was prepared. The hydrazine (64–65%) was added
directly to the microemulsion using the reducing agent to metal
molar ratio of 20. The reduction of metal ions was quick, and the
color of liquid quickly changed to black. Stirring was then contin-
ued for another 1 h. Then, carbon support (0.98 g) was introduced
and the system was kept under stirring for 1 h. Deposition of metal
nanoparticles on the support was carried out by introducing the
To prepare the particle size distribution diagram and estimation
of the average particle size (d), at least 100 particles (N) were man-
ually counted. The average metal particle diameter was calculated
as the average of individual diameters from the formula:
ꢀ
nidi
ꢀ
d =
ni
where ni is the number of particles having diameter di
The counting was carried out on electron micrographs regis-
tered starting from 100 000 to 200 000 magnifications, where metal
(Pd, Pd/Au, Au) particles contrasted very well with the support and
they were clearly detected. On these micrographs registered in BSE
(COMPO) mode the accuracy of the particle size scale was 0.5 nm.
HRTEM and STEM studies were performed on FEI Tecnai G2
transmission electron microscope operating at 200 kV equipped
with EDAX EDX and HAADF/STEM detectors. Samples for analy-
sis were prepared by placing a drop of the suspension of sample
in ethanol or THF onto a carbon-coated copper grid, followed by
evaporating the solvent.

4
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
2.3. Hydrogenation reactions
where NCAL,0 and NCAL are the moles of CAL initially present
in the reactor and the moles of CAL remaining at time t,
respectively; NHCAL, NHCOL and NPB are the numbers of moles
of hydrocinnamaldehyde (HCAL), saturated alcohol (HCOL) and
propyl benzene (PB) in the reaction mixture at the time t.
Hydrogenation experiments were carried out in an agi-
tated batch glass reactor at constant atmospheric pressure of
hydrogen at temperature of 22 ◦C following the methodology pre-
viously described [46,47]. Toluene was used as the solvent.
In the initial stage of hydrogenation experiment, the catalyst
and the solution were placed in the reactor in such a way that
there was no immediate contact between them. The catalyst was
activated “in situ” by passing nitrogen (20 min) and subsequently
hydrogen (30 min) at temperature of 22 ◦C before the hydrogena-
tion test. Then, the catalyst was contacted with the hydrogenated
solution and the experiment started. The progress of the reaction
was monitored by measuring the hydrogen consumed against reac-
tion time. Samples of solutions were withdrawn from the reactor
via a sampling tube at appropriate intervals of time and they were
analyzed by GC.
3. Results and discussion
3.1. Characterization of catalysts
All catalysts studied in this work were prepared using the
same microemulsion composition with the same value of param-
eter W = 5. Textural properties of studied catalysts are collected in
Table 1. As described before, carbon black Vulcan XC72 is used in
the present studies because of its advantageous textural charac-
teristics and hydrophobic surface. As shown in Table 1, the carbon
material has surface area of 228 m2/g and it exhibits an extended
mesoporous structure evidenced by high contribution of meso-
pores (1.72 cm3/g) relative to that of micropores (0.0484 cm3/g).
The pore size distribution diagram indicates the domination of
pores of 1.7–2.5 nm in diameter (Supplementary Fig. S1). These
textural features are consistent with the literature data [45,48,49].
Electron microscopic images (see below) indicate that the car-
bon material is composed of aggregated quasi spherical particles
with a size within 20–50 nm. These carbon particles are consti-
tuted of groups of parallel graphene planes separated by about
0.35–0.5 nm, which are structured into little crystalline phases.
They are observed as a very broad reflection centered at 2ꢀ = 43–44◦
in the XRD pattern of carbon material (Fig. 1).
Deposition of pre-formed metal particles results in distinct
changes in textural characteristics of carbon support. The specific
surface area is reduced and is accompanied by remarkable changes
in the porous structure (Supplementary Fig. S1). As the content
of Au, and thus the total metal content (Pd + Au) in the catalysts
increases, the surface area decreases. Simultaneously, the volume
of micropores decreases whereas that of mesopores does not essen-
tially change (Table 1). Pore distribution profiles indicate that apart
from some changes in the micro structure, distinct changes in a
mesoporous structure appear also. As XRD and microscopic data
show (see below) the metal particles in the PdAu/C catalysts are
of 5–8 nm in size, e.g. they are too large to be introduced inside
the micropores. Therefore, it may be assumed that the mesoporous
structure of the carbon material allows metal particles to adsorb
at the entrance of micropores with consequent blockage of micro-
pores. Similar effect was observed during preparation of Vulcan
XC72 supported electrocatalysts consisting of Pd–Pt nanoparticles
[48] or after modification with organic dopants [49].
Products analysis was performed with
a gas chromato-
graph PE Clarus 500 equipped with a flame ionization detector
under the following conditions: capillary column Elite-5 MS
(30 m × 0.25 mm × 0.25 ␮m coating) with helium as a carrier gas
(flow rate 1 ml/min) and injection temperature 250 ◦C. Product sep-
aration was obtained using temperature ramps: 65 ◦C for 1 min,
65–108 ◦C (15 ◦C/min), 108–115 (5 ◦C/min), 115–265 (15 ◦C/min)
to 265 ◦C hold for 3 min. Decane (Sigma–Aldrich) was used as the
standard.
Typically the hydrogenation test was carried out at 22 ◦C using
20 cm3 of cinnamaldehyde solution in toluene of c0 = 0.05 mol/dm3
and catalyst concentration 5 g/dm3. The same conditions were used
in hydrogenation of HCAL as the substrate. In hydrogenation of cin-
namalcohol (COL) reactant (c0 = 0.05 mol/dm3) the concentration of
catalyst was lower and equal to 0.25 g/dm3.
Shaking of the reactor was carried out at such a speed to ensure
that the rate of hydrogen uptake did not depend on agitation speed.
External diffusion resistance was ruled out, since no significant
variation in the catalytic results was observed by increasing the
agitation speed.
The conversion (C; %), selectivity to hydrocinnamaldehyde
[S(HCAL); %], saturated alcohol [S(HCOL); %] and propyl benzene
[S(PB); %] were calculated with the formulas
NCAL,0 − NCAL
CCAL
=
× 100%
(1)
(2)
NCAL,0
Ni
NHCAL + NHCOL + NPB
Si =
× 100%
Table 1
Physicochemical characterization of the catalysts, Au/Pd atomic ratio determined by EDS and STEM, and the initial rate of cinnamaldehyde hydrogenation.
***
Sample
Surface
Porosity [cm3/g]
Metal particle
size d [nm]
Au/Pd atomic ratio
Initial rate R
TOFIN
area [m2/g]
[mol CAL 10−5 min−1 g (cat) −1
]
[min−1
]
Total
1.77
Meso
1.722
Micro
SEM
XRD
EDS
–
STEM
Carbon
228
0.0484
–
–
–
Pd/C
6.7
5.5
4.8
5.2
5.3
7.5
5.5–6
5.5
22.7
22.1
21.4
18.5
12.1
7.8
1.21
1.07
0.95
0.70
0.36
0.15
PdAu-0.1 (Pd91Au9)*
204
188
179
1.41
1.42
1.40
1.37
1.39
1.38
0.0358
0.0263
0.0212
No data**
13/87
PdAu-0.2 (Pd83Au17
PdAu-0.4 (Pd71Au29
PdAu-0.8 (Pd56Au44
PdAu-1.8 (Pd36Au64
)
)
)
)
4–4.5 32/68
5.5–6 55/44; 54/46
6.5–7 63/37; 65/35;
64/36
166
138
1.34
1.38
1.32
1.37
0.0167
0.0066
62/38; 70/30
PdAu-2.1 (Pd32Au68
)
7.8
8.2
7–7.5 69/31; 66/34
66/34; 62/38;
74/26
4.0
0.07
–
Au/C
8–9
*
Nominal composition from the preparation.
Not determined because of very weak signal.
**
***
Calculation of TOFIN (min−1) from initial hydrogenation rate. TOF numbers were calculated on the basis of total loading of metals.

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
5
studied the monometallic Pd, Pt, Au and corresponding bimetal-
lic Pd–Au, Pt–Au and Pd–Pt nanoparticles (all the particles of size
in similar range). They were prepared by colloid-based method,
namely ethanol reduction of metal ions in the presence of PVP. The
obtained Pd particles gave the weakest XRD reflections whereas
those of Pt were more intense and the Au particles gave the most
distinct reflections. This effect has been also observed elsewhere
and attributed to more amorphous structure of palladium particles.
In the diffraction patterns of PdAu/C catalysts, there are no
diffractions arising from the crystalline Pd or Au components
(Fig. 1). The diffraction peaks locate at 2ꢀ intermediate between the
Au (1 1 1) and the Pd (1 1 1), which is generally accepted to be an
indication of PdAu alloy formation [5,8,39,50,51]. As the content of
Au grows the observed diffraction is progressively shifted to lower
angle relative to that of crystalline Pd. These XRD data indicate that
the PdAu particles retained original (fcc) crystalline structure and
gold is involved in alloying with palladium. A broad reflection at
about 2ꢀ = 43–44◦ associated with carbon can be seen in the pat-
terns of all catalysts. However, as the content of Au increases the
maximum of this broad diffraction is gradually shifted because of
its superimposition with that arising from the (2 0 0) planes.
All diffraction peaks arising from metal particles are broad thus
indicating the metal particles of size in nanoscale. The average par-
ticle size was calculated from the Pd (1 1 1) line broadening using
the Scherrer equation. The obtained average metal particle sizes
are given in Table 1.
The lattice parameter (a) is also calculated as described by other
authors [51,52] (Supplementary Fig. S2). The addition of Au leads
to an increase in the lattice parameter. Its growth is especially dis-
tinct at lower content of Au, up to Au/Pd = 0.8, after which relatively
small change occurs. For PdAu-0.8 and PdAu-1.5, the lattice param-
˚
˚
eter increases only slightly from a = 4.02 A to a = 4.06 A although
2-fold growth in the Au content occurs. The latter value of a is
˚
only slightly lower than a = 4.086 A, calculated for the monometallic
Au/C catalyst.
The SEM images of monometallic and bimetallic catalysts
together with corresponding particle size distributions diagrams
are displayed in Figs. 2–4.
The Au/Pd atomic ratio was determined by the EDS analysis
and the obtained data are collected in Table 1. They agree with
the nominal data for studied PdAu/C catalysts, evidencing effective
deposition of both metal components, Pd and Au, on the carbon
support. It can be observed from the micrographs registered at
low magnification (Fig. 2) that metal particles, even at high total
metal loading as in the PdAu-1.8 catalyst (2 wt% Pd + 6.7 wt% Au)
are well and almost uniformly distributed on the carbon sup-
port. In all catalysts the mono- as well as bimetallic particles all
are spherically shaped, nearly monodispersed and the individual
nanoparticles strongly predominate. However, some larger “white
spots” of shapes suggesting aggregation of few individual spherical
particles can be seen. This indicates that although aggregation of
particles takes place during the deposition of metal particles onto
carbon support, the contribution of this process is very low. The
particle size distribution diagram was determined by measuring
the size of over hundred particles from different areas in the SEM
micrographs. The aggregated particles were omitted in the particle
size analysis.
As shown in Fig. 3 Pd particles in the monometallic 2%Pd/C cat-
alyst are nearly monodispersed with size in narrow range 4–9 nm
giving the average Pd particles size d = 6.7 nm. The Au particles in
the monometallic 2%Au/C catalyst are of larger size, d = 8.2 nm, and
the size distribution is broader compared to Pd/C catalyst (Fig. 2).
The results are consistent with previous data revealing that at the
same parameter of “w” (at the same microemulsion composition)
the size of particles varied being to some extent determined by
the kind of metal Pd, Pt, Au [22]. The particle size distribution
Fig. 1. XRD diffraction pattern of carbon support and carbon supported Pd/C, Au/C
and PdAu/C catalysts.
The observed changes in textural properties (Table 1) sug-
gest penetration of metal particles inside the carbon carrier. This
could be related to the method of metal particles deposition. It
was performed by contacting support with the microemulsion
which predominantly consists of non-polar cyclohexane solvent. As
described before, hydrophobic character of support favors chemical
compatibility between the microemulsion and the support sur-
face thus making more effective wetting of the support by the
organic solvent and improves the distribution of the metal particles
throughout the support grains.
The XRD diffraction patterns of studied catalysts are displayed
in Fig. 1. In the diffraction pattern of 2%Au/C catalyst the intense
diffractions at 2ꢀ 38.16◦ and 44.3◦, indexed to the (1 1 1) and
(2 0 0) planes of face-centered cubic (fcc) crystalline Au, respec-
tively appear. The latter diffraction (44.3◦) is superimposed with
that of carbon support. In the case of 2%Pd/C catalyst, the diffrac-
tion located at 2ꢀ 40.16◦ indexed to (1 1 1) plane of face centered
cubic (fcc) crystalline Pd can be seen. Although in both these cat-
alysts metal loadings are the same, 2 wt%, the diffractions of gold
are much more intense compared to that of palladium. This phe-
nomenon is consistent with the results of Toshima et al. [4,5] who

6
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
Fig. 2. SEM images and metal particles size distribution diagrams of 2%Au/C and PdAu-1.8 catalysts registered at magnification 25 000 and 10 000, respectively.
diagrams indicate that the size of metal particles is dependent
on the Au/Pd ratio (Table 1). At low content of Au resulting in
Au/Pd < 1, the particles have the average size smaller compared
to that of monometallic Pd. When the Au/Pd molar ratio increases
above Au/Pd = 0.8 and the content of Au is equal or higher than that
of Pd, the size of particles gradually grows reaching 7.8 nm in the
PdAu-2.1 catalysts vs 8.2 nm in the Au/C (Figs. 2 and 3). Moreover,
at higher Au/Pd ratio, e.g. in Au-rich samples, the size distributions
of particles are remarkably broader. The particles of size in the
widest range and the largest average size (8.2 nm) can be seen in
the monometallic 2%Au/C catalyst.
Thus, the average size of the bimetallic particles in the stud-
ied catalysts depends on the content of Au (Au/Pd ratio) and
this relationship is displayed in Fig. 4. It should be pointed out
that the size of metal particles calculated from the microscopic
images agrees with those determined from the XRD diffractions.
The relationship shown in Fig. 4 is consistent with previous results
for the PdAu nanoparticles prepared by co-impregnation [39],
adsorption–reduction [50] as well as colloid-based [4,5,21] meth-
ods. The colloidal bimetallic PdAu particles of Au/Pd molar ratio
below 1 exhibited smaller size than the monometallic Pd or Au.
However, at higher Au/Pd ratio (when the content of Au increased)
the particles become larger reaching the size of monometallic Au
[4]. Similar size-effect was observed for other bimetallic particles
such as Pd–Pt, Pt–Au and it manifested especially distinctly for the
latter Pt–Au nanoparticles [4,5]. This effect was discussed to arise
from a specific role of individual metal components in the course
of metal particles formation including nucleation and aggregation
processes [50].
Selected PdAu/C catalysts are also subjected to the HRTEM stud-
ies to verify the alloy nature of the PdAu nanoparticles. Both the
morphology and the composition of individual particles, chosen
randomly, is also studied by the STEM attached to a high-resolution
HRTEM. No particles composed of Pd or Au only were observed by
the STEM in studied catalysts, with low as well as high Au/Pd ratios.
The representative HRTEM images of catalysts with low Au content
(PdAu-0.1, PdAu-0.4) are displayed in Fig. 5 and the micrographs of
Au-rich (PdAu-1.8, PdAu-2.1) catalysts are shown in Supplemen-
tary Fig. S3. It is observed that the particles are single crystals with
all the crystal lattices through the whole particle, indicating the
alloy form of Pd and Au in the PdAu-2.1 catalyst.
The STEM analysis confirms the co-existence of both met-
als in the particles, with the average composition of individual
particle which is roughly in agreement with the nominal con-
tents of Pd and Au used in the synthesis. The obtained average
data from the STEM analyses are collected in Table 1. How-
ever, the deviation of composition is observed among particles.
It might result from very weak detection signal as well as from
the detection errors because of very small size of the parti-
cles.

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
7
Fig. 3. SEM images and metal particles size distribution diagrams of 2%Pd/C and PdAu-0.2, PdAu-2.1 catalysts registered at magnification 200 000.
The catalysts were also subjected to the X-ray photoelectron
spectroscopy (XPS) experiments. This technique showed the Pd,
Au, C elements and traces of oxygen whereas no chloride was
ascertained in outermost surface of studied catalysts. The obtained
surface composition of Pd and Au (in at%) and the binding energies
of palladium Pd 3d5/2 and Au 4f7/2 peaks are collected in Table 2.
Here, the surface concentrations of palladium and gold estimated
by XPS is taken into discussion, although it is well known that
the depth of X-ray penetration might be up to 7–10 nm as it is to
some extent influences by chemical composition (atomic number)

8
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
80
60
40
20
0
80
PdAu-0.1
PdAu-0.4
60
d = 5.2 nm
d = 5.5 nm
40
20
0
3.0 4.0 4.9 5.7 6.5 7.4 8.2 9.0 9.8 11.5
d [nm]
3.0 4.0 4.9 5.7 6.5 7.4 8.2 9.0 9.8 11.5
d [nm]
80
60
40
20
0
PdAu-0.8
d = 5.3 nm
3.0 4.0 4.9 5.7 6.5 7.4 8.2 9.0 9.8 11.5
d [nm]
10
9
8
7
6
5
4
3
0
Pd
0.2 0.4 0.6 0.8
1
1.2 1.4 1.6 1.8
2
2.2 2.4 2.6
Au
Au/Pd [mol/mol]
Fig. 4. Metal particles size distribution of PdAu-0.1, PdAu-0.4 and PdAu-0.8 catalysts and the relation between metal particle size and content of Au (Au/Pd ratio).
of analyzed surface components. Nevertheless, the data provided
by this technique are widely considered to examine microstructure
of bimetallic particles-containing supported catalysts.
pointed out that during deposition of pre-formed metallic nanopar-
ticles on the carbon support the particles penetrated inside pore
structure of carbon material. This penetration effect is reflected in
the reduction of specific surface area accompanied by changes in
the porous structure (Supplementary Fig. S1). On the other hand,
the size of particles changed being determined by the Au/Pd ratio
(Fig. 4). At Au content corresponding to Au/Pd > 0.8, the size of
particles started to significantly grows. This could make their pen-
etration more difficult thus giving higher surface concentration of
Pd determined by XPS.
The PdAu/C catalysts were prepared in such way, that the
content of Pd was equal to 2 wt% whereas that of Au progres-
sively grows from 0.4 wt% in PdAu-0.1 sample up to 7.8 wt% in the
PdAu-2.1 catalyst. As shown in Table 2, the surface concentration of
Au grows in line with increasing the Au/Pd ratio. The surface con-
centration of Pd is low and it is within similar range (0.05–0.07 at%)
in the catalysts with lower Au/Pd ratio, up to PdAu-0.8. However,
in both Au-rich catalysts, PdAu-1.8 and PdAu-2.1, the Pd surface
concentration grows to 0.11–0.15 at% Pd, respectively. It should be
The XPS derived Au/Pd atomic ratio was calculated from the
intensities of the Au and Pd peaks. However, at high Au-content

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
9
Fig. 5. Micrographs of catalysts with low Au content, PdAu-0.1 and PdAu-0.4 and electron diffraction spectrum for PdAu-0.1 catalyst.
this analysis was to some extent complicated because of some
overlap between the Pd3d doublet and the Au 4d5/2 component.
Therefore, the sensitivity factors were introduced and the cal-
culated values are collected in Table 2. The Au/Pd atomic ratio
determined by XPS in outermost surface is higher than the nom-
inal value from the preparation. This shows the Pd-enrichment
which is observed within the whole Au content. Even if the con-
tent of Au is higher than that of Pd as in the Pd–Au-1.8 and
Pd–Au-2.1 catalysts the surface is enriched by Pd, what is evi-
denced by the Au/Pd molar ratio below 1. In the unsupported
PdAu colloidal particles prepared by microemulsion method the
surface of particles was also reported to be enriched by Pd
[21].
The XPS data demonstrate the binding energy of Pd 3d5/2 within
the range of 335–336 eV (Table 2). The peak component of binding
energy at ca. 335 eV which is characteristic of metallic Pd [53] domi-
nates in the spectra of all catalysts. In the spectrum of monometallic
Pd/C catalyst the only Pd state of binding energy 335.3 eV corre-
sponding to the Pd metal appears. In the spectra of all Au-containing
catalysts the binding energy of Pd-metal peak component is slightly
shifted (Supplementary Fig. S4). Moreover, the less intense peak
component at slightly higher energy of ca. 336–337 eV appears.

10
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
Table 2
XPS data: surface concentration (at%), Pd 3d5/2 and Au 4f7/2 binding energies (eV), and the Au/Pd atomic ratio.
Catalyst
C
O
Pd
Au
Au/Pd atomic ratio
Au 4f7/2 [eV]
Pd 3d5/2 [eV]
Pd/C
0.16
0.03
0.06
0.05
0.11
0.15
335.3
Pd–Au-0.1
Pd–Au-0.4
Pd–Au-0.8
Pd–Au-1.8
Pd–Au-2.1
97.05
97.71
95.51
96.97
97.07
2.13
1.84
3.99
2.33
2.35
Trace
0.01
0.02
0.10
0.13
84.4
83.9
84.6
84.6
84.5
336.07 (0.51)*
335.7 (0.53)
335.7 (0.68)
335.4 (0.59)
335.5 (0.57)
337.1 (0.49)
336.3 (0.47)
337.5 (0.32)
336.5 (0.41)
336.9 (0.43)
0.16
0.40
0.91
0.87
*
Contribution of energy state
The Pd binding energy values slightly above 336 eV are commonly
ascribed to the partially oxidized Pds+ species of lower electron den-
sity that Pd◦ or to the palladium in palladium oxide (336.5 eV) [54].
Thus, these spectral data could be an indication of Pd–Au electron
interaction. The binding energies registered for gold are ca. 84.3 eV
being typical of metallic Au [53].
The obtained Pd spectra are similar to those reported for
supported PdAu catalysts prepared by other colloid-based meth-
ods, namely by simultaneous reduction of palladium and gold by
ethanol in the presence of stabilizing polymer (PVP) [6,21].
100
80
60
40
20
0
3.2. Catalytic results
The catalytic hydrogenation of cinnamaldehyde (CAL) has been
chosen to evaluate the catalytic properties of PdAu/C catalysts
because this reaction is very sensitive to the surface properties
of catalysts. In general, CAL has two functional groups, C C and
0
20
40
60
80
100 120 140 160 180
t [min]
C
O, and their hydrogenation can lead to several products, namely
Pd
PdAu-0.8
PdAu-0.1
PdAu-1.8
PdAu-0.2
PdAu-2.1
PdAu-0.4
hydrocinnamaldehyde (HCAL), cinnamyl alcohol (COL), 3-phenyl-
1-propanol (HCOL) and the products of hydrogenolysis reaction,
␤-methyl, styrene and phenyl propane (PB) (Scheme 1). Palladium
is known to exhibit especially high reactivity toward the unsatu-
rated C C bond hydrogenation yielding hydrocinnamaldehyde as
the dominating product.
Fig. 6. Conversion of cinnamaldehyde against reaction time on monometallic Pd/C
and bimetallic PdAu/C catalysts.
In the presence of all studied catalysts, the hydrocinnamalde-
hyde (HCAL) and saturated alcohol (HCOL) are formed from the very
beginning of the reaction as the main products. Among hydrogenol-
ysis products, phenyl propane (PB) is detected but its content
is distinctly lower (2–8 mol%) compared to HCAL and HCOL. No
cinnamyl alcohol (COL) is detected at the present hydrogenation
conditions (toluene, 22 ◦C, atmospheric pressure of hydrogen) sim-
ilarly to what was previously reported on Pd catalysts [55–58]. No
other side products are detected.
Hammoudeh et al. [55] proved that in the hydrogenation of CAL
on Pd/SiO2 catalyst carried out at low temperature (25 ◦C) all satu-
rated alcohol formed does entirely come through the consecutive
hydrogenation of the cinnamyl alcohol. The experiments with COL
as the substrate showed 30-times faster COL hydrogenation than
the hydrogenation of CAL to cinnamyl alcohol [55,56]. A similar
reactivity was also reported by Cairns et al. [57] for Pd/SiO2 catalysts
obtained by hydrogen reduction of [Pd2Br4(PR3)2]. These catalysts
were unable to reduce HCAL to saturated alcohol in decalin solvent
at 135 ◦C whereas they readily converted cinnamyl alcohol to HCOL.
Based on these literature data, the control hydrogenation exper-
iments are performed on 2%Pd/C and bimetallic PdAu-1.8 using
HCAL as the substrate. On both catalysts, the hydrogenation of HCAL
at 22 ◦C does not produce any detectable amount of saturated alco-
hol. The concentration of HCAL in the reaction mixture does not
decrease and no consumption of hydrogen appears during catalytic
tests lasting up to 4–6 h. Thus, in agreement with [55–57] at present
mild reaction conditions (22 ◦C, atmospheric H2 pressure) satu-
rated alcohol HCOL is produced exclusively from the hydrogenation
of COL as an initial step in the reaction.
hydrogenation in terms of both, the rate and the selectivity control
toward C C/C O hydrogenation. The selectivity data are discussed
as the selectivity to either hydrocinnamaldehyde (HCAL) or fully
saturated alcohol (HCOL). Although the content of PB was very low
the selectivity to PB is also calculated.
Among all studied catalysts, the monometallic 2%Pd/C exhibits
the highest CAL conversion (Fig. 6) and the formation of HCAL
exceeds that of fully saturated alcohol (HCOL) (Supplementary Fig.
S5) giving the selectivity to HCAL as high as 65–70% (Fig. 7). This
high selectivity to HCAL is almost preserved vs CAL conversion
(Fig. 7). It is equal to ca. 70% in early stages of reaction and only
slightly decreases as the CAL is reacted attaining ca. 65% at almost
complete CAL conversion. For the 2%Au/C catalyst, the hydrogena-
tion reaction did not proceed at present experimental conditions.
The activity of present 2%Pd/C catalyst could be compared with
those of other Pd-catalysts in hydrogenation of CAL. Therefore the
value of TOF (s−1) expressed as the number of CAL moles converted
per second and per surface Pd atom is calculated. From the size of
Pd particles determined to be 6.7 nm (Table 1) the dispersion is cal-
culated to be 16.7% which gives the TOF = 0.13 s−1. This value could
be compared with the literature data collected in Table 3. Apart
from the TOF values also the conditions of CAL hydrogenation are
reported (solvent, temperature H2 pressure) because of their essen-
tial role in the activity. All the literature data concern reaction at
temperature of 60 ◦C or higher and at H2 pressure higher than 1 atm.
On the other hand, temperature of 25 ◦C and H2 pressure ca. 1 atm
are the conditions in present catalytic tests. Nevertheless, it can be
observed that TOF values are within wide range from 0.16 to 6.7
and they are determined not only by the metal particle size but
also by the type of support. In the case of Pd/C catalysts the TOF
values from 0.16 [59] up to 6.7 [60] can be observed. It should be
The data in Table 1 show that the content of Au in the cat-
alysts plays an important role in determining the course of CAL

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
11
80
70
60
50
40
30
20
stressed out that the selectivity control to C C/C O also remark-
ably varied. These carbon-induced effects have been discussed to
be result of various influence of carbon material on the Pd (Pt)
particles which is determined not only by textural and morphologi-
cal features (micro, meso-porosity, nanofibers, nanotubes) but also
by “carbon surface chemistry”. The presence of oxygen-containing
surface groups, their amount and the type (carboxyl, low or non-
acidic such as hydroxyl) were observed to have a strong impact on
the catalytic properties. According to Toebes et al. [64] the activity
enhancement by a factor of 25 was reached at the expense of selec-
tivity for the carbon nanofiber-supported Pt catalysts as a result
of the thermal removal of oxygen groups. The reason is explained
that the polarity changes of the catalyst surface due to removal
of oxygen species influence the activity by modifying the prefer-
ential adsorption modes of CAL. Furthermore, the electronic effect
induced by the carbon nanotubes due to differences in their cur-
vatures was also found to have a strong impact on the catalyst
performance [65]. The activity of present Vulcan XC72 supported
catalyst (TOF = 0.13 s−1) is lower compared to those of other cata-
lysts given in Table 3, however, in contrast to the latter catalysts,
this activity is determined at mild reaction conditions. Tempera-
ture increase from 20 to 60 ◦C results in ca. 3–4 fold increase in the
activity. Nevertheless, the activity of present Pd/C catalyst is only
0
10 20 30 40 50 60 70 80 90 100 110
Conversion [%]
Pd
PdAu-0.1
PdAu-1.8
PdAu-0.2
PdAu-2.1
PdAu-0.4
PdAu-0.8
70
slightly lower than that of Pd/CNT-TiO2 catalyst (TOF = 0.16 s−1
)
60
50
40
30
20
determined also in toluene solvent but at higher temperature (80 ◦C
and H2 pressure (2 MPa)) [59]. It can be assumed that the elec-
troactive features of Vulcan XC72 carbon may play a role similarly
to what was suggested by Sun et al. [41] from the comparison of
catalytic performance of Au-based catalysts on various supports,
among them Vulcan XC72 in CAL hydrogenation.
It is observed that as the Au content in the AuPd/C cata-
lysts increases the hydrogenation activity is more or less reduced
depending on the Au/Pd ratio. The relationships showing CAL con-
version vs reaction time (Fig. 6) indicate that low Au content
corresponding to Au/Pd < 0.8 does not significantly reduce activity
of catalysts. All these catalysts exhibit high CAL conversions. How-
ever, at higher Au content a decrease in the rate of CAL conversion
is distinct and the PdAu-2.1 catalyst with the highest Au/Pd ratio is
about 5-times less active than monometallic Pd/C.
0
10 20 30 40 50 60 70 80 90 100 110
Conversion [%]
Pd
PdAu-0.1
PdAu-1.8
PdAu-0.2
PdAu-2.1
PdAu-0.4
PdAu-0.8
It must be noticed that all the catalysts studied contain a fixed
percentage of 2 wt% Pd independently of the Au/Pd molar ratio.
This results in an overall increase in the total metal concentration
with growing Au content. In order to discuss a role of Au, initial
activities are normalized per mole of metal (TOFIN, min−1) in the
catalysts (Table 1) according to the procedure used in number of
studies [1,2,9,43,44]. The obtained specific activities indicate that
even the presence of a minor amount of Au (PdAu-0.1) reduces
the activity of pure Pd and subsequent increase in the Au con-
tent leads to gradual decrease in the Pd activity. A role of the Au
content in the reduction of Pd activity can be observed from the
relationship between the decrease of initial hydrogenation activ-
10
9
8
7
6
5
4
3
2
1
0
−1
ity (mol CAL min
g
−1) referred to the number of Au moles vs
cal
the content of Au (mol%) in the catalysts (Supplementary Fig. S6). A
very small amount of Au (9 mol%, PdAu-0.1) reduces Pd activity and
as the Au content grows the reduction in Pd activity progressively
grows up to the Au content ca. 40–45 mol%. At this composition cor-
responding to Au/Pd ∼0.8–1 molar ratio the decrease in Pd activity
is the highest. Larger amount of Au corresponding to Au/Pd > ∼1
also reduces activity of Pd but to lesser extent. It may be speculated
from this tendency that further growing of Au content at a fixed Pd
loading leading to Au/Pd ꢀ 1 may result in the PdAu particles of
activity comparable or even higher than that of Pd. This may sug-
gest a system consisting of Au in excess relative to that of Pd. In fact,
the nanostructured PtAu/SiO2 catalysts composed of fully dispersed
Pt-entities onto the surface of Au nanoparticles were observed to
0
10 20 30 40 50 60 70 80 90 100 110
Conversion [%]
Pd
PdAu-0.2
PdAu-2.1
PdAu-0.8
PdAu-1.8
Fig. 7. Selectivity to saturated aldehyde (HCAL), saturated alcohol (HCOL) and
propylobenzene (PB) vs conversion of cinnamaldehyde (CAL). Reaction carried out
on Pd/C and PdAu/C catalysts. For reaction conditions, see Fig. 6.

12
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
Table 3
The comparison of TOF data reported in the literature.
a
Conditions of CAL hydrogenation experiment
Catalyst
Pd [nm]
2.8–8.5
TOF [s−1
6.7
]
Ref.
2-Propanol/water; 60 ◦C; 4 MPa H2 pressure
Cyclohexane; 60 ◦C; 1 MPa H2 pressure
Pd/graphiteA
Pd/C activated carbon
[60]
[61]
2.8–3.3 90% HCAL;
irrespective of
Pd-particle size
Ethanol; 60 ◦C; 0.2 MPa H2 pressure
Toluene, 80 ◦C; 2 MPa H2 pressure
Pd/C, commercial
3.3
1.4 70% HCAL;
30%HCOL
0.25
0.23
0.20
0.16
0.21
0.3–0.5
0.3
[62]
[59]
0.5%Pd/FAC activated carbon
Pd/MC mesoporous carbon
Pd/CNF carbon nanofiber, powdered
Pd/CNF/TiO2 carbon nanofiber-wash coated TiO2-monolith-
Pd/CNF/TiO2 carbon nanofiber-washcoated TiO2 monolith
Pd/Al2O3
2.8
3.5
5.6
7.8
8.2
3
2-Propanol, 60 ◦C; 1 MPa H2
pressure
7
28
4
1.9
0.3
[63]
Pd/SiO2
54
3.4
a
The turnover frequency (TOF), number of CAL molecules converted per second and per surface Pd atom.
100
80
60
40
20
0
catalysts, the saturated alcohol (HCOL) was formed with selectiv-
ity higher than 98%. No isomerization to HCAL was noted, and the
formation of hydrogenolysis product, PB was also very low. The
monometallic 2%Pd/C catalyst has the highest activity while all
Au-containing catalysts are less active and at Au/Pd > 0.8 activity
drops drastically. Thus, the adsorption of isolated C C bonds of COL
reactant occurs on Pd as well as PdAu catalysts, but the presence of
Au strongly reduces reactivity to isolated C C bond hydrogenation,
especially in the case of Au-rich catalysts.
PdAu-1.8
PdAu-0.8
PdAu-0.4
PdAu-0.1
In next series of experiments, the hydrogenation was studied
using a mixture of CAL and unsaturated alcohol, COL (10 mol% COL
relative to CAL) on monometallic 2%Pd/C and Au-rich PdAu-1.8
catalyst. An affect of added COL on their reactivity can be com-
pared from graphs in Fig. 9. The hydrogen consumption curves and
the reactants distribution profiles show that on both studied cata-
lysts COL is preferentially reduced from the very beginning of the
experiment. This is evidenced by almost complete decrease of COL
content in the early stage of both reactions. It is also reflected by a
distinct break (indicated by the arrows in Fig. 9) on the hydrogen
consumption curves at the reaction time corresponding to the com-
plete conversion of COL. It is also observed that added COL reduces
the rate of CAL conversion on both Pd/C and PdAu-1.8 catalysts.
In the presence of added COL the selectivity to HCAL is greatly
diminished on Pd/C catalyst while it does not essentially change
on Au-rich PdAu-1.8 catalyst. Similar activity/selectivity effect due
to added COL observed on Ir/C catalyst was explained by a com-
petitive adsorption of COL with starting aldehyde (CAL) [67]. This
competitively adsorbed COL created a steric effect that inhibited
the planar adsorption of CAL (involving both C C and C O bonds)
and limited the formation of HCAL. A competitive adsorption of
unsaturated alcohol with the unsaturated aldehyde on Pt, Pd cat-
alysts was proved by the theoretical DFT calculation [68,69]. They
showed preferential adsorption of unsaturated alcohol through a
vertical atop geometry. On the other hand, saturated aldehyde can
not compete with the starting unsaturated aldehyde. Thus, COL
can alter the way in which the reactant (CAL) binds to the cata-
lyst surface. With increasing surface coverage of the COL (added or
formed during the reaction) the CAL becomes more hindered to flat
adsorptionthus preventingthe hydrogenationof C C bond yielding
saturated aldehyde. This effect can be clearly seen on Pd/C catalyst
(Fig. 9). However, there is no essential change in the selectivity to
HCAL on the PdAu-1.8 catalyst which itself offers distinctly lower
ability to HCAL formation because of less favorable planar mode of
CAL adsorption.
Pd/C
0
50
100
150
200
t [min]
Fig. 8. The influence of Au content in the PdAu/C catalysts on cinnamalcohol (COL)
hydrogenation. Reaction conditions: toluene, temperature 22 ◦C, catalyst concen-
tration 0.25 g/dm3, substrate COL concentration c0 = 0.05 mol/dm3.
be very active for the hydrogenation of ␣,␤-unsaturated carbonyl
compounds (CAL, furfural) [41,66].
An addition of Au has also an impact on the way CAL is reacted
(Fig. 7). In comparison with Pd/C, the changes in selectivity are not
so much strong on catalysts with low Au content, when Au/Pd is
below 0.8, and the hydrogenation of C C predominates over the
carbonyl group hydrogenation. However, with increasing Au/Pd
ratio from 0.1 to 0.4 a clear tendency reflecting in progressively
growing selectivity to saturated alcohol is observed. The selectivity
to saturated alcohol increases dramatically at the expense of that to
saturated aldehyde on Au-rich catalysts, when Au/Pd > 0.8. On both
Au-rich catalysts the selectivity to saturated alcohol (HCOL) attains
ca. 60% (vs 30% on Pd/C catalyst) whereas that to saturated aldehyde
HCAL is only 40% (compared to 65% on Pd/C). In their presence the
content of saturated alcohol (HCOL) exceeds that of saturated alde-
hyde (HCAL) during the whole CAL conversion (Supplementary Fig.
S5). The Au-content has also slight impact on the hydrogenolysis
reaction yielding propyl benzene (PB) (Fig. 7). On Au-rich catalysts
the ability to PB formation is somewhat smaller when compared
with small Au content – containing catalysts.
Next series of hydrogenation tests are performed using COL as
the substrate. They allow to compare reactivity of active sites in
monometallic Pd/C and PdAu/C catalysts for the hydrogenation of
isolated C C bond of COL reactant (Fig. 8). On all Pd/C and PdAu/C
Here, the stability of PdAu/C catalysts under CAL hydrogena-
tion reaction is also examined. In order to assess the stability
of catalysts, recycle tests were conducted using PdAu-0.1 and

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
13
PdAu-2.1 catalysts as follows: at the end of the first hydrogenation
run the catalyst was filtered off, washed with acetone-THF (1:1)
mixture and after drying in air was reused for the next reaction
under the same conditions. The reuse of both PdAu/C catalysts
recovered from the reaction mixture leads to almost reproducible
results in terms of both activity and selectivity (Supplementary
Fig. S7). The conversion remained stable during second cycle, and
the selectivity did not change essentially. Thus it is rational to
conclude about promising stability of studied PdAu/C catalysts
under used reaction conditions.
140
120
100
80
Pd/C
H2
60
Thus, the surface composition exhibits significant role in cat-
alytic behavior of PdAu/C catalysts not only in activity but also in
the C C/C O selectivity control in hydrogenation of CAL especially
in the case of Au-rich catalysts when the surface concentration of Au
is equal or even exceeds that of Pd (Table 2). 2%Au/C catalyst does
not exhibit activity at the present reaction conditions. The reason
could be the low capacity of gold to dissociate H2 which only occur
on gold surface defects (corners, kinks, etc.) [70]. Therefore, pal-
ladium could be considered to be essential in the hydrogenation
of CAL while the presence of gold assists the catalytic function of
palladium.
Similar effect, gradual decrease in the rate of C C hydrogena-
tion in 1,3-cyclooctadiene with growing Au content observed on
colloidal PdAu nanoparticles prepared by co-reduction (ethanol)
in the presence of PVP has been related to the presence of Au on
the surface since the activity in the C C hydrogenation could be
expected to be proportional to the number of surface Pd atoms
[3–5]. The surface composition of Au and Pd is according to the
authors determined by the nominal Au/Pd ratio with is decisive for
the type of microstructure existing in the PdAu particles. Simul-
taneous reduction (alcohol) at low temperature could produce
so-called core–shell structure resulting due to difference in the
reduction ability of Pd and Au. Since the reduction potential of
Au ions is more positive (E◦ (Au3+/Au) = + 1.0 V vs NHE) than that
of Pd ions (E◦ (Pd2+/Pd) = + 0.63 V vs NHE) the reduction of the Au
ions proceeds more easily than that of the Pd ions and the core
is Au-rich while surface shell is enriched by Pd. However, various
microstructures such as Au-core/Pd-shell, cluster-in-cluster or
random model could be formed as they are determined by the
Au/Pd ratio [5]. The most active catalyst was identified to be
Au/Pd = 1:4 (PdAu-0.25 in the present notation) with the surface
shell entirely composed of Pd atoms and a core composed of Au.
It exhibited enhanced activity due to electron interaction between
the Au-core and the surface shell Pd atoms. At higher Au content
the cluster-in-cluster or random models could be formed in which
both, Pd and Au atoms but in different relations could appear in
the surface shell. This implied that the presence of Au atoms on the
CAL + COL
40
COL
20
CAL
0
0
10 20 30 40 50 60 70 80 90 100
t [min]
140
120
100
80
PdAu-1.8
H2
60
CAL + COL
40
CAL
COL
20
0
0
10 20
30 40 50 60 70
80 90 100
t [min]
80
70
60
50
40
30
20
10
0
CAL
Pd/C
CAL + 10 % COL
CAL + 10 % COL
CAL
surface of colloidal AuPd particles resulted in lower activity for C
C
hydrogenation in 1,3-cyclooctadiene [4,5]. The same model has
been adopted to explain the mechanism of PdAu, PdPt and PtAu
nanoparticles formation by reverse “water-in-oil” microemul-
sion method [20–22]. Number of techniques demonstrated
no-complete “core–shell” structure of colloidal PdAu, PdPt, PtAu
particles formed at these conditions. In the case of PdAu particles,
both Pd and Au were detected in outer surface of colloidal particles
within wide range of composition. Hence, in view of such literature,
a microstructure of metal particles in the present PdAu/C catalysts
may be considered to be no-complete “core–shell” structure with
both Au and Pd in their outer surface. Consequently, as the surface
content of Au grows the hydrogenation activity of PdAu/C catalysts
gradually decreases. However, this growing surface content of Au
has an essential impact on the selectivity of CAL hydrogenation.
The selectivity to C O is remarkably promoted especially when the
Au content is comparable or higher than that of Pd. At Au/Pd > 1 the
HCAL formation is so strongly inhibited that the formation of satu-
rated alcohol is a dominating reaction. This may suggest a role of Au
PdAu-1.8
0
20
40
60
80
100
Conversion [%]
Fig. 9. The effect of added COL on the CAL hydrogenation in the presence of
Pd/C and PdAu-1.8 catalysts. Products distributions profiles and selectivity to sat-
urated aldehyde (HCAL). Reaction carried out on monometallic Pd/C and Au-rich
PdAu-1.8 catalysts. Reaction conditions: toluene, temperature 22 ◦C, catalyst con-
centration 5 g/dm3, CAL concentration c0 = 0.05 mol/dm3, initial COL concentration
c
0 = 0.005 mol/dm3.

14
T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
in the surface Au-rich catalysts in the geometry of CAL adsorption,
the phenomenon which is commonly taken into account to discuss
the selectivity effects theoretically studied by Delbecq et al. [68,69].
The aldehyde reactant can be adsorbed on a metal surface in a flat,
planar adsorption mode, parallel to the surface with both C C and
Pt-on-Au catalysts [40,41] and on a physical mixture of TiO2 sup-
ported Au and Pd catalysts [43] improved also strongly the activity,
but it exhibited tendency to facilitate the C C hydrogenation yield-
ing higher selectivity to saturated aldehyde (HCAL) [44].
In summary, the reverse “water-in-oil” microemulsion method
proves to be successful in the preparation of carbon supported
catalysts with dispersed and almost homogeneously distributed
PdAu nanoparticles with narrow size distributions. The PdAu par-
ticles in studied catalysts exhibited alloy structure and within the
whole Au/Pd ratios both components Pd and Au are detected in
their outer surface. The surface composition of these particles par-
ticularly in Au-rich catalysts played a role in the CAL molecule
adsorption resulting in enhanced selectivity to carbonyl group
hydrogenation.
On the other hand, it is well known that Pd–Au synergistic effect
observed by Yang et al. [39] is strongly dependent on the nano-
structure of bimetallic AuPd particles which may vary depending
on the method of preparation [72]. In particular thermal treatment
could play an important role as it facilitates an intermixing pro-
cess of Pd and Au components which may lead to restructuring of
surface composition of metal particles. It could also have a clear
impact on the chemical state and mutual interaction between the
components, e.g. Pd and Au [73].
In the present work the PdAu/C catalysts were neither cal-
cined nor hydrogen treated at high temperature after deposition
of pre-formed PdAu colloidal particles on the carbon support. The
preparation protocol described by Wu et al. [21] was preserved
and the PdAu/C catalysts were dried at temperature of 120 ◦C only.
It could be not excluded that thermal treatment would modify
microstructure of these PdAu/C catalysts because of an expected
thermally induced Pd–Au intermixing effect, similarly to what was
observed by Sarkany et al. [74] for colloidal PdAu nanoparticles
deposited on SiO2 and studied in the hydrogenation of acetylene.
The intermixing of Au and Pd upon thermal treatment significantly
enhanced the hydrogenation activity explained by thermally facil-
itated generation of gold diluted Pd–Au ensembles.
C
O bonds involved in adsorption and in a vertical (atop) geometry
via interaction with the surface by the oxygen atom (C O). The
first adsorption mode favors the hydrogenation of C C yielding
HCAL, while atop geometry promotes the formation of unsaturated
alcohol. Moreover, the CAL molecule contains a large phenyl group,
which can have a considerable steric effect. Because of repulsion
between metal and aromatic ring the parallel adsorption of CAL
molecule on metal surface could be to high extent hindered.
It is well known that the addition of second metal acting as a
modifier can result in an increasing as well as a decreasing activity.
The presence of ionic promoters directing the C O bond toward the
active sites, enhanced not only selectivity to unsaturated alcohol
but also the activity. Decreased activities have been ascribed to the
presence of high promoter concentrations resulting in a dramati-
cally decreasing hydrogenation capacity. For instance addition of
Ga, Sn and Ge improves selectivity whereas the activity decreased
[37]. The latter was also the case in the hydrogenation of CAL in
the presence of Sn-modified Pd/SiO2 [55] and Pd/C catalysts [71].
The decreased activity was related to lower number of active sites
because of Sn addition. Furthermore, as the content of Sn in the
Pd-Sn/SiO2 catalysts increased, their activity decreased accompa-
nied by a promotion of selectivity toward COL. From these results
the authors concluded about the changes in adsorption mode of
CAL due to electronic and geometric effects induced by the Sn.
The promotion of C O selectivity on present PdAu/C catalysts is
also accompanied by the reduced activity especially on the Au-rich
catalysts. At Au/Pd > 1 the HCAL formation is so strongly inhibited
that the formation of saturated alcohol is a dominating reaction
(Supplementary Fig. S5).
Therefore, the change of adsorption mode of CAL on Pd surface
induced by Au could be considered to be a main reason for the
enhanced selectivity toward COL. It can be supposed that when
the Au/Pd ratio grows, e.g. on the Au-rich surface the possibility of
flat-lying CAL adsorption mode becomes more and more inhibited.
Core-level XPS spectra indicate an electronic perturbation of pal-
ladium by gold (Supplementary Fig. S4) which may modify the Pd
sites both geometrically and electronically. However, these inter-
action may also provide new active sites which are able to active
the C O bond of CAL. The role of geometric dilution effect of Pd
atoms by alloyed Au could also be taken into consideration. Divid-
ing the Pd ensembles by Au into ensembles of smaller sizes would
prevent the CAL adsorption in the “parallel” mode that requires
multi-adsorption sites with consequent improving the selectivity
toward C O hydrogenation [55].
4. Conclusions
The liquid phase hydrogenation of cinnamaldehyde was stud-
ied on carbon (Vulcan XC72) supported PdAu catalysts with Au/Pd
atomic ratio from 0.1 up to 2.1 prepared by colloid-based reverse
“water-in-oil” microemulsion method. Accordingly, pre-formed
metal nanoparticles of controlled size with a narrow size distribu-
tion were deposited on the carbon support. In all catalysts the metal
particles are spherically shaped, nearly monodispersed and well
distributed throughout the carbon support. The size of metal parti-
cles was dependent on the Au-content expressed by the Au/Pd ratio.
The particles of low Au-content (Au/Pd < 0.8) are of smaller size than
Pd (6.7 nm), but the particle size increases and at Au/Pd = 2.1 it is
only slightly lower than that of Au (8.2 nm).
The content of Au (Au/Pd ratio) has impact on activity and selec-
tivity for the cinnamaldehyde hydrogenation. The C C group of
cinnamaldehyde was preferentially hydrogenated compared to the
carbonyl (C O) group in the presence of Pd/C catalyst. At low Au
content (Au/Pd < 0.8) the effect of Au was relatively weak observed
as a small decrease in activity with only small increase in selec-
tivity to C O hydrogenation. The effect of Au manifested strongly
on Au-rich catalysts (Au/Pd > 1) giving distinctly reduced reactiv-
ity to unsaturated double C C bond hydrogenation and selectivity
preference to the carbonyl group reduction C O. The change of
adsorption mode of CAL on the surface could be the main reason
of significantly limited C C adsorption on Au-rich catalysts. These
activity/selectivity effects induced by the presence of Au are dis-
cussed to arise from microstructure of PdAu alloy particles which
The metal particle size is another feature which could enhance
the selectivity toward unsaturated alcohol and larger particles of Pt,
Pd, Ru, Co, resulted in higher selectivity to C O hydrogenation [37].
The effect of metal particle size was related to the phenyl group,
sterically hindering the approach of the C C moiety to the surface
of large particle leading to the preferential adsorption of C O. This
selectivity improvement is due to the lower probability of the C
bond activation rather than to an increased activation of the C
C
O
bond. The size of metal particle increases in the PdAu/C catalysts
in particular for Au-rich catalysts, e.g. when the Au/Pd > 0.8. Thus,
it cannot be excluded that such metal particle size effect could also
contribute to the catalytic performance of studied PdAu/C catalysts.
The benefit brought by alloying Pd–Au in the PdAu/silica cat-
alysts with low Au content (Au/Pd molar ratio of 0.2:1) was
recognized to be enhanced activity as compared with Pd-catalyst
whereas it did not essentially change the selectivity of CAL hydro-
genation [39]. Hydrogen spillover effect acting on nanostructured

T. Szumełda et al. / Applied Catalysis A: General 487 (2014) 1–15
15
manifests distinctly in various surface concentrations of Au rela-
tive to that of Pd. It can be supposed that when the Au/Pd ratio
is high, e.g. on the Au-rich surface the possibility of cinnamalde-
hyde molecule adsorption in a “flat-lying CAL” geometry becomes
more and more inhibited. Core-level XPS spectra indicate electronic
perturbation of palladium by gold. It might be thought that the
Pd–Au interactions modify the Pd sites both geometrically and
electronically and it resulted in significantly limited C C bond
adsorption. However, these interaction may also provide new
active sites which are able to active the C O bond of CAL.
[25] S. Devarajan, P. Bera, S. Sampath, J. Colloid Interface Sci. 290 (2005) 117–129.
[26] U. Nylen, J.F. Delgado, S. Jaras, M. Boutonnet, Appl. Catal. A 262 (2004) 189–200.
[27] R. Touroude, P. Girard, G. Maire, J. Kizling, M. Boutonnet-Kizling, P. Stenius,
Colloid Surf. 67 (1992) 9–12.
[28] S.J. Chiang, B.J. Liaw, Y.Z. Chen, Appl. Catal. A 319 (2007) 144–152.
[29] B.A. Cheney, J.A. Lauterbach, J.G. Chen, Appl. Catal. A 394 (2011) 41–47.
[30] A. Di Wang, F. Villa, L. Porta, D.S. Prati, J. Phys. Chem. C 112 (2008) 8617–8622.
[31] D.I. Enache, J.K. Edwards, P. Landon, B. Solsona-Espira, A.F. Carley, A.A. Herz-
ing, M. Watanabe, C.J. Kiely, D.W. Knight, G.J. Hutchings, Science 311 (2006)
362–365.
[32] M. Bonarowska, J. Pielaszek, V.A. Semikolevov, Z. Karpin´ ski, J. Catal. 209 (2002)
528–538.
[33] A. Sarkany, A. Horvath, A. Beck, Appl. Catal. A 229 (2002) 117–125.
[34] A. Sarkany, O. Geszti, G. Safran, Appl. Catal. A 350 (2008) 157–163.
[35] A.C. Krauth, G.H. Bernstein, E.E. Wolf, Catal. Lett. 45 (1997) 177–186.
[36] G. Schmid, A. Lehnert, J.O. Halm, J.O. Bovin, Angew. Chem. Int. Ed. Engl. 30 (1991)
873–874.
Acknowledgements
[37] P. Gallezot, D. Richard, Catal. Rev. Sci. Eng. 40 (1998) 81–126.
[38] P. Maki-Arvela, J. Hajek, T. Salmi, D.Y. Murzin, Appl. Catal. A 292 (2005) 1–49.
[39] X. Yang, D. Chen, S. Liao, H. Song, Y. Li, Z. Fu, Y. Su, J. Catal. 291 (2012) 36–43.
[40] Y.C. Hong, K.Q. Sun, G.R. Zhang, R.Y. Zhong, B.Q. Xu, Chem. Commun. 47 (2011)
1300–1302.
[41] K.Q. Sun, Y.C. Hong, G.R. Zhang, B.Q. Xu, ACS Catal. 1 (2011) 1336–1346.
[42] H. Shi, N. Xu, D. Zhao, B.Q. Xu, Catal. Commun. 9 (2008) 1949–1954.
[43] R. Liu, Y. Yu, K. Yoshida, G. Li, H. Jiang, M. Zhang, F. Zhao, S. Fujita, M. Arai, J.
Catal. 269 (2010) 191–200.
This work was partially supported by the National Centre for
Research and Development, Poland, Grant no. SP/J/7/170071/12.
We are grateful to Dr Elz˙ bieta Bielan´ ska for taking SEM images and
Prof. Lidia Lityn´ ska-Dobrzyn´ ska for registering HRTEM images.
Appendix A. Supplementary data
[44] H. Gu, X. Xu, A. Chen, P. Ao, X. Yan, Catal. Commun. 41 (2013) 65–69.
[45] E. Antolini, E.R. Gonzales, Appl. Catal. A 365 (2009) 1–19.
[46] A. Drelinkiewicz, A. Waksmundzka, W. Makowski, J.W. Sobczak, A. Król, A.
Zieba, Catal. Lett. 94 (2004) 143–156.
[47] A. Drelinkiewicz, M. Hasik, J. Mol. Catal. A 177 (2001) 149–164.
[48] S. An, J.-H. Park, C.-H. Shin, J. Joo, E. Ramsaamy, J. Hwang, J. Lee, Carbon 49
(2011) 1108–1117.
Supplementary data associated with this article can be
found, in the online version, at http://dx.doi.org/10.1016/j.apcata.
2014.08.036.
References
[49] R. Sellin, J.-M. Clacens, C. Coutanceau, Carbon 48 (2010) 2244–2254.
[50] X. Wei, X.F. Yang, A.Q. Wang, L. Li., X.Y. Liu, T. Zhang, C.Y. Mou, J. Li, J. Phys.
Chem. C 116 (2012) 6222–6233.
[51] G. Zhang, Y. Wang, X. Wang, Y. Chen, Y. Zhou, Y. Tang, L. Lu, J. Bao, T. Lu, Appl.
Catal. B 102 (2011) 614–619.
[52] A.F. Lee, C.J. Baddeley, C. Hardacre, R.M. Ormerod, R.M. Lambert, J. Phys. Chem.
99 (1995) 6096–6102.
[53] NIST Standard Reference Database 20, Version 3.2 (Web Version), National
Institute of Standards and Technology, Gaithersburg, 2001.
[54] T. Kobayashi, Appl. Catal. B 30 (2001) 287–292.
[55] A. Hammoudeh, S. Mahmoud, J. Mol. Catal. A 203 (2003) 231–239.
[56] S. Mahmoud, A. Hammoudeh, S. Gharaibeh, J. Melsheimer, J. Mol. Catal. A 178
(2002) 161–167.
[57] G.R. Cairns, R.J. Cross, D. Stirling, J. Catal. 166 (1997) 89–97.
[58] S. Bhogeswararao, D. Srinivas, Catal. Lett. 140 (2010) 55–64.
[59] J. Zhu, Y. Jia, M. Li, M. Lu, J. Zhu, Ind. Eng. Chem. Res. 52 (2013) 1224–1233.
[60] A. Giroir-Fendler, D. Richard, P. Gallezot, Stud. Surf. Sci. Catal. 41 (1988)
171–178.
[61] A. Cabiac, T. Cacciaguerra, P. Trens, R. Durand, G. Delahay, A. Medevielle, D. Plee,
B. Coq, Appl. Catal. A 340 (2008) 229–235.
[62] W.M.S. Ide, B. Hao, M. Neurock, R.J. Davis, ACS Catal. 2 (2012) 671–683.
[63] M. Lasdaf, A.O.I. Krause, M. Linbad, M. Tiitta, T. Venalainen, Appl. Catal. A 241
(2003) 65–75.
[1] J.A. Lopez-Sanchez, N. Dimitratos, N. Glanville, L. Kesavan, C. Hammond, J.K.
Edwards, A.F. Carley, Ch.J. Kiely, G.J. Hutchings, Appl. Catal. A 291 (2011)
400–406.
[2] N.E. Kolli, L. Delannoy, C. Louis, J. Catal. 297 (2013) 79–92.
[3] N. Toshima, M. Harada, Y. Yamazaki, K. Asakura, J. Phys. Chem. 96 (1992)
9927–9933.
[4] T. Yonezawa, N. Toshima, J. Chem. Soc. Faraday Trans. 91 (1995) 4111–4120.
[5] N. Toshima, T. Yonezawa, N. J. Chem. (1998) 1179–1201.
[6] A.M. Venezia, V. La Oarola, D. Deganello, B. Pawelec, J.L.G. Fierro, J. Catal. 215
(2003) 317–325.
[7] A.M. Venezia, L.F. Liotta, G. Pantaleo, V. La Parola, G. Deganello, A. Beck, Z.
Koppany, K. Frey, D. Horvath, L. Guczi, Appl. Catal. A 251 (2003) 359–368.
[8] B. Pawelec, A.M. Venezia, V. La Parola, E. Cano-Serrano, J.M. Campos-Martin,
J.L.G. Fierro, Appl. Surf. Sci. 242 (2005) 380–391.
[9] J. Pritchard, L. Kesavan, M. Piccinini, Q. He, R. Tiruvalam, N. Dimitratos, J.A.
Lopez-Sanchez, A.F. Carley, J.K. Edwards, Ch.J. Kiely, G.J. Hutchibs, Langmuir 26
(2010) 16568–16577.
[10] S. Marx, F. Krumeich, A. Baiker, J. Phys. Chem. C 115 (2011) 8195–8205.
[11] M. Boutonnet, S. Logdberg, E.E. Svensson, Curr. Opin. Colloid Interface Sci. 13
(2008) 270–286.
[12] M.J. Schuger, K. Stickdorn, Chem. Rev. 95 (1995) 849–864.
[13] S. Erikson, U. Nylen, S. Rojas, M. Boutonnet, Appl. Catal. A: Gen. 265 (2004)
207–219.
[14] H.H. Ingelsten, J.-C. Beziat, K. Bergvist, A. Palmqvist, M. Skoglundh, H. Quiuhong,
L.K.L. Falk, K. Holmerg, Lamgmuir 18 (2002) 1811–1818.
[15] J. Rymes, G. Ehret, L. Hilaire, M. Boutonnet, K. Jiratova, Catal. Today 75 (2002)
297–303.
[16] A. Martinez, G. Prieto, Catal. Commun. 8 (2007) 1479–1486.
[17] A. Martinez, G. Prieto, J. Catal. 245 (2007) 470–476.
[18] R. Kosydar, M. Góral, J. Gurgul, A. Drelinkiewicz, Catal. Commun. 22 (2012)
58–67.
[19] M. Góral-Kurbiel, A. Drelinkiewicz, R. Kosydar, B. Debinska, P.J. Kulesza, J. Gur-
gul, Electrocatalysis 5 (2014) 23–40.
[20] M.-L. Wu, D.-H. Chen, T.-C. Huang, Chem. Mater. 13 (2001) 599–606.
[21] M.-L. Wu, D.-H. Chen, T.-C. Huang, Langmuir 17 (2001) 3877–3883.
[22] M.-L. Wu, D.-H. Chen, T.-C. Huang, J. Colloid Interface Sci. 243 (2001) 102–108.
[23] Z. Liu, J.Y. Lee, M. Han, W. Chen, L.M. Gan, J. Mater. Chem. 12 (2002) 2453–2458.
[24] L. Xiong, A. Manthiram, Electrochim. Acta 50 (2005) 2323–2329.
[64] M.L. Toebes, Y. Zhang, J. Hajek, T.A. Nijhuis, J.H. Bitter, A.J. van Dillen, D.Y.
Murzin, D.C. Koningsberger, K.P. de Jong, J. Catal. 226 (2004) 215–225.
[65] Z.-T. Liu, C.-X. Wang, Z.-W. Liu, J. Lu, Appl. Catal. A 344 (2008) 114–123.
[66] Y.-C. Hong, K.-Q. Sun, G.-R. Zhang, R.-Y. Zhong, B.-Q. Xu, Chem. Commun. 47
(2011) 1300–1302.
[67] J.P. Breen, R. Burch, J. Gomez-Lopez, K. Griffin, M. Hayes, Appl. Catal. A 268
(2004) 267–274.
[68] F. Delbecq, P. Sautet, J. Catal. 152 (1995) 217–236.
[69] F. Delbecq, P. Sautet, J. Catal. 211 (2002) 398–406.
[70] E. Bus, J.T. Miller, J.A. van Bakhoven, J. Phys. Chem. B 109 (2005) 14581–14587.
[71] J. Zhao, X. Xu, X. Li, J. Wang, Catal. Commun. 43 (2014) 97–102.
[72] A. Wang, X.Y. Liu, C.-Y. Mou, T. Zhang, J. Catal. (2013),
http://dx.doi.org/10.1016/j.jcat.2013.08.023.
[73] V. Ponec, Appl. Catal. A 149 (1997) 27–48.
[74] A. Sarkany, P. Hargittai, A. Horvath, Top. Catal. 46 (2007) 121–128.